Back to COVID-19 Visualisation

COVID-19 Commentary Archive

I’ve been posting commentary on Facebook about the “days to 1%” line in my Google Sheet linked above. Here is a sample of what I’ve written.

Boilerplate explanation paragraph: The number given with each country is my own way of boiling down your region or country's growth/doubling rate and number of people infected with COVID-19 into one number. It represents the number of days you have left at your current rate of infection (averaged over the last four days) before you reach 1%. It's not a forecast (your rate of change is likely to change before then, when public health authorities take positive or negative action), just one simple number that tells you how you're doing compared to others. I could have picked a bigger percentage than 1%, but you can't go too far before the exponential growth bends back down logistically, and it seemed as easy a place as any; most people can visualize the idea of 1 in 100. Starting on 2020-05-06, I stopped displaying these numbers for regions whose active case counts were known to be lower than an estimated previous peak value; in these cases, I show separately the number of days it would take the current 4-day average of exponential decay in active case counts to drop below half a person, with a floor of 14 days since the last new confirmed case, as well (starting 2020-05-14) as the current percentage of peak active cases.

2020-06-03

Comment on this post on Facebook.

today's daily chart as described below in text

Today's pre-peak days-to-one-percent infection numbers: Brazil (25→30), Alabama (40→40), California (46→48), U.S. (46→49), South Africa (50→57), Sweden (64→64), Mexico (76→74), Iran (95→93), India (93→94), Ontario (132→123), Indonesia (198→208), Poland (193→210), Israel (346→318), Spain (321→406), France (296→541), South Korea (1253→1172), Norway (1731→1891), Australia (1877→2116), China (63379→77785)

Compared to yesterday, here's how these regions are doing.

Much worse: Ontario, Israel, South Korea

Worse: Mexico, Iran

Better: Alabama, California, U.S., Sweden, India, Indonesia, Poland, Norway

Much better: Brazil, South Africa, Spain, France, Australia, China

Today's post-peak days-to-eradication numbers for regions of interest: New Zealand (3→2, 0%), Malta (51→39, 21.1%), Austria (75→83, 4.9%), Thailand (392→84, 4.1%), Denmark (405→174, 17.4%), Japan (247→207, 15.2%), Louisiana (158→220, 34.4%), Germany (406→251, 11.9%), Italy (305→309, 47.2%), Malaysia (6050→467, 49.8%), Singapore (615→506, 87.9%), Toronto (623→584, 63.5%), Romania (762→870, 65.7%), U.K. (1661→1946, 86.5%), Turkey (1573→2236, 44.3%), Canada (9369→4283, 99.0%), New York (4237→10310, 70.9%)

Days to eradication for other post-peak countries: Belize, Eritrea, Papua New Guinea, Saint Kitts and Nevis, Saint Lucia, Seychelles, Timor-Leste (eradicated); Montenegro (4, 0.0%); Taiwan (23, 3.4%); Vietnam (33, 19.0%); Niger (36, 9.2%); Croatia (46, 5.2%); Zambia (50, 27.1%); Morocco (54, 38.7%); Luxembourg (57, 2.8%); Chad (66, 43.8%); Mongolia (73, 72.7%); Switzerland (93, 3.2%); Slovakia (94, 14.8%); Equatorial Guinea (129, 79.9%); Eswatini (132, 70.5%); Saint Vincent and the Grenadines (142, 91.7%); Syria (174, 88.2%); Qatar (177, 78.4%); Bosnia and Herzegovina (223, 46.6%); Togo (224, 83.3%); Jordan (289, 79.7%); Guyana (311, 90.8%); Estonia (342, 19.3%); Sierra Leone (362, 84.1%); Panama (369, 88.9%); Tunisia (374, 13.6%); Georgia (389, 50.0%); Dominican Republic (396, 89.9%); Kosovo (423, 36.8%); Kuwait (435, 89.3%); Algeria (568, 91.5%); South Sudan (578, 92.2%); Lithuania (609, 49.5%); Saudi Arabia (620, 78.9%); Bulgaria (679, 83.5%); Hungary (795, 78.5%); Latvia (924, 66.9%); Serbia (1009, 64.0%); Uruguay (1223, 37.0%); Guinea (1372, 94.7%); Guinea-Bissau (1478, 70.3%); Tajikistan (1789, 96.9%); Gabon (1846, 98.3%); Belgium (2304, 68.8%); San Marino (3419, 63.6%); Ecuador (3606, 60.9%); Czechia (3638, 57.1%)

Days to eradication for other post-peak U.S. states and territories: MP (27, 45.5%); ND (109, 51.9%); MN (111, 72.1%); IN (151, 75.8%); VT (154, 9.6%); MI (165, 69.7%); MS (249, 75.5%); WY (358, 89.7%); ME (410, 89.2%); SD (659, 72.9%); NC (727, 79.3%); CT (1092, 85.1%); WV (1381, 98.0%); DC (1587, 93.3%); PA (1624, 73.8%); WI (1929, 97.4%); NJ (2051, 79.0%); DE (2674, 92.1%); KS (2782, 80.5%); RI (2804, 91.0%); MD (4264, 97.2%); IL (6455, 93.5%); NM (12963, 97.9%)

Days to eradication for other post-peak Canadian provinces and territories: NT, PE, YT (eradicated); NL (14, 1.0%); NS (14, 1.1%); SK (28, 15.8%); AB (54, 12.1%); MB (82, 6.2%); BC (249, 28.9%); QC (1099, 91.9%)

Brazil's improvement is part of its weekly reporting cycle. Week-on-week, it's still worsening, with the days-to-one-percent being 36 a week ago (and 30 now).

Iran has gone in a week from fewer than 2,000 new cases a day to more than 3,000 new cases a day.

Ontario had a third straight day of increasing new case numbers: 323, 326, 404, 446. There were only three days last month when we had as many as 446 new cases in a day, but today at least it could be part of the 700 new cases that were discovered yesterday. If the numbers go back down by the end of the week, we should know.

Toronto also has had three straight increasing new case days: 123, 150, 164, 175; and I would reason the same way about those rediscovered missing cases, some of which must be from Toronto. New cases were reported for the first time since neighbourhood records were published in three more neighbourhoods: East End-Danforth, Lawrence Park North, and Long Branch. That first one is two neighbourhoods away from ours, so it feels like it's getting closer. No changes in cases reported in 13 of 140 neighbourhoods: Alderwood, Bayview Village, Blake-Jones, Bridle Path-Sunnybrooke-York Mills, Danforth Village - Toronto, Eringate-Centennial-West Deane, Kingsway South, Lambton Baby Point, North Riverdale, Playter Estates-Danforth, Runnymede-Bloor West Village, Woodbine Corridor, Yonge-Eglinton.

Among the countries in the world with the ten largest numbers of active cases (U.S., Brazil, Russia, India, Peru, U.K., Chile, Spain, France, and Pakistan), the U.K. is alone in appearing to have active numbers that are trending downward. In the next ten (Italy, Bangladesh, Canada, Turkey, Iran, Qatar, Saudi Arabia, Belarus, Colombia, Egypt), Italy, Canada, Turkey, Qatar, and Saudi Arabia are trending downward. The first ten countries account for about 3/4 of the cases in the world, and then next ten about 1/9.

Today's graphic shows the 18 countries in the world with at least 1 in 1,000 of their population currently an active COVID-19 case, whose active case numbers are still increasing.

Stay safe.

2020-06-02

Comment on this post on Facebook.

today's daily chart as described below in text

Today's pre-peak days-to-one-percent infection numbers: Brazil (25→30), Alabama (40→40), California (46→48), U.S. (46→49), South Africa (50→57), Sweden (64→64), Mexico (76→74), Iran (95→93), India (93→94), Ontario (132→123), Indonesia (198→208), Poland (193→210), Israel (346→318), Spain (321→406), France (296→541), South Korea (1253→1172), Norway (1731→1891), Australia (1877→2116), China (63379→77785)

Compared to yesterday, here's how these regions are doing.

Much worse: Ontario, Israel, South Korea

Worse: Mexico, Iran

Better: Alabama, California, U.S., Sweden, India, Indonesia, Poland, Norway

Much better: Brazil, South Africa, Spain, France, Australia, China

Today's post-peak days-to-eradication numbers for regions of interest: New Zealand (3→2, 0%), Malta (51→39, 21.1%), Austria (75→83, 4.9%), Thailand (392→84, 4.1%), Denmark (405→174, 17.4%), Japan (247→207, 15.2%), Louisiana (158→220, 34.4%), Germany (406→251, 11.9%), Italy (305→309, 47.2%), Malaysia (6050→467, 49.8%), Singapore (615→506, 87.9%), Toronto (623→584, 63.5%), Romania (762→870, 65.7%), U.K. (1661→1946, 86.5%), Turkey (1573→2236, 44.3%), Canada (9369→4283, 99.0%), New York (4237→10310, 70.9%)

Days to eradication for other post-peak countries: Belize, Eritrea, Papua New Guinea, Saint Kitts and Nevis, Saint Lucia, Seychelles, Timor-Leste (eradicated); Montenegro (4, 0.0%); Taiwan (23, 3.4%); Vietnam (33, 19.0%); Niger (36, 9.2%); Croatia (46, 5.2%); Zambia (50, 27.1%); Morocco (54, 38.7%); Luxembourg (57, 2.8%); Chad (66, 43.8%); Mongolia (73, 72.7%); Switzerland (93, 3.2%); Slovakia (94, 14.8%); Equatorial Guinea (129, 79.9%); Eswatini (132, 70.5%); Saint Vincent and the Grenadines (142, 91.7%); Syria (174, 88.2%); Qatar (177, 78.4%); Bosnia and Herzegovina (223, 46.6%); Togo (224, 83.3%); Jordan (289, 79.7%); Guyana (311, 90.8%); Estonia (342, 19.3%); Sierra Leone (362, 84.1%); Panama (369, 88.9%); Tunisia (374, 13.6%); Georgia (389, 50.0%); Dominican Republic (396, 89.9%); Kosovo (423, 36.8%); Kuwait (435, 89.3%); Algeria (568, 91.5%); South Sudan (578, 92.2%); Lithuania (609, 49.5%); Saudi Arabia (620, 78.9%); Bulgaria (679, 83.5%); Hungary (795, 78.5%); Latvia (924, 66.9%); Serbia (1009, 64.0%); Uruguay (1223, 37.0%); Guinea (1372, 94.7%); Guinea-Bissau (1478, 70.3%); Tajikistan (1789, 96.9%); Gabon (1846, 98.3%); Belgium (2304, 68.8%); San Marino (3419, 63.6%); Ecuador (3606, 60.9%); Czechia (3638, 57.1%)

Days to eradication for other post-peak U.S. states and territories: MP (27, 45.5%); ND (109, 51.9%); MN (111, 72.1%); IN (151, 75.8%); VT (154, 9.6%); MI (165, 69.7%); MS (249, 75.5%); WY (358, 89.7%); ME (410, 89.2%); SD (659, 72.9%); NC (727, 79.3%); CT (1092, 85.1%); WV (1381, 98.0%); DC (1587, 93.3%); PA (1624, 73.8%); WI (1929, 97.4%); NJ (2051, 79.0%); DE (2674, 92.1%); KS (2782, 80.5%); RI (2804, 91.0%); MD (4264, 97.2%); IL (6455, 93.5%); NM (12963, 97.9%)

Days to eradication for other post-peak Canadian provinces and territories: NT, PE, YT (eradicated); NL (14, 1.0%); NS (14, 1.1%); SK (28, 15.8%); AB (54, 12.1%); MB (82, 6.2%); BC (249, 28.9%); QC (1099, 91.9%)

Brazil's improvement is part of its weekly reporting cycle. Week-on-week, it's still worsening, with the days-to-one-percent being 36 a week ago (and 30 now).

Iran has gone in a week from fewer than 2,000 new cases a day to more than 3,000 new cases a day.

Ontario had a third straight day of increasing new case numbers: 323, 326, 404, 446. There were only three days last month when we had as many as 446 new cases in a day, but today at least it could be part of the 700 new cases that were discovered yesterday. If the numbers go back down by the end of the week, we should know.

Toronto also has had three straight increasing new case days: 123, 150, 164, 175; and I would reason the same way about those rediscovered missing cases, some of which must be from Toronto. New cases were reported for the first time since neighbourhood records were published in three more neighbourhoods: East End-Danforth, Lawrence Park North, and Long Branch. That first one is two neighbourhoods away from ours, so it feels like it's getting closer. No changes in cases reported in 13 of 140 neighbourhoods: Alderwood, Bayview Village, Blake-Jones, Bridle Path-Sunnybrooke-York Mills, Danforth Village - Toronto, Eringate-Centennial-West Deane, Kingsway South, Lambton Baby Point, North Riverdale, Playter Estates-Danforth, Runnymede-Bloor West Village, Woodbine Corridor, Yonge-Eglinton.

Among the countries in the world with the ten largest numbers of active cases (U.S., Brazil, Russia, India, Peru, U.K., Chile, Spain, France, and Pakistan), the U.K. is alone in appearing to have active numbers that are trending downward. In the next ten (Italy, Bangladesh, Canada, Turkey, Iran, Qatar, Saudi Arabia, Belarus, Colombia, Egypt), Italy, Canada, Turkey, Qatar, and Saudi Arabia are trending downward. The first ten countries account for about 3/4 of the cases in the world, and then next ten about 1/9.

Today's graphic shows the 18 countries in the world with at least 1 in 1,000 of their population currently an active COVID-19 case, whose active case numbers are still increasing.

Stay safe.

2020-06-01

Comment on this post on Facebook.

today's daily chart as described below in text

Today's pre-peak days-to-one-percent infection numbers: Brazil (25→25), Alabama (40→40), California (46→46), U.S. (46→46), South Africa (50→50), Sweden (64→64), India (93→93), Iran (105→95), Poland (195→193), Indonesia (178→198), France (149→296), Spain (209→321), Israel (432→346), South Korea (1091→1253), Norway (1271→1731), Australia (1808→1877), China (77782→63379)

Compared to yesterday, here's how these regions are doing.

Much worse: Iran, Israel, China

Worse: Poland

Better: Brazil, Alabama, California, U.S., South Africa, Sweden, India, Australia

Much better: Indonesia, France, Spain, South Korea, Norway

Today's post-peak days-to-eradication numbers for regions of interest: New Zealand (4→3, 0%), Malta (48→51, 27.0%), Austria (65→75, 5.0%), Louisiana (new→158, 32.3%), Japan (172→247, 16.1%), Italy (237→305, 48.9%), Thailand (244→392, 4.1%), Denmark (296→405, 18.6%), Germany (514→406, 12.9%), Singapore (781→615, 89.0%), Toronto (new→623, 63.0%), Romania (387→762, 65.6%), Turkey (1236→1573, 44.6%), U.K. (1533→1661, 86.6%), New York (4145→4237, 70.7%), Malaysia (626→6050, 51.5%), Canada (18097→9369, 99.4%), Ontario (2966→24539, 80.8%), Mexico (new→56496, 96.1%)

Days to eradication for other post-peak countries: Belize, Dominica, Eritrea, Papua New Guinea, Saint Kitts and Nevis, Saint Lucia, Seychelles, Timor-Leste (eradicated); Montenegro (5, 0.0%); Iceland (14, 0.2%); Andorra (21, 3.8%); Slovenia (25, 0.9%); Taiwan (26, 3.4%); Niger (31, 9.6%); Gambia (37, 33.3%); Luxembourg (49, 2.9%); Zambia (50, 27.1%); Vietnam (51, 22.2%); Croatia (59, 6.2%); Chad (81, 46.2%); Slovakia (93, 15.2%); Switzerland (110, 3.1%); Uruguay (122, 37.3%); Morocco (126, 53.7%); Estonia (134, 20.1%); Cameroon (151, 78.7%); Tunisia (153, 13.4%); Sierra Leone (158, 78.2%); Eswatini (166, 72.7%); Mongolia (227, 72.7%); Guyana (228, 92.0%); Qatar (248, 80.9%); Jordan (254, 81.7%); Togo (295, 85.0%); Bosnia and Herzegovina (302, 48.0%); Panama (333, 83.7%); Saudi Arabia (353, 77.7%); Ecuador (358, 57.3%); South Sudan (364, 92.2%); Kosovo (423, 36.8%); Bulgaria (437, 83.4%); Lithuania (441, 50.9%); Syria (487, 94.7%); Dominican Republic (531, 90.1%); Hungary (539, 77.2%); Greece (636, 61.5%); Algeria (796, 93.5%); Serbia (1075, 64.0%); San Marino (1227, 63.6%); Kuwait (1256, 92.5%); Burkina Faso (1489, 23.1%); West Bank and Gaza (1489, 33.9%); Mozambique (2208, 96.0%); Gabon (2333, 92.7%); Belgium (2484, 69.1%); Czechia (3611, 56.7%)

Days to eradication for other post-peak U.S. states: MP (15, 36.4%); VT (63, 8.5%); ND (105, 55.9%); IN (155, 76.1%); MI (180, 68.9%); MN (301, 83.9%); MS (309, 71.0%); WY (626, 89.7%); ME (684, 94.4%); CT (739, 85.1%); PA (869, 73.4%); MA (1818, 84.7%); IL (2247, 92.6%); RI (2315, 91.0%); NJ (2580, 79.2%); DE (2631, 93.2%); NC (2993, 74.6%); DC (3403, 96.0%); IA (74327, 97.9%)

Days to eradication for other post-peak Canadian provinces: NT, PE, YT (eradicated); NS (34, 2.8%); MB (36, 5.2%); AB (55, 12.8%); SK (74, 22.5%); BC (356, 31.4%); QC (1515, 92.8%)

I had to reenter New Zealand's number manually, as they've reached a stage in their recovery that revealed a bug in my code. They have had no new cases now for 11 days, and have one last remaining active case. Well done.

Today's graphic celebrates New Zealand, by plotting its trajectory on a new cases per active case vs. active cases per population phase diagram, compared to Canada, Italy, Japan and the U.S.

CBC News reported, and Ontario Health did not dispute that approximately 700 positive test results at a western Toronto hospital network were not reported to public health authorities over the past several weeks, and will likely result in a technical bump in new cases here as the numbers enter the database. The test samples were collected at the William Osler Hospital Network in Etobicoke and Brampton, and sent to Mount Sinai Hospital for analysis. WOH staff mistakenly thought that MSH staff would report results to public health.

Numbers are in fact up already today in Ontario, but not in Toronto. 3,834 cases now in Ontario, up from yesterday's 3,783. In Toronto, we had a third straight day's decline, from 2,005 eventually to today's 1,880, 63% of our peak and dropping at a rate fast enough to see us theoretically clear of COVID-19 in a mere two years. In Ontario, we are back up over 400 new cases today for the first time in more than a week; in Toronto, new cases have risen for a second day, so the active cases will start rising too, unless recoveries and deaths rise miraculously too.

There was media concern about the possible effect of recent mass public protests against police violence against black people; I did not see any photos however of protesters who weren't masked and maintaining a safe social distance.

Of further concern was the inadequacy of the Ontario Dental Association's response to the provincial government's permission for them to reopen. There is reportedly no way for dentists to obtain N95 masks and other professional PPE, which are being reserved for hospital use. Dentists are also being instructed to screen patients by questionnaire, despite the demonstrated inability of questionnaires to detect asymptomatic patients who are shedding SARS-CoV-2.

There were also reports that a large majority of community-acquired cases of COVID-19 were being spread in illegally large family gatherings, and in workplaces with inadequate protections for employees; the chances of contracting the disease randomly on the street sound quite low.

Three cases showed up in Toronto neighbourhoods which had shown no recent COVID-19 activity: Leaside-Bennington, Milliken, and Steeles. 16 neighbourhoods have shown no activity since neighbourhood stats started being published last week. In the most dangerous neighbourhood in the city, Thistletown-Beaumond Heights, cumulative confirmed cases have risen from 32 to 48 in five days. (Anyone else who is scraping the Toronto neighbourhood COVID-19 data might want to take note that the spellings of four neighbourhood names changed today.)

Stay safe.

2020-05-31

Comment on this post on Facebook.

today's daily chart as described below in text

Today's pre-peak days-to-one-percent infection numbers: Louisiana (14→13), Brazil (24→25), Alabama (35→40), California (53→46), U.S. (48→46), South Africa (51→50), Sweden (64→64), Toronto (66→66), Mexico (67→71), India (95→93), Iran (105→105), France (201→149), Indonesia (175→178), Poland (169→195), Spain (187→209), Israel (432→432), South Korea (851→1091), Norway (1020→1271), Australia (1946→1808), China (213885→77782)

Compared to yesterday, here's how these regions are doing.

Much worse: California, France, Australia, China

Worse: U.S., India

The same: Louisiana, South Africa

Better: Brazil, Sweden, Toronto, Mexico, Iran, Indonesia, Israel

Much better: Alabama, Poland, Spain, South Korea, Norway

Today's post-peak days-to-eradication numbers for regions of interest: New Zealand (14→14, 0.1%), Malta (57→48, 27.7%), Austria (71→65, 5.0%), Japan (137→172, 17.0%), Italy (237→237, 49.7%), Thailand (new→244, 4.2%), Denmark (267→296, 19.2%), Romania (365→387, 66.2%), Germany (367→514, 13.1%), Malaysia (332→626, 52.0%), Singapore (2506→781, 91.5%), Turkey (1136→1236, 44.8%), U.K. (1717→1533, 87.1%), Ontario (2100→2966, 79.1%), New York (3612→4145, 70.7%), Canada (new→18097, 99.4%)

Days to eradication for other post-peak countries: Belize, Dominica, Eritrea, Papua New Guinea, Saint Lucia, Seychelles, Timor-Leste (eradicated); Saint Kitts and Nevis (1, 0.0%); Montenegro (6, 0.0%); Brunei (14, 0.9%); Iceland (14, 0.2%); Lesotho (14, 50.0%); Andorra (23, 4.5%); Cyprus (35, 27.5%); Niger (36, 10.8%); Gambia (37, 33.3%); Burkina Faso (41, 23.1%); Slovenia (42, 1.0%); Taiwan (57, 4.5%); Mongolia (65, 70.0%); Croatia (66, 6.7%); Luxembourg (70, 3.4%); Cameroon (80, 65.7%); Chad (89, 55.5%); Slovakia (96, 15.4%); Switzerland (116, 3.1%); Estonia (127, 20.1%); Sierra Leone (133, 82.0%); Finland (147, 48.3%); Uruguay (179, 38.3%); Saudi Arabia (221, 77.7%); Cuba (245, 18.0%); Eswatini (256, 71.2%); Morocco (274, 66.4%); Guinea-Bissau (275, 64.4%); Tunisia (282, 12.8%); Kosovo (286, 36.8%); Bosnia and Herzegovina (317, 49.3%); Panama (322, 76.0%); Lithuania (361, 50.5%); South Sudan (364, 92.2%); Hungary (460, 76.8%); Syria (479, 93.4%); Bulgaria (490, 84.6%); Qatar (492, 86.2%); Greece (577, 61.5%); Ecuador (588, 58.5%); Algeria (947, 94.7%); San Marino (1012, 63.6%); Dominican Republic (1018, 90.8%); Serbia (1248, 64.9%); Congo (Brazzaville) (2357, 98.9%); Belgium (3831, 69.4%); Czechia (9016, 57.9%); Moldova (29776, 99.9%)

Days to eradication for other post-peak U.S. states: MP (27, 45.5%); VT (75, 9.4%); HI (98, 8.8%); IN (184, 78.1%); MI (191, 68.4%); OK (227, 56.6%); ND (258, 64.1%); PA (363, 73.0%); NH (430, 73.5%); CT (586, 85.0%); ME (1332, 95.8%); DE (1389, 94.3%); ID (1581, 41.6%); MA (1690, 84.7%); MN (1848, 91.0%); IL (1982, 94.0%); RI (3733, 91.4%); WY (5624, 90.9%); DC (13242, 97.5%)

Days to eradication for other post-peak Canadian provinces: NT, PE, YT (eradicated); MB (36, 5.2%); NS (58, 3.2%); AB (189, 18.7%); BC (384, 31.8%); QC (2515, 94.0%)

There's a typo today in the CMAJ database, showing a negative number of active cases in Saskatchewan. Saskatchewan should still be on that last days to eradication list, but it's hard to tell absent accurate data.

Today's graphic shows those countries who have a lot of people still actively infected (more than 1 in 1,000), where that number is still increasing day by day: Oman, Armenia, Brazil, Djibouti, Peru, Chile, Sao Tome & Principe, Sweden, U.A.E., Spain, France, Maldives, Belarus, Bahrain, Portugal, Russia, Netherlands, Belgium, U.K., U.S., Kuwait. In this list, I'm reading roughly in rows left to right down the chart, so the ones higher up the list are infecting people faster the ones lower down the list. These look to me like countries where it would be critically important to stay indoors and socially distant, because there are a lot of infected people spreading quickly, and your chances of meeting someone and getting infected are highest.

In brighter news, as expected, Canada tentatively eased back over the top of the curve today. Back over, because it had looked like we had reached a peak active caseload of 33,574 on May 16th, but then things blew up and now we look like we're coming down from a slightly higher peak of 33,940. I hope this is the highest peak.

Thailand too had had a brief flare up, but is back on a downward track again, due to a small outbreak among Thai travellers returning from foreign countries.

In Toronto, there have been no reported changes in cumulative case numbers in 19 of the 140 neighbourhoods into which we are officially divided since the launch of official local statistics last Wednesday. I am very fortunate to be living in one of these relatively safe areas: Alderwood, Bayview Village, Blake-Jones, Bridle Path-Sunnybrooke-York Mills, Danforth Village - Toronto, East End-Danforth, Eringate-Centennial-West Deane, Kingsway South, Lambton Baby Point, Lawrence Park North, Leaside-Bennington, Long Branch, Milliken, North Riverdale, Playter Estates-Danforth, Runnymede-Bloor West Village, Steeles, Woodbine Corridor, Yonge-Eglinton. Total active cases in Toronto are down for a second day, from 1,964 to 1,905, still more than we had back on May 8th, but it's good that it's decreasing.

Ontario had its third consecutive daily decrease in active cases, bringing it all the way from 4,004 to 3,783 cases, or about where we were last Mother's Day, May 10th.

Stay safe.

https://www.poslfit.com/covid19

2020-05-30

Comment on this post on Facebook.

today's daily chart as described below in text

Today's pre-peak days-to-one-percent infection numbers: Louisiana (17→14), Brazil (28→24), Alabama (35→35), U.S. (56→48), South Africa (55→51), California (59→53), Sweden (98→64), Mexico (63→67), India (100→95), Iran (111→105), Canada (148→148), Poland (164→169), Indonesia (182→175), Spain (176→187), France (201→201), South Korea (878→851), Norway (1090→1020), Thailand (1948→2071)

Compared to yesterday, here's how these regions are doing.

Much worse: Louisiana, Brazil, U.S., South Africa, California, Sweden, India, Iran, Indonesia, South Korea, Norway

Better: Alabama, Mexico, Canada, Poland, Spain, France, Thailand

Today's post-peak days-to-eradication numbers for regions of interest: New Zealand (14→14, 0.1%), Malta (82→57, 31.1%), Austria (167→71, 5.3%), Japan (105→137, 17.5%), Italy (254→237, 51.6%), Denmark (328→267, 19.2%), Malaysia (new→332, 50.7%), Romania (350→365, 67.5%), Germany (459→367, 13.4%), Israel (307→637, 23.0%), China (254→751, 0.2%), Turkey (817→1136, 45.1%), U.K. (23590→1717, 87.7%), Ontario (new→2100, 80.6%), Singapore (1413→2506, 94.7%), New York (5132→3612, 70.9%), Toronto (new→4660, 65.8%), Australia (new→13041, 10.0%)

Days to eradication for other post-peak countries: Belize, Dominica, Eritrea, Papua New Guinea, Saint Lucia, Timor-Leste (eradicated); Seychelles (1, 0.0%); Saint Kitts and Nevis (2, 0.0%); Montenegro (7, 0.0%); Brunei (14, 0.9%); Lesotho (14, 50.0%); Laos (16, 23.1%); Slovenia (18, 1.1%); Gambia (21, 33.3%); Barbados (26, 17.3%); Andorra (28, 4.9%); Zambia (33, 43.2%); Cyprus (35, 27.5%); Burkina Faso (53, 23.1%); Taiwan (53, 5.3%); Ireland (57, 10.2%); Chad (61, 56.0%); Mongolia (69, 70.7%); Cameroon (80, 65.7%); Niger (87, 14.5%); Panama (94, 68.9%); Georgia (97, 48.7%); Croatia (105, 7.5%); Estonia (112, 20.0%); Slovakia (129, 16.6%); Switzerland (135, 3.7%); Cuba (145, 17.3%); Vietnam (159, 31.0%); Finland (164, 46.9%); Bhutan (170, 85.2%); Luxembourg (182, 4.1%); Uruguay (223, 38.6%); Morocco (245, 67.4%); Lithuania (271, 50.8%); Cabo Verde (290, 74.6%); Saudi Arabia (290, 83.6%); Bosnia and Herzegovina (406, 50.8%); Kosovo (423, 36.8%); Guinea-Bissau (451, 66.4%); Hungary (454, 76.7%); Bulgaria (461, 85.9%); Sierra Leone (487, 88.9%); San Marino (695, 63.8%); Greece (725, 62.2%); Czechia (834, 57.3%); Ecuador (1078, 57.6%); Algeria (1208, 97.2%); Syria (1513, 98.7%); Bahamas (1973, 65.9%); Serbia (3495, 66.0%); Belgium (4026, 69.5%)

Days to eradication for other post-peak U.S. states: MP (27, 45.5%); ID (59, 40.5%); OK (67, 54.0%); HI (69, 8.5%); VT (144, 10.1%); ND (147, 63.4%); MI (182, 66.0%); GU (379, 26.4%); PA (383, 75.2%); NH (474, 78.3%); CT (583, 87.1%); DE (821, 94.7%); MN (839, 94.3%); ME (843, 96.4%); WY (924, 88.9%); MA (930, 85.0%); TN (1014, 89.6%); IL (1707, 94.0%); KS (1796, 81.6%); OR (1796, 90.2%); DC (3511, 96.8%); AR (3737, 87.5%); RI (34095, 92.6%)

Days to eradication for other post-peak Canadian provinces: NT, PE, YT (eradicated); MB (20, 4.7%); SK (56, 26.3%); AB (171, 19.3%); BC (198, 31.8%); QC (4867, 94.7%)

Brazil continues to worsen significantly day by day. They recorded over 30,000 new cases for the first time, and their four-day average new case rate has slowly risen every day for the last four days, from 4.3% to 6.2%.

The United States extends to four days its run of increasing numbers of new cases reported daily, from 16,700 to 23,836.

Today's graph shows current trends on U.S. state new case rates per active case. If all were proceeding smoothly, the post-peak states ought to be seeing a gradual decline in new case rates per active case, or at least a low, steady, manageable number. Instead, in almost every state, COVID-19 is spreading more rapidly from each patient than it was before, resulting in daily increases over the last few days in the number of active cases, just like we had recently in Ontario when we started talking about reopening.

Sweden's numbers were finally updated, although they still lag by a few days. They're somewhat worse than when they last reported a week or so ago, but it's hard to say at a glance due to what looks like a strong weekly cycle of reporting variation.

Locally though, it's generally good but brief news tonight.

Toronto and Ontario have tipped back over onto the post-peak side, after trending back downward over the last four days in their active case counts. In the end, this second bump appears to have lasted just three weeks; it probably won't be the last one, but let's hope the rest will be more sporadic and smaller. Only 123 new cases in Toronto (323 in Ontario), which was fewer than the 156 official recovered cases. That was a near daily occurrence earlier (May 8-11, 13-18), but it's only the second time since then that we've got it right.

Quebec is also on the post-peak list, although with 4,867 active cases to Ontario's 3,933. That leaves almost every Canadian province and territory post-peak, with the exception of Newfoundland and Labrador, where after three weeks with no active cases a travel-related case spawned a small new outbreak, a similar situation reported here yesterday in New Brunswick. (And Nunavut, which isn't on either list, because there has never been a case of COVID-19 there.) I believe there's only one more active case in Canada then there was four days ago, so if all goes well the country itself should be post-peak tomorrow.

Stay safe.

2020-05-29

Comment on this post on Facebook.

today's daily chart as described below in text

Today's pre-peak days-to-one-percent infection numbers: Louisiana (17: no new data), Brazil (33→28), Alabama (35→35), South Africa (64→55), U.S. (59→56), California (50→59), Toronto (62→62), Mexico (65→63), Sweden (98: no new data), India (100→100), Iran (121→111), Ontario (132→139), Canada (132→148), Poland (162→164), Spain (262→176), Indonesia (195→182), France (213→201), Malaysia (289→358), South Korea (878→878), Norway (1090→1090), Australia (2461→1837), Thailand (2645→1948)

Compared to yesterday, here's how these regions are doing.

Much worse: Brazil, South Africa, Iran, Spain, Indonesia, France, Australia, Thailand

Worse: U.S., Mexico

Better: Alabama, Toronto, India, Ontario, Poland, South Korea, Norway

Much better: California, Canada, Malaysia

N/A: Louisiana, Sweden

Today's post-peak days-to-eradication numbers for regions of interest: New Zealand (14→14, 0.1%), Malta (127→82, 34.4%), Japan (91→105, 17.7%), Austria (168→167, 6.9%), China (130→254, 0.2%), Italy (278→254, 54.6%), Israel (183→307, 23.2%), Denmark (183→328, 20.5%), Romania (342→350, 68.6%), Germany (314→459, 14.0%), Turkey (888→817, 45.2%), Singapore (1295→1413, 95.2%), New York (10298→5132, 71.2%), U.K. (5130→23590, 88.6%)

Days to eradication for other post-peak regions: Belize (0, 0.0%), Eritrea (0, 0.0%), NT CAN (0, 0.0%), PE CAN (0, 0.0%), Papua New Guinea (0, 0.0%), Saint Lucia (0, 0.0%), Timor-Leste (0, 0.0%), YT CAN (0, 0.0%), Dominica (1, 0.0%), Seychelles (2, 0.0%), Saint Kitts and Nevis (3, 0.0%), Trinidad and Tobago (5, 0.0%), Montenegro (8, 0.0%), Brunei (14, 0.9%), Iceland (14, 0.1%), Lesotho (14, 50.0%), Western Sahara (14, 33.3%), Gambia (15, 33.3%), Laos (16, 23.1%), Slovenia (16, 1.1%), Barbados (26, 17.3%), Andorra (31, 6.8%), Zambia (33, 43.2%), Cyprus (37, 28.3%), Bhutan (48, 75.3%), Burkina Faso (53, 23.1%), HI USA (53, 8.2%), Ireland (54, 9.8%), ID USA (58, 40.2%), Taiwan (58, 5.7%), Chad (59, 62.5%), OK USA (62, 55.7%), SK CAN (68, 29.2%), MB CAN (69, 7.3%), ND USA (73, 66.6%), Niger (73, 15.3%), Croatia (75, 7.8%), Cameroon (84, 59.2%), Georgia (115, 53.0%), Cabo Verde (119, 68.4%), Estonia (121, 20.6%), Uruguay (123, 37.6%), Luxembourg (130, 3.9%), AB CAN (134, 19.7%), Finland (142, 44.7%), BC CAN (155, 31.8%), NH USA (161, 76.8%), Slovakia (172, 18.6%), Vietnam (189, 31.0%), Lithuania (192, 51.7%), Botswana (193, 80.0%), Tunisia (232, 14.3%), Morocco (254, 69.4%), Saudi Arabia (258, 84.6%), Comoros (303, 93.8%), VT USA (303, 10.9%), PA USA (334, 75.8%), San Marino (342, 64.1%), Latvia (389, 68.8%), Bosnia and Herzegovina (435, 52.7%), Guinea-Bissau (461, 71.7%), Bulgaria (462, 87.3%), SD USA (577, 76.3%), DE USA (583, 93.3%), Jordan (588, 86.6%), TN USA (609, 86.4%), Czechia (656, 57.6%), Greece (672, 62.3%), WY USA (677, 96.0%), CT USA (731, 88.4%), Switzerland (735, 4.2%), IL USA (764, 94.2%), ME USA (872, 95.7%), Mongolia (914, 95.9%), MA USA (1011, 85.9%), Algeria (1487, 97.2%), KS USA (1796, 81.6%), NM USA (1943, 98.2%), Bahamas (1973, 65.9%), Ecuador (2465, 61.6%), DC USA (2480, 95.4%), Hungary (2736, 81.6%), IA USA (3296, 97.4%), Serbia (3538, 66.3%), Somalia (3573, 94.2%), Belgium (4044, 70.1%), OR USA (5113, 90.2%), Sierra Leone (5330, 90.5%), MN USA (14922, 99.8%), RI USA (30170, 92.3%)

The City of Toronto has been publishing a breakdown since Wednesday of cumulative cases in each of 140 official neighbourhoods. The numbers seem a little noisy still, in that some neighbourhoods see numbers going up and down (when you would expect only up), and after the first day they added 2,000+ uncategorized cases, but it's a fantastic local resource. I can see that there have been no new cases in my neighbourhood during this time, nor in 46 other neighbourhoods including the ones immediately adjacent to us. The neighbourhood where the St. Lawrence Market is had one new case reported since Wednesday, bringing them to 102.

It makes me feel ashamed that the neighbourhoods that have the highest concentration of cases appear to be those with a lot of nursing homes, or residents of ethnicities overrepresented in the caregiver community.

The overall situation in Toronto continues to worsen, with active cases back over two thousand (2,005) for the first time since May 7th. Each active case is responsible for on average 0.09 new cases each day, and if it continues even at that rate, eventually even currently safe neighbourhoods will no longer be safe.

Ontario had an insignificant dip in its number of active cases today, from 4,004 to 3,997. Changes in active cases are the net difference between new infections, and recoveries + deaths. Since Ontario had seven more deaths today (41) than yesterday, I think the best that you could say about our situation is that we are staying steady; we are not clearly on the declining side of a bell curve.

Ontario Premier Doug Ford has reversed his long-standing position against a regional approach to reopening the province, against the wishes of his medical officer of health. It looks to me as though reopening has only worked elsewhere where the number of cases is down to a level that new cases can easily be identified and traced, and resources exist to rapidly identify, test and isolate contacts. I imagine that there are small towns in Ontario where these conditions have been met. Regional reopening only seems to work where there is a way to control people's movement between affected and unaffected regions; I presume this is what the provincial government is trying to figure out right now.

A good example of what goes wrong if you can't control interregional movement is what's going on in New Brunswick right now, where they had managed to get down to zero active cases for three days before a family doctor who travelled across the border to Quebec, did not self-isolate on return, and has since infected seven other patients, two in ICU. That's still a small enough number for public health officials to manage, as long as it doesn't happen too frequently.

Stay safe.

2020-05-28

Comment on this post on Facebook.

today's daily chart as described below in text

Today's pre-peak days-to-one-percent infection numbers: Brazil (38→33), Alabama (44→35), California (63→50), U.S. (56→59), Toronto (62→62), South Africa (63→64), Mexico (66→65), India (93→100), Iran (121→121), Canada (129→132), Ontario (124→132), Poland (178→162), Indonesia (205→195), France (943→213), Spain (380→262), Malaysia (255→289), South Korea (1115→878), Norway (1218→1090), Australia (3056→2461), Thailand (4719→2645)

Compared to yesterday, here's how these regions are doing.

Much worse: Brazil, Alabama, California, Poland, Indonesia, France, Spain, South Korea, Norway, Australia, Thailand

The same: Mexico

Better: U.S., Toronto, South Africa, India, Iran, Canada, Ontario

Much better: Malaysia

Today's post-peak days-to-eradication numbers for regions of interest: New Zealand (14→14, 0.1%), Japan (105→91, 18.3%), Malta (165→127, 39.9%), China (225→130, 0.2%), Austria (241→168, 7.2%), Denmark (311→183, 20.0%), Israel (176→183, 22.9%), Italy (369→278, 56.7%), Germany (255→314, 14.2%), Romania (685→342, 68.9%), Louisiana (277→344, 40.9%), Turkey (1223→888, 45.9%), Singapore (1612→1295, 95.1%), U.K. (6379→5130, 89.2%), New York (8776→10298, 71.5%)

Days to eradication for other post-peak regions: Belize (0, 0.0%), Eritrea (0, 0.0%), NT CAN (0, 0.0%), PE CAN (0, 0.0%), Papua New Guinea (0, 0.0%), Saint Lucia (0, 0.0%), Timor-Leste (0, 0.0%), YT CAN (0, 0.0%), Dominica (2, 0.0%), Seychelles (3, 0.0%), Saint Kitts and Nevis (4, 0.0%), Trinidad and Tobago (6, 0.0%), Montenegro (9, 0.0%), Brunei (14, 0.9%), Gambia (14, 41.7%), Lesotho (14, 50.0%), Slovenia (14, 1.3%), Western Sahara (14, 33.3%), Laos (16, 23.1%), Barbados (23, 17.3%), Andorra (26, 7.3%), Zambia (33, 43.2%), Cyprus (36, 28.1%), Taiwan (37, 5.3%), HI USA (51, 8.5%), ID USA (51, 37.9%), Chad (52, 62.0%), Ireland (54, 9.6%), SK CAN (54, 29.2%), Finland (66, 43.2%), MB CAN (69, 7.3%), OK USA (76, 58.9%), Croatia (84, 8.6%), ND USA (89, 72.6%), Georgia (92, 51.3%), Mongolia (94, 79.5%), Cabo Verde (95, 65.6%), BC CAN (108, 33.6%), Vietnam (109, 31.0%), Burkina Faso (110, 23.4%), Luxembourg (110, 4.2%), Bhutan (118, 79.0%), Switzerland (120, 4.0%), AB CAN (139, 20.9%), Niger (150, 17.3%), NH USA (155, 74.0%), Botswana (193, 80.0%), Slovakia (196, 19.3%), Bahamas (244, 63.7%), MS USA (250, 86.5%), Guinea-Bissau (252, 64.0%), Morocco (282, 69.6%), Comoros (303, 93.8%), Bosnia and Herzegovina (314, 52.6%), SD USA (330, 74.7%), Estonia (331, 23.9%), NC USA (333, 88.8%), Togo (333, 92.2%), Saudi Arabia (337, 87.7%), Ghana (352, 75.9%), El Salvador (395, 92.5%), Lithuania (404, 54.1%), Czechia (414, 57.2%), PA USA (417, 77.1%), Tunisia (428, 15.3%), San Marino (447, 64.8%), Latvia (457, 69.0%), Bulgaria (543, 89.3%), Sierra Leone (557, 92.3%), WY USA (591, 93.3%), Equatorial Guinea (702, 84.9%), TX USA (707, 92.9%), MN USA (762, 94.8%), Jordan (914, 89.8%), TN USA (924, 85.5%), DE USA (1039, 94.5%), Liberia (1045, 92.5%), Serbia (1097, 66.2%), IL USA (1107, 94.5%), Cuba (1306, 19.7%), MA USA (1401, 86.7%), RI USA (1618, 92.4%), OR USA (1630, 90.3%), CT USA (1827, 89.9%), ME USA (1867, 98.5%), Congo (Brazzaville) (1885, 86.5%), Greece (1980, 64.4%), Hungary (3170, 81.8%), Belgium (4848, 70.3%), UT USA (5102, 97.8%), Kosovo (5284, 39.0%), KS USA (12135, 80.1%), Guinea (50981, 99.9%)

Toronto and Ontario both rebounded today, Toronto gaining 10% of its active case count in new cases, ending back up at 1,982, a level we haven't seen since May 8th. More simply put, we had our first 200+ day after three days below 200 new cases. Ontario likewise is at 10% active per new, and now has 4,004 cases, also where we were on May 8th. Ontario didn't quite get back up to 400, seeing only 383 newly confirmed cases.

The City of Toronto has started publishing maps showing the cumulative infection rate by neighbourhood, which appears to depend on income, number of nursing homes, and random chance. If you use the interactive tool they provide to filter out institutional outbreaks, you can see that income becomes the dominant factor. The rate ranges widely across the city, from 45 cases per 100,000 in St. Andrew-Windfields to 1,525 in Humber Heights-Westmount. Our neighbourhood is at the low end, the yahoos on the Danforth notwithstanding, at 103; the city as a whole currently averages about 370.

The situation in Brazil continues to worsen. They are still at the point where they are setting daily records for new cases (26,417 today), and about a thousand people a day are dying.

A while ago, I started looking at delay-adjusted case fatality ratios for my regions of interest. As some of them are reaching the end of their curve (when the delay adjustment converges), it's interesting to see how different the CFRs are. France tops the list at 19-20% (depending on delay), followed by Italy at 14-15%, with countries like Canada (8-9%) and Spain (9-10%) in the middle, and countries like South Korea (2.4%) and Singapore (0.1%) at the bottom. Clearly, France and Italy are tremendously underreporting their cases (even more when you look at their excess mortality stats); Singapore is likely fortunate in that their outbreak has been largely confined to younger foreign workers who are unlikely to die.

Stay safe.

2020-05-27

Comment on this post on Facebook.

today's daily chart as described below in text

Today's pre-peak days-to-one-percent infection numbers: Brazil (41→38), Alabama (37→44), U.S. (51→56), Toronto (56→62), California (62→63), South Africa (61→63), Mexico (65→66), India (89→93), Sweden (98→98), Iran (124→121), Ontario (113→124), Canada (113→129), Poland (166→178), Indonesia (177→205), Malaysia (236→255), Spain (392→380), South Korea (1713→1115), Norway (1193→1218)

Compared to yesterday, here's how these regions are doing.

Much worse: Spain, South Korea

Worse: Brazil, Iran

Better: California, South Africa, Mexico, India, Poland, Malaysia, Norway

Much better: Alabama, U.S., Toronto, Ontario, Canada, Indonesia

N/A: Sweden

Today's post-peak days-to-eradication numbers for regions of interest: New Zealand (52→14, 0.9%), Japan (103→105, 20.5%), Malta (537→165, 42.1%), Israel (157→176, 23.3%), China (585→225, 0.2%), Austria (348→241, 7.7%), Germany (298→255, 14.1%), Louisiana (205→277, 39.3%), Denmark (308→311, 21.1%), Australia (366→369, 9.9%), Italy (407→369, 60.2%), Thailand (103→654, 4.5%), Romania (479→685, 72.7%), Turkey (1184→1223, 46.5%), Singapore (1871→1612, 96.4%), France (2923→3235, 50.3%), U.K. (16729→6379, 89.9%), New York (14397→8776, 71.5%)

Days to eradication for other post-peak regions: Belize (0, 0.0%), NT CAN (0, 0.0%), PE CAN (0, 0.0%), Papua New Guinea (0, 0.0%), Saint Lucia (0, 0.0%), YT CAN (0, 0.0%), Eritrea (1, 0.0%), Timor-Leste (1, 0.0%), Dominica (3, 0.0%), Seychelles (4, 0.0%), Saint Kitts and Nevis (5, 0.0%), Trinidad and Tobago (7, 0.0%), Montenegro (10, 0.0%), Brunei (14, 1.9%), Gambia (14, 41.7%), Lesotho (14, 62.5%), NL CAN (14, 1.0%), Slovenia (14, 1.3%), Laos (16, 23.1%), Barbados (23, 17.3%), Taiwan (32, 5.7%), Zambia (33, 43.2%), Andorra (36, 8.5%), Jordan (38, 50.6%), Ireland (54, 9.3%), ID USA (68, 42.3%), MB CAN (69, 7.3%), Switzerland (73, 3.9%), SK CAN (76, 32.5%), HI USA (87, 10.4%), ND USA (87, 71.5%), Croatia (91, 9.0%), GU USA (91, 24.2%), Chad (93, 73.0%), Luxembourg (102, 4.5%), Georgia (105, 55.7%), Vietnam (109, 31.0%), OK USA (112, 64.2%), BC CAN (114, 34.0%), Slovakia (114, 19.3%), Tunisia (133, 13.8%), NS CAN (145, 4.1%), Niger (149, 18.2%), MT USA (155, 7.0%), AB CAN (159, 21.7%), Eswatini (168, 77.3%), Togo (176, 86.5%), Bhutan (195, 82.7%), Cabo Verde (207, 73.9%), MS USA (212, 80.3%), West Bank and Gaza (224, 30.3%), NH USA (232, 79.1%), Finland (235, 59.4%), NC USA (255, 81.9%), Guinea-Bissau (259, 71.2%), WY USA (294, 91.3%), Estonia (299, 24.1%), Bosnia and Herzegovina (328, 53.8%), Bahamas (329, 64.8%), AK USA (338, 19.4%), Cuba (367, 19.8%), Latvia (406, 68.3%), TX USA (417, 90.3%), Guyana (461, 94.3%), Bulgaria (477, 89.9%), Mongolia (510, 95.5%), Lithuania (518, 54.3%), Morocco (568, 75.0%), Czechia (628, 58.2%), Ghana (677, 78.2%), MN USA (681, 92.9%), UT USA (682, 95.0%), SD USA (719, 74.4%), PA USA (738, 82.0%), Somalia (834, 88.9%), Saudi Arabia (835, 94.3%), KS USA (857, 80.1%), RI USA (891, 92.4%), El Salvador (1005, 94.5%), OR USA (1139, 90.2%), Guinea (1584, 94.5%), San Marino (1759, 65.1%), MA USA (1909, 86.9%), DE USA (2011, 96.7%), Paraguay (2022, 59.3%), Greece (2369, 64.7%), Hungary (2543, 82.2%), Belgium (2803, 70.2%), ME USA (4065, 97.9%), CT USA (5140, 91.2%), Serbia (8865, 66.8%)

Toronto had a fifth consecutive day with fewer new cases, gradually down from 258 on May 22nd to 152 today. We had our second day in nine days where the number of active cases decreased: on May 18th when the current upturn started, we've risen up to 1,931, which is about where we were on May 8th. Two more days of this, and we may return to the post-peak list.

Ontario had its first decline in active cases since May 16th (3,412), dropping from yesterday's 4,110 back down to 3,956.

In both Toronto and Ontario, the ratio of new cases to active cases is around 7-8%, suggesting that we are close to or below equilibrium.

The United States still leads the world in active cases, with about 1.2M by JHU numbers, which say that the U.S. has seen about 1.7M cases of whom 0.4M have recovered and 0.1M have died. Two weeks ago, these numbers were 1.1M = 1.4M - 0.2M - 0.1M. As I see it, there's a problem here, which is that the U.S. is reporting that fewer than 16% of active cases from two weeks ago have either recovered or died. The most likely explanation to me is that the recovered statistics are drastically underreported.

That's why when I graph active cases, I assume that at least 75% of active cases lagged 14 days have recovered, and accept the official number only where it is higher.

By this measure, the United States is still the world leader by a wide margin, but has only 653,000 confirmed or probable cases that should still be active. The U.K. is either in fourth place with 96,000 (my estimate) or second with 228,000 (JHU). Russia is in third place either way (188K or 227K), and Brazil rounds out the fourth either way with its plausibly reported 208K. The next ten are (in different orders, but this is mine) India, Peru, Spain, Italy, France, Chile.

I note with some concern South Korea's switch back to the pre-peak list today, after an increase of 34 to 735 in their active case count today pushed them back up to where they were more than a week ago. Not that long ago, they had spent four weeks with their new per active ratio below 1%; today it is 5%, reportedly due to the ongoing effects of the nightclub outbreak at the beginning of the month.

Today's graphic charts these top ten nations and shows the seven-day forward trends for their rate of new cases per active, which is improving for Peru and Spain, and worsening for the rest.

Stay safe.

2020-05-26

Comment on this post on Facebook.

today's daily chart as described below in text

Today's pre-peak days-to-one-percent infection numbers: Alabama* (44→37), Brazil (36→41), U.S. (48→51), Toronto (49→56), South Africa (61→61), California (62→62), Mexico (65→65), India (87→89), Sweden* (98→98), Canada (113→113), Ontario (102→113), Iran (115→124), Poland (161→166), Indonesia (159→177), Malaysia (304→236), Spain (238→392), Norway (1050→1193)

* See commentary

Compared to yesterday, here's how these regions are doing.

Much worse: Alabama, Malaysia

Better: U.S., South Africa, California, Mexico, India, Sweden, Canada, Iran, Poland

Much better: Brazil, Toronto, Ontario, Indonesia, Spain, Norway

Today's post-peak days-to-eradication numbers for regions of interest: New Zealand (63→52, 2.3%), Japan (101→103, 22.2%), Thailand (49→103, 4.1%), Israel (151→157, 24.3%), Louisiana (220→205, 37.0%), Germany (290→298, 14.9%), Denmark (234→308, 21.4%), Austria (386→348, 7.8%), Australia (307→366, 10.0%), Italy (477→407, 62.6%), Romania (525→479, 74.6%), Malta (537→537, 44.4%), China (819→585, 0.2%), Turkey (1712→1184, 46.9%), Singapore (40557→1871, 96.2%), France (2708→2923, 50.0%), South Korea (576→5094, 10.2%), New York (7254→14397, 71.9%), U.K. (2356→16729, 90.2%)

Days to eradication for other post-peak regions: Belize (0, 0.0%), NT CAN (0, 0.0%), PE CAN (0, 0.0%), Papua New Guinea (0, 0.0%), Saint Lucia (0, 0.0%), YT CAN (0, 0.0%), Eritrea (2, 0.0%), Timor-Leste (2, 0.0%), Dominica (4, 0.0%), Seychelles (5, 0.0%), Saint Kitts and Nevis (6, 0.0%), Trinidad and Tobago (8, 0.0%), Montenegro (11, 0.0%), NL CAN (14, 1.0%), Gambia (19, 50.0%), Brunei (25, 2.8%), NS CAN (28, 3.6%), Andorra (35, 8.5%), Jordan (35, 49.7%), Slovenia (36, 2.1%), Taiwan (39, 6.8%), HI USA (42, 10.7%), GU USA (47, 27.5%), Croatia (67, 9.1%), Tunisia (71, 13.8%), Slovakia (78, 19.7%), SK CAN (80, 36.8%), Switzerland (85, 4.5%), West Bank and Gaza (93, 28.0%), Paraguay (104, 58.5%), Niger (107, 18.2%), Luxembourg (109, 4.6%), ND USA (113, 76.8%), MB CAN (118, 8.3%), BC CAN (136, 36.0%), Cuba (137, 20.3%), Georgia (142, 61.4%), AB CAN (151, 22.8%), Guinea-Bissau (164, 70.6%), Togo (178, 87.4%), Finland (192, 56.5%), MS USA (213, 74.4%), NH USA (216, 83.7%), Bosnia and Herzegovina (229, 54.4%), Eswatini (232, 72.0%), Congo (Brazzaville) (235, 84.3%), MI USA (298, 82.6%), Estonia (305, 24.6%), MT USA (321, 7.4%), Morocco (338, 77.2%), TX USA (395, 89.9%), Somalia (398, 90.8%), Chad (409, 83.8%), NC USA (413, 77.7%), Mongolia (414, 94.5%), Guyana (461, 94.3%), Lithuania (490, 56.0%), PA USA (509, 84.0%), Rwanda (509, 67.5%), Vietnam (526, 34.8%), WY USA (576, 91.3%), Ghana (578, 78.7%), Cote d'Ivoire (618, 95.2%), Bulgaria (674, 91.7%), KS USA (709, 83.2%), San Marino (723, 66.0%), OR USA (811, 91.9%), KY USA (1021, 86.2%), Greece (1153, 64.8%), Latvia (1243, 67.6%), Moldova (1286, 95.6%), RI USA (1363, 92.7%), Belgium (1893, 70.2%), NE USA (2106, 95.0%), OK USA (2343, 82.9%), Saudi Arabia (2523, 97.0%), NJ USA (2532, 79.7%), UT USA (2602, 97.2%), Uruguay (2876, 42.6%), Hungary (4913, 82.1%), Serbia (5732, 66.6%), Czechia (6460, 59.7%), MA USA (7821, 89.2%), Senegal (10057, 98.7%)

We had a mixed day in Toronto today. On the bright side, we had only 161 new cases, which is better than the six days of 200+ that we recently had, and I think we might have a few more sub-200 days before the effects of the most recent oppositional behaviour are seen. People are tired of staying safe, and it's really hard to explain to someone who hasn't lost a loved one, or who isn't close to someone who is at risk due to age or preexisting medical condition, how much the whole country is counting on them to do their part to keep any asymptomatic infection they may have to themselves. Our new cases today were 8% of our active total, which is still high enough to suggest an R_t value greater than one, or an epidemic that is continuing to worsen. And we were up 36 to 1,978 active cases today, which is a level that we hadn't previously seen since May 7th, two thirds of the peak of close to 3,000 that we set on April 23rd. Our hospital numbers are drifting downward for now, though they are sure to come back up, but any break our healthcare workers can get is a good one. Only seven people were recorded as having died of COVID-19 today. Toronto's four-day average increase in active cases is down to 1.36%, and coming down fast enough that we might switch over to the post-peak list again in a day or two. So, mixed.

In Ontario, where the Armed Forces have been brought in to take over the worst run nursing homes, a military report was scathing about the conditions troops found. Premier Ford acknowledged that the situation was serious and needed to change; he did not reportedly draw a connection to the efforts by his government and the previous Liberal government to eliminate oversight of nursing homes in the province.

Not sure what happened to the people who were updating the Swedish Wikipedia page from whence I was fetching my data. They haven't posted anything since May 20th. I hope they're okay. Alabama is several days behind in reporting too.

New Zealand has now gone a sixth straight day with no new cases; Malta managed its first day with no new cases in a couple of weeks.

Canada and the United States have seen four straight days of decreasing new case numbers. Toronto four days, New York State three days.

Stay safe.

2020-05-25

Comment on this post on Facebook.

today's daily chart as described below in text

Today's pre-peak days-to-one-percent infection numbers: Brazil (32→36), U.S. (47→48), Toronto (46→49), South Africa (58→61), California (63→62), Mexico (61→65), India (88→87), Sweden (98→98), Ontario (101→102), Canada (109→113), Iran (109→115), Indonesia (129→159), Poland (149→161), Spain (169→238), Malaysia (458→304)

Compared to yesterday, here's how these regions are doing.

Much worse: Malaysia

The same: California, India

Better: U.S., Toronto, South Africa, Mexico, Sweden, Ontario, Canada, Iran, Poland

Much better: Brazil, Indonesia, Spain

Today's post-peak days-to-eradication numbers for regions of interest: Norway (15→15, 10.1%), Thailand (54→49, 3.9%), New Zealand (151→63, 2.4%), Japan (61→101, 24.1%), Israel (145→151, 25.8%), Louisiana (new→220, 35.7%), Denmark (225→234, 22.4%), Germany (316→290, 15.2%), Australia (454→307, 9.9%), Austria (636→386, 8.1%), Italy (451→477, 65.4%), Romania (531→525, 76.1%), Malta (new→537, 44.4%), South Korea (2086→576, 9.9%), China (2487→819, 0.2%), Alabama (470→1045, 96.5%), Turkey (2592→1712, 47.7%), U.K. (2072→2356, 88.8%), France (2032→2708, 50.8%), New York (3841→7254, 71.9%), Singapore (5721→40557, 98.0%)

Days to eradication for other post-peak regions: Belize (0, 0.0%), NT CAN (0, 0.0%), PE CAN (0, 0.0%), Papua New Guinea (0, 0.0%), YT CAN (0, 0.0%), Saint Lucia (1, 0.0%), Eritrea (3, 0.0%), Timor-Leste (3, 0.0%), Dominica (5, 0.0%), Seychelles (6, 0.0%), Saint Kitts and Nevis (7, 0.0%), Trinidad and Tobago (9, 0.0%), Montenegro (12, 0.0%), Taiwan (24, 7.2%), Brunei (25, 2.8%), NL CAN (25, 1.6%), Gambia (30, 58.3%), NS CAN (34, 4.1%), GU USA (36, 25.3%), Slovenia (44, 2.3%), Andorra (48, 11.5%), Croatia (54, 10.2%), HI USA (54, 11.0%), Paraguay (62, 58.1%), Tunisia (63, 15.6%), Cuba (69, 19.0%), SK CAN (76, 38.8%), Switzerland (82, 4.4%), Luxembourg (88, 4.6%), Slovakia (89, 21.3%), West Bank and Gaza (111, 28.9%), Tanzania (140, 42.8%), AB CAN (150, 24.4%), Georgia (160, 64.8%), BC CAN (180, 37.2%), Congo (Brazzaville) (185, 84.3%), Ghana (190, 79.0%), Finland (204, 55.3%), Estonia (211, 25.1%), Eswatini (229, 72.7%), Guinea-Bissau (233, 76.2%), MS USA (241, 69.2%), MB CAN (247, 8.8%), Lithuania (253, 59.4%), Vietnam (262, 34.2%), MI USA (265, 81.6%), Niger (285, 20.2%), PA USA (286, 86.0%), Tajikistan (320, 90.4%), VT USA (326, 11.6%), Bosnia and Herzegovina (331, 56.2%), MT USA (341, 7.7%), Rwanda (424, 69.6%), Togo (440, 92.2%), Morocco (511, 79.2%), Guyana (529, 96.1%), OR USA (539, 90.8%), TX USA (626, 92.3%), Moldova (665, 93.5%), NH USA (751, 93.6%), Somalia (1026, 95.0%), RI USA (1058, 92.4%), DC USA (1079, 96.8%), NC USA (1297, 76.2%), Bulgaria (1528, 93.1%), NE USA (1534, 94.6%), Belgium (1760, 70.8%), Mongolia (1868, 98.2%), NJ USA (2099, 79.6%), CO USA (2100, 89.3%), Maldives (2727, 96.8%), KS USA (2799, 83.2%), Greece (3568, 65.1%), MA USA (4182, 89.8%), Czechia (4506, 60.7%), Netherlands (40525, 68.6%)

Six days in a row of 200+ people contracting COVID-19 in Toronto and 400+ in Ontario, thanks to relaxed public attitudes and fewer restrictions on business operations in the province. Today, the premier announced an indefinite delay on further lifting of restrictions, such as permitting people to meet in groups of more than five. Although the federal government recommends the use of face coverings where social distancing is not possible, neither the province nor the city have made it mandatory, despite frequent calls to do so.

Surgical masks catch all exhaled coronaviruses, as a widely publicized paper in Nature Medicine reported several weeks ago. (For the sake of safety, they ran the tests with similarly sized coronaviruses to SARS-CoV-2, but not the actual novel coronavirus that is killing us now.) They also catch all influenza viruses, and no rhinoviruses. So if everyone wore masks all the time in public, then no one would catch COVID-19 or the flu from a stranger in public. Of course, even in the 30-50 countries in the world that have some sort of requirement on the covering of the face in all public situations, none exclude ineffectual loose fabrics from meeting the letter of the requirement, and enforcement varies but is never perfect. So they're not a panacea, or a replacement for washing hands, social distancing, or quarantine. And if they give anyone much of a personal sense of security, they're missing the point. But it's something that people can be encouraged to do that's helpful.

There are a lot of asymptomatic carriers of COVID-19 in the community. You could be one of them, and exhaling an infectious dose of SARS-CoV-2 with each breath. Unless you are being tested daily, you have no way of telling, and even then a mistake could be made. So don't take chances with the lives of the strangers around you: they could be someone that you need in your life, or someone who lives with them, or just someone whose life should have value to you.

Today's charts show countries that have universal public face covering requirements vs. those that don't. The ones that do tend not to not to have very high rates of new cases per active case; they also tend not to have extremely low rates. My intuitive guess, unsupported by evidence, is that this is because enough people are wearing proper masks properly to slow down contagion, and enough people are blithely wearing useless masks improperly to keep it going.

Stay safe.

2020-05-24

Comment on this post on Facebook.

today's daily chart as described below in text

Today's pre-peak days-to-one-percent infection numbers: Louisiana (15→16), Brazil (30→29), Toronto (41→43), U.S. (51→49), Mexico (63→62), South Africa (67→62), California (69→69), India (91→88), Sweden (98→98), Ontario (103→105), Iran (108→108), Canada (113→109), Indonesia (136→119), Malta (116→123), Poland (133→148), Spain (152→158)

Compared to yesterday, here's how these regions are doing.

Much worse: South Africa, Canada, Indonesia

Worse: U.S., India

The same: Brazil, Mexico

Better: Louisiana, Toronto, California, Sweden, Ontario, Iran, Malta, Spain

Much better: Poland

Today's post-peak days-to-eradication numbers for regions of interest: Norway (14→15, 9.7%), Thailand (39→35, 4.7%), New Zealand (34→61, 2.9%), Japan (89→70, 28.0%), Austria (116→133, 8.7%), Israel (154→148, 28.2%), Denmark (141→159, 23.2%), Germany (482→261, 16.5%), Alabama (207→299, 84.4%), South Korea (133→309, 10.4%), Italy (406→388, 68.3%), Australia (345→390, 10.6%), Romania (1013→537, 76.8%), China (610→610, 0.2%), Malaysia (737→613, 44.6%), France (788→679, 50.4%), U.K. (626→707, 90.5%), Turkey (2097→2039, 48.2%), New York (3029→2927, 72.0%), Singapore (5119→43241, 98.9%)

Days to eradication for other post-peak regions: Belize (0, 0.0%), NT CAN (0, 0.0%), PE CAN (0, 0.0%), Papua New Guinea (0, 0.0%), YT CAN (0, 0.0%), Mauritius (1, 0.0%), Western Sahara (1, 0.0%), Saint Lucia (3, 0.0%), Eritrea (5, 0.0%), Timor-Leste (5, 0.0%), Dominica (7, 0.0%), Seychelles (8, 0.0%), Saint Kitts and Nevis (9, 0.0%), Trinidad and Tobago (11, 0.0%), Montenegro (14, 0.5%), NL CAN (14, 1.6%), Monaco (21, 6.5%), Tunisia (30, 16.0%), Estonia (31, 26.2%), Holy See (32, 48.0%), MB CAN (33, 8.8%), Taiwan (36, 8.7%), NS CAN (40, 4.7%), Grenada (41, 41.7%), Uganda (42, 58.7%), Andorra (55, 13.6%), Cuba (57, 21.1%), Paraguay (62, 60.1%), SK CAN (62, 42.1%), Ireland (64, 16.5%), Croatia (66, 11.4%), Slovakia (71, 23.7%), HI USA (86, 12.6%), Sao Tome and Principe (88, 41.5%), Slovenia (118, 3.1%), Switzerland (132, 5.7%), Niger (138, 21.0%), Tanzania (140, 42.8%), Georgia (168, 69.5%), San Marino (182, 65.9%), VT USA (184, 11.2%), Guinea-Bissau (204, 79.4%), Lithuania (204, 57.2%), Ghana (207, 82.3%), Uruguay (225, 41.6%), Bosnia and Herzegovina (230, 58.6%), MI USA (235, 79.1%), Morocco (261, 79.6%), Burma (289, 54.5%), KY USA (329, 89.7%), PA USA (332, 86.9%), Luxembourg (349, 5.5%), BC CAN (366, 42.3%), Vietnam (377, 36.7%), Sri Lanka (407, 76.4%), KS USA (415, 83.5%), WA USA (504, 88.9%), OK USA (560, 83.5%), Burkina Faso (633, 28.0%), OR USA (634, 92.8%), Bahrain (661, 94.7%), TN USA (675, 85.8%), AK USA (710, 20.5%), SD USA (788, 77.7%), NE USA (809, 92.8%), Eswatini (851, 87.7%), Somalia (868, 92.2%), Czechia (1010, 61.4%), Belgium (1350, 71.2%), MA USA (1381, 89.0%), Portugal (1662, 80.8%), Greece (1686, 65.6%), NH USA (2043, 90.8%), Rwanda (2080, 69.6%), Bulgaria (2144, 95.8%), DC USA (2274, 95.5%), DE USA (2318, 98.5%), CO USA (2398, 90.5%), NJ USA (3001, 79.8%), Cabo Verde (3096, 82.7%), WY USA (3120, 99.2%), Netherlands (3520, 68.2%), CT USA (3523, 91.9%), Dominican Republic (3917, 95.1%), Albania (4112, 60.6%), Cyprus (4310, 51.8%), QC CAN (20426, 95.2%)

New Zealand's days to eradication backing off from 34 to 61 is the result of a big drop in active cases five days ago working its way out of the four-day mean; the situation has not changed substantially. Japan's decrease in days to eradication was a result of a steep drop from 89 to 70 remaining active cases.

I had two reasons to feel somewhat optimistic about Toronto today. One was that while driving to pick up groceries (curbside seafood from the excellent Mike's Fish Market at the St. Lawrence Market), I saw that most people outside of my neighbourhood were fairly compliant with Canadian Chief Public Health Officer Theresa Tam's instructions to wear face coverings when not able to maintain a safe social distance. I'll just make a point of not shopping on the Danforth for the foreseeable future.

The other was that active cases increased by only nine to 1,883 today after rising 447 over the previous four days. It's unfortunately more pertinent to observe that we had our fifth consecutive day of 200+ new cases after six days with far fewer, and that if the ratio of newly confirmed cases to current active cases is decreasing, it's much more because the active cases are increasing than because the new cases are decreasing. The reason it looks better (and is so labelled above) is because we had many more people than usual declared recovered today, subtracted from the active case total. On the bright side, it means that there are only 1,883 new cases and not closer to 2,000 to spread further contagion.

You can say more or less the same about Ontario's numbers today. Three days straight now of 400+ cases, but only a net increase of six new active cases in this province, because 379 were declared cured and 27 died.

Another type of darkly positive news is that excess mortality numbers in Europe have reached the point where they are clearly showing that they peaked back in early April. It takes three weeks for numbers to finish reporting, so the last reasonably firm numbers are for Week 17 (April 20-26). In all parts of the continent participating in Euromomo.eu, mortality numbers peaked in Week 14 or 15 (March 30-April 12). In England, they reached only 42.34 standard deviations above seasonal means rather than the forecast 44.10, and have since dropped to 30.11 standard deviations.

In Spain, they reached 34.23 standard deviations, and I am mystified by a forecast that they will be at -4.14 (i.e., below the mean) this week. I could imagine that if a lot more people who were close to death were killed by COVID-19, that might cause a rebound effect reducing mortality for a short period of time, but four sigmas seems implausibly large.

Stay safe.

2020-05-23

Comment on this post on Facebook.

today's daily chart as described below in text

Today's pre-peak days-to-one-percent infection numbers: Louisiana (15→16), Brazil (30→29), Toronto (41→43), U.S. (51→49), Mexico (63→62), South Africa (67→62), California (69→69), India (91→88), Sweden (98→98), Ontario (103→105), Iran (108→108), Canada (113→109), Indonesia (136→119), Malta (116→123), Poland (133→148), Spain (152→158)

Compared to yesterday, here's how these regions are doing.

Much worse: South Africa, Canada, Indonesia

Worse: U.S., India

The same: Brazil, Mexico

Better: Louisiana, Toronto, California, Sweden, Ontario, Iran, Malta, Spain

Much better: Poland

Today's post-peak days-to-eradication numbers for regions of interest: Norway (14→15, 9.7%), Thailand (39→35, 4.7%), New Zealand (34→61, 2.9%), Japan (89→70, 28.0%), Austria (116→133, 8.7%), Israel (154→148, 28.2%), Denmark (141→159, 23.2%), Germany (482→261, 16.5%), Alabama (207→299, 84.4%), South Korea (133→309, 10.4%), Italy (406→388, 68.3%), Australia (345→390, 10.6%), Romania (1013→537, 76.8%), China (610→610, 0.2%), Malaysia (737→613, 44.6%), France (788→679, 50.4%), U.K. (626→707, 90.5%), Turkey (2097→2039, 48.2%), New York (3029→2927, 72.0%), Singapore (5119→43241, 98.9%)

Days to eradication for other post-peak regions: Belize (0, 0.0%), NT CAN (0, 0.0%), PE CAN (0, 0.0%), Papua New Guinea (0, 0.0%), YT CAN (0, 0.0%), Mauritius (1, 0.0%), Western Sahara (1, 0.0%), Saint Lucia (3, 0.0%), Eritrea (5, 0.0%), Timor-Leste (5, 0.0%), Dominica (7, 0.0%), Seychelles (8, 0.0%), Saint Kitts and Nevis (9, 0.0%), Trinidad and Tobago (11, 0.0%), Montenegro (14, 0.5%), NL CAN (14, 1.6%), Monaco (21, 6.5%), Tunisia (30, 16.0%), Estonia (31, 26.2%), Holy See (32, 48.0%), MB CAN (33, 8.8%), Taiwan (36, 8.7%), NS CAN (40, 4.7%), Grenada (41, 41.7%), Uganda (42, 58.7%), Andorra (55, 13.6%), Cuba (57, 21.1%), Paraguay (62, 60.1%), SK CAN (62, 42.1%), Ireland (64, 16.5%), Croatia (66, 11.4%), Slovakia (71, 23.7%), HI USA (86, 12.6%), Sao Tome and Principe (88, 41.5%), Slovenia (118, 3.1%), Switzerland (132, 5.7%), Niger (138, 21.0%), Tanzania (140, 42.8%), Georgia (168, 69.5%), San Marino (182, 65.9%), VT USA (184, 11.2%), Guinea-Bissau (204, 79.4%), Lithuania (204, 57.2%), Ghana (207, 82.3%), Uruguay (225, 41.6%), Bosnia and Herzegovina (230, 58.6%), MI USA (235, 79.1%), Morocco (261, 79.6%), Burma (289, 54.5%), KY USA (329, 89.7%), PA USA (332, 86.9%), Luxembourg (349, 5.5%), BC CAN (366, 42.3%), Vietnam (377, 36.7%), Sri Lanka (407, 76.4%), KS USA (415, 83.5%), WA USA (504, 88.9%), OK USA (560, 83.5%), Burkina Faso (633, 28.0%), OR USA (634, 92.8%), Bahrain (661, 94.7%), TN USA (675, 85.8%), AK USA (710, 20.5%), SD USA (788, 77.7%), NE USA (809, 92.8%), Eswatini (851, 87.7%), Somalia (868, 92.2%), Czechia (1010, 61.4%), Belgium (1350, 71.2%), MA USA (1381, 89.0%), Portugal (1662, 80.8%), Greece (1686, 65.6%), NH USA (2043, 90.8%), Rwanda (2080, 69.6%), Bulgaria (2144, 95.8%), DC USA (2274, 95.5%), DE USA (2318, 98.5%), CO USA (2398, 90.5%), NJ USA (3001, 79.8%), Cabo Verde (3096, 82.7%), WY USA (3120, 99.2%), Netherlands (3520, 68.2%), CT USA (3523, 91.9%), Dominican Republic (3917, 95.1%), Albania (4112, 60.6%), Cyprus (4310, 51.8%), QC CAN (20426, 95.2%)

New Zealand's days to eradication backing off from 34 to 61 is the result of a big drop in active cases five days ago working its way out of the four-day mean; the situation has not changed substantially. Japan's decrease in days to eradication was a result of a steep drop from 89 to 70 remaining active cases.

I had two reasons to feel somewhat optimistic about Toronto today. One was that while driving to pick up groceries (curbside seafood from the excellent Mike's Fish Market at the St. Lawrence Market), I saw that most people outside of my neighbourhood were fairly compliant with Canadian Chief Public Health Officer Theresa Tam's instructions to wear face coverings when not able to maintain a safe social distance. I'll just make a point of not shopping on the Danforth for the foreseeable future.

The other was that active cases increased by only nine to 1,883 today after rising 447 over the previous four days. It's unfortunately more pertinent to observe that we had our fifth consecutive day of 200+ new cases after six days with far fewer, and that if the ratio of newly confirmed cases to current active cases is decreasing, it's much more because the active cases are increasing than because the new cases are decreasing. The reason it looks better (and is so labelled above) is because we had many more people than usual declared recovered today, subtracted from the active case total. On the bright side, it means that there are only 1,883 new cases and not closer to 2,000 to spread further contagion.

You can say more or less the same about Ontario's numbers today. Three days straight now of 400+ cases, but only a net increase of six new active cases in this province, because 379 were declared cured and 27 died.

Another type of darkly positive news is that excess mortality numbers in Europe have reached the point where they are clearly showing that they peaked back in early April. It takes three weeks for numbers to finish reporting, so the last reasonably firm numbers are for Week 17 (April 20-26). In all parts of the continent participating in Euromomo.eu, mortality numbers peaked in Week 14 or 15 (March 30-April 12). In England, they reached only 42.34 standard deviations above seasonal means rather than the forecast 44.10, and have since dropped to 30.11 standard deviations.

In Spain, they reached 34.23 standard deviations, and I am mystified by a forecast that they will be at -4.14 (i.e., below the mean) this week. I could imagine that if a lot more people who were close to death were killed by COVID-19, that might cause a rebound effect reducing mortality for a short period of time, but four sigmas seems implausibly large.

Stay safe.

2020-05-22

Comment on this post on Facebook.

today's daily chart as described below in text

Today's pre-peak days-to-one-percent infection numbers: Louisiana (17→15), Brazil (35→30), Toronto (49→41), U.S. (54→51), Mexico (64→63), South Africa (65→67), California (69→69), India (91→91), Sweden (98→98), Ontario (111→103), Iran (108→108), Canada (111→113), Malta (106→116), Poland (140→133), Indonesia (139→136), Spain (237→152)

Boilerplate explanation paragraph: The number given with each country is my own way of boiling down your region or country's growth/doubling rate and number of people infected with COVID-19 into one number. It represents the number of days you have left at your current rate of infection (averaged over the last four days) before you reach 1%. It's not a forecast (your rate of change is likely to change before then, when public health authorities take positive or negative action), just one simple number that tells you how you're doing compared to others. I could have picked a bigger percentage than 1%, but you can't go too far before the exponential growth bends back down logistically, and it seemed as easy a place as any; most people can visualize the idea of 1 in 100. Starting on 2020-05-06, I stopped displaying these numbers for regions whose active case counts were known to be lower than an estimated previous peak value; in these cases, I show separately the number of days it would take the current 4-day average of exponential decay in active case counts to drop below half a person, with a floor of 14 days since the last new confirmed case, as well (starting 2020-05-14) as the current percentage of peak active cases.

Compared to yesterday, here's how these regions are doing.

Much worse: Brazil, Toronto, Ontario, Poland, Spain

Worse: Louisiana, U.S., Indonesia

The same: Mexico

Better: South Africa, California, India, Sweden, Iran, Canada

Much better: Malta

Today's post-peak days-to-eradication numbers for regions of interest: Norway (976→14, 9.3%), New Zealand (34, no new data, 3.0%), Thailand (63→39, 4.9%), Japan (89→89, 30.6%), Austria (120→116, 8.5%), South Korea (128→133, 10.3%), Denmark (184→141, 23.6%), Israel (144→154, 30.0%), Alabama (new→207, 78.7%), Australia (318→345, 10.8%), Italy (467→406, 70.1%), Germany (330→482, 17.0%), China (new→610, 0.2%), U.K. (621→626, 90.5%), Malaysia (598→737, 44.8%), France (1069→788, 50.8%), Romania (8446→1013, 80.6%), Turkey (1515→2097, 48.7%), New York (2635→3029, 72.1%), Singapore (2112→5119, 98.3%)

Days to eradication for other post-peak regions: Belize (0, 0.0%), NT CAN (0, 0.0%), PE CAN (0, 0.0%), Papua New Guinea (0, 0.0%), YT CAN (0, 0.0%), Mauritius (2, 0.0%), Western Sahara (2, 0.0%), Saint Lucia (4, 0.0%), Eritrea (6, 0.0%), Timor-Leste (6, 0.0%), Dominica (8, 0.0%), Seychelles (9, 0.0%), Saint Kitts and Nevis (10, 0.0%), Trinidad and Tobago (12, 0.0%), Iceland (14, 0.2%), Monaco (14, 4.9%), Montenegro (14, 0.5%), NL CAN (14, 2.1%), Sao Tome and Principe (24, 41.5%), Estonia (32, 26.6%), Holy See (32, 48.0%), Tunisia (35, 18.2%), Uganda (37, 52.5%), MB CAN (39, 9.3%), Grenada (41, 41.7%), NS CAN (44, 6.2%), Andorra (51, 13.9%), Taiwan (53, 9.8%), Ireland (61, 15.9%), Cuba (64, 24.0%), Croatia (66, 12.5%), Slovenia (74, 3.1%), SK CAN (76, 47.4%), Bahamas (86, 61.5%), Slovakia (96, 26.5%), Niger (101, 22.2%), AK USA (118, 19.4%), VT USA (122, 11.4%), Tanzania (140, 42.8%), Georgia (142, 72.5%), Switzerland (145, 6.3%), Kosovo (164, 36.4%), AB CAN (165, 27.7%), Eswatini (169, 78.8%), Uruguay (172, 42.9%), Paraguay (196, 75.0%), San Marino (206, 68.2%), TN USA (207, 83.2%), Bosnia and Herzegovina (209, 61.4%), SD USA (213, 74.6%), Lithuania (243, 59.1%), Albania (249, 59.9%), Latvia (249, 69.0%), Syria (265, 64.9%), Bhutan (279, 95.2%), Vietnam (279, 36.1%), BC CAN (332, 43.2%), Burma (333, 53.5%), Tajikistan (341, 77.2%), OK USA (345, 84.0%), WA USA (370, 88.1%), Guinea-Bissau (372, 83.9%), Morocco (407, 85.4%), KS USA (472, 87.5%), Czechia (507, 60.0%), PA USA (532, 91.3%), Ghana (580, 83.7%), Burkina Faso (633, 28.0%), Cabo Verde (633, 80.8%), WY USA (753, 95.2%), CT USA (794, 92.2%), KY USA (887, 89.2%), OR USA (935, 95.6%), CO USA (979, 87.8%), IA USA (987, 96.2%), Bulgaria (1062, 95.8%), Guinea (1178, 97.3%), Portugal (1235, 79.4%), Belgium (2079, 71.8%), Netherlands (2236, 68.4%), Cyprus (2364, 51.9%), Luxembourg (2746, 5.5%), NJ USA (3352, 81.2%), Greece (4092, 66.5%), Mongolia (4732, 99.5%), QC CAN (5935, 94.8%)

At the risk of sounding repetitious, we did a great job in Toronto and in Ontario of bringing the COVID-19 pandemic locally to heel, so that at the beginning of this week, we had only 1,427 (Toronto) active cases out of a total of 6,493, with that number dropping steadily at 3%, setting us up to eradicate the disease by early 2021. Then, because of warmer weather and relaxed government messaging, everyone decided that they were done with social distancing and facial coverings. As I posted elsewhere today, a quick survey of the first 100 pedestrians I saw walking on the Danforth, the crowded local high street, showed 7 wearing face masks. 8 if you include the one who dangled their mask under their chin. We've now had four days of record-setting increases, and have climbed to 1,874 with no sign of slowing down. I am expecting a little decline in the percentage increase though for two or three days, because I think people stayed in during bad weather on the weekend. Then it'll keep soaring until stricter restrictions are brought back in place. The exercise has set us back two weeks so far (that is, today's active case count is what it was two weeks ago), and that'll get worse before it gets better.

If you're in urban Ontario, please keep yourself and fellow residents safe by staying home except for urgent errands, wearing a mask and maintaining your distance if you do go out, and washing your hands thoroughly when you come home.

Meanwhile in Malta, there was only one new case added to their active total of 125 today, the lowest in two weeks, despite continued intense testing efforts. Malta ranks fourth among countries of at least 100,000 people in per capita COVID-19 testing, behind Iceland, U.A.E. and Bahrain, and ahead of Luxembourg, Lithuania, Denmark, Mauritius and Cyprus. Canada and the U.S. are at the bottom of the first quartile, with a per capita testing rate of less than a third of the leaders.

Turkey marks its seventh consecutive day of daily decreases in new cases, the longest run among my regions of interest. (New York has the longest run of increases at three days, but is probably part of a weekly reporting cycle.) Thailand and China reported no new cases today. Testing revealed one new case in New Zealand yesterday, a recent rarity, but not considered serious because it was acquired from a household member in isolation.

Today's graphic shows how all of the countries in the world are doing at bringing down their active case numbers.

Stay safe.

2020-05-21

Comment on this post on Facebook.

today's daily chart as described below in text

Today's pre-peak days-to-one-percent infection numbers: Brazil (35→35), Alabama (39→39), Toronto (56→49), U.S. (53→54), Mexico (68→64), South Africa (62→65), California (72→69), India (90→91), Sweden (98→98), Malta (127→106), Iran (108→108), Canada (107→111), Ontario (116→111), Indonesia (166→139), Poland (151→140), China (85520→100617)

Compared to yesterday, here's how these regions are doing.

Much worse: Toronto, Mexico, Malta, Ontario, Indonesia, Poland

Worse: California

Better: Brazil, Alabama, U.S., South Africa, India, Sweden, Iran, Canada

Much better: China

Today's post-peak days-to-eradication numbers for regions of interest: New Zealand (40→34, 3.0%), Thailand (88→63, 5.8%), Japan (145→89, 33.5%), Austria (133→120, 8.8%), South Korea (133→128, 10.4%), Israel (183→144, 32.2%), Denmark (188→184, 25.5%), Louisiana (113→191, 42.4%), Australia (213→318, 10.9%), Germany (314→330, 17.5%), Italy (611→467, 72.1%), Malaysia (653→598, 44.3%), U.K. (821→621, 90.7%), Norway (817→976, 57.6%), France (1088→1069, 51.7%), Spain (1073→1314, 54.8%), Turkey (1074→1515, 49.0%), Singapore (new→2112, 98.1%), New York (2621→2635, 72.4%), Romania (1470→8446, 81.7%)

Days to eradication for other post-peak regions: Belize (0, 0.0%), NT CAN (0, 0.0%), PE CAN (0, 0.0%), Papua New Guinea (0, 0.0%), YT CAN (0, 0.0%), Mauritius (3, 0.0%), Western Sahara (3, 0.0%), Saint Lucia (5, 0.0%), Eritrea (7, 0.0%), Timor-Leste (7, 0.0%), Dominica (9, 0.0%), Seychelles (10, 0.0%), Saint Kitts and Nevis (11, 0.0%), Trinidad and Tobago (13, 0.0%), Benin (14, 11.8%), Iceland (14, 0.3%), Monaco (14, 4.9%), Montenegro (14, 0.5%), NL CAN (14, 2.1%), Slovenia (14, 3.1%), Sao Tome and Principe (31, 48.4%), Holy See (32, 48.0%), Uganda (32, 44.6%), NS CAN (34, 6.2%), Estonia (36, 28.1%), Grenada (41, 41.7%), MB CAN (44, 9.3%), Tunisia (52, 21.6%), Ireland (57, 15.0%), Niger (69, 21.8%), SK CAN (73, 50.7%), Andorra (78, 16.9%), Bahamas (86, 61.5%), Cuba (87, 26.5%), Slovakia (88, 27.7%), Switzerland (101, 6.2%), AK USA (111, 18.4%), Croatia (125, 15.2%), Albania (129, 57.8%), Kosovo (159, 36.3%), Eswatini (174, 79.7%), BC CAN (177, 42.8%), Bosnia and Herzegovina (177, 61.1%), San Marino (187, 67.4%), Guinea-Bissau (192, 86.0%), Georgia (199, 75.2%), Burkina Faso (200, 28.3%), VT USA (204, 12.3%), Taiwan (207, 13.2%), Uruguay (213, 44.6%), SD USA (215, 75.9%), AB CAN (217, 29.6%), TN USA (268, 82.5%), MS USA (278, 85.2%), Bhutan (279, 95.2%), Morocco (350, 84.7%), Lebanon (388, 84.1%), Cabo Verde (402, 82.1%), KY USA (442, 89.9%), HI USA (447, 15.1%), Latvia (454, 72.1%), Luxembourg (493, 5.8%), WA USA (518, 90.2%), Vietnam (561, 36.7%), Czechia (750, 61.1%), Bulgaria (773, 95.0%), KS USA (791, 84.3%), MO USA (969, 93.3%), WY USA (1010, 96.0%), Cyprus (1125, 51.4%), Dominican Republic (1135, 90.9%), WI USA (1253, 97.1%), NE USA (1269, 96.9%), Greece (1270, 65.6%), OK USA (1304, 84.5%), OR USA (1361, 96.3%), NC USA (1454, 97.4%), Burma (1760, 54.1%), Portugal (1986, 80.5%), Netherlands (2785, 68.7%), Malawi (2792, 99.4%), Belgium (4122, 72.5%), IA USA (5041, 97.5%), CT USA (7606, 92.4%), CO USA (9356, 91.0%), MA USA (10265, 90.7%), Zambia (15645, 88.7%), Jamaica (15648, 49.3%), Lithuania (26572, 66.1%), Saudi Arabia (39312, 99.9%), NJ USA (78205, 81.4%), Ghana (113481, 95.1%)

I am unhappily unsurprised to see that last week's relaxed attitudes toward social distancing in Toronto and Ontario had exactly the effects that you would expect this week.

Toronto saw 228 new confirmed cases, 200+ for the third straight day. The last time that happened was May 6-8.

Ontario saw 413 new confirmed cases, the second time in three days that we've broken 400. The last time that happened was May 6-8.

Toronto had 14 new confirmed cases for every 100 active cases; in Ontario at large it was 11%. Toronto's 4-day average increase in active cases is up to 4%, the worst it's been since we started tracking them. Ontario's is at 1.9%, also the worst ever.

If there's one thing that surprises me, it's that we haven't yet started to see worse numbers in the so-called cottage country surrounding the city. Mobile phone operators have reported much larger number of people leaving the city much earlier in the year than usual, but many regions have also taken steps to keep seasonal residents from reopening their summer homes, to help preserve scarce rural health resources.

The situation in Quebec might be heading in the same direction. Their per capita level of active infection is seven times Ontario's, and has started increasing again as of today.

As a result of the situation in these two most populated provinces, Canada as a whole is now seeing its four-day mean of active cases trending back upward after two days in negative territory.

I suppose that's enough negativity for now.

Estonia caught my eye showing up on the days to eradication list above, at 36 days, based on the Johns Hopkins University dataset. I know that they have a low rate of spread now, of one or two percent per active case, and only a few hundred active cases; what made me wonder was why they hadn't been on the list yesterday. It turns out that there's a problem with how the recovered figures are being gathered by JHU. Official reports list the number of closed cases, and as a separate category the number of cases that have gone 28 days without treatment or death (and are therefore presumed cured). Today's recovery number includes both types; past days included only the first number. I've submitted a bug report.

Japan is seeing a steady and rapid reduction in the active case count, driven largely by a very small daily new case rate of about 1% (of active cases - remember, the corresponding figure in Toronto is 14%).

New Zealand is down to 28 active cases and saw only one new case all week; they are well on course to eradication.

Iceland is down to 3 active cases, and is averaging one new confirmed case per week. They're a good indication of what a good end to the epidemic might look like.

Stay safe.

2020-05-20

Comment on this post on Facebook.

today's daily chart as described below in text

Today's pre-peak days-to-one-percent infection numbers: Brazil (37→35), Singapore (39→36), Alabama (39→39), U.S. (50→53), Toronto (69→56), South Africa (56→62), Mexico (67→68), California (68→72), India (90→90), Sweden (58→98), Iran (115→108), Ontario (115→116), Malta (129→127), Poland (177→151), Indonesia (175→166), China (71263→85520)

Compared to yesterday, here's how these regions are doing.

Much worse: Toronto, Iran, Poland, Indonesia

Worse: Brazil, Singapore, Malta

Better: Alabama, U.S., Mexico, California, India, Ontario

Much better: South Africa, Sweden, China

Today's post-peak days-to-eradication numbers for regions of interest: New Zealand (51→40, 3.2%), Thailand (new→88, 6.2%), Louisiana (109→113, 35.9%), Austria (new→133, 9.0%), South Korea (175→133, 10.5%), Japan (147→145, 45.5%), Israel (146→183, 34.7%), Denmark (167→188, 26.8%), Australia (422→213, 10.8%), Germany (353→314, 18.3%), Italy (987→611, 74.2%), Malaysia (457→653, 45.8%), Norway (975→817, 57.7%), U.K. (4895→821, 92.2%), Spain (new→1073, 54.3%), Turkey (902→1074, 49.1%), France (2008→1088, 52.0%), Romania (2044→1470, 82.2%), New York (2125→2621, 72.8%), Canada (5083→2847, 98.5%)

Days to eradication for other post-peak regions: Belize (0, 0.0%), NT CAN (0, 0.0%), Papua New Guinea (0, 0.0%), YT CAN (0, 0.0%), PE CAN (1, 0.0%), Mauritius (4, 0.0%), Western Sahara (4, 0.0%), Saint Lucia (6, 0.0%), Cambodia (8, 0.0%), Eritrea (8, 0.0%), Timor-Leste (8, 0.0%), NB CAN (9, 0.0%), Dominica (10, 0.0%), Seychelles (11, 0.0%), Saint Kitts and Nevis (12, 0.0%), Benin (14, 22.7%), Monaco (14, 4.9%), NL CAN (14, 2.1%), Namibia (14, 15.4%), Slovenia (14, 3.3%), Iceland (21, 0.4%), Montenegro (25, 1.4%), Holy See (32, 48.0%), NS CAN (34, 6.9%), Sao Tome and Principe (36, 52.7%), Grenada (41, 41.7%), Ireland (55, 14.5%), MP USA (64, 54.5%), Andorra (72, 16.9%), Tunisia (73, 25.3%), Taiwan (74, 13.2%), Burkina Faso (75, 29.9%), Croatia (77, 15.0%), Bahamas (86, 61.5%), Slovakia (88, 28.7%), SK CAN (89, 56.9%), HI USA (92, 14.5%), Cuba (94, 29.2%), Switzerland (104, 6.7%), Niger (120, 24.3%), VT USA (127, 11.9%), Kosovo (145, 34.3%), Bosnia and Herzegovina (148, 64.2%), Uruguay (183, 45.5%), MB CAN (184, 11.9%), BC CAN (228, 44.2%), MS USA (228, 79.9%), TN USA (257, 82.3%), SD USA (270, 79.5%), AB CAN (273, 31.0%), San Marino (278, 70.4%), Georgia (282, 75.8%), Guinea (296, 87.2%), Albania (327, 60.6%), Ecuador (378, 45.1%), Dominican Republic (429, 91.6%), Morocco (462, 88.0%), WA USA (489, 90.5%), Guinea-Bissau (506, 91.9%), Latvia (526, 70.2%), Cabo Verde (612, 87.7%), North Macedonia (645, 47.2%), NC USA (715, 89.1%), OK USA (806, 80.5%), Eswatini (871, 89.4%), QC CAN (905, 94.8%), Cyprus (1052, 52.1%), MO USA (1058, 94.8%), Lithuania (1127, 64.0%), IA USA (1388, 94.3%), Bulgaria (1395, 94.9%), KY USA (1622, 89.5%), Czechia (1727, 62.7%), KS USA (1761, 90.9%), PA USA (1808, 96.3%), NE USA (1814, 96.5%), Netherlands (2357, 69.1%), OR USA (2414, 98.7%), NJ USA (2922, 81.7%), Guyana (4169, 96.1%), MN USA (5338, 95.7%), MA USA (5504, 91.2%), Sri Lanka (15520, 78.3%), Greece (20322, 66.3%), Hungary (22668, 80.7%), Ghana (113481, 95.1%)

As I commented earlier, today’s numbers for Toronto and Ontario bear out my observations from yesterday; we are rapidly regressing, and are back to where we were a week ago in active cases. Each active confirmed case in Toronto resulted in an average of 0.17 new confirmed cases today, almost as high as yesterday’s record. Toronto had dipped down to 1,427 active confirmed cases two days ago, and is now back up to 1,678. The four-day trend in active cases is now positive, indicating that we are headed back up toward the 1% infection mark and not down to eradication.

In the past week, Toronto has seen 1,185 new cases. The city categorizes 722 (61%) of these as close contact with another case, 345 (29%) as community-acquired, 63 (5%) as travel-related, 34 (3%) as institutional, and 21 (2%) as healthcare. The 2.1:1 ratio of close contact cases to community cases matches the cumulative ratio, even though they accounted for 90% of recent new cases compared to 75% historically. Among the other three categories, institutional cases were proportionately up a little, healthcare down a lot, and travel up a lot. I'm surprised to learn that people are still traveling to Toronto and importing new cases of COVID-19.

The best way that I can see this playing out in the short term is if enough people are concerned enough at the sudden uptick here that they start being a lot more careful, and then pat themselves on the back a little prematurely when the effect of Sunday and Monday's bad weather keeps the numbers down.

Canadians were asked today to wear masks or face coverings for additional protection where physical distancing is difficult to maintain. I did see most fellow customers at the local Loblaws supermarket wearing masks, although some teenagers were making a point of hooking masks on their ears but dangling the mask under their chins. Kids.

Today's graphic shows how Ontario is now the outlier among Canadian provinces and territories, in seeing both increasing numbers of active cases, and at a rate that is increasing as well.

In happier news, Turkey saw a fifth straight day of declines in new cases; today's figure of 972 marked the first time since Marwch 25th that there were fewer than 1,000 new cases there in a day.

19 American states and territories are now listed above as post-peak, climbing back up from a recent low of 12 two days ago.

New Zealand reported a fourth consecutive day with no new cases. Four days ago, four historic cases were added to the total, but they were long since resolved. They are down to 30 remaining active cases, and might well be COVID-19 free in a month or two.

Stay safe.

2020-05-19

Comment on this post on Facebook.

today's daily chart as described below in text

Today's pre-peak days-to-one-percent infection numbers: Brazil (35→37), Alabama (50→39), Singapore (33→39), U.S. (47→50), South Africa (63→56), Sweden (58→58), Mexico (67→67), California (63→68), Toronto (82→69), India (95→90), Spain (103→103), Iran (114→115), Ontario (114→115), Malta (131→129), Indonesia (170→175), Poland (172→177), Austria (591→659), Thailand (5051→8219)

Compared to yesterday, here's how these regions are doing.

Much worse: Alabama, South Africa, Toronto, India

Worse: Malta

Better: Brazil, U.S., Sweden, Mexico, California, Spain, Iran, Ontario, Indonesia, Poland

Much better: Singapore, Austria, Thailand

Today's post-peak days-to-eradication numbers for regions of interest: New Zealand (49→51, 3.8%), Louisiana (104→109, 34.5%), Israel (202→146, 35.4%), Japan (146→147, 45.5%), Denmark (127→167, 28.1%), South Korea (394→175, 11.4%), Germany (208→353, 19.2%), Australia (524→422, 11.4%), Malaysia (228→457, 46.9%), China (87→511, 0.2%), Turkey (769→902, 49.3%), Norway (2800→975, 58.1%), Italy (1605→987, 77.0%), France (8863→2008, 53.9%), Romania (213→2044, 82.3%), New York (2138→2125, 73.2%), U.K. (2524→4895, 97.0%), Canada (121908→5083, 98.5%)

Days to eradication for other post-peak regions: Belize (0, 0.0%), NT CAN (0, 0.0%), Papua New Guinea (0, 0.0%), YT CAN (0, 0.0%), PE CAN (2, 0.0%), Mauritius (5, 0.0%), Western Sahara (5, 0.0%), Saint Lucia (7, 0.0%), Cambodia (9, 0.0%), Eritrea (9, 0.0%), Timor-Leste (9, 0.0%), NB CAN (10, 0.0%), Dominica (11, 0.0%), Seychelles (12, 0.0%), Saint Kitts and Nevis (13, 0.0%), Benin (14, 17.6%), Iceland (14, 0.3%), Slovenia (14, 3.6%), Montenegro (25, 1.4%), NS CAN (26, 6.9%), Sao Tome and Principe (36, 52.7%), Brunei (37, 3.8%), NL CAN (42, 3.6%), Croatia (55, 15.9%), Burkina Faso (57, 29.0%), West Bank and Gaza (60, 19.7%), SK CAN (61, 58.9%), Taiwan (62, 13.2%), Gambia (66, 83.3%), Andorra (74, 19.3%), Switzerland (76, 7.2%), AB CAN (77, 23.5%), Bahamas (87, 64.3%), Grenada (89, 52.1%), VT USA (108, 12.6%), Cuba (115, 31.8%), Niger (121, 24.5%), Cabo Verde (131, 83.4%), HI USA (137, 15.5%), Kosovo (145, 34.3%), Slovakia (160, 33.3%), Uruguay (160, 45.9%), GU USA (184, 25.3%), Jamaica (188, 48.5%), Bosnia and Herzegovina (193, 66.2%), MB CAN (213, 13.5%), MS USA (232, 74.6%), Tunisia (246, 31.8%), SD USA (256, 80.8%), BC CAN (260, 45.3%), MT USA (265, 6.3%), Ecuador (295, 42.2%), Liberia (301, 80.2%), NH USA (404, 92.3%), Kyrgyzstan (422, 79.6%), WA USA (449, 90.6%), Luxembourg (497, 5.9%), KS USA (525, 91.0%), Serbia (614, 65.3%), Morocco (781, 90.7%), Angola (790, 90.1%), Georgia (799, 80.2%), Cyprus (850, 52.1%), MO USA (872, 94.6%), IA USA (874, 94.1%), TN USA (937, 90.8%), North Macedonia (1028, 47.3%), QC CAN (1035, 95.4%), Albania (1044, 60.9%), Uzbekistan (1086, 38.5%), NJ USA (1235, 81.1%), Netherlands (1383, 69.0%), San Marino (1390, 75.1%), NE USA (1619, 97.2%), Rwanda (1686, 70.3%), Dominican Republic (2254, 96.0%), Hungary (3166, 80.3%), Paraguay (4013, 91.2%), OR USA (5170, 97.6%), Ghana (5690, 97.6%), Guinea (166410, 87.2%), NC USA (283062, 84.4%)

Based on Johns Hopkins University Data, San Marino, Benin, and Qatar are well past the 1% confirmed infection mark, and the following countries are within a month of it: Nicaragua (9), Kuwait (14), Mauritania (15), Chile (23), Bahrain (24), Tajikistan (28), Peru (29), Burundi (30).

There are various ways of expressing how rapidly an epidemic is spreading. There's the effective reproduction number, which gives on average how many other people each infected person infects: it's hard to estimate accurately, but tells you whether the epidemic is worsening (R_e > 1) or dying out (R_e < 1). There is the new confirmed case percentage, which is easy to calculate, but becomes less meaningful once you have a significant number of recovered or dead people who aren't infectious anymore, as the denominator of the ratio keeps growing and makes a low percentage look better but be less comparable with earlier percentages. There is the daily new case count: just as easy to calculate, as long as you have a functioning public health system and good testing, but it tends to make for sensationalistic media sound bites as new peaks are reached even as the percentage rate is decreasing.

What I like right now for ease of calculation and understanding is the number of new cases each day as a percentage of the remaining active cases. This is closely related to the effective reproduction number, and also gives you a direct sense of how well active cases are being contained (or perhaps, how many untested cases remain in the community).

Today's graphic plots all the countries of the world according to this rate, and what the level of active infection is.

In Ontario, the government announced last week that limited reopening would begin today, which (together with sunny weather last week) led to a substantial increase in the number of people gathering in public places. You would think that this would lead to increased new cases about now.

In Ontario, our new cases per remaining active case jumped from 8.9% to 12.5%, the highest it's been in a month. The 427 new cases themselves are not an unusually large number for recent weeks; there was just no reason for us to have so many per active case after we had driven our active count down so quickly and so far. That pushes our four-day average change in active cases back into positive territory, and moves us from the post-peak list above back to the pre-peak list.

In Toronto, our new cases per remaining active case nearly tripled, from 6.3% to 18.4%, the highest it's been since we started tracking recovered and active cases over a month ago. That pushes us back up from 48% of peak active to 52%, and tomorrow we will almost certainly move back from the post-peak list to the pre-peak list.

If the situation continues, Canada will move back onto pre-peak status as well. I'm glad the Canadian government was able to negotiate an extended closure of our U.S. border today; I think the numbers show that we could be better at public education about the ongoing risks of the epidemic.

Stay safe.

2020-05-18

Comment on this post on Facebook.

today's daily chart as described below in text

Today's pre-peak days-to-one-percent infection numbers: Singapore (27→33), Brazil (35), U.S. (47), Alabama (47→50), Sweden (58), California (63), South Africa (63), Mexico (64→67), India (95), Iran (118→114), Malta (103→131), Indonesia (160→170), Poland (163→172)

Compared to yesterday, here's how these regions are doing.

Much worse: Iran

Better: Brazil, U.S., Alabama, Sweden, California, South Africa, Mexico, India, Indonesia, Poland

Much better: Singapore, Malta

Today's post-peak days-to-eradication numbers for regions of interest: New Zealand (49, no new data, 4.8%), China (87, no new data, 0.2%), Louisiana (new→104, 32.8%), Denmark (127, no new data, 30.0%), Japan (146, no new data, 49.5%), Israel (202, no new data, 40.9%), Germany (208, no new data, 19.8%), Romania (213, no new data, 82.1%), Malaysia (228, no new data, 46.6%), Toronto (271→281, 47.8%), South Korea (394, no new data, 13.1%), Australia (524, no new data, 11.9%), Spain (671, no new data, 56.7%), Turkey (769, no new data, 50.4%), Ontario (1020→1336, 69.0%), Italy (1605, no new data, 79.7%), Austria (1706, no new data, 11.2%), New York (1829→2138, 73.3%), U.K. (2524, no new data, 98.1%), Thailand (2538, no new data, 8.0%), Norway (2800, no new data, 59.7%), France (8863, no new data, 54.0%), Canada (121908, no new data, 98.2%)

Days to eradication for other post-peak regions: NT CAN (0, 0.0%), Papua New Guinea (0, 0.0%), Suriname (0, 0.0%), YT CAN (0, 0.0%), Belize (1, 0.0%), PE CAN (3, 0.0%), Mauritius (7, 0.0%), Western Sahara (7, 0.0%), Saint Lucia (9, 0.0%), Cambodia (11, 0.0%), Eritrea (11, 0.0%), NB CAN (11, 0.0%), Timor-Leste (11, 0.0%), Dominica (13, 0.0%), Iceland (14, 0.5%), Montenegro (14, 1.9%), Saint Vincent and the Grenadines (14, 27.3%), Brunei (21, 3.8%), Fiji (25, 24.5%), Namibia (25, 23.1%), Comoros (30, 70.0%), NS CAN (39, 9.0%), Taiwan (41, 14.4%), Holy See (44, 60.0%), West Bank and Gaza (50, 20.2%), Andorra (56, 21.9%), Croatia (63, 18.3%), Gambia (66, 83.3%), SK CAN (73, 62.7%), Cuba (83, 35.1%), Burkina Faso (84, 31.5%), Finland (91, 48.7%), Tunisia (91, 32.7%), NL CAN (94, 4.2%), Luxembourg (95, 6.1%), Rwanda (96, 67.5%), VT USA (98, 12.8%), Jamaica (101, 49.4%), MB CAN (110, 13.5%), Slovakia (111, 36.6%), Angola (117, 78.4%), HI USA (124, 15.8%), Switzerland (174, 8.4%), Bosnia and Herzegovina (201, 71.8%), MS USA (238, 69.3%), Paraguay (246, 88.2%), BC CAN (248, 46.7%), Liberia (260, 80.2%), SD USA (346, 86.1%), AB CAN (348, 33.2%), GU USA (360, 25.3%), Maldives (438, 93.6%), Uzbekistan (453, 37.8%), Ecuador (484, 45.4%), Libya (514, 41.3%), NH USA (534, 89.2%), Serbia (536, 66.0%), MO USA (616, 94.3%), Morocco (618, 93.5%), Cyprus (623, 52.6%), Uruguay (696, 49.5%), NE USA (709, 94.8%), Ghana (718, 95.1%), Slovenia (820, 55.1%), QC CAN (914, 95.5%), Kazakhstan (919, 93.1%), Netherlands (1044, 69.7%), Georgia (1064, 85.0%), Hungary (1136, 80.5%), DC USA (1200, 96.8%), NJ USA (1298, 82.3%), Dominican Republic (1370, 94.0%), San Marino (1638, 76.6%), Burma (1784, 54.7%), Bahamas (2283, 74.2%), Belgium (3189, 73.2%), Niger (3487, 29.8%), GA USA (3589, 98.8%), NC USA (4937, 77.0%), Estonia (5527, 56.6%)

Iran recorded the largest number of confirmed new cases today (2,294) since early April. Iran hit a low daily new case number of 802 on May 2nd, two weeks after low-risk activities were permitted to resume subject to social distancing measures, and their new cases have steadily worked their way back up.

I'm going to try to go to bed a little earlier than usual tonight. If I do, then some numbers won't yet have been updated, so I'll end up discussing them tomorrow: Brazil, U.S., Sweden, California, South Africa, India. Some of these countries are a little irregular with their updates, so it could even be the day after.

Malta was back down to only five new cases today. A lot of businesses will reopen in Malta on Friday, subject to new rules; I hope they are able to get their regular operations back underway without further outbreaks, and look forward to seeing how they rise to the challenge of adapting their tourism economy to the post-pandemic world.

Singapore had its first day under 400 in a month. Although Singapore has had a rough time with the ongoing outbreaks in its foreign worker dormitories, it's worth noting that they have by far the lowest reported death rate among my areas of interest, by at least an order of magnitude. The next lowest is Malta, and I have to think that in each case, it's a combination of the most affected populations being younger (foreign workers and refugees more than institutionalise delderly), and that both countries are able to test large numbers of potentially infected residents to find the asymptomatic carriers. The former factor reduces the numerator, the latter increases the denominator in the deaths / cases ratio.

Ontario's active case total crept back up from 3,412 to 3,415 for the first time in a week, but its four-day average rate still points downward.

Quebec appears to have finally hit the top of its curve as of yesterday, having had just barely over 20,000 active cases four days ago, creeping down to just over 19,000 today. Thanks to Quebec's success, Canada has also seen two days of decreases in active cases, and its four-day trend turned negative yesterday.

In Toronto, the active case total dropped again, from 1,471 to 1,427, at the same steady rate that it's seen over the last four days. If nothing changes, we would have another nine or ten months to eradication. One number that is gradually drifting up is the proportion of hospitalised cases, which was 10% of active cases four weeks ago, 20% two weeks ago, and has now hit 29%. Are these severe cases accumulating in hospitals, or are we not testing enough cases in the community, or what?

As forecast, Toronto celebrated a cool, wet Victoria Day. I heard someone setting off a lot of big fireworks in the local schoolyard, but aside from that it looked like there were very few people out and about. Sadly, the forecast is for hot, summery weather by next weekend.

Stay safe.

2020-05-17

Comment on this post on Facebook.

today's daily chart as described below in text

Today's pre-peak days-to-one-percent infection numbers: Louisiana (21→21), Singapore (28→27), Brazil (33→35), U.S. (49→47), Alabama (51→47), Sweden (58→58), California (65→63), South Africa (63→63), Mexico (64→64), India (99→95), Malta (116→103), Iran (118→118), Indonesia (142→160), Poland (155→163)

Compared to yesterday, here's how these regions are doing.

Much worse: Alabama, India, Malta

Worse: U.S., California

The same: Singapore

Better: Louisiana, Brazil, Sweden, South Africa, Mexico, Iran, Poland

Much better: Indonesia

Today's post-peak days-to-eradication numbers for regions of interest: New Zealand (36→49, 4.8%), China (63→87, 0.2%), Denmark (151→127, 30.0%), Japan (407→146, 49.5%), Israel (193→202, 40.9%), Germany (228→208, 19.8%), Romania (239→213, 82.1%), Malaysia (255→228, 46.6%), Toronto (276→271, 49.3%), South Korea (265→394, 13.1%), Australia (562→524, 11.9%), Spain (508→671, 56.7%), Turkey (583→769, 50.4%), Ontario (652→1020, 68.9%), Italy (1402→1605, 79.7%), Austria (241→1706, 11.2%), New York (1556→1829, 73.7%), U.K. (2398→2524, 98.1%), Thailand (61→2538, 8.0%), Norway (2556→2800, 59.7%), France (new→8863, 54.0%), Canada (new→121908, 98.2%)

Days to eradication for other post-peak regions: NT CAN (0, 0.0%), Papua New Guinea (0, 0.0%), Suriname (0, 0.0%), YT CAN (0, 0.0%), Belize (1, 0.0%), PE CAN (4, 0.0%), Mauritius (7, 0.0%), Western Sahara (7, 0.0%), Saint Lucia (9, 0.0%), Timor-Leste (11, 0.0%), Eritrea (11, 0.0%), Cambodia (11, 0.0%), NB CAN (12, 0.0%), Dominica (13, 0.0%), Montenegro (14, 1.9%), Iceland (14, 0.5%), Saint Vincent and the Grenadines (14, 27.3%), Brunei (21, 3.8%), NS CAN (23, 10.1%), Fiji (25, 24.5%), Namibia (25, 23.1%), Comoros (30, 70.0%), Taiwan (41, 14.4%), Holy See (44, 60.0%), West Bank and Gaza (50, 20.2%), NL CAN (50, 4.2%), Andorra (56, 21.9%), Croatia (63, 18.3%), Gambia (66, 83.3%), MB CAN (73, 13.0%), Cuba (83, 35.1%), SK CAN (84, 67.9%), Burkina Faso (84, 31.5%), Tunisia (91, 32.7%), Finland (91, 48.7%), Luxembourg (95, 6.1%), Rwanda (96, 67.5%), Jamaica (101, 49.4%), Slovakia (111, 36.6%), Angola (117, 78.4%), VT USA (149, 13.7%), Switzerland (174, 8.4%), HI USA (193, 15.8%), Bosnia and Herzegovina (201, 71.8%), AB CAN (237, 34.0%), Paraguay (246, 88.2%), Liberia (260, 80.2%), KS USA (287, 88.2%), BC CAN (314, 49.5%), SD USA (371, 87.5%), Maldives (438, 93.6%), Uzbekistan (453, 37.8%), Ecuador (484, 45.4%), Libya (514, 41.3%), Serbia (536, 66.0%), Morocco (618, 93.5%), WV USA (620, 85.7%), Cyprus (623, 52.6%), Uruguay (696, 49.5%), Ghana (718, 95.1%), NJ USA (790, 81.4%), Slovenia (820, 55.1%), DC USA (843, 95.9%), QC CAN (856, 94.8%), Kazakhstan (919, 93.1%), Netherlands (1044, 69.7%), Georgia (1064, 85.0%), Hungary (1136, 80.5%), Dominican Republic (1370, 94.0%), San Marino (1638, 76.6%), Burma (1784, 54.7%), UT USA (1939, 97.8%), MO USA (2240, 96.9%), Bahamas (2283, 74.2%), RI USA (2503, 93.9%), Belgium (3189, 73.2%), Niger (3487, 29.8%), MA USA (4433, 91.1%), Estonia (5527, 56.6%), PA USA (7860, 98.2%), TN USA (8186, 98.8%), MI USA (10137, 89.3%)

Malta had seven new cases today, with the oldest being 55 years old, at least five of whom were community-acquired cases detected due to expanded testing. It seems clear that the virus is widespread in the community, and that Malta has a lot of work to do to locate all and isolate all remaining cases.

Canada's active cases are slightly below four days ago now, so they move from the days-to-one-percent-infection list to the very bottom of the days-to-eradication list. If our active cases continue to decline at the same rate, we will be rid of the last one in the year 2354.

Ontario numbers are dropping slowly, with 23 fewer active cases today for a remaining total of 3,412, compared to 21 fewer yesterday.

Toronto's numbers are also continuing their steady drift downward: for the first time since the stats were posted, fewer than 5% of total cases to date remain in hospital; ICU occupancy is under 90 for the first time in a month; active cases are below 1,500 for the first time in more than a month, which is less than half the peak active cases for the first time since peak. Cool, rainy weather seems to have kept people home for much of the day, which may help delay our second wave.

U.S. numbers are better today than yesterday, but the four-day mean still has three bad days to finish clearing before it's days-to-one-percent starts showing significant improvement. Before that though, there's a reasonable possibility that the U.S. may move over to the days-to-eradication list, as today is the first day when I estimated that active cases in the United States dropped. It's a very rough estimate, based on very noisy data, but it looks to me like there were about 200 fewer cases out of over 600,000 active cases today than there were yesterday, which is the first drop since the pandemic started. When the active case numbers show a downward four-day trend, it'll switch lists.

Today's graphic shows estimated active case number changes vs. ratio of active cases to total population for positive role models Germany, New Zealand and South Korea, and local contenders Canada and the United States.

Stay safe.

2020-05-16

Comment on this post on Facebook.

today's daily chart as described below in text

Today's pre-peak days-to-one-percent infection numbers: Louisiana (21→21), Singapore (24→28), Brazil (33→33), U.S. (53→49), Alabama (51→51), Sweden (55→58), South Africa (63→63), Mexico (62→64), California (68→65), U.K. (69→69), India (104→99), Canada (103→103), Malta (159→116), Iran (127→118), Indonesia (141→142), Poland (124→155), France (381→424)

Compared to yesterday, here's how these regions are doing.

Much worse: U.S., India, Malta, Iran

Worse: California

Better: Louisiana, Brazil, Alabama, Sweden, South Africa, Mexico, U.K., Canada, Indonesia

Much better: Singapore, Poland, France

Today's post-peak days-to-eradication numbers for regions of interest: New Zealand (39→36, 4.8%), Thailand (62→61, 7.9%), China (57→63, 0.2%), Denmark (129→151, 31.5%), Israel (217→193, 41.9%), Germany (203→228, 20.9%), Romania (226→239, 84.4%), Austria (176→241, 11.2%), Malaysia (221→255, 48.0%), South Korea (346→265, 13.1%), Toronto (829→276, 51.1%), Japan (2391→407, 58.3%), Spain (505→508, 56.7%), Australia (213→562, 12.3%), Turkey (440→583, 51.1%), Ontario (854→652, 69.4%), Italy (1690→1402, 80.1%), New York (1650→1556, 74.2%), Norway (1599→2556, 60.1%)

Days to eradication for other post-peak regions: Suriname (0, 0.0%), NT CAN (0, 0.0%), YT CAN (0, 0.0%), Papua New Guinea (1, 0.0%), Belize (2, 0.0%), PE CAN (5, 0.0%), Mauritius (8, 0.0%), Western Sahara (8, 0.0%), Saint Lucia (10, 0.0%), Timor-Leste (12, 0.0%), Cambodia (12, 0.0%), Eritrea (12, 0.0%), NB CAN (13, 0.0%), Iceland (14, 0.5%), Saint Vincent and the Grenadines (14, 27.3%), Montenegro (14, 1.9%), Burundi (14, 53.6%), Brunei (21, 3.8%), Barbados (23, 23.1%), NS CAN (25, 11.2%), Fiji (25, 24.5%), Namibia (25, 23.1%), Monaco (27, 8.1%), Comoros (30, 70.0%), Tanzania (35, 50.2%), Holy See (44, 60.0%), West Bank and Gaza (49, 20.6%), NL CAN (50, 4.2%), Andorra (52, 22.3%), Taiwan (55, 16.7%), Luxembourg (62, 5.7%), Syria (62, 43.2%), Croatia (68, 20.3%), MB CAN (73, 13.0%), Jamaica (73, 49.8%), Cuba (79, 38.0%), Rwanda (81, 69.4%), Tunisia (82, 34.4%), Finland (83, 45.9%), Ireland (83, 26.2%), Slovakia (84, 38.0%), SK CAN (96, 72.7%), Liberia (108, 82.1%), Switzerland (112, 9.0%), Angola (117, 78.4%), Gambia (126, 83.3%), Grenada (131, 52.1%), AB CAN (135, 34.7%), VT USA (165, 13.8%), BC CAN (235, 49.5%), Bosnia and Herzegovina (270, 78.0%), Paraguay (272, 90.3%), Eswatini (283, 91.7%), Serbia (328, 65.3%), Niger (351, 29.2%), Uruguay (370, 51.5%), Georgia (511, 82.7%), Libya (514, 41.3%), Guyana (532, 96.5%), Slovenia (546, 54.9%), SD USA (548, 89.2%), Kazakhstan (585, 92.1%), MI USA (635, 86.1%), Morocco (657, 94.8%), AK USA (667, 19.4%), Hungary (680, 80.8%), Maldives (732, 92.3%), Cyprus (734, 53.4%), NJ USA (816, 83.0%), KS USA (934, 92.8%), San Marino (1022, 76.8%), RI USA (1024, 92.2%), OK USA (1146, 83.5%), Netherlands (1192, 70.4%), Belgium (1773, 73.2%), WV USA (2006, 85.7%), United Kingdom (2398, 97.8%), GA USA (2597, 98.0%), Estonia (2620, 56.2%), DC USA (2708, 98.3%), OH USA (2930, 98.6%), NV USA (3245, 97.4%), Lithuania (3776, 65.6%), MA USA (4277, 91.9%), CO USA (5828, 90.8%), TN USA (6257, 99.1%), PA USA (6865, 98.6%), Ecuador (8143, 50.1%)

Calculating days-to-one-percent infection numbers based on today's Johns Hopkins University data, there are 2 countries (San Marino and Qatar) that are (like New York) past 1% confirmed cases, and another 12 that are on track to pass that mark within 30 days: Kuwait, Andorra (15); Mauritania, Bahrain, Mongolia (18); Chile (22); Central African Republic (24); Peru (25); Gabon (26); Tajikistan (28); Singapore (29); Oman (30).

Because half of these countries are in Asia, today's graphic shows how all the countries in Asia and Oceania are doing, plotting as usual the 4-day average daily new case rate against the ratio of confirmed cases to total population.

Singapore had its first sub-2% new case increase since May 11, with 465 newly confirmed cases, along with their first new death during that time.

Brazil had its first decrease in new cases today in five days, with only 14,919 today compared to 15,305 confirmed yesteray.

Malta had a third straight day of double-digit new cases, reportedly due to increased testing. Only five of the 14 new cases were symptomatic, and today's 1,727 tests (or 0.4% of the population, compared say to 0.1% here in Ontario) set a national record.

391 new cases in Ontario today, fewer than yesterday's 428 but more than any other number in the past week. Active cases dropped slightly from 3,456 to 3,435, still 61% of peak.

129 new cases in Toronto today was the fewest in over a month. Active cases dropped 41 to 1,525, or 51% of their peak, and our days-to-eradication is under a year for the first time in almost a week.

On a brief shopping expedition this holiday afternoon in Toronto, I observed that about 20% of pedestrians were wearing masks, and very few were trying to remain distant. It was the first beautiful, sunny, warm day of spring, the lilacs and daffodils are blooming, everyone was out enjoying it, and a week from now a lot more people will be sick, and some will die.

Stay safe.

2020-05-15

Comment on this post on Facebook.

today's daily chart as described below in text

Today's pre-peak days-to-one-percent infection numbers: Louisiana (23→21), Singapore (27→24), Brazil (50→33), Alabama (47→51), U.S. (57→53), Sweden (59→55), Mexico (70→62), South Africa (64→63), California (73→68), U.K. (65→69), Canada (102→103), India (96→104), Poland (127→124), Iran (135→127), Indonesia (156→141), Malta (176→159), France (464→381)

Compared to yesterday, here's how these regions are doing.

Much worse: Brazil, U.S., Sweden, Mexico, California, Iran, Indonesia, Malta, France

Worse: Louisiana, Singapore, Poland

The same: South Africa

Better: U.K., Canada, India

Much better: Alabama

Today's post-peak days-to-eradication numbers for regions of interest: New Zealand (40→39, 5.3%), China (49→57, 0.2%), Thailand (62→62, 7.9%), Denmark (154→129, 33.7%), Austria (148→176, 10.8%), Germany (207→203, 21.6%), Australia (185→213, 12.2%), Israel (152→217, 44.9%), Malaysia (272→221, 50.2%), Romania (263→226, 83.8%), South Korea (351→346, 13.5%), Turkey (323→440, 51.8%), Spain (1034→505, 58.0%), Toronto (1162→829, 52.4%), Ontario (951→854, 70.0%), Norway (1223→1599, 60.1%), New York (1259→1650, 74.9%), Italy (1823→1690, 80.7%), Japan (1473→2391, 58.1%)

Days to eradication for other post-peak regions: YT CAN (0, 0.0%), NT CAN (0, 0.0%), Suriname (1, 0.0%), Papua New Guinea (2, 0.0%), Belize (3, 0.0%), PE CAN (6, 0.0%), Mauritius (9, 0.0%), Western Sahara (9, 0.0%), Saint Lucia (11, 0.0%), Timor-Leste (13, 0.0%), Eritrea (13, 0.0%), Cambodia (13, 0.0%), Montenegro (14, 1.9%), Saint Vincent and the Grenadines (14, 27.3%), VI USA (14, 6.8%), NB CAN (14, 1.4%), NS CAN (16, 13.1%), Iceland (20, 0.9%), Holy See (24, 60.0%), Namibia (25, 23.1%), Fiji (25, 24.5%), Monaco (27, 8.1%), Barbados (30, 25.0%), Comoros (30, 70.0%), NL CAN (31, 4.7%), Tanzania (35, 50.2%), Syria (42, 39.6%), Botswana (45, 34.3%), Finland (50, 43.4%), Brunei (51, 4.7%), Taiwan (52, 17.4%), Andorra (57, 25.4%), MB CAN (72, 14.5%), Tunisia (73, 35.0%), Cuba (74, 39.6%), Luxembourg (75, 6.1%), BC CAN (78, 50.1%), Ireland (81, 25.5%), Rwanda (82, 71.2%), AK USA (83, 17.9%), AB CAN (87, 34.3%), West Bank and Gaza (88, 26.6%), Slovakia (96, 38.9%), Jamaica (110, 54.9%), Croatia (114, 24.2%), Liberia (136, 85.8%), Panama (150, 71.0%), Grenada (153, 58.3%), Eswatini (162, 83.7%), Libya (164, 39.0%), Bosnia and Herzegovina (174, 76.9%), Latvia (181, 67.4%), SK CAN (191, 84.2%), MT USA (284, 6.7%), Maldives (309, 88.8%), Switzerland (318, 10.7%), OK USA (325, 83.2%), SD USA (336, 91.1%), Georgia (338, 82.7%), Cyprus (402, 53.6%), Paraguay (420, 91.8%), Uruguay (422, 52.8%), SC USA (425, 92.2%), VT USA (441, 15.1%), Serbia (472, 69.2%), WV USA (517, 86.6%), Hungary (533, 81.1%), Slovenia (559, 55.4%), Albania (617, 62.3%), San Marino (715, 76.4%), NJ USA (716, 84.2%), Burma (784, 55.6%), Kazakhstan (788, 95.7%), Morocco (845, 94.8%), CO USA (852, 90.5%), Sri Lanka (1083, 76.0%), RI USA (1212, 92.7%), DC USA (1305, 96.4%), Ecuador (1315, 48.2%), MI USA (1342, 90.7%), Estonia (1348, 56.2%), Belgium (1350, 73.3%), MA USA (1716, 91.8%), CT USA (2001, 87.2%), Netherlands (2135, 71.1%), KS USA (2363, 97.5%), United Kingdom (3247, 98.0%), Portugal (4260, 78.0%), PA USA (4963, 98.7%), OH USA (4964, 98.0%), Benin (6343, 99.6%)

Today's graphic shows all the regions of interest, with their current trends, plotting the four-day average new case rate vs. proportion of population confirmed infected.

The American state of Wisconsin celebrated yesterday a third straight day of 5% or fewer of active cases infecting another patient, and two days of less than 1% growth in the active case total, with a State Supreme Court ruling that the government does not have the authority to lock down its citizens. This was the second time that Wisconsin was able to get the rate of transmission down below 5%; the first was just before a rally of about 1,500 protesters, 72 of whom became infected, boosting the rate briefly back above 8%. Let's see how they're doing in a couple of weeks.

The U.S. numbers are starting to creep back up again for this and similar reasons. Their new case total for the day was back up to where it was six days ago, and their four-day average new case rate was trending back upward for the first time in six days as well. On the bright side, eighteen states, territories or districts show signs of active case decline.

Brazil's daily new case rate has risen in four days from 3.5% to 8.6%, during which time more than 50,000 new cases have pushed their total to over 220,000 confirmed.

Louisiana's daily increase is not so bad (averaging 1.55%), but it's first on the list because they're close to 1% of their population confirmed infected, and they haven't yet seen a decline in their active numbers.

The Canadian province of Nova Scotia reported two days ago that all new cases that day were associated with the Northwood Nursing Home, which now accounts for 42 of the province's 117 remaining active cases, and 45 of its deaths.

Ontario reported 78 cases that should have been included in yesterday's tally were incorrectly delayed to today instead. We continue to average a daily decline in active cases of about 2%, which would take about a year to eradicate the disease from our province if we stay the course; we have instead started reopening businesses while still restricting socialization outside of immediate household members.

Toronto's active case total inched down a little further from 1,598 to 1,566. I went to the St. Lawrence Market to go grocery shopping this afternoon, and saw it more crowded than it had been since our lockdown began; the store clerks were all run ragged. About half the shoppers were wearing masks, and most but not all of the clerks. I'm not sure what to hope for this Victoria Day weekend, when our city traditionally splits between those who drive out to the cottage and those who plant their gardens; I guess it's just for everyone to maintain social distance wherever they are.

Stay safe.

2020-05-14

Comment on this post on Facebook.

today's daily chart as described below in text

Today's pre-peak days-to-one-percent infection numbers: Singapore (26→27), Alabama (47→47), Brazil (50→50), U.S. (54→57), Sweden (70→59), South Africa (63→64), U.K. (65→65), California (67→73), India (96→96), Canada (102→102), Poland (131→127), Iran (144→135), Indonesia (168→156), Malta (252→176)

Compared to yesterday, here's how these regions are doing.

Much worse: Sweden, Poland, Iran, Indonesia, Malta

Better: Singapore, Alabama, Brazil, U.S., South Africa, U.K., India, Canada

Much better: California

Today's post-peak days-to-eradication numbers for regions of interest: New Zealand (43→40, 6.0%), China (48→49, 0.2%), Thailand (68→62, 7.7%), Austria (163→148, 11.0%), Israel (205→152, 45.6%), Denmark (221→154, 35.7%), Australia (171→185, 12.4%), Germany (270→207, 22.4%), Romania (1805→263, 85.8%), Malaysia (282→272, 52.2%), Turkey (283→323, 52.4%), South Korea (767→351, 13.7%), Louisiana (274→637, 47.0%), Ontario (719→951, 70.9%), Spain (1217→1034, 58.9%), Toronto (10747→1162, 53.5%), Norway (1003→1223, 60.4%), New York (1043→1259, 75.0%), France (1109→1322, 54.5%), Japan (1894→1473, 62.6%), Italy (1426→1823, 81.6%), Mexico (new→4249, 99.1%)

Days to eradication for other post-peak regions: NT CAN (0, 0.0%), YT CAN (0, 0.0%), Suriname (2, 0.0%), Papua New Guinea (3, 0.0%), Belize (4, 0.0%), PE CAN (7, 0.0%), Mauritius (10, 0.0%), Western Sahara (10, 0.0%), Saint Lucia (12, 0.0%), Eritrea (14, 3.3%), VI USA (14, 6.8%), Montenegro (14, 2.8%), Monaco (14, 8.1%), Cambodia (14, 1.1%), NS CAN (16, 14.2%), Zimbabwe (18, 30.8%), Holy See (24, 60.0%), Comoros (30, 70.0%), Barbados (30, 25.0%), Iceland (31, 1.1%), Saint Vincent and the Grenadines (35, 45.5%), West Bank and Gaza (35, 28.9%), Tanzania (35, 50.2%), NL CAN (36, 5.2%), Grenada (45, 50.0%), MT USA (47, 5.3%), Taiwan (60, 18.9%), BC CAN (70, 51.9%), AB CAN (71, 36.2%), Rwanda (72, 74.4%), Andorra (72, 27.3%), Luxembourg (75, 6.6%), Ireland (79, 24.5%), Tunisia (87, 40.4%), Cuba (88, 43.3%), AK USA (102, 19.4%), Central African Republic (107, 80.5%), Jamaica (109, 56.5%), Slovakia (116, 40.9%), Botswana (124, 34.3%), Croatia (137, 26.0%), Switzerland (141, 10.4%), SK CAN (146, 85.2%), Panama (146, 67.5%), MS USA (163, 80.3%), Eswatini (165, 84.8%), MB CAN (172, 15.5%), VT USA (190, 15.6%), Liberia (197, 84.9%), NC USA (207, 83.2%), HI USA (230, 18.3%), ND USA (280, 81.2%), Sri Lanka (282, 77.8%), Uruguay (291, 52.1%), Albania (301, 59.9%), Bosnia and Herzegovina (320, 82.1%), OK USA (342, 84.7%), Serbia (353, 68.0%), SC USA (355, 90.7%), Finland (382, 72.4%), WV USA (417, 84.5%), Maldives (428, 84.5%), Mali (434, 83.2%), Cyprus (443, 54.1%), Hungary (473, 82.9%), Uzbekistan (488, 38.1%), North Macedonia (490, 48.7%), Burma (494, 54.7%), MA USA (570, 88.3%), KS USA (577, 92.0%), Latvia (595, 73.7%), Slovenia (602, 55.8%), Ecuador (682, 48.6%), NJ USA (842, 85.3%), Georgia (862, 88.2%), Estonia (887, 55.7%), Guinea (920, 96.6%), Belgium (1311, 73.4%), RI USA (1450, 93.6%), Bulgaria (1522, 96.1%), CO USA (1547, 90.8%), San Marino (1671, 77.9%), Niger (1709, 29.2%), SD USA (1737, 94.2%), MN USA (1749, 96.2%), Netherlands (2322, 71.9%), Kazakhstan (2384, 97.6%), UT USA (2543, 98.6%), Portugal (2564, 77.7%), PA USA (2611, 98.4%), IN USA (2723, 98.0%), MI USA (3061, 91.9%), Benin (6343, 99.6%), Greece (16353, 64.3%)

Sweden's 773 new cases today were within 5 of their record set back on April 29th, and yesterday's 147 deaths was their worst day so far. Their estimated recovered cases are only about 45% of the current total, so there are still a lot of confirmed cases out there causing new infections.

Malta's worsening today is due to 14 new cases being reported, the most since April 11th. Half of the new cases are related to healthcare: healthcare patients, workers, or close contacts of workers.

At time of writing, the worst days-to-one-percent in the Americas are (see chart): 29 Chile, 37 Peru, 49 Brazil, 56 Bolivia, 56 Haiti, 58 U.S.A., 59 El Salvador, 60 Guatemala, 70 Colombia, 71 Nicaragua, 72 Honduras, 73 Dominican Republic, 76 Mexico, 82 Panama, 92 Guyana, 96 Argentina. The other 23 are over 100. Mexico has had a wobbly path up the curve; it's showing a possible peak after relative stability over four days, but it has done this several times before.

Toronto's active case count dropped today to 1,598, its lowest mark since April 10th. These included 153 new cases, or 9.2% of yesterday's active case count. Active cases are very slowly coming down, at about 1% per day averaged over four days, in large part due to how high the new case rate still is. I think we've been fortunate in keeping our hospital transmission down, and that we learned our lessons well enough from SARS that when cases do turn up they are swiftly dealt with. That's not so much the case with community transmission, and of course nursing homes.

Ontario active cases also dropped to a recent new low of 3,492, dropping at about 2% daily but with 7% of active cases causing a new case.

It was announced in Toronto that our street will be one of a network of designated Quiet Streets, where pedestrian use of the roadway will be prioritized, to let people walk around the neighbourhood more easily while maintaining social distancing on narrow sidewalks.

Stay safe.

2020-05-13

Comment on this post on Facebook.

today's daily chart as described below in text

Today's pre-peak days-to-one-percent infection numbers: Singapore (26→26), Alabama (54→47), Brazil (46→50), U.S. (51→54), South Africa (65→63), U.K. (62→65), California (62→67), Sweden (52→70), Mexico (72→76), India (95→96), Canada (90→102), Poland (129→131), Iran (144→144), Indonesia (178→168), Malta (267→252)

Compared to yesterday, here's how these regions are doing.

Much worse: Alabama, Indonesia, Malta

Worse: South Africa

Better: Singapore, U.S., U.K., California, Mexico, India, Poland, Iran

Much better: Brazil, Sweden, Canada

Today's post-peak days-to-eradication numbers: New Zealand (60→43), China (30→48), Thailand (new→68), Austria (291→163), Australia (214→171), Israel (212→205), Denmark (182→221), Germany (264→270), Louisiana (242→274), Malaysia (307→282), Turkey (278→283), Ontario (1007→719), South Korea (3850→767), Norway (1364→1003), New York (631→1043), France (532→1109), Spain (912→1217), Italy (1436→1426), Romania (new→1805), Japan (755→1894), Toronto (147→10747)

Of the regions I'm tracking, the one that is doing the worst by far is Brazil, where the daily new case rate rose for a third day in a row (but whose 4-day average is still being pulled down by an anomalously low number from three days ago) and are back over 6%. Brazil will see 200,000 cumulative cases and 100,000 active cases in the next two days. Next are India and Mexico, both at about 5%.

Today's graphic shows the Johns Hopkins University data for all countries in the Americas. From right to left, the leaders in most per capita infections are the United States, Peru, Panama, Chile, Ecuador, Dominican Republic, and Brazil. Everyone else is doing at least half an order of magnitude better at controlling or underreporting the outbreak.

A 53-year-old Maltese heart surgeon has become the sixth person to succumb to COVID-19 in that country.

All three staff members who contracted COVID-19 at my aunts' nursing home (Castleview Wychwood Towers) have returned to work after successfully testing negative twice.

Toronto saw a significant number of new recoveries (215 for a total of 5,655) and is back down to 1,655 active cases, or about where it was three days ago. Its day number jumped two orders of magnitude because the moving average dropped a much worse day four days ago. Yes, this is one of the reasons why my statistician friends prefer Gaussian filters, which dampen the effect of the oldest and newest data.

Ontario likewise saw a significant number of recoveries (414 for a total of 14,080) and is at a new low of 3,626 active cases.

New Zealand has gone three days in a row now without any new cases, and its active case count has dropped back below 100 for the first time.

Stay safe.

2020-05-12

Comment on this post on Facebook.

today's daily chart as described below in text

Today's pre-peak days-to-one-percent infection numbers: Singapore (27→26), Brazil (46→46), U.S. (51→51), Sweden (52→52), Alabama (51→54), U.K. (57→62), California (57→62), South Africa (62→65), Mexico (70→72), Canada (84→90), India (93→95), Poland (152→129), Iran (144→144), Romania (139→156), Indonesia (191→178), Malta (266→267), Thailand (2838→2838)

Compared to yesterday, here's how these regions are doing.

Much worse: Poland, Indonesia

The same: Singapore

Better: Alabama, South Africa, Mexico, Canada, India, Iran, Malta

Much better: U.K., California, Romania

Not yet reporting: Brazil, Iran, Sweden (not since May 8th, may need to change source), Thailand, U.S.

Today's post-peak days-to-eradication numbers: China (33→30), New Zealand (45→60), Toronto (147→147), Denmark (273→182), Israel (203→212), Australia (257→214), Louisiana (233→242), Germany (805→264), Turkey (283→278), Austria (168→291), Malaysia (618→307), France (811→532), New York (511→631), Japan (218→755), Spain (new→912), Ontario (571→1007), Norway (1653→1364), Italy (1089→1436), South Korea (430→3850)

Today's graphic shows all the countries of the world, plotted horizontally according to the estimated active infected proportion of their population, and vertically according to the rate at which that is changing.

Danish fatalities dropped from a cumulative 533 to 527; this is due to a review that reclassified 12 deaths as not likely due to COVID-19. There were actually only six COVID-19 deaths in Denmark today.

Italy reported its first upturn in new cases in a week, almost doubling from yesterday's 744 to today's 1,402. Over 1,000 of them are reportedly in the hardest-hit region of Lombardy. Italy started reopening from lockdown a week ago, on May 4th and still has 81,266 active cases.

Poland had 595 cases, the first time it has exceeded 500 since April 19th. Poland reopened public gathering places such as shopping centres and museums, subject to social distancing rules, on May 4th.

South Korea is still seeing gradually decreasing double-digit new cases, down to 26 now in the aftermath of a superspreader nightclub event the weekend before last.

Statistics Canada says that they will release a report on recent excess mortality later this month; they usually wait at least a year before publishing mortality figures. As I learned back in December, reporting a death in Canada involves a doctor faxing a form to a provincial office, which can take months to enter it into a database. In the interim, the bereaved make do as best as they can with a document provided by a funeral home confirming that they looked after the deceased. Eventually, an official, numbered death certificate may be ordered from the provincial government; it consists of a digitized version of the original fax, numbered and certified.

Ontario active cases were slightly up for the first time in five days to 3,791, not as high as two days previous. There were 361 new cases (the most in three days), less 56 deaths and 260 recoveries for a net increase of 45 active cases. The active case number is still trending downward over the last four days.

Toronto likewise has seen its active cases jump back up to 1,713, more than where they were three days ago, but still only 57% of peak. There were 218 new cases (the most in four days, and 13% of the previous day's active case total), less 33 deaths and 100 recoveries for a net increase of 85 active cases. It's not a day-of-the-week thing, as Tuesdays are historically our lowest new case days; I have to think it's got something to do with all the people I see going out without their masks.

New York State repoted a fifth consecutive day of decreases in new cases, and today's 1,430 is the lowest since March 18th; I estimate though that at least a third of their 338,485 confirmed cases are still active.

Stay safe.

2020-05-11

Comment on this post on Facebook.

today's daily chart as described below in text

Today's pre-peak days-to-one-percent infection numbers: Singapore (24→27), Brazil (39→46), Alabama (80→51), U.S. (46→51), Spain (53→51), Sweden (52→52), U.K. (51→57), California (57→57), South Africa (76→62), Mexico (61→70), Canada (79→84), India (90→93), Romania (119→139), Iran (136→144), Poland (153→152), Indonesia (175→191)

Compared to yesterday, here's how these regions are doing.

Much worse: Alabama, South Africa

Worse: Spain

The same: Poland

Better: Sweden, California, Canada, India, Iran, Indonesia

Much better: Singapore, Brazil, U.S., U.K., Mexico, Romania

Today's post-peak days-to-eradication numbers: China (36→33), New Zealand (50→45), Toronto (233→147), Austria (241→168), Israel (293→203), Japan (249→218), Louisiana (new→233), Australia (255→257), Denmark (228→273), Turkey (268→283), Malta (88→320), South Korea (288→430), New York (529→511), Ontario (987→571), Malaysia (525→618), Germany (804→805), France (1446→811), Italy (1091→1089), Norway (2571→1653), Thailand (273→1898)

Today's graphic: regions of interest that have made it past their peak, and how they got there on the traditional daily new case percentage vs. proportion of population infected state diagram.

South Africa's situation is worsening rapidly, increasing over five days from 236 new cases to a record 637 new cases today; its four-day average new case rate has increased for four days straight, to 6.7%. Alabama is also worsening day by day, with a new case rate above 4% for the first time since April 21st.

Many parts of European countries are lifting some restrictions today as well. In France, the country is split into red zones (such as Paris) and green zones (anywhere with low population density), with different rules for each: travel certificates are now only needed for rush hour in Paris and trips of more than 100 km everywhere; primary schools and daycares are reopening; shops outside of malls may reopen; gatherings of fewer than 10 people are permitted. In Spain (but not in Madrid or Barcelona), people may gather in groups of up to 10, and restaurants may reopen subject to social distancing. In Germany, households may pair up to meet with each other (as is the case in some Atlantic Canadian provinces), all remaining shops that had been closed may now reopen subject to social distancing, and schools have reopened for primary students and those needing to take exams.

India saw a record 4,213 new cases today, breaking a three-day streak of decreasing new case numbers, but is still trending downward with its most recent four-day average rate at 6.1%.

There was only one new symptomatic case in China today, after two days in the teens due to what is being reported as hundreds of asymptomatic (and therefore not officially counted) cases in one residential complex in Wuhan.

South Korea is anticipating difficulty in tracing contacts of the current outbreak, due to stigmatization of the LGBTQ community that are thought to be the most involved. It would be easy for some people in more socially liberal countries to point fingers, except that we are all finding that the pandemic is taking a disproportionately heavy toll on whichever of our fellow human beings we have chosen to devalue: the criminals, the elderly, the female, the homeless, the immigrants, the LGBTQ, the poor, the racialized, the stateless, the unbelievers, the undocumented, etc. And that burden is carried by society as a whole, because of how difficult the disease is to contain and treat.

Quebec reopens public schools today. Their active case numbers reached a new peak of about 19,700 today, up 100 in each of the last two days.

Manitoba has two new cases, close contacts of a recent outbreak at a trucking company; they are now back up to 35 active cases.

Saskatchewan dropped ten active cases to 199 today, its first significant decrease; it may have reached its peak.

Toronto has tied its previous record with a fourth consecutive drop in active cases, and at 1,628 is at 54.5% of peak - almost halfway there!

Ontario has unsurprisingly also had four straight decreases in active cases, and is now at 3,746 remaining confirmed active cases, or 66% of peak.

Stay safe.

2020-05-10

Comment on this post on Facebook.

today's daily chart as described below in text

Today's pre-peak days-to-one-percent infection numbers: Singapore (25→24), Louisiana (38→34), Brazil (34→39), U.S. (45→46), U.K. (45→51), Sweden (52→52), California (53→57), Mexico (58→61), South Africa (76→76), Canada (79→79), Alabama (80→80), India (89→90), Romania (114→119), Iran (140→136), Poland (155→153), Indonesia (173→175)

Compared to yesterday, here's how these regions are doing.

Much worse: Louisiana, Iran

Worse: Poland

The same: Singapore

Better: U.S., Sweden, California, Mexico, South Africa, Canada, Alabama, India, Romania, Indonesia

Much better: Brazil, U.K.

Today's post-peak days-to-eradication numbers: China (34→36), New Zealand (54→50), Malta (78→88), Denmark (186→228), Toronto (312→233), Austria (154→241), Japan (203→249), Australia (213→255), Turkey (249→268), Thailand (154→273), South Korea (161→288), Israel (253→293), Malaysia (332→525), New York (577→529), Germany (216→804), Ontario (1304→987), Italy (1059→1091), Spain (997→1192), France (new→1446), Norway (985→2571)

Today's graph shows all the regions of interest that haven't yet clearly reached their active case peak, including their past history.

I took a broader look at the New York numbers, and accounted for their weekly cyclical reporting pattern, where Sunday, Monday and Tuesday have fewer than average new cases (Monday has 11% of the week), and the rest of the days have more (Friday has 19%). The state claims that active case numbers are still on the rise, but this can only be true if they're the only place in the world where at least 75% of confirmed cases don't end up dead or recovered within two weeks. Or maybe everyone else is wrong in guessing that if someone doesn't come back for a checkup within two weeks, they can be removed from the active case list, and New York is checking more carefully. It seems more likely that they are just using a very conservative criterion for recoveries, and are therefore showing a lot more active cases than they actually have. My best guess now is that their active case totals have been drifting slowly downward for two weeks. This must be helping the outlook for the nation as a whole, with the state accounting for about a quarter of the total cases.

South Korea and China are experiencing small outbreaks, and are worth following to give an idea of what any densely populated country has in store after they mostly eradicate SARS-CoV-2. China has gone from 1 or 2 cases a day back up into the teens; South Korea from 2 or 3 up into the 30s.

I missed two interesting stories in France last week. On a technical note, they had a huge spike on May 6th due reportedly to a new testing lab coming online with about 5,000 backdated positive tests. At about the same time, and news reports were not clear whether it was the same lab, it was announced that testing of frozen samples taken on December 27th identified a man who is now the earliest known case COVID-19 in France, and may help to explain the spread of the disease in that country. Neither the man nor his wife travelled, but his wife worked at a supermarket near Paris-CDG airport that is popular with travellers, so the supposition is that someone flying from China to France stopped to get their groceries from her and inadvertently started the epidemic in France. I expect there will be more to come in this story, and look forward to seeing if more detailed genetic analysis of the sample can tell more about its origins.

Ontario has had a third straight day of decreases in its active case count, now 3,832 or 67% of its peak. Toronto has also had three consecutive decreases, and is now at 1,643 or 55% of peak. Right now, depending on the day there is for every 10-14 active cases one new case. Hospital occupancy is still slowly increasing, but ICU numbers are steady.

Stay safe.

2020-05-09

Comment on this post on Facebook.

today's daily chart as described below in text

Today's pre-peak days-to-one-percent infection numbers: New York (-60→-59), Singapore (26→25), Brazil (36→34), Louisiana (38→38), U.S. (45→45), U.K. (45→45), Sweden (52→52), California (61→53), Mexico (63→58), South Africa (88→76), Canada (78→79), Alabama (64→80), India (80→89), Romania (112→114), Iran (140→140), France (130→145), Poland (138→155), Indonesia (173→173)

Much worse: California, Mexico, South Africa

Worse: New York, Brazil

The same: Singapore

Better: Louisiana, U.S., U.K., Sweden, Canada, Romania, Iran, Indonesia

Much better: Alabama, India, France, Poland

Today's post-peak days-to-eradication numbers: China (60→34), New Zealand (46→54), Malta (115→78), Austria (125→154), Thailand (127→154), South Korea (138→161), Denmark (195→186), Japan (236→203), Australia (137→213), Germany (200→216), Turkey (219→249), Israel (203→253), Toronto (new→312), Malaysia (268→332), Norway (1326→985), Spain (1285→997), Italy (986→1059), Ontario (new→1304)

California takes a long time to publish its results, so today's changes reflect an increase from 1,799 to 2,049 in daily case numbers in the two-day period ending on May 8. Their numbers are going up and down, with no clear sign of a trend, even when smoothed by four or eight days.

Mexico likewise has shown active case counts ranging from 5,000 to 10,000 just in this week, but again with no clear trend.

Canada is doing slightly better overall. Quebec and Saskatchewan are the only provinces with significant, increasing active case counts, increasing at about 2% and 5% daily; New Brunswick was COVID-19-free for three days, but now is back up to two active cases. The rest of the provinces have seen their active case counts coming down from their peaks. Today's graphic shows each province and territory, its history, and where they are now on cumulative confirmed case increase rate vs. proportion infected.

After two consecutive days of rapid decline, Toronto's active case count has dropped to a new post-peak low of 1,660 or 56% of its peak of an estimated 2,986 from April 23rd, and takes us back to a level last seen on April 11th. That's a huge drop since 2,028 two days ago, and consists of less 45 deaths and 660 recoveries less 337 new cases. [Note that I have re-estimated that peak of 2,986 upward from 2,700 today, by revising the assumption that everyone recovers after 14 days to say that just 75% of patients do. This change didn't affect the date of the peak, just its magnitude.] That gives us a days-to-eradication number for the city at last, of 312, which is better than none at all.

Ontario has had two days of less rapid decline in its active case count, from 4,075 to 3,962, but enough to get a days-to-eradication of 1,304.

I am having continuing issues with the Alabama data. The state publishes cumulative case counts that you can fetch by hovering over a graph, and probably by writing a script to scrape the data that is feeding the graph. They don't do the same for deaths, but just show the most recent total, which can change over the course of a day at irregular intervals; there is no obvious way to recover older data. They do not track recoveries at all. I had switched to using the state data directly, after volunteers missed too many days in copying it to Wikipedia (the deaths information was therefore irretrievable), but noticed today that the state has gone back as far as April 2nd and retroactively altered (ret-conned) cumulative case counts, requiring that I rekey and check all the data since then if I want to keep tracking them. I am mildly annoyed, and will likely make Alabama be the first region I've had to remove from my list of regions of interest; if I do so, people who want to follow Alabama in my charts can still do so by selecting the COVIDTracking.com dataset. I'll wait a day or two to do so though, as I've found that sometimes, large-scale changes of this sort are mistaken, and are reversed as soon as they are detected. I have even deeper respect for the volunteers at COVIDTracking.com now, who spend their days carefully searching for data for all American states and territories, cross-checking official and unofficial reports, and then making it available for everyone for free.

U.S. case numbers published online have also been revised generally slightly upward dating backward a few weeks, presumably as additional reports are located and verified. There is a possibility that new cases were down every so slightly yesterday, from 27,535 to 27,518, but this is well within the range of errors that appear to be being corrected going forward. As I wrote yesterday, New York is looking like it is on the steadiest path to recovery, but its data is still very noisy and their active case numbers are still slowly increasing. Its days-past-1% number worsened today for the first time in about a week. There are about 25,500 cases accounted for in all the state data so far today, but the total hasn't been confirmed or published at my source.

Japan's daily new cases, which had surprisingly dropped over Golden Week, were back up a little bit from 84 to around 100 (plus or minus two) in the last two days, which is a level that they hadn't seen consistently since the last week of March. Their recovered case numbers are not plausible (half of their confirmed cases from 14 days ago), but if we estimate them at 75% with 14-day lag, they are down to two-thirds of their active case load, with a little less than 6,000 to go.

Stay safe.

2020-05-08

Comment on this post on Facebook.

today's daily chart as described below in text

Today's pre-peak days-to-one-percent infection numbers: New York (-59→-60), Brazil (38→36), Louisiana (38→38), U.K. (45→45), U.S. (50→45), Toronto (48→46), Sweden (85→52), California (64→61), Alabama (71→64), Canada (81→78), India (80→80), South Africa (88→88), Ontario (96→91), Romania (107→112), France (131→130), Poland (137→138), Iran (147→140), Indonesia (162→173)

Compared to yesterday, here's how these regions are doing.

Much worse: U.S., Sweden, Alabama, Ontario, Iran

Worse: Brazil, Toronto, California, Canada

The same: France

Better: New York, Louisiana, U.K., India, South Africa, Romania, Poland, Indonesia

Today's post-peak days-to-eradication numbers: New Zealand (49→46), China (46→60), Malta (112→115), Austria (157→125), Thailand (359→127), Australia (168→137), South Korea (148→138), Denmark (331→195), Germany (130→200), Israel (57→203), Turkey (196→219), Japan (236), Malaysia (276→268), Mexico (293), Italy (986), Spain (1285), Norway (1326), Singapore (4415)

Some countries (Italy, Japan, Mexico, Spain, Norway, Singapore) were incorrectly reported as not being post-peak yesterday, because of an error in the estimation of recovered patient numbers in the situation where the reported numbers are missing or impossible. They are listed today without prior numbers.

Japan is down to 69% of its peak after two weeks, which suggests encouragingly that the Japanese generally heeded social distancing measures during their recent Golden Week holidays.

Mexico is at 87% of peak after a few days, but oscillating wildly and it's not clear that this is its final peak.

Italy is down to about 86% of its peak, but it's taken five weeks to get there. Spain is at 65% after three weeks, Norway is at 62% after five weeks, Singapore is at 95% after one week. All four are of course dealing with very large active case loads even if they are on their way down.

Today's graphic shows a selection of American states, plotting their daily average changes in active cases vs. proportion of population infected.

At the right side of the chart, a curious artifact in New York's data shows that they have some sort of weekly glitch in reporting results, I presume some Sunday results getting delayed to Monday. They have about 1% of their population actively infected, increasing at 2% daily from over 200,000 now.

Alabama, California, and Texas had all been doing well, but have started sliding back up over the past several days. California is now increasing the most quickly of them at 6% daily, even though Alabama (4%) an Texas (2%) have been publicly criticized for reopening too rapidly.

Toronto's active case count dropped for the first time in six days, from 2,028 to 1,865, less than 70% of our peak and a little less than where we were the day before yesterday. (It bugs me sometimes that we stopped saying "ereyesterday" more than a century ago.)

Ontario's active case count dropped insignificantly from 4,075 to 4,068, drawing perplexity from chief medical officer Dr. David Williams, who wondered "if people are being less than consistent in their physical distancing, and if they were in close contact, if they were not wearing a facial covering in those few moments" Based on my anecdotal observations on brief, necessary shopping expeditions this week, I would say yes. People aren't being consistent in their physical distancing, especially in shops, and many people are still not covering their faces. Williams adds "You still have to do that or we're going to be in this plateau for quite a while." We'll remain on a plateau as long as behaviours and restrictions remain the same. If restrictions are eased, we'll start climbing again.

Stay safe.

2020-05-07

Comment on this post on Facebook.

today's daily chart as described below in text

Today's pre-peak days-to-one-percent infection numbers: New York (-59→-59), Singapore (29→28), Louisiana (40→38), Brazil (43→38), U.K. (48→45), Toronto (54→48), U.S. (46→50), Spain (75→62), California (64→64), Mexico (69→65), Alabama (51→71), India (84→80), Canada (64→81), Sweden (233→85), South Africa (83→88), Ontario (93→96), Romania (102→107), France (138→131), Poland (134→137), Iran (192→147), Indonesia (157→162), Italy (171→171), Norway (344→331), Japan (428→428)

Compared to yesterday, here's how these regions are doing.

Much worse: Brazil, Toronto, Spain, Mexico, India, Sweden, France, Iran, Norway

Worse: Louisiana, U.K.

The same: Singapore

Better: New York, California, South Africa, Ontario, Romania, Poland, Indonesia, Italy, Japan

Much better: U.S., Alabama, Canada

Today's post-peak days-to-eradication numbers: China (45→46), New Zealand (57→49), Israel (61→57), Malta (120→112), Germany (130→130), South Korea (171→148), Austria (148→157), Australia (194→168), Turkey (217→196), Malaysia (425→276), Denmark (627→331), Thailand (590→359)

We return today to the active vs. active graph, showing best guesses as to daily rate of change in active cases, plotted against proportion of population actively infected. The guessing part involves things like smoothing rates over four days, and assuming that in the worst case unreported scenario, 75% of patients recover two weeks after diagnosis. If you're above the midline, you're worsening; below, and you're improving. If you're to the right, you have a lot of active cases per capita; and if you're to the left, you have relatively few.

Outliers in the area of doing poorly and worsening: San Marino (1 in 71 people are actively infected), Holy See (1 in 80), Qatar (1 in 174), Singapore (1 in 305), United States (1 in 336), United Kingdom (1 in 385), Belgium (1 in 386).

Outliers in the area of having gotten themselves into a terrible situation but now rapidly clearing their backlog of active cases: Spain (1 in 699 and dropping 2% daily), Ireland (1 in 1,275 and 13%), Israel (1 in 1,643 and 4%), Turkey (1 in 1,791 and 6%), Germany (1 in 4,120 and 8%).

The CMAJ dataset has worse daily active numbers for the province of Ontario than I am seeing on the official Ministry of Health and Long-Term Care website, but the latter is a slightly confusing mix of Ministry numbers and iPHIS database numbers. To be fair to their administrators, the problem isn't that they are presenting two conflicting sets of data for the same population, but that the two sets exist and haven't been politically reconciled. That probably speaks to long-term issues in public health administration in the province. The MHLTC numbers suggest that Ontario is still barely seeing an average decrease in active cases, of about 1.5% daily, for a days-to-eradication of 589 days; CMAJ says the numbers have been increasing for too many days. In either case, I think we're a little past the peak, but not coming down quickly yet.

In Toronto, a day after Mayor John Tory shifted the city's message from the chief medical officer's "stay home" to the mayor's "stay apart", we have had a fifth consecutive day of increases in active cases, rising from 1,767 (65% of our peak) to now 2,208 (75% of peak) at an average daily rate of 3%, giving us about a week before we get back up to our earlier peak levels. Hospital beds allocated for COVID-19 are at their maximum to date at 397. Our new case number today (249) was 12% of the number of active cases, indicating that spread is not yet being contained well. It's worth reiterating that according to city statistics, half the people who are testing positive are presumably low-risk individuals under the age of 50; they're certainly the ones that I see in abundance now on the streets not being particularly careful about social distancing.

At my aunts' nursing home, they are now reporting three confirmed cases among staff, and none among residents. They're also making arrangements for patients to receive deliveries for the first time in weeks, to let people send their mothers flowers for Mother's Day. It's a nice gesture, but it makes me wish there were an Aunts' Day on the calendar.

In the province of Quebec, officials are starting to push back previously announced aggressive deadlines for reopening the province for business, leading to relief mixed with a good deal of confusion. Given that they have over 16,000 active cases, are gaining 4% more daily, and that their hospitals are at or over capacity because their long-term care network has failed, I'm glad that they're doing the right thing now.

Stay safe.

2020-05-06

Comment on this post on Facebook.

today's daily chart as described below in text

Today's pre-peak days-to-one-percent infection numbers: New York (-49→-59), Singapore (33→29), Louisiana (40→40), Brazil (52→43), U.S. (46→46), U.K. (52→48), Alabama (53→51), Toronto (58→54), California (68→64), Canada (58→64), Spain (81→75), South Africa (72→83), India (84→84), Romania (109→102), Poland (136→134), France (306→138), Indonesia (161→157), Italy (157→171), Iran (192→192), Sweden (109→233), Norway (380→344), Japan (326→428)

Compared to yesterday, here's how these regions are doing.

Much worse: Singapore, Brazil, U.K., Toronto, California, Spain, Romania, France, Indonesia, Norway

Worse: Alabama, Poland

Better: Louisiana, U.S., India, Italy, Iran

Much better: New York, Canada, South Africa, Sweden, Japan

Today's post-peak days-to-eradication numbers: China (56→45), New Zealand (78→57), Israel (67→61), Malta (256→120), Germany (216→130), Austria (220→148), South Korea (216→171), Australia (236→194), Turkey (414→217), Ontario (396→396), Malaysia (1176→425), Mexico (new→565), Thailand (new→590), Denmark (682→627)

In Ontario, the Ministry of Health and Long-Term Care reported 71 new deaths in long-term care for a new total of 1,074, while the provincial iPHIS database reported a drop due to recategorization by 45 cases to 653 deaths.

Toronto has seen its third straight day of increases in active cases, rising over that time from 1,767 to 1,888 cases, still down from a peak of 2,700 on April 23rd, but trending badly. ICU and intubation cases are still trending down, because they lag active cases. Toronto doesn't make it onto the new second list today because its numbers are back on the rise, and what with the mayor's announcement today that people should stop thinking about staying home and just make sure to keep their distance when they're out, I'm not expecting our active cases to decrease in the short-term future.

I spent more time thinking about coding than commentary today, so that's it for now.

Today's graphic: new cases per current active case count, vs. percentage of population confirmed with active infections, examining cases causing concern (Singapore, Brazil, Spain, United States, United Kingdom) interest (Canada, Japan), and admiration (New Zealand).

Stay safe.

2020-05-05

Comment on this post on Facebook.

today's daily chart as described below in text

Today's numbers: New York (-43→-49), Singapore (29→33), Louisiana (31→40), U.S. (39→46), U.K. (46→52), Brazil (51→52), Alabama (56→53), Toronto (51→58), Canada (55→58), California (56→68), Mexico (61→69), South Africa (72→72), Spain (69→81), India (95→84), Ontario (83→86), Sweden (109→109), Romania (107→109), Denmark (130→132), Turkey (125→133), Poland (157→136), Italy (135→157), Indonesia (161→161), Iran (206→192), Malta (294→294), France (372→306), Malaysia (276→312), Japan (267→326), Norway (410→380), Germany (314→385), Israel (225→577), Austria (1346→1000), Australia (1352→1189), Thailand (1960→2312), South Korea (5553→6410), New Zealand (1874→6893), China (244157→213640)

Compared to yesterday, here's how these regions are doing.

Much worse: India, Poland, France, Norway

Worse: Alabama

Better: Brazil, Canada, South Africa, Ontario, Sweden, Romania, Indonesia

Much better: New York, Singapore, Louisiana, U.S., U.K., Toronto, California, Mexico, Spain, Italy, Japan, Thailand

On the recovery side of the active case curve so don't worry about the above: Iran, Austria, Australia, China, Denmark, Turkey, Malta, Malaysia, Germany, Israel, South Korea, New Zealand

I've recalculated days-to-eradication figures for each country, but the days look a little noisy for the sake of clarity, and I'll likely have to smooth the recent trend data a little more. I'll keep an eye on it for a few more days.

In the meantime, Kristen reminded me about the CMAJ dataset that gives recovery data (and a lot more) for all Canadian provinces and territories. I'm still interested in finding analogous data for the United States, and am surprised that it's not available. Late at night, it's easy to weave conspiracy theories about why this might be, but it's more likely that not enough American states have gotten close enough to peak COVID-19 activity to have the interest or energy to track what happens next.

Today's chart shows where each Canadian province and territory is doing in terms of new daily cases as a percentage of currently active cases, and active cases as a proportion of population.

British Columbia is dropping its active case count at 1.3% daily, and is at this rate 508 days to eradication.

Alberta 4.7% daily and 175 days;

Manitoba 0.6% and 646 days;

Ontario 2.6% and 338 days;

Nova Scotia 3.0% and 203 days;

Prince Edward Island 9.6% and 14 days;

Newfoundland and Labrador 12.8% and 25 days.

New Brunswick 15.9% and 4 days: this requires a little more explanation. New Brunswick had a new case pop up today in Fredericton after sixteen days without a new case; the cause is still being investigated, but appears to be unrelated to that of a COVID-19-positive man who flew from Toronto to Moncton last week and then left the province. So New Brunswick will have at least one case on their books for two weeks or so. I should also be putting a floor on the days to eradication at 14 days since the last new case.

Quebec is problematic for two reasons. Their reported recoveries do not make sense, in that they are only half of the confirmed cases from two weeks ago. Even if I rounded them up to 75%, they are only one of two parts of Canada where active cases are increasing daily (4.6%) instead of decreasing. It's hard to understand why they are reopening the province for business, and it makes me nervous, especially for my friends and family who live in or near that province.

The other part of Canada where active cases are on the rise is Saskatchewan, which is seeing an 11% daily increase in new cases. They too are experimenting with reopening businesses. I regret my earlier optimism about this.

The case that was reported earlier in Nunavut turned out to be a false positive; they remain the only province or territory in Canada with no history of COVID-19.

Northwest Territories have been COVID-19-free for more than a week.

It's been three weeks since the last confirmed case in Yukon, and their Chief Medical Officer said that it's still premature to ease restrictions.

Lastly, I am saddened by a large outbreak reported at the Toronto Western Hospital today, involving four units, 79 staff and 19 patients. I have the deepest respect for the work done at TWH; I wouldn't be able to see to type this but for emergency retinal surgery there a couple of years ago. And because they are known for their neurology department, my mother spent a lot of time in their care during her last decade with dementia, including being treated for pneumonia during the last month of her life. Now, the hospital operates a large unit devoted to treating COVID-19, which must be terrifying for those involved, and an inspiration to the rest of us.

Stay safe.

2020-05-04

Comment on this post on Facebook.

today's daily chart as described below in text

Today's numbers: New York (-36→-43), Singapore (30→29), Louisiana (28→31), U.S. (39→39), U.K. (42→46), Brazil (47→51), Toronto (43→51), Canada (61→55), Alabama (64→56), California (56→56), Mexico (58→61), Spain (60→69), South Africa (75→72), Ontario (78→83), India (99→95), Romania (114→107), Sweden (59→109), Turkey (109→125), Denmark (129→130), Italy (123→135), Poland (166→157), Indonesia (162→161), Iran (218→206), Israel (288→225), Japan (62→267), Malaysia (272→276), Malta (314→294), Germany (241→314), France (298→372), Norway (390→410), Austria (624→1346), Australia (1768→1352), New Zealand (1874→1874), Thailand (2930→1960), South Korea (4625→5553), China (94943→244157)

Compared to yesterday, here's how these regions are doing.

Much worse: Canada, Alabama, India, Romania, Poland, Iran, Israel, Malta, Australia, Thailand

Worse: South Africa

The same: Singapore, Indonesia

Better: U.S., California, Mexico, Ontario, Denmark, Malaysia, Norway, New Zealand

Much better: New York, Louisiana, U.K., Brazil, Toronto, Spain, Sweden, Turkey, Italy, Japan, Germany, France, Austria, South Korea, China

Today's graphic shows the countries who have the highest per capita rates of active COVID-19 infections, plotted against the rate at which this is increasing.

Ontario has 370 new cases today, 500 recoveries, and 84 deaths (36 of which were in long-term care). We are at our ninth day since peak active cases, and down to 72.6% of peak with 4,118 currently active cases (these numbers are better than what I was reporting yesterday, because I had not correctly accounted for deaths). That number has come down by 500 + 84 - 370 = 214 since yesterday: 40% from deaths, and 60% from the fact that fewer people are contracting COVID-19 than two weeks ago.

Toronto skipped a day's reporting yesterday to synchronize something with Ontario. We now have 1,807 currently active cases, which is 67% of our peak from 11 days ago.

Quebec has decided to delay reopening stores in Montreal by a week, to May 18th. Given that the number of active cases in the provinces is still rising rapidly daily, this sounds a little better than a terrible idea.

The American National Institutes of Health reportedly cancelled funding on Friday for the New York research nonprofit EcoHealth Alliance, that has funded $3.7 million in research into the risks of zoonotic transmission of coronaviruses from bats to humans, and discovered inter alia that 3% of residents in rural Yunnan province in China have antibodies to coronaviruses similar to SARS. I hope that other countries are willing to step up to the challenge of saving our world from coronaviruses.

I was thinking a little more about what happens on the good side of the active case curve tonight. Active case numbers gradually decrease; sometimes faster and sometimes slower as civil restrictions are relaxed. Sometimes they rise back up a bit, before people come to their senses. I'm guessing in some parts of the world, the curve will drop back down again because everyone who can get sick has gotten sick, although in this case there probably won't be enough healthy people left to count who's sick and who's not.

Thoughts like this, along with my documented personal quirk of viewing too many things in the world as countdowns measured in days, leads me to propose as a second-half statistic to watch the number of days left until an epidemic is locally eradicated. To be precise, let's take the current average rate of decline in the active case count, and project that straight forward until we reach less than half a person infected. Yes, I know that it would make some people more comfortable if I extrapolated using Bayesian methods to measure error, and Gaussian filters rather than just a geometric mean, and a compartmental model instead of the pretense of exponential decay. I'm still toying with these in my spare time, but they're not ready for production yet, and from what I've seen any reduction in error that they might provide would be vastly overshadowed by unforeseeable changes to local social distancing norms.

Using the current JHU data then, I calculate the active case count as the confirmed cases less the sum of the deaths and recoveries. If the deaths and recoveries add up to less than 75% of the confirmed cases from 14 days previous, I round them up to that amount to more closely conform with definitions of recovery in other countries (80% of cases are serious enough to be diagnosed but mild enough to be cured within 14 days; reduce 80% to 75% just in case something is locally different). Calculate the four-day moving average rate of change in active case counts, and exclude the 100 or so countries whose active case counts are still increasing. Then use the same log calculation as for the days to 1% to figure out who's doing better than whom, by calculating how many days that average rate of change would take (if it remained constant) to bring them down to less than half an infected person.

Once again, this is not a pointless attempt to forecast what is happening in these countries. It's an attempt to try to explain what the current rate of decline in active cases, and the current number of active cases, mean together in any given country. Clear? Here are the figures.

Within a month (makes me think case counts are being underreported): Belize 3, Trinidad and Tobago 5, Fiji 7, Monaco 8, Eswatini 9, Antigua and Barbuda 12, Timor-Leste 12, Cambodia 14, Mauritius 14, Montenegro 15, Brunei 17, Syria 23, Nicaragua 25, Nepal 27, Guinea 30.

Within three months (most of these look plausible to me): Iceland 34, Tanzania 54, Malta 56, Luxembourg 67, North Macedonia 70, China 71, Israel 79, Switzerland 88. Saint Vincent and the Grenadines 94.

Within a year: New Zealand 97, Niger 97, Croatia 99, Burkina Faso 120, Albania 142, Andorra 183, Uzbekistan 199, Germany 200, Kyrgyzstan 208, Thailand 212, South Korea 228, Austria 233, Djibouti 257, Finland 276, Uruguay 299.

Longer than a year (these look like countries that have just had a rough four days of it, and will get better soon): Australia 399, Rwanda 506, Cuba 599, Taiwan 700, Denmark 985, Iran 1040, Mali 8523.

Looking for your country and haven't found it yet? It's still on the upward side of the curve. Canada's active cases are increasing daily at 1.8%; the U.S. at 4.1%; the U.K. at 4.8%.; etc. Ontario would be about 50 days. Toronto had a bad day and bounced back up, but would have been at 119 days two days ago.

Stay safe.

2020-05-03

Comment on this post on Facebook.

today's daily chart as described below in text

Today's numbers: New York (-33→-36), Louisiana (28→28), Singapore (29→30), U.S. (40→39), U.K. (42→42), Toronto (43→43), Brazil (42→47), California (55→56), Mexico (59→58), Sweden (59→59), Spain (46→60), Canada (61→61), Japan (321→62), Alabama (94→64), South Africa (76→75), Ontario (81→78), India (102→99), Turkey (94→109), Romania (118→114), Italy (111→123), Denmark (118→129), Indonesia (169→162), Poland (148→166), Iran (211→218), Germany (202→241), Malaysia (292→272), Israel (235→288), France (203→298), Malta (436→314), Norway (294→390), Austria (488→624), Australia (1866→1768), New Zealand (1585→1874), Thailand (2298→2930), South Korea (5941→4625), China (89942→94943)

Compared to yesterday, here's how these regions are doing.

Much worse: Japan, Alabama, Romania, Indonesia, Malaysia, Malta, Australia, South Korea

Worse: Ontario, India

The same: U.S., Mexico, South Africa

Better: Louisiana, Singapore, U.K., California, Sweden, Canada, Iran, China

Much better: New York, Brazil, Spain, Turkey, Italy, Denmark, Poland, Germany, Israel, France, Norway, Austria, New Zealand, Thailand

N/A: Toronto

Toronto is not reporting data today because they "are reconciling [their] reporting cycle to align with the provincial approach"; their data reporting will resume tomorrow. As of yesterday's data, we were 9 days past our peak active case count, and down to 67% of that peak.

Ontario had its fifth consecutive day of decreased active cases, down now to 5,548 or 86% of its peak of 6,486. Daily new cases are at 2.5% of the total cases to date, or about 7.8% of the remaining active cases.

Canada had daily new cases amounting to 3.5% of the cumulative total cases to date, but 6.1% of the remaining active cases.

The United States had daily new cases equal to 2.8% of the cumulative total cases to date, but 3.5% of the remaining confirmed active cases.

As things get better, the former statistics become less meaningful, as recovered cases do not contribute causally to new infections. From a data visualization point-of-view, the problem was that early on, no one was reporting recovered case totals because (inter alia) no one was sure who would recover and when. As it became clear that two weeks was a reasonable threshold for declaring someone recovered, countries started to publish statistics based on this threshold and the Johns Hopkins University data now includes this.

So I've added a new series for visualizing daily new cases as a percentage of remaining confirmed active cases. You can select it on the website as "JHU Global New vs Active", and today's graphic shows an example. I can't show some countries yet (France, Netherlands, Norway, Poland, Singapore, Sweden, United Kingdom), because JHU is not showing significant numbers of recovered patients for them, making it impossible to estimate the active infected population. In theory, the number of recovered patients and deaths should add up to at least as much as the total number of cases two weeks previously. Not surprisingly, the United States leads in per capita active cases. Mexico and Brazil have the highest adjusted new case rates. Iceland is doing as well as South Korea or China in limiting new cases.

Stay safe.

2020-05-02

Comment on this post on Facebook.

Today's numbers: New York (-34→-33), Louisiana (33→28), Singapore (28→29), U.S. (40→40), Brazil (39→42), U.K. (43→42), Toronto (43→43), Spain (45→46), California (56→55), Sweden (59→59), Mexico (57→59), Canada (64→61), South Africa (85→76), Ontario (79→81), Turkey (89→94), Alabama (102→94), India (108→102), Italy (109→111), Denmark (106→118), Romania (106→118), Poland (140→148), Indonesia (150→169), Germany (189→202), France (167→203), Iran (194→211), Israel (196→235), Malaysia (376→292), Norway (280→294), Japan (277→321), Malta (254→436), Austria (468→488), New Zealand (1583→1585), Australia (2062→1866), Thailand (2215→2298), South Korea (5941→5941), China (43812→89942)

Compared to yesterday, here's how these regions are doing.

Much worse: Louisiana, South Africa, Alabama, India, Malaysia, Australia

Worse: New York, Canada

The same: U.K., California

Better: Singapore, U.S., Toronto, Spain, Sweden, Mexico, Ontario, Turkey, Italy, Poland, Germany, Iran, Norway, Austria, New Zealand, Thailand, South Korea

Much better: Brazil, Denmark, Romania, Indonesia, France, Israel, Japan, Malta, China

Toronto hospitalised cases are up for a second day in a row, 345 to 350 to 375, but our total active cases have dropped back below where they were two days ago, 1,793 to 1,844 to 1,767. 44% of new cases in the last two days have been in institutions (hospitals and long-term care facilities), including ten new institutions, as recorded in the iPHIS database.

Ontario active cases are down for a fourth day in a row, 6,447 to 5,729. Ontario reports that 44% of COVID-19 deaths province-wide in the iPHIS database, but that in the more inclusive numbers as reported to the Ministry of Health and Long-Term Care, before being verified for classification into iPHIS, 77% of deaths were actually institutional patients.

If we assume reasonably that Toronto is representative of Ontario, then about three hundred of our 398 COVID-19 deaths to date have been vulnerable elderly and disabled residents of institutions where they were supposed to be receiving appropriate care and support.

With not much change to report among the countries the first few regions on the list above, that are either in a state of disaster or hurtling toward it, let's take a look at a different visualisation for a change. Today's graphic shows only those countries which (in the Johns Hopkins University data) have showed more than a day's decrease in active (not cumulative total) cases, and illustrates five facts about each: the name of the country, the continental region (keyed by colour), how many active cases remain (bubble size), how many days it's been since the country had its peak active cases (logarithmic vertical axis), and what proportion of that peak number are still active. So big bubbles are worse than small ones, because they represent active cases; and the top right corner is where you want your country to be, beside China. If you're in the top left like Iraq, it's been a while since you hit your peak, but you haven't made much progress since. If you're in the bottom left like Moldova, you're just barely past what you hope is the top of your active case curve. If you're in the bottom right like Timor-Leste, then you probably never had a large number of simultaneous active cases to begin with, but you're having trouble getting rid of the last ones.

And of course, if you're not on the chart at all, then you're not over the top of the curve yet, so stay at home, wash your hands, and be patient. Toronto would be at (1,96%); and Ontario would be at (4,89%), about halfway between France and Albania.

Stay safe.

2020-05-01

Comment on this post on Facebook.

today's daily chart as described below in text

Today's numbers: New York (-33→-34), Singapore (29→28), Louisiana (45→33), Brazil (40→39), U.S. (45→40), U.K. (47→43), Toronto (46→43), Spain (49→45), California (57→56), Mexico (62→57), Sweden (56→59), Canada (62→64), Ontario (78→79), South Africa (92→85), Turkey (89→89), Alabama (71→102), Romania (106→106), Denmark (111→106), India (108→108), Italy (112→109), Poland (131→140), Indonesia (172→150), France (167→167), Germany (221→189), Iran (194→194), Israel (208→196), Malta (254→254), Japan (271→277), Norway (280→280), Malaysia (422→376), Austria (434→468), New Zealand (2054→1583), Australia (2305→2062), Thailand (2003→2215), South Korea (4617→5941), China (38832→43812)

Compared to yesterday, here's how these regions are doing.

Much worse: Louisiana, U.S., U.K., Spain, Mexico, South Africa, Denmark, Indonesia, Germany, Israel, Malaysia, New Zealand, Australia

Worse: Toronto, Italy

The same: Singapore, Brazil, California

Better: New York, Sweden, Canada, Ontario, Turkey, Romania, India, Poland, France, Iran, Malta, Japan, Norway, Austria

Much better: Alabama, Thailand, South Korea, China

I had a fun but full evening tonight, eating takeout ramen and playing Mario Kart with family, then playing board games online (Stone Age and Istanbul) with old friends until the wee hours; this report will be a little shorter than usual.

Comoros and Tajikistan have started showing up on the JHU data due to the arrival of confirmed cases of COVID-19 in those two countries; I pushed a minor update to my webpage to plot them correctly.

Regarding yesterday's brief discussion of euromomo.eu excess mortality statistics, it's worth reading this week's Euromomo bulletin, which states that Ireland is not in fact showing significant negative mortality, but is overwhelmed by the pandemic to the extent that they are unable to report statistics as usual.

Toronto has backslid slightly, and I admit to some concern that as warmer weather returns this spring, with downtown temperatures forecast to finally rise above 20°C, everyone is going to forget about COVID-19 and go out and infect each other. There were certainly a lot of people out in the neighbourhood today disregarding our chief medical officer's request to stay at home except for urgent errands. It doesn't help that there has been conflicting messaging encouraging people to cheer up by going out and getting exercise in fresh air. Anyway, after three days of declining active case counts, we are up a little to 1,844 from 1,793 yesterday and 2,006 four days ago. Likewise, there are five more people in hospital (350), one more in ICU (112), and one more intubated (88). Here's hoping it's just a blip.

I have made a significant change to the webpage, which is illustrated in today's graphic. The earlier displays of growth rate vs. per capita infection were intended to show the neat, arching progression of each region's epidemic, with the current growth rate easier to see than on a case vs. time plot. I have added a new data series that shows the same statistics, but based on the JHU estimates for remaining active cases in each region (that is, total cases - deaths - recoveries). I am not yet sure whether I am going to keep the series in its current form, and know that if I do I'll need to change its rendering somewhat to reduce confusion. The main possibility I see for confusion is that there is no longer a clear timewise progression, and if consecutive points happen to be distant, there is no visual connection between them.

But you can observe in today's chart that the active case count can (and ideally should) have a negative growth rate, and that the overall shape of the curve should be that of a backslanting capital J. In the chart shown, China has a complete curve, New Zealand and Australia are almost complete and almost identical, Germany is headed in the right direction, France is slowing down, Canada isn't quite there yet, and the U.K. and U.S are almost identical and still on the upswing.

Stay safe.

2020-04-30

Comment on this post on Facebook.

today's daily chart as described below in text

Today's numbers: New York (-29→-33, continuing to improve due to decreasing new case rate), Singapore (24→29), Brazil (45→40), Louisiana (49→45), U.S. (43→45), Toronto (47→46), U.K. (52→47), Spain (47→49), Sweden (71→56), California (64→57), Mexico (68→62), Canada (62→62), Alabama (67→71), Ontario (77→78), Turkey (90→89), South Africa (92→92), Romania (93→106), India (100→108), Denmark (114→111), Italy (105→112), Poland (119→131), France (193→167), Indonesia (178→172), Iran (185→194), Israel (273→208), Germany (207→221), Malta (287→254), Japan (236→271), Norway (316→280), Malaysia (459→422), Austria (457→434), Thailand (1597→2003), New Zealand (3423→2054), Australia (1896→2305), South Korea (4489→4617), China (48814→38832)

Compared to yesterday, here's how these regions are doing.

Much worse: Brazil, Louisiana, U.K., Sweden, California, Mexico, France, Indonesia, Israel, Malta, Norway, Malaysia, Austria, New Zealand, China

Worse: Denmark

The same: Toronto, Turkey

Better: U.S., Spain, Canada, Alabama (data continues to be highly suspect, and I am considering removing this region from my list), Ontario, South Africa, India, Italy, Iran, Germany, South Korea

Much better: New York, Singapore, Romania, Poland, Japan, Thailand, Australia

Of the "much worse"s, Brazil and the U.K. are the only ones that cause me significant concern. Brazil's daily numbers have been steadily and significantly rising over the past few days, and their active case curve is the bumpiest of all. England worries me because according to this week's excess mortality statistics posted at euromomo.eu, they have by far the highest excess mortality z-score, with an estimated 44.10 standard deviations above the expected number of deaths for this time of year, as of three weeks ago (the last date for which reasonably firm figures are available), and a forecast of >40 after that. (Pause for brief teachable moment where I talk with the family about means and standard deviations.)

In Toronto, we had our third straight day of declines in active cases, dropping over that period from 2,006 to 1,793. We also had our first decline in hospital bed occupancy in almost two weeks, from 354 to 345.

In Ontario, people aged 60+ have accounted for 44% of confirmed cases and 95% of deaths; people in nursing homes 25% and 80%. In Quebec, people aged 60+ have accounted for 43% of confirmed cases and 97% of deaths; people in nursing homes 24% and 80%. These numbers are remarkably similar, and likely reflect similar culture, climate, architecture and population density.

Ontario, where the first case of COVID-19 was identified, has 40% of Canada's population, 40% of its COVID-19 cases, and 29% of its new cases today. Quebec, with 22% of Canada's population, now has 52% of Canada's cases, and 58% of the new cases today.

You would think then that if one province were to announce dates for reopening after the successful control of the epidemic, it would be Ontario, but no. We are still waiting for two to four weeks of steady decrease in our new cases, availability of excess hospital capacity everywhere, and science.

In Quebec, they have announced dates for when intraprovincial roadblocks would be lifted, when stores outside of malls can reopen, when the construction industry and all parts of its supply chain can reopen, when manufacturing can resume at half and then full capacity, and when schools and daycares can open. They are looking at having everything open again by May 11, with the exception of schools in Montreal, and the remaining half of manufacturing capacity. There is significant opposition to the reopening plan from doctors, teachers, neighbouring provinces, and pretty much everyone who understands what a second wave of COVID-19 would do to Quebec.

In other Canadian news, the first case of COVID-19 was reported in the territory of Nunavut, which had been the last remaining pandemic-free region in our country. They now have an infection rate of 1 case in 39,000.

Today's graphic: a comparison of how all the Canadian provinces and territories are doing, using JHU data.

Stay safe.

2020-04-29

Comment on this post on Facebook.

today's daily chart as described below in text

Today's numbers: New York (-20→-29, continuing to improve as their new case rate drops), Singapore (25→24), U.S. (37→43), Brazil (44→45), Toronto (53→47), Spain (51→47), Louisiana (50→49), U.K. (49→52), Canada (62→62), California (64→64), Alabama (63→67), Mexico (66→68), Sweden (80→71), Ontario (71→77), Turkey (90→90), South Africa (113→92), Romania (103→93), India (100→100), Italy (102→105), Denmark (99→114), Poland (120→119), Indonesia (154→178), Iran (181→185), France (195→193), Germany (201→207), Japan (173→236), Israel (212→273), Malta (391→287), Norway (379→316), Austria (434→457), Malaysia (577→459), Thailand (753→1597), Australia (1723→1896), New Zealand (1575→3423), South Korea (3861→4489), China (40675→48814)

Compared to yesterday, here's how these regions are doing.

Much worse: Toronto, Spain, Sweden, South Africa, Romania, Malta, Norway, Malaysia

Worse: France

The same: Singapore, Louisiana, Poland

Better: Brazil, U.K., Canada, California, Alabama (I wouldn't put much credence in this one; the maintainers of the AL Wikipedia page have been doing a very inconsistent job this week), Mexico, Ontario, Turkey, India, Italy, Iran, Germany, Austria

Much better: New York, U.S., Denmark, Indonesia, Japan, Israel, Thailand, Australia, New Zealand, South Korea, China

Spain had its first day of more than 4,000 new cases (4,771) in six days, hence much worse. The others in that category are due to more gradual changes.

If you're looking at inferred case numbers from reported deaths, the big jump in U.K. deaths today is a technical one, resulting from the incorporation of about 4,000 nursing home COVID-19 deaths into the total national statistics, which had previously excluded them. Deaths there now total 26,097. There has been no similar change in reporting of their new cases, which are ticking along at about 2.5% per day now.

Although the new case rate in Toronto has been creeping back upward over the last few days, and we saw more than 200 new confirmed cases for the first time in a week today, overall our numbers are suggesting that we are at the top of the curve. Active confirmed cases are down for the second consecutive day, to 1,860 from 2,006 two days ago. Our hospitalisation rates are still creeping up though, as are ICU and intubation numbers. One area in which we continue to do poorly is in management of outbreaks in nursing homes: one third of the new cases in the last two days have been at nursing homes, although they are likely being tested more thoroughly than the general public.

I celebrated the warmish weather today (10°C) by going to get the winter tires on my car changed for summer ones, using Toyota's contactless service programme. You book your appointment online or by phone, drop your key off in an envelope, pay by credit card, and pick your car up when it's ready without ever having to have an in-person social interaction. The tricky part was getting to and from the appointment, which involved taking a bicycle in the car, and a long ride up and down Broadview Avenue, about which my muscles will remind me tomorrow morning.

Active unresolved cases in Ontario dropped from 6,417 to 6,116 today, to the lowest level since April 22nd, aided by the lowest number of new cases (371), since April 7th. A few more weeks of this, and I can get a haircut.

United States new case numbers are continuing to drop daily, although tonight's likely won't be published before I post this. As of this morning, new cases had dropped for four straight days, from 36,138 to 20,945. Based on Johns Hopkins University data, most U.S. states and territories have a better four-day average new case rate now than they did four days ago; the exceptions from worst to best by change in rate are: ND PR DE IN VI VT IA MN. North Dakota has the worst overall change, moving from 4.5% to 7.3%, and now has 991 cases. Indiana has the highest total case count among those that are worsening, with 16,588 after rising from 4.0% to 4.9%. Delaware has the worst current average new case rate at 7.4%, with 4,575 cases. At the other end of the spectrum, Wyoming is the most improved state, going from 10.5% to 3.2%; Guam has the best current rate (0.0%), followed by Hawaii (0.3%), Montana (0.4%), Alaska (0.9%) and Vermont (1.0%).

At a county level, among counties with at least 10,000 reported cases, two have worsened in their four-day average new case rate: Nassau, NY (slightly up to 1.7%) and Rockland, NY (up from 1.6% to 3.2%); the rest have improved, with the most improved being Los Angeles, CA (7.6% to 3.2%), Suffolk, MA (7.4% to 3.1%), and Middlesex, MA (7.5% to 3.5%).

New York State's numbers today (4,585) are higher than the last two days, but still low enough compared to the 5-figure number five days ago that the trend is downward.

Stay safe.

2020-04-28

Comment on this post on Facebook.

today's daily chart as described below in text

Today's numbers: New York (-16→-20 - confusing, yes, but this is a good thing: the new case rate has dropped quickly so that 1% looks much further back in the rearview mirror now), Singapore (21→25), U.S. (37→37), Brazil (47→44), U.K. (44→49), Louisiana (38→50), Spain (38→51), Toronto (53→53), Canada (58→62), Alabama (57→63), California (59→64), Mexico (61→66), Ontario (65→71), Sweden (75→80), Turkey (82→90), Denmark (101→99), India (108→100), Italy (92→102), Romania (97→103), South Africa (100→113), Poland (111→120), Indonesia (145→154), Japan (120→173), Iran (178→181), France (109→195), Germany (145→201), Israel (158→212), Norway (303→379), Malta (857→391), Austria (389→434), Malaysia (422→577), Thailand (685→753), New Zealand (1277→1575), Australia (1633→1723), South Korea (3770→3861), China (53378→40675)

Compared to yesterday, here's how these regions are doing.

Much worse: India [I think this is an artifact due to a source data issue yesterday], Malta, China

Worse: Brazil, Denmark

Better: U.S. [no data available yet at time of writing], Toronto, Canada, Mexico, Sweden, Romania, Poland, Indonesia, Iran, Australia, South Korea

Much better: New York, Singapore, U.K., Louisiana, Spain, Alabama, California, Ontario, Turkey, Italy, South Africa, Japan, France, Germany, Israel, Norway, Austria, Malaysia, Thailand, New Zealand

Today's graphic shows all the countries of the world divided into four continental regions: the Americas (affected moderately and spreading moderately), Europe (affected severely and spreading slowly), Africa (affected slightly and spreading quickly), and Asia/Oceania (same as Americas, but more outliers). I'll let the charts speak for themselves and turn to my particular regions of interest.

Malta is listed as much worse, because two days after having no new cases, they suddenly have eight. That's few enough to list and trace, so they should be okay, but we should expect more to come. One was an English expat whose work colleagues are now exposed, another a healthcare worker who thinks she isolated before being contagious, a third self-misdiagnosed allergies and has been active in the community for three weeks; the remaining five are asylum seekers in a so-called reception centre.

China reported 22 new cases, the most it's seen in over a week. It still hasn't confirmed the rumoured new cases in Harbin, although Heilongjiang province has a two-week daily average of about 0.8% new cases compared to the national average of 0.05%.

Brazil's confirmed numbers fluctuate significantly from day to day, between 3,000 and 6,000 new cases. They currently claim a total of about 72,000 cases, which looks like it's underestimated by an order of magnitude based on reported deaths and estimated mortality.

Toronto started including estimated full numbers of recovered, previously confirmed cases yesterday, so that gives us one data point: unresolved cases are down from 2,006 yesterday to 1,893, or a 6% decline. That's a significant drop, and is grounds for mild optimism. We also have the lowest number of intubated cases (82, down from 86) in a week, and ICU bed occupancy has trended downward for four days. Overall hospitalisations are still trending upward though, at 319 compared to 281 a week ago, and for the past five days we've seen 130-180 new cases a day without much of a trend.

Ontario has announced criteria for phased reopening, and Ottawa announced general agreement at a national level as to how to coordinate reopening among neighbouring provinces that are in very different situations in the pandemic. In the first stage, some workplaces would be able to reopen if they can meet current public health guidelines, essential gatherings of a limited number of people could resume, and some open spaces would be reopened. In the second and third stages, each of these three types of restrictions would be progressively lifted, with minimal restrictions in the third stage. As to when each new stage could begin, there would have to be two to four weeks of consistent decline in new daily case numbers (we are currently on a 0-day streak), hospitals would have to be operating at below capacity for acute and critical care and have a surplus of ventilators and PPE, testing practices would have to permit the identification and recommended isolation of 90% of new cases within 24 hours of exposure, and ongoing random testing of vulnerable populations.

New York State Governor Andrew Cuomo announced yesterday that further analysis of last week's random antibody testing showed that 14.9% of state residents and 24.7% of New York City residents have been infected with SARS-CoV-2. Given that the confirmed new case rate has slowed down since last week from 2% to 1%, the infected proportion in New York City is still likely well under 30%, and might end up well under 40%. This is good news and in a small way, bad news. It's good that the worst affected place in the world is not getting worse very quickly any more, but part of the decrease is due to the well known epidemiological process that slows spread down when there are fewer surviving susceptible patients remaining. The increase in new cases each day is proportional to both the number of contagious cases and to the number of susceptible cases. So if only 75% of the city population is susceptible, that slows the transmission down to 75% of what it was back when 100% were susceptible. What people hope for with herd immunity is that if say you can get the susceptible population down to half of the total population, by infecting and/or killing the other half; then when the disease returns, it will only spread half as much, and it will be twice as easy to contain.

Cuomo added that they are planning on establishing isolation centres as in Asian countries that have been more successful in fighting COVID-19, so that when people do acquire COVID-19, they don't have to infect all of their family members, roommates, caregivers, people who share their building ventilation systems, etc. Good for him.

The United Kingdom has had fewer new cases each day now for a fourth day, dropping from 5,386 to 3,996. Likewise Spain, from 6,740 to 2,706. New York has had three such days; no other tracked region has as many, except possibly the as yet unreported United States.

The highest current 4-day average new case rates are Brazil (7.9%), Mexico (6.8%), India (6.3%), Singapore (5.5%); the rest are all under 5%, and a dozen are under 1%.

Stay safe.

2020-04-27

Comment on this post on Facebook.

today's daily chart as described below in text

Today's numbers: New York (-14→-16), Singapore (19→21), U.S. (33→37), Louisiana (38→38), Spain (N/A→38), U.K. (43→44), Brazil (47→47), Toronto (44→53), Alabama (52→57), Canada (54→58), California (38→59), Mexico (54→61), Ontario (57→65), Sweden (57→75), Turkey (74→82), Italy (83→92), Romania (89→97), South Africa (87→100), Denmark (94→101), India (99→108), France (151→109), Poland (105→111), Japan (120→120), Indonesia (127→145), Germany (133→145), Israel (110→158), Iran (175→178), Norway (231→303), Austria (374→389), Malaysia (366→422), Thailand (654→685), Malta (1070→857), New Zealand (1133→1277), Australia (1552→1633), South Korea (4605→3770), China (53375→53378)

Compared to yesterday, here's how these regions are doing.

Much worse: France, Malta, South Korea

Better: Louisiana, U.K., Brazil, Canada, Denmark, Poland, Japan, Germany, Iran, Austria, Thailand, Australia, China

Much better: New York, Singapore, U.S., Toronto, Alabama, California, Mexico, Ontario, Sweden, Turkey, Italy, Romania, South Africa, India, Indonesia, Israel, Norway, Malaysia, New Zealand

N/A: Spain

Today's graphic: most of the current trends on regions I'm tracking, omitting Spain (see below) and Sweden (due to ongoing concerns about methodology).

Toronto Public Health reported today that the number of recovered cases in Toronto soared from 364 yesterday to 2,670 today. I am 90% sure of what happened, but have been unable to find documentation online proving my hunch. Many jurisidictions have adopted the practice that if a patient tests positive for COVID-19, and is alive but no longer receiving treatment two weeks later, then they are declared recovered without waiting for one or more test results to confirm it. You could call these probable recovereds*, but be sure to put the asterisk at the end so that Scrabble players know that it's not a valid word. Two weeks ago, there were 2,362 cases, which is close enough to the 2,306 increase in recovered cases that the difference can easily be imagined to be deceased or still hospitalised cases.

The Spanish found the methodological changes yesterday too confusing, so they've reversed, and yesterday's numbers are ret-conned (not yet good in Scrabble, but has an anagram). The authorities should probably not have started including random antibody testing in with the main confirmed case count, but now that they've done so, they're stuck doing it for the rest of the pandemic.

I was looking in more detail at who is where on the active case curve, after I realized that I could scrape that data from most of my online sources, and rig up a primitive SIRD model to guesstimate in the remaining cases. The numbers at the top of this post tell you how bad things are, but not whether you're likely past the worst of it. Now, many regions make an informed but somewhat arbitrary decision to declare infected patients recovered if they've gone two weeks without dying or needing hospital care. So if you assume that most of the active cases from two weeks ago are going to recover, you can estimate how many active cases there are now just based on how many total confirmed cases are reported and how many deaths. It's not very precise, and it's not unusual for it to be off by 10-20% from officially published active case counts where available, but the errors tend to be systematic and drift slowly, so I think modelled active case counts are a reasonable basis for guessing which side of the curve a region is on.

Aside: the curve doesn't have to look like a bell. If it does, it'll won't look like a very smooth one. The official active case curve for British Columbia, for example, looks like part of a drawing by The Little Prince of a boa constrictor digesting an elephant (boa serpens elephantem concoquens).

Here's how I would group the regions I'm tracking according to their active case behaviour.

Rapidly climbing still: Brazil

Could be approaching the top: Austria, Canada, California, India, Indonesia, Japan, Poland, Romania, South Africa, Spain, United States

Moving erratically at what might be the top: Alabama, France, Mexico, Singapore, Sweden (see previous comments about episode data), Toronto, United Kingdom

Starting to come back down: Italy, Ontario, Turkey

Back down to a fraction of peak: Australia (19%), China (0%), Denmark (60%, but wavering), Germany (51%), Israel (84%), Iran (45%), Louisiana (41%), Malta (49%), Malaysia (68%, but moving very slowly), New York (78%), New Zealand (26%), Norway (40%), South Korea (22%), Thailand (18%),

Stay safe.

2020-04-26

Comment on this post on Facebook.

today's daily chart as described below in text

Today's numbers: Spain (32→N/A), New York (-13→-14), Singapore (18→19), U.S. (33→33), Louisiana (35→38), California (36→38), U.K. (43→43), Toronto (39→44), Brazil (47→47), Alabama (52→52), Canada (51→54), Mexico (48→54), Ontario (54→57), Sweden (50→57), Turkey (69→74), Italy (75→83), South Africa (85→87), Romania (82→89), Denmark (82→94), India (97→99), Poland (105→105), Israel (75→110), Japan (104→120), Indonesia (123→127), Germany (130→133), France (120→151), Iran (172→175), Norway (191→231), Malaysia (347→366), Austria (348→374), Thailand (652→654), Malta (855→1070), New Zealand (926→1133), Australia (1923→1552), South Korea (4873→4605), China (43792→53375)

Compared to yesterday, here's how these regions are doing.

Much worse: Australia, South Korea

Same: New York

Better: Singapore, U.S., California, U.K., Brazil, Alabama, Canada, Ontario, Turkey, South Africa, Romania, India, Poland, Indonesia, Germany, Iran, Malaysia, Austria, Thailand

Much better: Louisiana, Toronto, Mexico, Sweden, Italy, Denmark, Israel, Japan, France, Norway, Malta, New Zealand, China

See below: Spain

The official proportion of long-term care deaths in Ontario is dropping, and is down to 54.5%, due apparently to some reclassification of deaths resulting in 15 cases being removed from the LTC deaths list today. There's actually, confusingly, another official count which shows 200 more LTC deaths or 78%. My understanding is that the latter is the result of the Ministry of LTC checking with LTC facilities for current numbers, whereas the former is how many people have had time to log in the integrated public health information system database. A note in the official statistics says to trust the 78% as being more complete.

The total number of active unresolved cases in Ontario is in fact down for the first time today (as I had hoped it would yesterday), down a little from 6,486 back to the previous day's 6,432. I note for comparison with other regions where active case counts are not easy to find, that this is happening on a day when the new case numbers dropped from 3.52% to 3.12% and the four-day average from 4.50% to 4.19%.

There are no immediate plans to loosen social distancing restrictions here, as it looks like we have barely reached the top of the curve, and we could still go like British Columbia or Alberta and lose control again. Premier Doug Ford may also be guided in this by the recently announced diagnosis of his own 95-year-old mother-in-law with COVID-19, while she is a resident in a nursing home that has already lost 12 other residents to COVID-19.

Speaking of nonagenarian Ontarians (try saying that quickly), a 96-year-old close relative of mine in Toronto (most of you will know who I'm talking about) broke two teeth yesterday and would normally be on his way to the dentist for emergency treatment. Instead, I've emailed his dentist and expect to hear back shortly that as long as he doesn't develop any further symptoms (pain, infection, fever, weight loss), he should wait until the clinic reopens.

I was talking to a school administrator yesterday who said that his custodial staff had been required to remove all first aid kits, hand sanitizer, PPE, etc. from his school soon after it was closed. It made his job difficult afterward when he was instructed to share the school's supply of laptops and tablets with students who needed them at home, because they didn't even leave him a pair of gloves. When schools do reopen in some form, there's going to be a significant effort involved in restocking everything.

On April 19th, Spain started including positive antibody tests in its case count, which accounted for 7,365 of that day's 4,258 [sic] new cases. No explanation at the time was given for a reduction by 6,212 in the number of active PCR-confirmed cases, although one could charitably suppose that there was an overlap between the two categories. Today, they changed their counting methodology again, saying that it was misleading to include antibody tests in the daily totals, because the patients could have been exposed much earlier than the test date. Subtracting yesterday's 17,854 positive antibody tests resulted in the total case count appearing to decrease substantially, from 223,759 to 207,634. Taking it into account would give an increase of 1,729 (the so-called Ramanujan's number), which would be the lowest real increase in over a month, and less than half of the previous day's 3,995. So all this by way of saying that I do not have a lot of confidence in the official numbers being produced by the Spanish Ministry of Health, although I have sympathy for their task of having to make sense of reports from all the regional offices. Still, the reported number of active cases has been on a gradual decline for six days, coming down from 90,746 to 85,712, and the number of recoveries is increasing apace. It looks like Spain may be over the hump.

New reports in Singapore say that the apparent downtown in new cases is just because the government has given up on contact tracing and even systematic testing of foreign workers that are the hardest hit by the current outbreak. Instead, anyone with respiratory symptoms is being sent to a community isolation facility and not counted toward confirmed totals.

In New Zealand, 1 new confirmed and 4 probable cases were reported, but 6 past confirmed cases were somehow rescinded, resulting in a net drop of 1 cumulative case. Their active case numbers have been working their way steadily down, from a high of 929 on April 6th back when their 4-day average new case rate was still over 8% to 328 yesterday and 289 today. Hospitalisations are down from a high of over 20 to just 7 today.

For the first time in over a month, no new cases have been reported in Malta today, thanks to a strict lockdown policy requiring about a quarter of the population (all vulnerable citizens in specific categories, such as the immunocompromised and seniors, including even the president) to stay home. A discussion in comments yesterday raised the possibility of developing herd immunity in the general population while quarantining the vulnerable. I said that I thought it would be difficult to get 40% of the whole population infected while keeping 40% at home; the Maltese example shows that a better possible outcome of vulnerable quarantining is the eradication of the disease.

Today's graphic: one-week trends for all of these regions.

Stay safe.

2020-04-25

Comment on this post on Facebook.

today's daily chart as described below in text

Today's numbers: New York (-14.5→-13.1), Singapore (15→18), Spain (32→32), U.S. (35→33), Louisiana (36→35), California (36→36), Toronto (38→39), U.K. (44→43), Brazil (55→47), Mexico (48→48), Sweden (47→50), Canada (49→51), Alabama (51→52), Ontario (51→54), Turkey (59→69), Italy (73→75), Israel (76→75), Denmark (88→82), Romania (76→82), South Africa (80→85), India (90→97), Japan (104→104), Poland (112→105), France (102→120), Indonesia (119→123), Germany (136→130), Iran (165→172), Norway (200→191), Malaysia (337→347), Austria (377→348), Thailand (999→652), Malta (264→855), New Zealand (1270→926), Australia (1713→1923), South Korea (4731→4873), China (29439→43792)

Compared to yesterday, here's how these regions are doing.

Much worse: Brazil, Denmark, Poland, Germany, Norway, Austria, Thailand, New Zealand

Worse: New York, U.S.

The same: Louisiana, U.K., Israel

Better: Spain, California, Toronto, Mexico, Sweden, Canada, Alabama, Ontario, Italy, Romania, South Africa, India, Japan, Indonesia, Iran, Malaysia, South Korea

Much better: Singapore, Turkey, France, Malta, Australia, China

New Zealand is much worse today than really good, so still pretty good. Their daily new cases have gradually drifted up from a low of three back up to today's nine. Yes, that's a lot worse than three, but if we had that few here we would be partying in each other's houses.

New York continues to wander through unexpected, uncharted mathematical waters like the value of a future contract for an obsolete commodity. Its statistics is going back closer to zero as its daily new case rate increases now for the fourth consecutive day (raw daily), or third day (smoothed). As of today, New York overtook Toronto's daily new case rate, climbing to 3.89% ahead of Toronto's 3.17%. For the fourth consecutive day, New York saw more cases confirmed, rising over that period from 4,178 to 10,553 new cases for a cumulative total now of 282,143 confirmed cases, with total cases including asymptomatic ones likely ten times as many based on last week's random antibody tests.

Some people talk about the possibility of “herd immunity” slowing the spread of COVID-19 once 40% of the population is infected. How close is New York State to that threshold? Assuming they're at about 14.5% infected now, then the answer would be log(40/14.5)/log(1.0389) = 27 days at today's new case rate. The death toll would rise to hundreds of thousands, and the disease might die down over a period of months or years, if there were any reason to suspect that recovering from COVID-19 leaves you immune, which reports from China and South Korea suggest is not the case.

If we calculated days to 40% instead of to 1%, using confirmed cases, here is how everyone would rank: Singapore (63), Mexico (87), California State (93), Brazil (95), New York State (116), Toronto (116), U.S.A. (129), Ontario (138), Canada (140), Alabama State (144), U.K. (148), South Africa (150), India (154), Sweden (157), Japan (188), Romania (188), Turkey (192), Spain (193), Indonesia (202), Poland (214), Israel (234), Denmark (244), Louisiana State (263), Italy (321), France (382), Germany (417), Iran (456), Norway (550), Malaysia (665), Austria (1,067), Thailand (1,091), New Zealand (1,904), Malta (2,170), Australia (3,891), South Korea (9,522), China (75,121, or about two centuries).

If we uniformly guess that confirmed cases underreport actual infections by an order of magnitude, here is how the rankings would change: New York State (35), Singapore (35), California State (58), Mexico (63), Brazil (65), Toronto (68), U.S.A. (69), U.K. (82), Canada (85), Ontario (85), Alabama State (87), Sweden (90), Spain (93), South Africa (109), Turkey (115), India (118), Louisiana State (121), Romania (122), Japan (135), Israel (135), Denmark (143), Poland (146), Indonesia (152), Italy (168), France (218), Germany (238), Iran (279), Norway (326), Malaysia (467), Austria (618), Thailand (817), New Zealand (1,294), Malta (1,349), Australia (2,663), South Korea (6,620), China (55,565).

Closer to home, Ontario had its lowest increase in unresolved cases, and continues to show signs of being very close to the peak of that curve. There were 54 more cases today, compared to 233 the previous day, and 175 the day before that. Also, although it's not a particularly meaningful statistic, today for the first time patients who have been declared to have recovered from COVID-19 in Ontario (6,698) outnumber those who are still fighting it (6,486). Deaths today decreased slightly for the third day in a row: 54, 50, 48 (of which in Toronto, 12, 16, 15). That's still a lot of dead people, and it seems like most people know someone who knows someone who has died.

Even closer, one of my father's older female neighbours had COVID-19 recently, based on typical symptoms, although she did not qualify for testing. She says she is mostly recovered.

Singapore's new case numbers the last few days have decreased 1,037, 897, 618, so it looks like their foreign worker outbreak issue is coming under control.

Today's graphic: side-by-side comparison of the reported state of American states vs. the state of all countries in the world.

Stay safe.

2020-04-24

Comment on this post on Facebook.

today's daily chart as described below in text

Today's numbers: New York (-15.0→-14.5), Singapore (12→15), Spain (45→32), U.S. (39→35), Louisiana (32→36), California (45→36), Toronto (38→38), U.K. (45→44), Sweden (54→47), Mexico (55→48), Canada (46→49), Alabama (51→51), Ontario (50→51), Brazil (61→55), Turkey (51→59), Italy (78→73), Romania (82→76), Israel (76→76), South Africa (89→80), Denmark (88→88), India (96→90), France (98→102), Japan (133→104), Poland (116→112), Indonesia (140→119), Germany (166→136), Iran (159→165), Norway (203→200), Malta (233→264), Malaysia (415→337), Austria (399→377), Thailand (999→999), New Zealand (1267→1270), Australia (1741→1713), South Korea (4867→4731), China (24742→29439)

Compared to yesterday, here's how these regions are doing.

Much worse: Spain, U.S., California, Sweden, Mexico, Brazil, Italy, Romania, South Africa, India, Japan, Poland, Indonesia, Germany, Malaysia, Austria, Australia, South Korea

Worse: Norway, New York

The same: U.K.

Better: Toronto, Canada, Alabama, Ontario, Israel, Denmark, France, Iran, Thailand, New Zealand

Much better: Singapore, Louisiana, Turkey, Malta, China

New York had its first "worse" day in several days, with a large uptick in new cases, which have increased daily now for the third consecutive day, from 4,178 new cases on April 21st to 8,130 new cases on April 24th. New York state health officials also released yesterday results of random antibody sampling conducted over the past week, which indicated that 13.9% (or roughly 2.7 million) of the surviving population has been infected. If we assume that anyone who was sampled and is going to die will do so in about 7 days, that should be when the confirmed death count hits about 20,000. According to the New York Times' excellent ongoing analysis of excess mortality around the world, COVID-19 mortality appears to be being underreported by about 25% in New York, so let's call that number 25,000. That makes a death rate of about 1%, which is at the low end of what areas have reported.

Just based on what I see in the media, which my past travels to the United States have always made clear to me are a terrible way to judge a country, it seems as though many Americans are not complying with social distancing measures, because their circumstances or their personal desire for freedom prevent them. Where the U.S. managed to do a good job of rapidly bringing their new case rates down below 5%, their daily rates have been risen daily for three days, and are back above 4% for the first time in a week. That rate needs to come down to and stay somewhere around 1-3% in order for the active case curve to flatten, and instead it's straightening back out again. Today's new case number of 36,138 was their largest to date.

Ontario started reporting the daily cumulative number of deaths in long-term care (LTC, nursing homes) on April 19th. A brief discussion in their daily report acknowledges that the numbers are hard to reconcile, because they are coming from different sources, using different methodologies, at different rates. This has resulted in the reported new deaths in LTC exceeding the total new deaths reported in the province for the last two days; today we had 105 out of 50 new deaths being in nursing homes. This has had the effect or rapidly increasing the proportion of the total deaths that have taken place in nursing homes, which has risen from 43% at the start of the week now to 61%.

Still, it's worth noting that in many parts of the world, deaths in nursing homes are not being counted as COVID-19 deaths at all, due to historical reporting practices.

In Ontario today, some media were reporting that our so-called resolved cases now exceeded unresolved cases by a significant margin, with the resolved cases representing 52.4% of the total. A resolved case is considered resolved for this purpose if a patient is deceased, or not hospitalised and at least 14 days past their symptom onset or investigated episode date. While it is somewhat good news that this percentage has crept up from 47.8% to 52.4% over the course of a week, we still have more unresolved cases (6,432) than resolved surviving cases (6,324). What's happened in the past week is that 2,246 have recovered, 285 people have died, while 3,994 new cases have been confirmed; this gives an increase of 3994-2246-285 = 1463 in the number of unresolved cases this week.

Another concerning issue is that these numbers, 285 and 2246, tell us that 11.3% of patients whose cases were resolved in the past week did so by dying. This is consistent with mortality demographics published by the same authority, showing that 67% (or 95%) of the deaths have been patients aged 80+ (or 60+), and that 17% (or 11.9%) of that age 80+ (or 60+) patient cohort has died.

On a personal note, my aunts say that in their nursing home, they are back to being confined 24 hours a day to their rooms. When staff enter, they are fully dressed in personal protective equipment (face shield, mask, gown, gloves, although I doubt they have enough to change their PPE over the course of the day, let alone between patients. I haven't heard any updates from their nursing home since they first reported one patient and one staff member confirmed with COVID-19 on Monday. My aunts say that local rumour has it that the cases were on a floor above them (so somewhere between the 5th and 7th).

My aunts' nursing home, whose outbreak started four days ago, is still not counted among the official statistics of affected nursing homes.

Several countries appear to be well past the top of their active case curve. Australia only started publishing their recovery numbers on April 5th, when the active case count was starting to drop at a fairly high new case rate level of 3.6%. Malta's started to clearly drop when their new case rate was at about 2%. Germany's active case count has come gradually since April 12th (new case rate 3.3%). For Germany, that means dropping from 60,000 sick to 45,000 sick; for Australia, it's brought the number down from about 3,300 to 1,500.

I have not found any indication in U.S. data that active cases are starting to taper off anywhere.

The active case count situation in Canada is complicated and regionalized. Because interprovincial travel is restricted, different provinces are at very different stages in the progress of the epidemic. Ontario, Quebec and Nova Scotia, with about 2/3 of the population of the country, are still nowhere near the top of the curve, with Ontario doing somewhat better than Quebec and Nova Scotia. British Columbia and Alberta, with another 1/4 of the population, managed to flatten the curve in mid-March and early April respectively, but have since let it go and it's back on the rise again. The other provinces and territories are all well past the peak of the curve. Manitoba is about 40% down from peak, Saskatchewan and New Brunswick are at about half peak, and the other provinces much better than that. All of the provinces that are doing well managed to get things turned around at the beginning of April. Well done.

Today's graphic shows the different trajectories of these five countries: Malta, Australia, Germany, Canada, and the United States.

Singapore is back under 1,000 new cases today. South Korea went a second straight day with no fatalities. Well done to them too.

Stay safe.

2020-04-23

Comment on this post on Facebook.

today's daily chart as described below in text

Today's numbers: New York (-13→-15), Singapore (13→12), Louisiana (35→32), Toronto (41→38), U.S. (40→39), Spain (46→45), California (48→45), U.K. (41→45), Canada (48→46), Ontario (50→50), Turkey (47→51), Alabama (61→51), Sweden (70→54), Mexico (56→55), Brazil (68→61), Israel (81→76), Italy (75→78), Romania (84→82), Denmark (89→88), South Africa (94→89), India (91→96), France (109→98), Poland (96→116), Japan (95→133), Indonesia (138→140), Iran (148→159), Germany (146→166), Norway (239→203), Malta (233→233), Austria (422→399), Malaysia (387→415), Thailand (657→999), New Zealand (1010→1267), Australia (1136→1741), South Korea (4033→4867), China (27097→24742)

Compared to yesterday, here's how these regions are doing.

Much worse: Alabama, Sweden, Brazil, Israel, South Africa, France, Norway, Austria, China

Worse: Louisiana, Toronto, California, Canada, Romania

The same: Singapore, U.S., Spain, Mexico, Denmark

Better: Ontario, Italy, India, Indonesia, Iran, Malta, Malaysia

Much better: New York (confusing, yes, but their new case rate is still steadily decreasing), U.K., Turkey, Poland, Japan, Germany, Thailand, New Zealand, Australia, South Korea

I made three changes to the main webpage today, adding the full Johns Hopkins University data (as JHU Global) with subnational regions like Canadian provinces combined into countries, reducing the size of past flag markers on the graph to 40% of what they were before to slightly reduce visual clutter, and adding the days-to-1-% figures to the current day's marker popup title. Today's graphic therefore shows the current position in the phase diagram of each country in the world. Here are some highlights.

The countries with the highest proportion of their population confirmed infected are all European, and they start small: San Marino, Vatican City, Andorra, Luxembourg, Iceland; then Spain, Belgium, Ireland.

The country with the lowest proportion (of confirmed cases in population) are: Yemen, Angola, Papua New Guinea, Burundi, Ethiopia.

The countries with the highest four-day average new case rates are: Maldives, Burundi (22%), Sudan, Somalia (21%).

Countries with low positive DT1% indices not listed above: Andorra (10), Qatar (16), Maldives (18), Belarus (24), Russia (28), Saudi Arabia (30).

Singapore is categorized as the same, because the 1,000 or so new cases a day that it has seen for the past four days is now what this primitive model is expecting, so another day of it is just more of the same.

Toronto hasn't yet reached the peak of the curve. Hospitalisations are still creeping up (three more to 293 today), ICU bed occupancy is stable, but one more patient has been intubated (now 86 total). We're close, but we're also still confirming 200+ cases a day, and not testing who knows how many more. Ontario likewise is seeing 500-600 new cases. Parts of the country that are less affected are talking about gradually reopening businesses; I think they're doing the right thing, and hope it goes well for them. Saskatchewan, for instance at time of writing, saw only 13 new cases in the last four-day interval over which I'm calculating averages [originally reported as 13 cases in a day, corrected by Risa Horowitz], which is a manageable number to trace.

I am curious as to what's going on in Harbin. Various reports say that a student returning from New York set off a new outbreak that has resulted in a lockdown, some American reports blame the Russians. The weird thing to me is that the increased numbers are not yet showing up in Chinese statistics anywhere.

No one died of COVID-19 in South Korea today, and only six new cases were confirmed!

Stay safe.

2020-04-22

Comment on this post on Facebook.

today's daily chart as described below in text

Today's numbers: New York (-11→-13), Singapore (12→13), Louisiana (35→35), U.S. (38→40), U.K. (38→41), Toronto (42→41), Spain (49→46), Turkey (44→47), California (53→48), Canada (49→48), Ontario (49→50), Mexico (61→56), Alabama (61→61), Brazil (67→68), Sweden (70→70), Italy (74→75), Israel (114→81), Romania (89→84), Denmark (95→89), India (87→91), South Africa (94→94), Japan (98→95), Poland (90→96), France (98→109), Indonesia (128→138), Germany (116→146), Iran (142→148), Malta (199→233), Norway (224→239), Malaysia (378→387), Austria (413→422), Thailand (546→657), New Zealand (694→1010), Australia (780→1136), South Korea (4031→4033), China (24738→27097)

Compared to yesterday, here's how these regions are doing.

Much worse: California, Mexico, Israel, Romania, Denmark

Worse: Spain, Japan

The same: Toronto, Canada

Better: Singapore, Louisiana, U.S., Turkey, Ontario, Alabama, Brazil, Sweden, Italy, India, South Africa, Poland, Indonesia, Iran, Norway, Malaysia, Austria, South Korea, China

Much better: New York, U.K., France, Germany, Malta, Thailand, New Zealand, Australia

Today's graph shows all European nations who have at least 1 in 1,000 of their population confirmed with COVID-19, excluding Andorra, the Holy See, and San Marino because of their small populations.

Euromomo has replaced their dated static graphs with more modern interactive tools at https://www.euromomo.eu/graphs-and-maps/ that are a delight to play with. You can easily tell, for example, that three weeks ago, Spain was experiencing mortality that was 18.03 standard deviations above the mean, and England two weeks ago was at least 18.49. As noted earlier, these numbers tend to increase for two or three weeks as death reports make their way into the databases. It's also worth taking a look at the excess mortality map for week 13 in 2020, which shows how things are worst in Spain, Italy and the Netherlands, very bad in England and France, and bad in Scotland, Wales, Sweden and Switzerland.

The Japanese government is asking its people to reduce contact through social distancing, explaining that they should avoid crowded locations, not queue up too closely, minimize travel and outings even during the upcoming Golden Week holidays, and generally reduce interpersonal contact 80%. Specifically, they published a list of ten key recommendations: visit relatives online rather than in person, shop alone at off-peak hours, go jogging only in small groups and at off-peak hours, order non-urgent items online, move social drinking to online platforms, seek medical consultations online, work out online using videos, eat takeout or home-cooked food, work from home unless providing an essential service, and to wear a mask when talking to people in person.

Japan's new cases have trended steadily back down to below 5%, but they had a onetime increase in 47% of their cumulative death toll today, due to the inclusion of 80 cases previously categorized as presumptive.

Toronto has started publishing data about institutional outbreaks. 139 of Toronto's 210 deaths to date have been in institutions: retirement homes, long-term care facilities, or hospitals. Three long-term care facilities account for 81 of the 139. Their cases account for 1,345 of the 4,069 cases in the city.

I have personal knowledge however that the 1,345 is incomplete because the list does not include the nursing home where my aunts live, which has at least one confirmed staff case and one confirmed resident case, based on mass communication from the nursing home director. The communication also indicated that current policy is that hospitals may not transfer patients into nursing homes, and that nursing home staff may no longer work at more than one institution.

I'm trying to make sense of the widely reported USC study that estimated that 320,000 or 4% of Los Angeles County residents tested positive for coronavirus antibodies as of April 10-11. Mortality statistics suggest that the county has about half of the affected population of the state, which as of April 11th reported 21,794 confirmed cases statewide and looks like it might have 1,700 fatalities by April 25th, two weeks after the study date. If we say that 850 of those 1,700 are in L.A. county, that gives a mortality rate among antibody-positive cases of a quarter of a percent, which is a lot lower than any estimates for mortality among symptomatic patients. If say we accepted one popular figure of 2% mortality for that group, that would mean that out of every eight people that are infected, only one develops symptoms. Or if 1%, then one in four. Either way, it helps to explain why this disease is so dangerous: there are a lot of people who are carrying it unknowingly, and who may be spreading it in the community.

New cases in the United States are up for the second day in a row, and back after a brief lull above the 30,000 and 3.5% marks. Its four-day average is still trending downward, but barely.

New York is down to about 2% daily new cases and 3% daily deaths, or about 5,000 and 500 respectively.

Canada had a second consecutive day under 5% (both one-day and four-day).

Singapore had its third straight day of more than 1,000 new cases, but it's been fewer every day.

Of all the regions that I'm tracking, Malta has the fewest new cases today at 1, followed by New Zealand at 3. Looking back further over the last three days, New Zealand has had only 3+3+5=11 to Malta's 1+12+4. If you want a good news story about COVID-19, look online for coverage about the positive approach that the New Zealand government has taken.

Stay safe.

2020-04-21

Comment on this post on Facebook.

today's daily chart as described below in text

Today's numbers: New York (-9→-11), Singapore (13→12), Louisiana (30→35), U.K. (35→38), U.S. (38→38), Toronto (34→42), Turkey (44→44), Ontario (47→49), Canada (46→49), Spain (43→49), California (61→53), Alabama (69→61), Mexico (60→61), Brazil (55→67), Sweden (62→70), Italy (69→74), India (89→87), Romania (83→89), Poland (77→90), South Africa (88→94), Denmark (92→95), France (136→98), Japan (98→98), Israel (108→114), Germany (94→116), Indonesia (118→128), Iran (135→142), Malta (216→199), Norway (183→224), Malaysia (356→378), Austria (348→413), Thailand (502→546), New Zealand (556→694), Australia (627→780), South Korea (3439→4031), China (25855→24738)

Compared to yesterday, here's how these regions are doing.

Much worse: California, Alabama, France, Malta, China

Worse: India

The same: Singapore

Better: U.S., Turkey, Ontario, Canada, Mexico, Italy, Romania, South Africa, Denmark, Japan, Israel, Indonesia, Iran, Malaysia, Thailand

Much better: New York, Louisiana, U.K., Toronto, Spain, Brazil, Sweden, Poland, Germany, Norway, Austria, New Zealand, Australia, South Korea

I've added the Johns Hopkins University data for all African countries today, as shown in today's graphic. That's Djibouti on the right, at 7% daily new cases and 0.1% infection. Sudan at the top left, streaking upward at 34% from 1 in 400,000. Fading into the top right is Eswatini at 18% from 1 in 37,000, crossing paths in four or five days with Somalia at 25% from 1 in 56,000. At the bottom left are Burundi and South Sudan, with a handful of confirmed cases each but no new ones in a week.

Julian West and Russell Honeybun (and before them several people including I think Jonathan Haruni) asked why the series in yesterday's graph looked bimodal, Bactrian, or two-humped. I've spent some time looking into this, and believe that it is just apophenia or pareidolia.

These graphs show the four-day average of new case rates, so a smoothed version of the new case rate statistic. Because they are smoothed, any bumps that survive are noticeable; this is the point of smoothing, and I use it here because a bump in this graph means that something went wrong and was then fixed. Some graphs have only one prominent bump: Italy and Poland. Some graphs are just a mess of bumps: Sweden, United Kingdom. And some graphs have two prominent bumps, about which more below. If a curve has one bump, we can comfortably match it against our memories of ballistic arcs. If a curve has a lot of bumps, we can mentally throw it away for being messy. If it has just two bumps though, it piques our hopeful attention. Here are some examples that I've taken a closer look at.

Canada had a peak on March 17 (31%), a trough on March 22 (20%) and a peak March 26 (29%), and no later extrema. This appears to be the result of reporting methodological change in Quebec on March 23rd, when the province began accepting hospital tests as evidence. So here, what went wrong was just that the authorities had to change their measuring criteria midstream.

India: March 23 (26%), March 29 (14%), April 4 (22%). This appears to be due to two factors. Testing was extremely limited until March 21st, resulting in a wave of backlogged cases that were confirmed rapidly when private labs came online. In the week before and after that date, social distancing measures were enacted, which eventually caused a genuine downturn in the graph.

Australia: March 4 (19%), March 8 (11%), March 16 (25%). This was because of a sudden bump on March 4th due to the repatriation of passengers from the cruise ship, Diamond Princess, who accounted for all but one of the new cases that day.

Malaysia: March 6 (30%), March 10 (12%), March 17 (36%). There was a large cluster of cases associated with a religious event that took place on March 11th.

Spain: February 28 (83%), March 7 (26%), March 11 (52%). There was a cluster of cases early on associated with funerals in Basque country, that once identified and isolated temporarily brought the rate under control.

United States: March 4-8 (43-47%), March 17 (27%), March 21 (41%). I think what happened here is that the United States identified the early threat of travel, and took measures to control it, but didn't respond quickly enough to the later threat of domestic contagion.

So I think broadly speaking, we can categorize prominent bumps in these graphs into methodological changes (including new categories of cases, or suddenly increasing testing), and comparatively serious outbreaks that are brought under control. In the methodological case, the bump is an artifact, and subsides once the new cases no longer contribute to recent percentage increases. In the case of outbreaks, in order for a bump to actually look like a bump and not a suddenly steep climb, there needs to be an adequate response. This happens more often early on, when total case numbers are low, and patients who come from a cruise ship or funeral are few enough to be counted and contact-traced. To have a downward turn later on requires either serious social distancing efforts, or the development of herd immunity in surviving members of the population.

2020-04-20

Comment on this post on Facebook.

today's daily chart as described below in text

Today's numbers: New York (-7→-9), Singapore (15→13), Louisiana (31→30), Toronto (30→34), U.K. (34→35), U.S. (32→38), Spain (36→43), Turkey (42→44), Canada (47→46), Ontario (47→47), Brazil (51→55), Mexico (58→60), California (48→61), Sweden (54→62), Italy (61→69), Alabama (55→69), Poland (73→77), Romania (64→83), South Africa (90→88), India (90→89), Denmark (82→92), Germany (83→94), Japan (78→98), Israel (100→108), Indonesia (98→118), Iran (126→135), France (122→136), Norway (166→183), Malta (144→216), Austria (281→348), Malaysia (269→356), Thailand (489→502), New Zealand (511→556), Australia (594→627), South Korea (2703→3439), China (4482→25855)

Compared to yesterday, here's how these regions are doing.

Worse: Singapore, South Africa

The same: Louisiana, Canada, India

Better: U.K., Turkey, Ontario, Brazil, Mexico, Poland, Israel, Iran, Thailand, New Zealand, Australia

Much better: New York, Toronto, U.S., Spain, California, Sweden, Italy, Alabama, Romania, Denmark, Germany, Japan, Indonesia, France, Norway, Malta, Austria, Malaysia, South Korea, China

Today's graphic is a cautionary tale showing the history to date of Canada, Japan, New York (state), and the United States. Canada, New York and the United States have reached about the level that Japan reached a month ago, when it looked like it was on its way to defeating COVID-19 just by maintaining a polite Asian distance from it. Then things went wrong, and it's take a month to bring the new case rate back down to the same neighbourhood, during which time the total number of cases has increased by an order of magnitude, from 1 in 100,000 to 1 in 10,000. This is not the time to relax social distancing measures.

Singapore continues to race toward 1%, with 1,426 new cases today after only 596 yesterday. As before, the vast majority (1,369, or 96%) of the new cases are from foreign worker dormitories, and most of the new cases are asymptomatic ones who were required to submit to testing while being isolated after potential contact with other cases. The number of community transmission cases outside of interactions with foreign workers is reportedly only about 20 per day.

Toronto hospitalisations increased slightly, from 241 to 262, after remaining steady for several days. Intubations were up one from 81 to 82. New cases overall are down for the third day in a row, and at 136 almost half of the 264 that we saw four days ago. Also showing a similar steady decrease in new cases: Austria, China, Denmark, Iran, Italy, New York State, Romania.

Ontario current hospitalised and intubated cases are down slightly for the third day in a row, and everyone is continuing to talk about how we may be reaching the top of the curve. (Unless of course they're talking about senseless gun violence or oil futures.)

My aunts' nursing home reported their first confirmed case of COVID-19 today, and says that they will switch to an even stricter protocol. I'm not sure what that entails: maybe keeping everyone in their rooms as much as possible? The case was a staff member, which is probably worse than if it were a resident, as staff members have more contact with other residents. My aunt told me today that there were some very sick people in the lunchroom today. This makes 78 institutions in Toronto that have confirmed a case of COVID-19.

As many media have reported, nursing homes here are not designed to contain diseases like COVID-19. My mom's nursing home had private bathrooms, but my aunts' does not, so anyone who uses a toilet (which admittedly is a much smaller proportion than in the general population) has to share a communal bathroom with everyone else in their wing. They also don't have the staffing levels to safely care for residents when they are all in their individual rooms.

New York State's confirmed cases increased by less than 2% for the first time today. California's was barely above 2%, at 2.13%. Mexico saw its first day in a week in which the number of new cases didn't increase. So good news all around, but no time to relax. The parachute has slowed our rate of descent, but we shouldn't take it off yet.

Stay safe.

2020-04-19

Comment on this post on Facebook.

today's daily chart as described below in text

Today's numbers: New York (-5→-7), Singapore (15→15), Toronto (32→30), Louisiana (29→31), U.S. (39→32), U.K. (36→34), Spain (32→36), Turkey (40→42), Canada (46→47), Ontario (47→47), California (49→48), Brazil (44→51), Sweden (51→54), Alabama (50→55), Mexico (63→58), Italy (62→61), Romania (62→64), Poland (78→73), Japan (78→78), Denmark (78→82), Germany (82→83), India (99→90), South Africa (92→90), Indonesia (95→98), Israel (86→100), France (93→122), Iran (121→126), Malta (122→144), Norway (134→166), Malaysia (265→269), Austria (219→281), Thailand (493→489), New Zealand (440→511), Australia (569→594), South Korea (2352→2703), China (4322→4482)

Compared to yesterday, here's how these regions are doing.

Much worse: U.S., Mexico, Poland, India, Thailand

Worse: Toronto, U.K., South Africa

The same: California, Italy

Better: Singapore, Turkey, Canada, Ontario, Sweden, Romania, Japan, Denmark, Germany, Indonesia, Iran, Malaysia, Australia, China

Much better: New York, Louisiana, Spain, Brazil, Alabama, Israel, France, Malta, Norway, Austria, New Zealand, South Korea

U.S. numbers are as usual preliminary, based on incomplete state reports.

The official Spanish death statistics have taken an unusual turn, with 20,590 being confirmed dead of COVID-19 today and 20,639 having been reported dead yesterday. No, this is not the long-feared start of the apocalyptic zombie phase of the pandemic, but it's still something quite serious. National statistics were being compiled by the ministry of health based on reports from autonomous communities (the equivalent here of provinces or states), and one of them (Catalonia) had recently started including nursing home deaths in its statistics, contrary to national policy. When Catalonian nursing home deaths were subtracted, this resulted in a net decrease in the cumulative number of deaths to date.

Early on, there was a lot of talk about how testing issues meant that confirmed cases were not an accurate indicator of the severity of an outbreak, and some recommended counting confirmed deaths instead. It has subsequently become clear that confirmed deaths are also not a particularly accurate indicator. Early on, there weren't enough deaths to be statistically significant, and of course deaths lag infections by a variable number of days typically in the weeks. Now however we have the opposite problem, in that even though there are enough people dying to be able to treat the process statistically, there are too many for authorities to accurately confirm. When someone dies in a nursing home, there isn't anything more that you can do for them. Papers are signed, and attention turns to the survivors. Nursing homes, even poorly run ones, are well aware of the risks of contagious respiratory diseases, and are already doing everything that they can to contain and minimize outbreaks. Testing the dead so that the living know how concerned they should be is a luxury.

In the interim, those countries that actively track all-cause mortality statistics have some sense of how bad things are. These too though will have to be gone over in the end to distinguish a presumed reduction in motor vehicle fatalities and other once common causes of death, and a possible increase in fatalities from medical conditions that might not be receiving usual care, from increased deaths due directly to respiratory infections. The Spanish ministry of health publishes statistics (see Momo Spain link in the Google Sheet) on all-cause mortality daily. As usual, reporting delays create an artifact in the data making it look as though deaths are steeply declining in recent days, but these numbers should increase as reports make their way to Madrid. If we look back to the last time that deaths looked stable, in the week ending on April 4th, all-cause mortality was 16,341, expected deaths were 7,794.5, and confirmed COVID-19 deaths were 5,965. This leaves an unaccounted for excess mortality of 2,581.5 during that week (43% of the reported COVID-19 deaths), not considering the presumed other changes to the normal patterns of mortality. So as in England either COVID-19 deaths are being undercounted by at least 43%, or there is something killing Spaniards that is at least 43% as deadly as COVID-19.

Where patient needs have not yet surpassed hospital capacity, hospitalisations can be a good indicator of the progress of an epidemic. They're certainly a good indicator of a community's capacity to cope with an epidemic. Today's graphic is based on official reports from the City of Toronto, and shows on a logarithmic scale the number of COVID-19 hospitalisations by day, and how many of those are in ICU beds. Looking at the chart, there is reason to be optimistic that we may have reached the top of the curve over the last five days. Or what New York Mayor Cuomo calls "halftime". If we can bring the situation under control with current social distancing measures, it should not take too much longer before we can gradually start reopening things.

Singapore's cases have dropped barely back under 10% again to 9.95%, after a week in which they rose almost up to 20%. In absolute numeric terms, their 596 new cases today is the median daily number over the past week.

New York's new confirmed cases were under 3% for the first time since the start of the pandemic.

Alabama, which reported an unusual sub-1% increase in new cases yesterday, is back up to its more typical 4% today. Still, it's trending steadily downward.

France has had two days out of three now with less than 1% new cases.

Malta had only one new case today. Not much room for improvement!

Aside from Singapore (and I continue to trust that their situation is rapidly being brought under control), every region that I'm monitoring is doing okay to well. No region is showing anything more than a slight upward trend in the four-day average daily new case percentage; most are decreasing. Two thirds of the regions had fewer cases in absolute numbers than yesterday; the ones that still showed increases were: Alabama, India, Indonesia, Israel, Malaysia, Mexico, Ontario, Poland, South Korea, Spain, Turkey, United Kingdom, United States.

Stay safe.

2020-04-21

Comment on this post on Facebook.

today's daily chart as described below in text

Today's numbers: New York (-9→-11), Singapore (13→12), Louisiana (30→35), U.K. (35→38), U.S. (38→38), Toronto (34→42), Turkey (44→44), Ontario (47→49), Canada (46→49), Spain (43→49), California (61→53), Alabama (69→61), Mexico (60→61), Brazil (55→67), Sweden (62→70), Italy (69→74), India (89→87), Romania (83→89), Poland (77→90), South Africa (88→94), Denmark (92→95), France (136→98), Japan (98→98), Israel (108→114), Germany (94→116), Indonesia (118→128), Iran (135→142), Malta (216→199), Norway (183→224), Malaysia (356→378), Austria (348→413), Thailand (502→546), New Zealand (556→694), Australia (627→780), South Korea (3439→4031), China (25855→24738)

Compared to yesterday, here's how these regions are doing.

Much worse: California, Alabama, France, Malta, China

Worse: India

The same: Singapore

Better: U.S., Turkey, Ontario, Canada, Mexico, Italy, Romania, South Africa, Denmark, Japan, Israel, Indonesia, Iran, Malaysia, Thailand

Much better: New York, Louisiana, U.K., Toronto, Spain, Brazil, Sweden, Poland, Germany, Norway, Austria, New Zealand, Australia, South Korea

I've added the Johns Hopkins University data for all African countries today, as shown in today's graphic. That's Djibouti on the right, at 7% daily new cases and 0.1% infection. Sudan at the top left, streaking upward at 34% from 1 in 400,000. Fading into the top right is Eswatini at 18% from 1 in 37,000, crossing paths in four or five days with Somalia at 25% from 1 in 56,000. At the bottom left are Burundi and South Sudan, with a handful of confirmed cases each but no new ones in a week.

Julian West and Russell Honeybun (and before them several people including I think Jonathan Haruni) asked why the series in yesterday's graph looked bimodal, Bactrian, or two-humped. I've spent some time looking into this, and believe that it is just apophenia or pareidolia.

These graphs show the four-day average of new case rates, so a smoothed version of the new case rate statistic. Because they are smoothed, any bumps that survive are noticeable; this is the point of smoothing, and I use it here because a bump in this graph means that something went wrong and was then fixed. Some graphs have only one prominent bump: Italy and Poland. Some graphs are just a mess of bumps: Sweden, United Kingdom. And some graphs have two prominent bumps, about which more below. If a curve has one bump, we can comfortably match it against our memories of ballistic arcs. If a curve has a lot of bumps, we can mentally throw it away for being messy. If it has just two bumps though, it piques our hopeful attention. Here are some examples that I've taken a closer look at.

Canada had a peak on March 17 (31%), a trough on March 22 (20%) and a peak March 26 (29%), and no later extrema. This appears to be the result of reporting methodological change in Quebec on March 23rd, when the province began accepting hospital tests as evidence. So here, what went wrong was just that the authorities had to change their measuring criteria midstream.

India: March 23 (26%), March 29 (14%), April 4 (22%). This appears to be due to two factors. Testing was extremely limited until March 21st, resulting in a wave of backlogged cases that were confirmed rapidly when private labs came online. In the week before and after that date, social distancing measures were enacted, which eventually caused a genuine downturn in the graph.

Australia: March 4 (19%), March 8 (11%), March 16 (25%). This was because of a sudden bump on March 4th due to the repatriation of passengers from the cruise ship, Diamond Princess, who accounted for all but one of the new cases that day.

Malaysia: March 6 (30%), March 10 (12%), March 17 (36%). There was a large cluster of cases associated with a religious event that took place on March 11th.

Spain: February 28 (83%), March 7 (26%), March 11 (52%). There was a cluster of cases early on associated with funerals in Basque country, that once identified and isolated temporarily brought the rate under control.

United States: March 4-8 (43-47%), March 17 (27%), March 21 (41%). I think what happened here is that the United States identified the early threat of travel, and took measures to control it, but didn't respond quickly enough to the later threat of domestic contagion.

So I think broadly speaking, we can categorize prominent bumps in these graphs into methodological changes (including new categories of cases, or suddenly increasing testing), and comparatively serious outbreaks that are brought under control. In the methodological case, the bump is an artifact, and subsides once the new cases no longer contribute to recent percentage increases. In the case of outbreaks, in order for a bump to actually look like a bump and not a suddenly steep climb, there needs to be an adequate response. This happens more often early on, when total case numbers are low, and patients who come from a cruise ship or funeral are few enough to be counted and contact-traced. To have a downward turn later on requires either serious social distancing efforts, or the development of herd immunity in surviving members of the population.

2020-04-20

Comment on this post on Facebook.

today's daily chart as described below in text

Today's numbers: New York (-7→-9), Singapore (15→13), Louisiana (31→30), Toronto (30→34), U.K. (34→35), U.S. (32→38), Spain (36→43), Turkey (42→44), Canada (47→46), Ontario (47→47), Brazil (51→55), Mexico (58→60), California (48→61), Sweden (54→62), Italy (61→69), Alabama (55→69), Poland (73→77), Romania (64→83), South Africa (90→88), India (90→89), Denmark (82→92), Germany (83→94), Japan (78→98), Israel (100→108), Indonesia (98→118), Iran (126→135), France (122→136), Norway (166→183), Malta (144→216), Austria (281→348), Malaysia (269→356), Thailand (489→502), New Zealand (511→556), Australia (594→627), South Korea (2703→3439), China (4482→25855)

Compared to yesterday, here's how these regions are doing.

Worse: Singapore, South Africa

The same: Louisiana, Canada, India

Better: U.K., Turkey, Ontario, Brazil, Mexico, Poland, Israel, Iran, Thailand, New Zealand, Australia

Much better: New York, Toronto, U.S., Spain, California, Sweden, Italy, Alabama, Romania, Denmark, Germany, Japan, Indonesia, France, Norway, Malta, Austria, Malaysia, South Korea, China

Today's graphic is a cautionary tale showing the history to date of Canada, Japan, New York (state), and the United States. Canada, New York and the United States have reached about the level that Japan reached a month ago, when it looked like it was on its way to defeating COVID-19 just by maintaining a polite Asian distance from it. Then things went wrong, and it's take a month to bring the new case rate back down to the same neighbourhood, during which time the total number of cases has increased by an order of magnitude, from 1 in 100,000 to 1 in 10,000. This is not the time to relax social distancing measures.

Singapore continues to race toward 1%, with 1,426 new cases today after only 596 yesterday. As before, the vast majority (1,369, or 96%) of the new cases are from foreign worker dormitories, and most of the new cases are asymptomatic ones who were required to submit to testing while being isolated after potential contact with other cases. The number of community transmission cases outside of interactions with foreign workers is reportedly only about 20 per day.

Toronto hospitalisations increased slightly, from 241 to 262, after remaining steady for several days. Intubations were up one from 81 to 82. New cases overall are down for the third day in a row, and at 136 almost half of the 264 that we saw four days ago. Also showing a similar steady decrease in new cases: Austria, China, Denmark, Iran, Italy, New York State, Romania.

Ontario current hospitalised and intubated cases are down slightly for the third day in a row, and everyone is continuing to talk about how we may be reaching the top of the curve. (Unless of course they're talking about senseless gun violence or oil futures.)

My aunts' nursing home reported their first confirmed case of COVID-19 today, and says that they will switch to an even stricter protocol. I'm not sure what that entails: maybe keeping everyone in their rooms as much as possible? The case was a staff member, which is probably worse than if it were a resident, as staff members have more contact with other residents. My aunt told me today that there were some very sick people in the lunchroom today. This makes 78 institutions in Toronto that have confirmed a case of COVID-19.

As many media have reported, nursing homes here are not designed to contain diseases like COVID-19. My mom's nursing home had private bathrooms, but my aunts' does not, so anyone who uses a toilet (which admittedly is a much smaller proportion than in the general population) has to share a communal bathroom with everyone else in their wing. They also don't have the staffing levels to safely care for residents when they are all in their individual rooms.

New York State's confirmed cases increased by less than 2% for the first time today. California's was barely above 2%, at 2.13%. Mexico saw its first day in a week in which the number of new cases didn't increase. So good news all around, but no time to relax. The parachute has slowed our rate of descent, but we shouldn't take it off yet.

Stay safe.

2020-04-19

Comment on this post on Facebook.

today's daily chart as described below in text

Today's numbers: New York (-5→-7), Singapore (15→15), Toronto (32→30), Louisiana (29→31), U.S. (39→32), U.K. (36→34), Spain (32→36), Turkey (40→42), Canada (46→47), Ontario (47→47), California (49→48), Brazil (44→51), Sweden (51→54), Alabama (50→55), Mexico (63→58), Italy (62→61), Romania (62→64), Poland (78→73), Japan (78→78), Denmark (78→82), Germany (82→83), India (99→90), South Africa (92→90), Indonesia (95→98), Israel (86→100), France (93→122), Iran (121→126), Malta (122→144), Norway (134→166), Malaysia (265→269), Austria (219→281), Thailand (493→489), New Zealand (440→511), Australia (569→594), South Korea (2352→2703), China (4322→4482)

Compared to yesterday, here's how these regions are doing.

Much worse: U.S., Mexico, Poland, India, Thailand

Worse: Toronto, U.K., South Africa

The same: California, Italy

Better: Singapore, Turkey, Canada, Ontario, Sweden, Romania, Japan, Denmark, Germany, Indonesia, Iran, Malaysia, Australia, China

Much better: New York, Louisiana, Spain, Brazil, Alabama, Israel, France, Malta, Norway, Austria, New Zealand, South Korea

U.S. numbers are as usual preliminary, based on incomplete state reports.

The official Spanish death statistics have taken an unusual turn, with 20,590 being confirmed dead of COVID-19 today and 20,639 having been reported dead yesterday. No, this is not the long-feared start of the apocalyptic zombie phase of the pandemic, but it's still something quite serious. National statistics were being compiled by the ministry of health based on reports from autonomous communities (the equivalent here of provinces or states), and one of them (Catalonia) had recently started including nursing home deaths in its statistics, contrary to national policy. When Catalonian nursing home deaths were subtracted, this resulted in a net decrease in the cumulative number of deaths to date.

Early on, there was a lot of talk about how testing issues meant that confirmed cases were not an accurate indicator of the severity of an outbreak, and some recommended counting confirmed deaths instead. It has subsequently become clear that confirmed deaths are also not a particularly accurate indicator. Early on, there weren't enough deaths to be statistically significant, and of course deaths lag infections by a variable number of days typically in the weeks. Now however we have the opposite problem, in that even though there are enough people dying to be able to treat the process statistically, there are too many for authorities to accurately confirm. When someone dies in a nursing home, there isn't anything more that you can do for them. Papers are signed, and attention turns to the survivors. Nursing homes, even poorly run ones, are well aware of the risks of contagious respiratory diseases, and are already doing everything that they can to contain and minimize outbreaks. Testing the dead so that the living know how concerned they should be is a luxury.

In the interim, those countries that actively track all-cause mortality statistics have some sense of how bad things are. These too though will have to be gone over in the end to distinguish a presumed reduction in motor vehicle fatalities and other once common causes of death, and a possible increase in fatalities from medical conditions that might not be receiving usual care, from increased deaths due directly to respiratory infections. The Spanish ministry of health publishes statistics (see Momo Spain link in the Google Sheet) on all-cause mortality daily. As usual, reporting delays create an artifact in the data making it look as though deaths are steeply declining in recent days, but these numbers should increase as reports make their way to Madrid. If we look back to the last time that deaths looked stable, in the week ending on April 4th, all-cause mortality was 16,341, expected deaths were 7,794.5, and confirmed COVID-19 deaths were 5,965. This leaves an unaccounted for excess mortality of 2,581.5 during that week (43% of the reported COVID-19 deaths), not considering the presumed other changes to the normal patterns of mortality. So as in England either COVID-19 deaths are being undercounted by at least 43%, or there is something killing Spaniards that is at least 43% as deadly as COVID-19.

Where patient needs have not yet surpassed hospital capacity, hospitalisations can be a good indicator of the progress of an epidemic. They're certainly a good indicator of a community's capacity to cope with an epidemic. Today's graphic is based on official reports from the City of Toronto, and shows on a logarithmic scale the number of COVID-19 hospitalisations by day, and how many of those are in ICU beds. Looking at the chart, there is reason to be optimistic that we may have reached the top of the curve over the last five days. Or what New York Mayor Cuomo calls "halftime". If we can bring the situation under control with current social distancing measures, it should not take too much longer before we can gradually start reopening things.

Singapore's cases have dropped barely back under 10% again to 9.95%, after a week in which they rose almost up to 20%. In absolute numeric terms, their 596 new cases today is the median daily number over the past week.

New York's new confirmed cases were under 3% for the first time since the start of the pandemic.

Alabama, which reported an unusual sub-1% increase in new cases yesterday, is back up to its more typical 4% today. Still, it's trending steadily downward.

France has had two days out of three now with less than 1% new cases.

Malta had only one new case today. Not much room for improvement!

Aside from Singapore (and I continue to trust that their situation is rapidly being brought under control), every region that I'm monitoring is doing okay to well. No region is showing anything more than a slight upward trend in the four-day average daily new case percentage; most are decreasing. Two thirds of the regions had fewer cases in absolute numbers than yesterday; the ones that still showed increases were: Alabama, India, Indonesia, Israel, Malaysia, Mexico, Ontario, Poland, South Korea, Spain, Turkey, United Kingdom, United States.

Stay safe.

2020-04-18

Comment on this post on Facebook.

today's daily chart as described below in text

Today's numbers: New York (-4→-5), Singapore (18→15), Louisiana (29→29), Toronto (32→32), Spain (31→32), U.K. (35→36), U.S. (33→39), Turkey (38→40), Brazil (43→44), Canada (48→46), Ontario (45→47), California (51→49), Alabama (46→50), Sweden (54→51), Italy (64→62), Romania (65→62), Mexico (66→63), Japan (84→78), Poland (81→78), Denmark (74→78), Germany (83→82), Israel (80→86), South Africa (106→92), France (67→93), Indonesia (94→95), India (96→99), Iran (116→121), Malta (104→122), Norway (148→134), Austria (188→219), Malaysia (191→265), New Zealand (351→440), Thailand (484→493), Australia (570→569), South Korea (1847→2352), China (4014→4322)

Compared to yesterday, here's how these regions are doing.

Much worse: Singapore, Japan, South Africa, Norway

Worse: Canada, California, Sweden, Italy, Romania, Mexico, Poland

The same: New York, Germany, Australia

Better: Louisiana, Toronto, Spain, U.K., Turkey, Brazil, Ontario, Denmark, Israel, Indonesia, India, Iran, Thailand, China

Much better: U.S., Alabama, France, Malta, Austria, Malaysia, New Zealand, South Korea

U.S. numbers are estimated low, as we are still waiting for one region to report as of when I type this.

I've added Johns Hopkins University data for Asia to the website, and use it to illustrate today's post. Qatar, followed by Israel, have the highest per capita rates of infection, but Israel's outbreak is mostly under control, whereas Qatar's are still rapidly increasing. Singapore could overtake Qatar in six days, if they do not figure out what to do with foreign worker dormitories; I am reasonably optimistic that they will manage to test and isolate as needed, but additional testing will push their numbers up until they identify all cases. Singapore's 4-day average new case percentages have however increased in four of the past five days, including today.

There are also some smaller countries where any number of new cases is a rapid percentage increase; in this category I include Papua-New Guinea, Maldives, and Timor-Leste.

I have made a technical change to the rendering of all the little flags. When a lot of data is on the graph, the rendering was taking long enough to make Chrome complain about its impact on user experience. Each country is now rendered separately, so in a complicated graph in a slow browser you may see them appearing one after another.

Japan is not actually too much worse than yesterday; their daily new case numbers have levelled off again at 6-7%. About 5-10 people are dying each day, sometimes more, sometimes less. It's the sort of situation where people can get used to what's going on and stop taking things seriously enough. I noticed here in Toronto that it seemed as though there were a lot more people out running errands today than in past days. Maybe it's because it was a nice day, or a Saturday, but it was the first time in weeks that I checked for traffic using the Waze app. At least officials here (and I think in Japan) continue to treat the threat seriously, and aren't talking about relaxing restrictions as they are in many American states.

South Africa had its worst percentage increase in confirmed cases since March 27th, when they saw 9% more cases today after a long run of days in the 3-7% range. A reported cause of at least half of the new cases is an outbreak in the Eastern Cape.

Regions that have improving their 4-day average new case percentages daily for more than a week: Iran, New York, Turkey, South Korea, New Zealand.

Regions that have a less than 1% raw new case percentage, and how long they have maintained this level: Malta (1 day), New Zealand (2 days), Austria (3 days), Australia (7 days), South Korea (since April 1), China (since February 21). Sweden's data would put them on this list too, except that I know that they are using episode dates rather than report dates, so their most recent number is always a small increment. Alabama likewise looks like some sort of glitch: I'd like to believe that they've gone in successive days from 6% to 9% to 3% to less than 1%, but that last seems a little unlikely.

Stay safe.

2020-04-17

Comment on this post on Facebook.

today's daily chart as described below in text

Today's numbers: New York (-3→-4), Singapore (18→18), Louisiana (32→29), Spain (36→31), Toronto (38→32), U.S. (35→33), U.K. (37→35), Turkey (37→38), Brazil (53→43), Ontario (47→45), Alabama (41→46), Canada (53→48), California (59→51), Sweden (59→54), Italy (66→64), Romania (64→65), Mexico (71→66), France (55→67), Denmark (79→74), Israel (60→80), Poland (91→81), Germany (103→83), Japan (84→84), Indonesia (94→94), India (77→96), Malta (115→104), South Africa (103→106), Iran (113→116), Norway (164→148), Austria (186→188), Malaysia (163→191), New Zealand (325→351), Thailand (479→484), Australia (600→570), South Korea (1675→1847), China (8205→4014)

Compared to yesterday, here's how these regions are doing.

Much worse: Spain, Toronto, Brazil, Canada, California, Sweden, Mexico, Denmark, Poland, Germany, Malta, Norway, Australia, China

Worse: Louisiana, U.S., U.K., Ontario, Italy

Better: New York, Singapore, Turkey, Romania, Japan, Indonesia, South Africa, Iran, Austria, New Zealand, Thailand

Much better: Alabama, France, Israel, India, Malaysia, South Korea

A significant number of late-reporting state cases would have pushed the U.S. number yesterday from 35 to 34, had they arrived in time.

Swedish figures have been revised back to March 26th (deaths) and March 30th (cases). Apparently Swedish Wikipedia editors are following their public health authorities in trying to report cases by episode date rather than by report date, which differs from just about every other region in the world. This generally has the effect of making it continually seem as though the epidemic is plateauing, because almost no one tests positive the day after they first contract the illness.

The CDC shifted its counting methodology on April 14th to combine confirmed and probable cases, rather than reporting only confirmed cases. It's taking a while for states to adopt the CDC's recommendations in this respect, so if you see an unusual bump in a state that has recently been doing well, there's a good chance that this is all it is. Confirmed cases are those where an actual laboratory test has confirmed that a patient is infected with the coronavirus responsible for COVID-19; a probable case has a technical definition but in the end comes down to someone's best judgement in the possible absent of lab tests. Toronto switched over to the more inclusive system on April 10th; New York City (which I do not track) has been using it for a while, but New York State (which I do track) has not yet adopted it.

New York State reports a slight downturn in total hospitalizations, as well as patients in ICU beds and being intubated; there are still about 2,000 new cases being diagnosed daily. I got them from watching a little of Gov. Cuomo's daily briefing, to get the correct numbers after noticing that whoever updated the Wikipedia page today was using egregiously inconsistent numbers; it's possible that they added in unofficial probable numbers. Cuomo has a reassuring stage presence.

Toronto's numbers are somewhat worse today. 264 new cases compared to 211 (confirmed and probable) the day before, a continuing worsening trend in the percentage of cases found in institutionalised patients, up now to 7.8%.

Brazil jumped from 2105 new cases yesterday to 3796 today. Spain from 4289 to 5891. Canada 1316 to 1727. (If we look at the situation in Canada by province though, there are provinces where the daily new case rate has been brought rapidly down to or below 1%, such as Manitoba, Saskatchewan, New Brunswick, Prince Edward Island and Newfoundland and Labrador.)

Stay safe.

http://www.poslfit.com/covid19

2020-04-16

Comment on this post on Facebook.

today's daily chart as described below in text

Today's numbers: New York (-2→-3), Singapore (23→18), Louisiana (33→32), U.S. (35→35), Spain (37→36), Turkey (35→37), U.K. (35→37), Toronto (38→38), Alabama (41→41), Ontario (48→47), Canada (53→53), Brazil (55→53), France (59→55), California (66→59), Sweden (65→59), Israel (54→60), Romania (70→64), Italy (64→66), Mexico (66→71), India (73→77), Denmark (80→79), Japan (83→84), Poland (89→91), Indonesia (87→94), South Africa (103→103), Germany (103→103), Iran (110→113), Malta (133→115), Malaysia (147→163), Norway (164→164), Austria (190→186), New Zealand (272→325), Thailand (459→479), Australia (599→600), South Korea (1622→1675), China (5873→8205)

Compared to yesterday, here's how these regions are doing.

Much worse: Singapore, France, California, Sweden, Romania, Malta, Austria

Worse: Brazil

The same: New York, Louisiana, Spain, Ontario, Denmark

Better: U.S., Turkey, U.K., Toronto, Alabama, Canada, Italy, Mexico, India, Japan, Poland, Indonesia, South Africa, Germany, Iran, Norway, Thailand, Australia, South Korea

Much better: Israel, Malaysia, New Zealand, China

I was wondering how the fact that people are generally staying at home and not engaging in dangerous behaviour like driving or attacking strangers might offset the additional death rate due to COVID-19. Public Health England is estimating that in the week ending on April 10th ("week 15"), deaths in England and Wales from all causes would be somewhat in excess of double the seasonal average. For the preceding week, they reported that there were 16,387 deaths, which was 6,082 more than average. Of these deaths, 3,475 were attributed wholly or in part to COVID-19, which suggests to me that COVID-19 deaths are either being underreported by 75%, or there is something that is 75% as dangerous as COVID-19 that is going around killing the English.

Euromomo says that for the week ending on April 5th, 20+ nations had a collective average of about 35% in excess of average, but with huge national variation. Spain is worst off by far, followed by Belgium, France, Italy and the Netherlands. Their graphing software doesn't seem to be designed for this situation, but as far as I can tell, Spain is more than 30 standard deviations above the mean, and the other four only about 20 standard deviations.

Various friends in Japan have told me that the situation is very different from here: there is no lockdown mentality, and aside from kids staying home from school and hospitals slowly seeing more COVID-19 patients, people aren't doing much to change their social behaviour. To be fair, that behaviour was a lot less touchy-feely to begin with than it is here, and a lot more hygiene conscious, but it doesn't seem to be working well.

U.S. President Donald Trump's suggestion that the Canadian border would be "one of the early borders to be released" met with mixed reactions here in Canada. Prime Minister Justin Trudeau politely responded that there is still "a significant amount of time" before non-commercial cross-border traffic can resume; Ontario Premier Doug Ford said more bluntly, "[...] absolutely not! I don't want [Americans] in Ontario." I think we're a long way before we can risk reopening our international borders again; we have to get to the point where interprovincial borders can reopen first.

I have some concerns about the quality of the Swedish data, which frequently sees changes being backdated several days. This is reportedly due to backdating of official data posted by the Swedish public health agency. France

California saw a huge uptick in new cases today, with a record 1,758 new cases, after hovering around 1,000-1,100 most recent days.

I am worried for friends in Singapore, where the government's early efforts at contact tracing and isolation lived up to their reputation for efficiency; now though, they appera to be on the verge of losing control of the situation. 728 new cases today is almost as much as the previous two days combined, and even if 654 are from foreign worker dormitories, it must be putting a strain on healthcare resources. I did not expect to see them in second place on my list now.

Singapore is now the only region on my list with a 4-day average new case rate above 10% (14.99%). Next highest is India (9%), then Brazil and Mexico (8%); kudos to Australia and Austria for staying below 1%.

2020-04-15

Comment on this post on Facebook.

today's daily chart as described below in text

Today's numbers: New York (-1→-2), Singapore (26→23), Louisiana (28→33), Turkey (31→35), U.S. (30→35), U.K. (27→35), Spain (41→37), Toronto (33→38), Alabama (40→41), Ontario (48→48), Canada (53→53), Israel (55→54), Brazil (70→55), France (56→59), Italy (55→64), Sweden (67→65), Mexico (64→66), California (63→66), Romania (58→70), India (61→73), Denmark (78→80), Japan (68→83), Indonesia (79→87), Poland (83→89), Germany (83→103), South Africa (118→103), Iran (103→110), Malta (87→133), Malaysia (121→147), Norway (175→164), Austria (153→190), New Zealand (259→272), Thailand (405→459), Australia (467→599), South Korea (1473→1622), China (4959→5873)

Compared to yesterday, here's how these regions are doing.

Much worse: New York (but see below), Spain, Brazil, South Africa, Norway

Worse: Singapore, Sweden

The same: Israel

Better: Alabama, Ontario, Canada, France, Mexico, California, Denmark, Poland, Iran, New Zealand

Much better: Louisiana, Turkey, U.S., U.K., Toronto, Italy, Romania, India, Japan, Indonesia, Germany, Malta, Malaysia, Austria, Thailand, Australia, South Korea, China

I added JHU data for Europe to the website, and today's graphic shows all European countries and their seven-day trends. Belarus has the highest daily new case rate (14%), and the smallest countries such as San Marino, the Holy See, and Andorra not surprisingly have the highest per capita case counts. Luxembourg and Iceland are next, but Liechtenstein and Malta are seeing the case numbers more typical of larger countries. The larger countries that look worst off are Spain, Belgium and Ireland. Lowest confirmed case count is in Kosovo, followed by Georgia, Ukraine and Bulgaria.

I may need to rethink the negative metric above for New York, as it's not actually too much worse, seeing more of a continuing very gradual lessening in the rate at which it is worsening. The thing though is that when the new case rate eventually starts dropping significantly, it will look the 1% day was in the more distant past than it actually was. Hmm.

Spain did actually have a terrible day, with 6,599 cases after three days below 4,000. Brazil likewise leapt from several days below 2,000 up to 3,058, breaking 3,000 for the first time.

Toronto is considering closing some major downtown streets to cars, because no one drives on them anymore, and making pedestrian traffic flow one-way. Two more nursing-home outbreaks for a total of 71. On average in Ontario, one nursing home outbreak consists of 13 cases, and results in 1.5 fatalities. Sporadic case sources shifting slightly toward close contact (34 to 36%) and away from travel, community and institutional.

Ontario announced today that nursing home staff would be banned from working at multiple institutions for the next 14 days. The government continues to face criticism for past reduction of funding of public healthcare (including long-term care), and for its inopportune reduction in nursing home inspections when it came to power two years ago.

Canada lost its 1,000th patient to the pandemic today, but the chief public health officer is reminding everyone correctly that the new case rate has come way down and continues to drop, hovering now around 5%. The federal government announced that emergency benefits would now be available to those who are earning up to $1,000 per month (up from $0), as well as to seasonal workers and those who had recently become unemployed when the pandemic hit. It is also negotiating with provincial governments to top up essential worker salaries to ensure adequate staffing levels under restricted working conditions.

Stay safe.

2020-04-14

Today's numbers: New York (0→-1), Singapore (28→26), U.K. (26→27), Louisiana (23→28), U.S. (30→30), Turkey (29→31), Toronto (23→33), Alabama (33→40), Spain (39→41), Ontario (35→48), Canada (51→53), Israel (57→55), Italy (51→55), France (60→56), Romania (55→58), India (67→61), California (58→63), Mexico (59→64), Sweden (68→67), Japan (45→68), Brazil (55→70), Denmark (78→78), Indonesia (79→79), Poland (73→83), Germany (72→83), Malta (78→87), Iran (94→103), South Africa (138→118), Malaysia (129→121), Austria (117→153), Norway (133→175), New Zealand (227→259), Thailand (358→405), Australia (356→467), South Korea (1431→1473), China (4956→4959)

Compared to yesterday, here's how these regions are doing.

Much worse: France, India, South Africa, Malaysia

Worse: Singapore, Israel

The same: New York, Sweden

Better: U.K., U.S. (but see below), Spain, Canada, Italy, Romania, Denmark, Indonesia, South Korea, China

Much better: Louisiana, Turkey, Toronto, Alabama, Ontario, California, Mexico, Japan, Brazil, Poland, Germany, Malta, Iran, Austria, Norway, New Zealand, Thailand, Australia

I have added another data source, showing all the countries in the Americas (except for Canada, due to a technical issue to be corrected later) based on the Johns Hopkins University data. At a glance, Panama is racing to become the next United States. Ecuador looks next worse, with fewer infections but a greater rate of transmission trending upward. Belize techically has the highest new case rate, but it's based on a tiny amount of data. Likewise, Nicaragua has the lowest proportion of confirmed cases, based on a miniscule amount of data. Most major countries seem currently on track to bringing things under control at around the 1 in 10,000 mark, but we'll see.

I also added controls yesterday to adjust the X and Y scales on the graph. I'll work on arranging the controls more tidily later.

New York is now at -1. That is, it's been a day since it reached the 1% target that it should never have reached. Its 4-day average has just dropped below 4% for the first time, and the end might be approaching. Daily hospital deaths continue apace in the 700s daily as they have for seven of the last eight days.

France had 5,497 new confirmed cases outside of nursing homes today, more than double yesterday's 2,673 or triple the day before's 1,613. India had more than 1,000 new confirmed cases for the first time today. So, both much worse. South Africa and Malaysia were not quite as bad.

The death count in Japan has been revised downward dating back to April 4th, reducing yesterday's 143 to a 109. The stated reason is to bring the numbers in line with WHO statistics. I checked the Japanese Ministry of Health announcement, which agrees with the new lower value. The larger numbers are appearing in reports from the public broadcaster, NHK.

Germany showed a significant dip yesterday (more so in the RKI data that I use for historical reasons than in the JHU data), and a smaller rebound today. Averaged out over four-day periods, it just shows a gradual descent below the 3% threshold.

The CBC is reporting that the number of COVID-19 patients in Ontario ICUs has been stable at slightly above 250 for the last six days, and is far below what models predicted for this point, crediting effective compliance with social distancing measures. Ontario's new case numbers were in the 400s for the fourth day, but at 483 the highest during that period. In percentages, it's a pretty steady 6% growth. Premier Ford announced that schools would not reopen on May 4th, but assured students that their academic year would not be cancelled. Students are currently working online where possible (with school boards scrambling to ensure connectivity and laptops for all students), and have been told that work done online during these times can raise but not lower their final marks.

Toronto's institutional outbreaks are increasing; we are at 69 now, and the proportion of confirmed cases is up from 4.9% yesterday to 5.2% today. Toronto's 4-day average into the single digits at 7.7% for the first time since April 4th. Let's hope for no more rebounds.

The U.S. totals are late being released today, and were not available when I calculated the above numbers. [Postscript: the new numbers came in while I was posting this, and show the U.S. improving from 30 days to 34 days.]

Regions whose 4-day average has steadily increased for the last ten days or more: Canada, Australia, Spain, Denmark, Iran, New York State, Turkey, South Korea, New Zealand, California.

Stay safe.

http://www.poslfi.com/covid19

2020-04-13

Today's numbers: New York (0→0), Toronto (20→23), Louisiana (17→23), U.K. (25→26), Singapore (28→28), Turkey (27→29), U.S. (27→30), Alabama (28→33), Ontario (31→35), Spain (35→39), Japan (45→45), Italy (47→51), Canada (47→51), Brazil (55→55), Romania (49→55), Israel (61→57), California (46→58), Mexico (59→59), France (52→60), India (66→67), Sweden (41→68), Germany (57→72), Poland (65→73), Malta (44→78), Denmark (67→78), Indonesia (72→79), Iran (94→94), Austria (96→117), Malaysia (132→129), Norway (164→133), South Africa (137→138), New Zealand (168→227), Australia (312→356), Thailand (303→358), South Korea (1429→1431), China (5743→4956)

Compared to yesterday, here's how these regions are doing.

Much worse: Israel, Norway, China

Worse: Malaysia

Better: New York, U.K., Singapore, Japan, Brazil, Mexico, India, Iran, South Africa, South Korea

Much better: Toronto, Louisiana, Turkey, U.S., Alabama, Ontario, Spain, Italy, Canada, Romania, California, France, Sweden, Germany, Poland, Malta, Denmark, Indonesia, Austria, New Zealand, Australia, Thailand

New York State is still hovering at the 1% mark in confirmed cases, and they had their lowest gain since March 25th, at 6,337 new cases. Also their lowest new deaths in a week, at 671. I'm hoping that's the start of a trend and not just random noise; the daily deaths had been in the 700s for a week. The new death percentage rate did drop daily for a week, down to 7% from 15%; but this is because the base on which they are calculated - the cumulative death count - gradually doubled during that timespan.

U.S. new cases dropped below 25,000 for the first time in more than a week.

I have added a new data source labelled JHU Canada. It gives a breakdown of the situation in Canada by province, using the acclaimed Johns Hopkins University database. Thanks to David Boys for encouraging me to include other provinces and territories. And for those with a fine eye for detail, the territory of Nunavut is omitted because there are no confirmed cases there. Let's keep it that way.

Japan saw its first double-digit death day, with 21 deaths reported. The new case count of 622 for the day was down though for the first time in several days, hence the "Better" rating.

Israel has been seeing 300-400 new cases a day or about 3.5% new cases daily; today's was somewhat worse than the recent trend, which had been overall steady for a week.

Norway had a day four days ago when they reported only two new cases rather than their typical 100 or so. That's working its way out of the average today, causing a "Much worse".

China's numbers are gradually starting to drift upward again. Ten days ago, they were seeing a steady 30-40 cases a day, now it's up to about 100 cases a day. I wonder if they will need to shut things down again, and if so, when.

In Toronto, 5% of new cases arise in institutions, and we have about 0.5% of our population in long-term care beds. There have been several outbreaks in LTC facilities in the city, and gradually tightening restrictions on their operations.

Australia has managed a second consecutive day of sub-1% new cases, as confusingly has Austria. The only other regions reporting under 1% today are Germany (first time), China (for a long time now), South Korea (two weeks), and Sweden (first time).

Stay safe.

2020-04-12

Today's numbers: New York (1→0), Louisiana (16→17), Toronto (20→20), U.K. (24→25), U.S. (25→27), Turkey (26→27), Singapore (29→28), Alabama (29→28), Ontario (29→31), Spain (30→35), Sweden (33→41), Malta (44→44), Japan (63→45), California (45→46), Italy (48→47), Canada (42→47), Romania (46→49), France (43→52), Brazil (45→55), Germany (52→57), Mexico (55→59), Israel (59→61), Poland (60→65), India (66→66), Denmark (54→67), Indonesia (78→72), Iran (88→94), Austria (79→96), Malaysia (127→132), South Africa (153→137), Norway (131→164), New Zealand (152→168), Thailand (206→303), Australia (240→312), South Korea (1269→1429), China (6771→5743)

Compared to yesterday, here's how these regions are doing.

Much worse: Japan, Indonesia, South Africa, China

The same: New York, Singapore, Alabama, Italy

Better: Louisiana, Toronto, U.K., Turkey, Malta, California, Romania, Mexico, Israel, India, Iran, Malaysia

Much better: U.S., Ontario, Spain, Sweden, Canada, France, Brazil, Germany, Poland, Denmark, Austria, Norway, New Zealand, Thailand, Australia, South Korea

So here we are at the end of this part of the story. New York has reached a target that I didn't think anyone could: they have 0.97% of their population testing positive for COVID-19, or within rounding error of one percent. In densely populated areas, everyone is exposed, everyone is likely to get infected, and tens of thousands of people will die. Confirmed COVID-19 deaths are accumulating at a steady 800 or so per day in the state, but according to the New York Times, that number represents how many patients are dying each day in hospital following a COVID-19 diagnosis. It does not include people who have tested positive for COVID-19 but then subsequently died at home, or those who were not able to get tested. In the first eight days of April, the state recorded about 3,500 deaths in New York City, while the city counted 1,891 people found dead at home apparently of COVID-19. I imagine this practice is not unique to New York.

New reported cases in Japan are up for the sixth day in a row, going from 252 up to 743. The government there is still weighing the economic and political costs of shutdown vs. death. Japan was an early outlier in a good way, but now in a bad way, as they are second only to Singapore on my list of regions that have worsened significantly over the past four days.

Today's Toronto death numbers remain unknown, apparently due to a technical issue. Newspapers are reporting on concerns about possible spread among the homeless and institutionalized, but there do not yet seem to be numbers to indicate a current problem. Anecdotally, a homeless member of the Toronto Scrabble Club had no trouble being getting tested a few weeks ago, as a condition of entry into a shelter. And the city is publishing daily statistics on institutional cases, which are so far relatively low.

The city has been visibly enforcing physical distancing rules, and police and bylaw enforcement officers have been highly visible downtown. They report cautioning about 250 people a day (which doesn't seem like many to me), issuing an average of 5 tickets a day (again), and reportedly towed four vehicles. It is not clear to me how towing a vehicle helps fight COVID-19. The mayor says that the rate of complaints about physical distancing violations has dropped dramatically over the last few days. I've also been seeing a lot of social workers out talking to homeless people, making sure that they are doiong okay and that they understand the situation, and am grateful for what they do.

New reported cases are down for the fourth day in a row in Canada, and there is hope that the next number will be less than a thousand.

The U.S. is also down for four consecutive days, and they are heading toward 25,000 new cases a day.

Their smoothed 4-day average new rate has been trending downward since March 21st.

Unless there's been a reporting error, Australia's new case percentage dropped below 1% for the first time today; they have been trending downward since March 24th.

The most improved listed region over the past four days has been Mexico, followed closely by Denmark.

Stay safe.

2020-04-11

Today's numbers: New York (2→1), Louisiana (14→16), Toronto (20→20), U.K. (29→24), U.S. (25→25), Turkey (26→26), Alabama (27→29), Singapore (31→29), Ontario (29→29), Spain (29→30), Sweden (33→33), Canada (40→42), France (40→43), Malta (28→44), Brazil (38→45), California (39→45), Romania (48→46), Italy (53→48), Germany (47→52), Denmark (43→54), Mexico (51→55), Israel (57→59), Poland (55→60), Japan (63→63), India (68→66), Indonesia (78→78), Austria (76→79), Iran (84→88), Malaysia (126→127), Norway (94→131), New Zealand (117→152), South Africa (132→153), Thailand (209→206), Australia (219→240), South Korea (1087→1269), China (7941→6771)

Compared to yesterday, here's how these regions are doing.

Much worse: U.K., Italy, China

Worse: Singapore, Romania, India, Thailand

The same: New York

Better: Toronto, U.S., Turkey, Ontario, Spain, Sweden, Canada, Mexico, Israel, Japan, Indonesia, Austria, Iran, Malaysia

Much better: Louisiana, Alabama, France, Malta, Brazil, California, Germany, Denmark, Poland, Norway, New Zealand, South Africa, Australia, South Korea

I've changed the user interface slightly in order to add a third data possible data source: COVIDTracking.com, which offers current data on all U.S. states. Thanks to Jonathan Haruni for pointing it out to me.

There was a news story about a nursing home in Dorval, a suburb of Montreal. It's sad. Don't read it unless you think you're feeling too happy. It's especially concerning next to advice from the EU saying that the elderly will likely need to remain quarantined for a year.

In the United Kingdom, the sudden addition of an unexpected four thousand or so confirmed cases is (according to the gov.uk) the result of adding in an additional set of tests that had been going on since March 25th, but that had been omitted from the official totals. Up until now, we had only been seeing what were described as "swab testing in PHE labs and NHS hospitals for those with a medical need and the most critical workers and their families"; we also now see "swab testing for key workers and their households". In the newly added category, which appears to choose people solely on the basis of their social role and not COVID-19 symptoms, and where one would imagine that the test subjects have been more careful about contagion than an average person, 35% of those tested were positive.

In Italy, the daily new case rate has been below 4% for a week, but it has been trending slightly higher over the last few days. Since they already have 150,000+ cases, that means that on a bad day they are seeing more than 4,500 new cases, and today was a bad day.

China had its first day above 0.1% since March 23rd, which in their case means almost 100 new cases. Even in a country as tightly controlled as China, it's proving difficult to completely eradicate the virus.

New York State has about 19,440,000 people. More than 180,000 of them have tested positive so far for COVID-19.

Of the countries I'm tracking, France has the most suspiciously low confirmed case rate, about 18 times lower than their death rate suggests.

Toronto's new case rate was back down to 9% today, but it's four-day average of 14% is the highest on my list, due to the bad two previous days.

Stay safe.

http://www.poslfit.com

2020-04-10

Today's numbers: New York (3→2), Louisiana (11→14), Toronto (30→20), U.S. (25→25), Turkey (27→26), Alabama (26→27), Malta (27→28), Spain (29→29), U.K. (30→29), Ontario (36→29), Singapore (36→31), Sweden (33→33), Brazil (40→38), California (40→39), France (40→40), Canada (40→40), Denmark (37→43), Germany (52→47), Romania (49→48), Mexico (50→51), Italy (53→53), Poland (55→55), Israel (60→57), Japan (57→63), India (63→68), Austria (78→76), Indonesia (73→78), Iran (79→84), Norway (108→94), New Zealand (98→117), Malaysia (120→126), South Africa (147→132), Thailand (204→209), Australia (213→219), South Korea (973→1087), China (8497→7941)

Compared to yesterday, here's how these regions are doing.

Much worse: Toronto, Ontario, Singapore, Germany, Norway, South Africa, China

Worse: Brazil, Israel, Austria

The same: New York, Turkey, U.K., California, Romania

Better: U.S., Alabama, Malta, Spain, Sweden, France, Canada, Mexico, Italy, Poland, India, Indonesia, Iran, Malaysia, Thailand, Australia

Much better: Louisiana, Denmark, Japan, New Zealand, South Korea

I have two days left to figure out how to report New York when it reaches the 1% mark that seemed so unlikely when I started this. As I say above, it's thankfully just an arbitrary point somewhat short of where exponential growth theoretically ought to stop. In practice, it turns out that when close to 800 people a day are dying in a pandemic, it's hard to be sure you're finding and counting all the ones that were infected with the virus. Assuming they are all being found, there must be at least 10% of the population being infected at this point, and probably many more.

Even at levels of infection much lower than New York, there is systematic underreporting of COVID-19 deaths, because the usual reporting systems have not adapted to an era in which it's important to know exactly how many people died of what each day. In Ontario, when my mother passed away in December, it took me two months to get an official death certificate issued, and I had been told to expect three months. Under normal circumstances, people aren't in a big hurry when they are bereaved. The Guardian reported (h/t Tony Leah) on April 9th that in the U.K., the Office for National Statistics does not count COVID-19 deaths in nursing homes until death certificates are officially registered weeks later, resulting in an underreporting by at least an order of magnitude. As I wrote earlier, in France, where the nursing home death count is more readily available, it is still not being included in the official national total.

On a smaller scale, we had a systematic change here in Toronto, where the city had up until today been reporting confirmed and probable cases separately; I had been using the confirmed numbers. The probable cases are no longer separately identified, so there's a one-day bump that will take four days to even out.

In Japan, the NHK reported that 70% of new cases are among those younger than 50 (who are half the population), and that there is a strong suspicion of community transmission in restaurants and bars that still remain open.

In Canada, I heard a discussion of how it can be hard in a liberal democracy to suddenly impose strict controls on movement and behaviour, no matter how much they might be necessary to save lives. While I think Canadians are generally fairly obedient and law-abiding, what we are more is conformist. We like to do what everyone else is doing, because things run more smoothly that way. People ignore speed limits because it's customary to do so. They didn't used to wear seatbelts, because it wasn't customary; now it is, so we do. But getting everyone used to the idea of social distancing and isolation takes time and sustained effort, and there are still a small number of people who aren't getting the idea, and that's all it takes to keep the pandemic spreading. And because we live in a liberal democracy, it doesn't matter if the government makes it illegal to meet in groups of six or more until everyone gets the idea that this is a good thing to do, because the government does not have the resources to arrest the whole nonconforming population.

Let's see if I can find some more positive things to end on. The daily new case rate, smoothed over 4 days, has been declining steadily in Canada and Israel for 9 or 10 days; New Zealand and Turkey for 2 weeks, Australia and Spain for 2.5 weeks; Italy, New York and the U.S. for about three weeks. And the only place where that new case rate is over 15% is Toronto, and that only on a technicality. So we've come a long way since we had doubling periods of two or three days. Let's keep up the good work.

Stay safe.

2020-04-09

Today's numbers: New York (4→3), Louisiana (13→11), U.S. (23→25), Alabama (29→26), Turkey (26→27), Malta (33→27), Spain (28→29), U.K. (26→30), Toronto (26→30), Sweden (40→33), Singapore (46→36), Ontario (37→36), Denmark (34→37), Brazil (45→40), California (33→40), France (47→40), Canada (37→40), Romania (54→49), Mexico (46→50), Germany (52→52), Italy (52→53), Poland (47→55), Japan (60→57), Israel (55→60), India (58→63), Indonesia (79→73), Austria (80→78), Iran (70→79), New Zealand (83→98), Norway (94→108), Malaysia (104→120), South Africa (152→147), Thailand (166→204), Australia (150→213), South Korea (866→973), China (8624→8497)

Compared to yesterday, here's how these regions are doing.

Much worse: Malta, Sweden, Singapore, Brazil, France, Romania, Indonesia, South Africa, China

Worse: Louisiana, Alabama, Japan, Austria

The same: New York, Ontario

Better: Turkey, Spain, Germany, Italy

Much better: U.S., U.K., Toronto, Denmark, California, Canada, Mexico, Poland, Israel, India, Iran, New Zealand, Norway, Malaysia, Thailand, Australia, South Korea

Another 10,000 confirmed cases in New York State, and 800 confirmed deaths. Both numbers still rising day by day, even though they no longer test for coronavirus postmortem. The new case rate is holding steady at 7% daily. If the death total were accurate, and they had 2% fatality after two weeks, then their confirmed cases should represent only a tenth of those infected, meaning one person in 12 has it. By comparison, Wuhan ended up confirming only 1 in 221 of their population as having COVID-19.

Ontario, also categorized as "the same" above, has a month to go before it gets to this point, and every week our premier and mayors are putting into effect new measures to control spread. The one thing that I wish we had done by now or could do soon is something I think British Columbia has already done, which is to forbid staff from working in more than one nursing home. It's only a matter of time here before there are major outbreaks in our long-term care facilities, and the emailed status updates from my aunts' home are getting more and more serious. Thankfully no cases in theirs yet, but residents are confined to their rooms except for bathroom breaks, and there have been no visitors or deliveries permitted for weeks.

Almost 200 new cases in Toronto today. We're getting about 10% a day, same as Ontario, but we have more good days and bad days, because of our smaller numbers. Today was a bad day.

The daily rate has gone down in Canada for the 8th consecutive day, but is still over 7%. So we are rapidly improving, but still worsening as quickly as New York. Better than Japan now, which is hovering around 10%, and shocks me.

Malta had a big jump in their numbers, in part due to widespread random screening. That's also a good/bad news situation, with random checks turning up infected workers in nursing homes and hospitals.

Norway looks like a glitch. I can't believe they only had two new cases today; I am guessing that something broke down in the reporting system and that the missing cases will show up tomorrow.

On the other hand, the huge increase in Singapore today is reportedly genuine, and the government is stepping in to try to separate and isolate infected foreigners in dormitories.

Stay safe.

2020-04-08

Today's numbers: New York (4→4), Louisiana (9→13), U.S. (23→23), Toronto (32→26), Turkey (26→26), U.K. (27→26), Spain (27→28), Alabama (31→29), California (33→33), Malta (30→33), Denmark (33→34), Ontario (37→37), Canada (32→37), Sweden (45→40), Brazil (48→45), Singapore (51→46), Mexico (49→46), Poland (48→47), France (44→47), Italy (48→52), Germany (50→52), Romania (46→54), Israel (new→55), India (50→58), Japan (58→60), Iran (64→70), Indonesia (86→79), Austria (83→80), New Zealand (61→83), Norway (93→94), Malaysia (101→104), Australia (153→150), South Africa (155→152), Thailand (173→166), South Korea (703→866), China (10365→8624)

Compared to yesterday, here's how these regions are doing.

Much worse: Toronto, Sweden, Singapore, Indonesia, Thailand, China

Worse: Alabama, Brazil, Mexico, Austria, Australia, South Africa

The same: U.K., Poland

Better: New York, U.S., Turkey, Spain, California, Denmark, Ontario, France, Germany, Japan, Norway, Malaysia

Much better: Louisiana, Malta, Canada, Italy, Romania, India, Iran, New Zealand, South Korea

Sorry for the delay in posting these results. They are from about 18 hours ago (so last night in local time). I needed some extra sleep last night due to a family issue (now satisfactorily resolved); I should be able to post again tonight.

The premier of our province of Ontario was expressing frustration yesterday with the low rate of COVID-19 testing taking place here, and mentioned that we had been testing 2,000-3,000 daily (and detecting about 400 cases a day now for about a week and a half); he said he would like to see that number increased to about 13,000 per day. This would increase the testing rate from 0.01~0.02% of the population to 0.09% daily.

I heard on the Japanese news that the government is finally asking entertainment venues such as pachinko parlours to close. Better late than never.

The new case rate in Louisiana is dropping fastest of all the ones I'm watching. This is good news, but also a little late, in that they are still seeing about a thousand new confirmed cases daily.

Stay safe.

2020-04-07

Today's numbers: New York (5→4), Louisiana (9→9), U.S. (22→23), Turkey (26→26), U.K. (24→27), Spain (26→27), Malta (57→30), Alabama (29→31), Toronto (34→32), Canada (32→32), California (30→33), Denmark (31→33), Ontario (32→37), France (38→44), Sweden (38→45), Romania (39→46), Poland (new→48), Brazil (49→48), Italy (43→48), Mexico (52→49), Germany (46→50), India (58→50), Singapore (55→51), Japan (55→58), New Zealand (51→61), Iran (58→64), Austria (75→83), Indonesia (85→86), Norway (78→93), Malaysia (90→101), Australia (156→153), South Africa (164→155), Thailand (136→173), South Korea (592→703), China (14074→10365)

Compared to yesterday, here's how these regions are doing.

Much worse: Malta, India, Singapore, South Africa, China

Worse: Toronto, Mexico, Australia

The same: New York, Brazil

Better: Louisiana, U.S., Turkey, Spain, Canada, Denmark, Japan, Indonesia

Much better: U.K., Alabama, California, Ontario, France, Sweden, Romania, Italy, Germany, New Zealand, Iran, Austria, Norway, Malaysia, Thailand, South Korea

I've added Poland to the list, and taken Jonathan Haruni's suggestion to have the 7-day trend in the graph follow the recent trend in the new case rate rather than keeping that rate steady. I've also changed the categorization above based on Phil Willow's suggestion, so that it's now based on both difference and ratio rather than just difference.

Malta had its largest single-day increase in cases, with 52 new cases bringing them to 293. In exponential growth, one expects every day's new case number to be the largest increase to date, but this one was noteworthy insofar as the previous record was 19. The Health Minister played down its importance, saying that the low numbers in recent days were not what had been predicted, and that this new number put them back on the model track. I would perhaps have tried to play up the fact that the 52 new confirmed cases came from 825 tests, indicating that there is a good likelihood that they are catching all the cases in the country. And if they keep testing 0.2% of the population per day, they should be able to stay on top of things for a while yet.

France continues to pretend that nursing home cases are different from other cases. France also has the highest discrepancy between 2% of the two-week-old case numbers and current death totals. Those numbers should be roughly comparable, but are off by a factor of more than 20 in France. The only other region on my list over 20 in that category is Louisana.

Many of the regions are experiencing slight upticks in new cases, after encouraging downward runs. Both look like statistical noise, which is why the smoothed data is there. In many cases, slight ups and downs must just be vagaries of reporting timing.

Most regions are now well past where China ended up in per capita infections. Several are still exceeding China's worst daily records of 149 deaths in one day: Germany (173), Italy (604), Spain (704), U.K. (786), France (1417), U.S.A. (2017, of which 731 in N.Y.)

Most regions are however developing new cases more slowly now than they were four days ago. To be clear, that doesn't mean that they are doing better, but rather that things are not getting worse as quickly as they were before. The exceptions are in order of decreasing difference in new case rate: Malta, Toronto, Mexico, Japan, Singapore, South Africa. Of these, the only one where the current average new case rate is greater than 10% is Mexico at 11%.

The biggest drops (improvements) over that period of time were Louisiana (down 14% from 26% to 12%), California (down 6%), and Canada, the U.K. and Sweden (all down 5%). Keep up the good work.

Stay distant, clean and safe.

2020-04-06

Today's numbers: New York (5→5), Louisiana (7→9), U.S. (19→22), U.K. (22→24), Spain (22→26), Turkey (25→26), Alabama (28→29), California (27→30), Denmark (31→31), Ontario (28→32), Canada (27→32), Toronto (30→34), France (42→38), Sweden (33→38), Romania (new→39), Italy (40→43), Germany (34→46), Brazil (43→49), New Zealand (46→51), Mexico (58→52), Japan (75→55), Singapore (56→55), Malta (65→57), Iran (49→58), India (49→58), Austria (new→75), Norway (57→78), Indonesia (93→85), Malaysia (78→90), Thailand (114→136), Australia (77→156), South Africa (129→164), South Korea (514→592), China (14188→14074) Compared to yesterday, here's how these regions are doing. Much worse: France, Mexico, Japan, Malta, Indonesia, China The same: Singapore Better: New York, Turkey, Alabama, Denmark Much better: Louisiana, U.S., U.K., Spain, California, Ontario, Canada, Toronto, Sweden, Italy, Germany, Brazil, New Zealand, Iran, India, Norway, Malaysia, Thailand, Australia, South Africa, South Korea I have to keep this short today, as it's been a busy day with the kids back to schoolwork if not school, and an unexpectedly difficult time foraging for groceries. France appears to be alone in inexplicably categorizing nursing home cases (of which they have more than 23,000) separately from their main case total. One of my sources started adding them in today, but pulled them back out. I imagine that they'll end up being put together in a few days' time, which will make make for a big one-day jump. I added Austria and Romania today. No further progress on adding all the JHU data as a separate feed. Stay safe.

2020-04-05

Today's numbers: New York (5→5), Louisiana (6→7), U.S. (19→19), U.K. (22→22), Spain (19→22), Turkey (25→25), Canada (27→27), California (25→27), Ontario (24→28), Alabama (29→28), Toronto (53→30), Denmark (30→31), Sweden (28→33), Germany (29→34), Italy (39→40), France (33→42), Brazil (36→43), New Zealand (40→46), India (43→49), Iran (49→49), Singapore (62→56), Norway (50→57), Mexico (59→58), Malta (54→65), Japan (49→75), Australia (66→77), Malaysia (78→78), Indonesia (91→93), Thailand (103→114), South Africa (150→129), South Korea (450→514), China (14677→14188)

Compared to yesterday, here's how these regions are doing.

Much worse: Toronto, Singapore, South Africa, China

The same: Alabama, Mexico

Better: New York, Louisiana, U.S., U.K., Turkey, Canada, Denmark, Italy, Iran, Malaysia

Much better: Spain, California, Ontario, Sweden, Germany, France, Brazil, New Zealand, India, Norway, Malta, Japan, Australia, Indonesia, Thailand, South Korea

When I started calculating these stats, I didn't actually expect anyone to reach 1%. Yesterday, I had to extend my graph to the right from the 1% mark up to 10% to accommodate projections and inferences for New York and Louisiana (and for the inferences, that's not even enough).

Some of my friends are buying into the White House messaging about misreporting numbers in a prescient attempt to make the U.S. look bad, as though it'll be okay for the U.S. to lose tens of thousands of its residents as long as China lost more. It reminded me about a story that my father told (I'll have to get the details from him when I next see him) about speaking with a high-ranked U.S. military officer toward the end of the Vietnam War. My father remarked on the wide discrepancies between U.S. casualties as reported nightly by American and North Vietnamese authorities, and asked if he was correct in assuming that both sides were exaggerating in their own favour. The officer stared at him for a moment, then said "The Vietnamese have no reason to lie."

Toronto's numbers yesterday showed a net loss in active infected cases, which turned out (as I later commented) to be just a blip, presumably due to late reporting of most of yesterday's cases. So it's better to think not of 11 new cases yesterday being great and 203 today being terrible, but the 214 together representing a 12% daily increase, which is not terrible but not fantastic. Every few days, when I go out to buy or deliver (to my father) groceries or medicine, there are fewer people outside, and it's turning things around on a scale of weeks. There will be more positive blips followed by negative ones, but eventually the good news will become constant. I hope it's soon.

Tomorrow marks back to school for our two school-age sons, but in some as yet unclear online form. Fingers crossed for that too. Sure glad I reminded both boys to remove all perishables from their lockers three weeks ago.

Fewer than 6,000 new cases in Spain today for the first time in two weeks. They've inflected, let's hope for the last time.

Italy has been under 5% daily for a week, which sounds like better news than it is. They have at least twice as many infected cases as China did when they were at that growth rate, and the Italian death rate suggests an inference of 3-6 times more, so they could be following a trajectory with numbers that are 6-12 times more than the Chinese. Also, there aren't any other countries that have stopped at the 5% mark without dropping straight down to about 1%, so either they still have a problem with controlling spread or the number being reported is dominated by testing capacity.

Still looking at how/whether to switch to JHU data. One obstacle that I'm facing is that I'm not seeing Toronto data in their dataset, and I'd like to keep it for obvious personal reasons.

Stay safe.

2020-04-04

Today's numbers: New York (6→5), Louisiana (6→6), Spain (18→19), U.S. (19→19), U.K. (21→22), Ontario (24→24), Turkey (22→25), California (new→25), Canada (26→27), Sweden (27→28), Alabama (30→29), Germany (28→29), Denmark (30→30), France (26→33), Brazil (32→36), Italy (40→39), New Zealand (41→40), India (48→43), Japan (73→49), Iran (44→49), Norway (49→50), Toronto (45→53), Malta (49→54), Mexico (61→59), Singapore (66→62), Australia (66→66), Malaysia (77→78), Indonesia (85→91), Thailand (90→103), South Africa (189→150), South Korea (423→450), China (13944→14677)

Compared to yesterday, here's how these regions are doing.

Much worse: India, Japan, Singapore, South Africa

Worse: Mexico

The same: New York, Alabama, Italy, New Zealand

Better: Louisiana, Spain, U.S., U.K., Ontario, Canada, Sweden, Germany, Denmark, Norway, Australia, Malaysia

Much better: Turkey, France, Brazil, Iran, Toronto, Malta, Indonesia, Thailand, South Korea, China

If you are clicking through to my web site that plots growth rate against infected proportion, I've added a new feature today that might require explanation. It is often reported that there is something wrong with the numbers being reported: the reported case numbers are incomplete due to testing or other methodology, the reported death numbers are incomplete, or the death rate seems to vary from country to country. To address this, I've created a new set of data series for those countries for which I have death numbers, by applying the assumptions that the death rate is 2% and that it takes 14 days to go from confirmed diagnosis to death. You can view either the confirmed figures that I've been using until now, or these newly inferred figures, by tapping the appropriate buttons in the legend. For example, if the death total today is 4% of the infected total two weeks ago, then we can infer that the infected total is twice what has been confirmed.

Active confirmed cases in Toronto were actually down today for what I believe is the first time: the official total of 823 includes 25 deaths and 66 recoveries, leaving 732 active; yesterday was 812 - 13 - 56 = 743. We do have 11 more patients in ICU beds today, along with another 25 in other hospital beds, as well as 203 patients considered probable but not yet confirmed; but all this shutting down and social distancing seems to be working.

Various sources reported that the French numbers are missing from their confirmed total 17,827 cases in nursing homes (EHPAD en français). My understanding is that the French are publishing only those cases where COVID-19 has been confirmed by PCR testing, and that there are a substantial number of probable cases in nursing homes that are being systematically omitted. Watch for a big jump in their numbers if they change their protocol.

Spain has had two days of down around 7,000 confirmed new cases, and may actually finally be inflecting after 12 days in the doldrums.

France, Spain and Turkey have the highest discrepancies between their two-week-old confirmed cases and current deaths with a factor of around 25, compared to 20 for the United States, 9 for Canada, 2.5 for Germany, and none for South Korea and Singapore.

Canada, France, Indonesia, Italy, Spain, and Sweden all had fewer reported deaths today than yesterday.

I have added California to the list. They seem to be doing okay, especially for an American state.

Stay safe.

http://poslfit.com/covid19

2020-04-03

Today's numbers: New York (7→6), Louisiana (7→6), Spain (17→18), U.S. (19→19), U.K. (22→21), Turkey (23→22), Ontario (21→24), France (25→26), Canada (24→26), Sweden (30→27), Germany (25→28), Denmark (33→30), Alabama (new→30), Brazil (36→32), Italy (40→40), New Zealand (new→41), Iran (41→44), Toronto (50→45), India (50→48), Norway (49→49), Malta (51→49), Mexico (56→61), Australia (61→66), Singapore (73→66), Japan (73→73), Malaysia (80→77), Indonesia (88→85), Thailand (79→90), South Africa (181→189), South Korea (387→423), China (11244→13944)

Compared to yesterday, here's how these regions are doing.

Much worse: Brazil, Toronto, Singapore

Worse: Sweden, Denmark, India, Malta, Malaysia, Indonesia

The same: New York, Louisiana, U.K., Turkey

Better: Spain, U.S., France, Italy, Norway, Japan

Much better: Ontario, Canada, Germany, Iran, Mexico, Australia, Thailand, South Africa, South Korea, China

Toronto got its new confirmed case rate down below 10% for four days in a row, and now two days at 11%. In absolute numbers, less than 100 each day, but 100 new cases a day is still much too high. We have 89 patients hospitalized, including 42 in ICU beds.

Spain is down to 7,134, several hundred lower than usual for the past week and a half. If we assume that the death rate is 2% and average time from diagnosis to death is 14 days, then the case count is being underreported by a factor of 26. So 119,199 should be three million, and 1 in 15 people in the country should already be infected.

Italy's 4,585 new cases on a short, two-day trend downward 100 cases/day.

France has the widest variation in daily new cases among the countries I'm tracking. 4611, 2599, 4376, 7578, 4861, 2116, 5233 in the last seven days. Groups of people that large should be better at obeying a law of large numbers.

Japan still finding 200+ new cases a day. The working holiday students at our favourite local Japanese restaurant all managed to make it home before the borders closed, but are starting to question their choice.

The U.S. diagnosed 32,182 new cases today, more than 3,000 more than yesterday, but a lower percentage increase than the day before, a pyrrhic improvement.

They have a total of 7,007 deaths, increasing at 24% daily. That's a doubling time of 3.2 days, and leaves about 23 days before there are a million dead.

Louisiana appears to have the most underreporting (based on the 2%/14-day argument): numbers there should be off by a factor of 34.

I welcome New Zealand and Alabama to the list. I've git-pulled the JHU data, and am looking into how best to integrate it into my page, so there may not be too many more individual additions.

Stay safe.

2020-04-02

Today's numbers: Louisiana (12→7), New York (7→7), Spain (17→17), U.S. (19→19), Ontario (22→21), U.K. (23→22), Turkey (21→23), Canada (27→24), Germany (26→25), France (24→25), Sweden (33→30), Denmark (35→33), Brazil (41→36), Italy (38→40), Iran (39→41), Norway (50→49), India (71→50), Toronto (63→50), Malta (56→51), Mexico (56→56), Australia (44→61), Singapore (73→73), Japan (68→73), Thailand (68→79), Malaysia (83→80), Indonesia (79→88), South Africa (161→181), South Korea (392→387), China (1795→11244)

Compared to yesterday, here's how these regions are doing.

Much worse: Louisiana, Brazil, India, Toronto, Malta, South Korea

Worse: Canada, Sweden, Denmark, Malaysia

The same: Ontario, U.K., Germany, Norway

Better: New York, Spain, U.S., France, Mexico, Singapore

Much better: Turkey, Italy, Iran, Australia, Japan, Thailand, Indonesia, South Africa, China

So in fact, someone jumped the gun in reporting a lot of extra cases for China, and they've now taken them away, so China looks better again. Is everything okay in China? Were they reporting real numbers? (No, they were reporting integers! #lateNightMathDadJokes). I don't think it matters what their reporting method was, as long as it was more or less consistent throughout their epidemic. It gives us an idea of how the curve can end in one extreme case.

I'm accumulating death figures for more numbers, and thinking about optionally displaying estimated "real" case numbers based on the assumptions that death numbers are accurate (how can they be, what country that loses a thousand lives a day would risk lives at this point making sure the death numbers are 100% correct?), that the actual death rate is 2% (for some value of two), and that the mean time from diagnosis to death is 14 days (for some value of 14; I chose 14 because I read it in a paper that I'm now going to have to hunt down). That would multiply the case rate in South Africa 3, Japan by a factor of 3; Toronto 4; Denmark 5; Canada and Iran 8; Mexico and Italy 15; New York 20; Brazil, France and U.S. 25; Indonesia and Spain 28; Louisiana 40; Turkey 50. South Korea would be unchanged, which sounds plausible. Singapore would actually be cut in half, which suggests a methodological issue. :) I don't have data on the other countries, which is one big reason that I haven't done this yet. And yes, I should switch to using the Johns Hopkins data, but I haven't had a day free to set that up yet.

74 new confirmed cases in Toronto today push us back up over 10% for the first time in five days. Stricter measures announced today requiring two-metre separation in public parks, and asking people to stay home where possible should help keep us below 10%.

Spain's latest new case count of 7,947 continues their 8-days-out-of-9 streak of finding eight thousand new cases a day, to the nearest thousand. Hard to think it's coincidence. Italy is doing something weirder, where they've been between 4,000 and 6,600 for sixteen days, but without a clear pattern: it's trending downward at maybe 55 cases a day, but with a big sinusoidal pattern overlaid with a period of about a week. Given the number of regions I'm following, it could just be pareidolia (or apophenia, depending on whether the curve looks like a snake or a sine wave).

Stay safe.

2020-04-01

Today's numbers: New York (7→7), Louisiana (14→12), Spain (17→17), U.S. (19→19), Turkey (19→21), Ontario (25→22), U.K. (24→23), France (22→24), Germany (27→26), Canada (32→27), Sweden (34→33), Denmark (36→35), Italy (35→38), Iran (36→39), Brazil (46→41), Australia (44→44), Norway (47→50), Mexico (53→56), Malta (69→56), Toronto (49→63), Thailand (68→68), Japan (64→68), India (63→71), Singapore (70→73), Indonesia (79→79), Malaysia (77→83), South Africa (167→161), South Korea (374→392), China (10534→1795)

Compared to yesterday, here's how these regions are doing.

Much worse: Canada, Brazil, Malta, South Africa, China

Worse: Louisiana, Ontario

The same: U.K., Germany, Sweden, Denmark

Better: New York, Spain, U.S., Australia, Thailand, Indonesia

Much better: Turkey, France, Italy, Iran, Norway, Mexico, Toronto, Japan, India, Singapore, Malaysia, South Korea

Canada's not actually much worse; there was an erroneous case count posted that gave us a false much better day followed by a false much worse day. We are staying steady at about the 25-27 day mark, which is better than being closer to what my brother calls Doomsday, but not great. Our daily new case counts have flattened out over the last three days in the narrow range of 1125-1146. One would think that this might mean too limited testing, but the provincial breakdowns show much noisier variations in different regions that suggests that this is not the case. Still, the messaging from all levels of government is not encouraging, with the earliest date that anything might reopen being pushed back to May 4.

I read that China is now expanding testing significantly, so their big bump today is likely asymptomatic cases that randomly tested positive. A lot of people are complaining that this makes the earlier figures misleading. I'm not so sure: each country decides how it's going to count numbers, and will typically change its mind at least once. Here in Toronto, they've decided to start announcing "probable" cases that aren't included in the numbers I'm showing, which are "confirmed". In China, they did the same thing back on February 10th, which did end up inflating their total diagnosed counts by an amount comparable to Toronto's current 25% (probable as a proportion of confirmed).

Spain has now reported 8,000 new cases a day (to the nearest thousand) for seven out of the last eight days. As I've said before, I think this is testing capacity. I've started tracking death rates for some countries, including Spain, where the total deaths are increasing daily by 10% (not actually bad compared the the U.S.' 25%). You have to believe that either the mortality rate is increasing, or the diagnosis rate is decreasing.

In Spain, they lost 656 lives to COVID-19 on March 25th, and 923 on April 1st; over that time their cumulative death toll increased an average of 14% daily. In the U.S., those numbers were 225 and 1,007, and a 32% daily increase. That American mortality rate cannot mathematically continue at 32% forever: there's a hard limit of 40 days, at which point, everyone will be dead. Epidemiologically speaking, the rate should come down long before then, because the number of susceptible people left in the population will have dwindled. But do stay home if you can, and stop spreading the disease.

2020-03-31

Today's numbers: New York (7→7), Louisiana (18→14), Spain (16→17), U.S. (16→19), Turkey (16→19), France (26→22), U.K. (21→24), Ontario (26→25), Germany (24→27), Canada (26→32), Sweden (new→34), Italy (31→35), Iran (35→36), Denmark (39→36), Australia (39→44), Brazil (54→46), Norway (35→47), Toronto (31→49), Mexico (45→53), India (73→63), Japan (72→64), Thailand (65→68), Malta (91→69), Singapore (66→70), Malaysia (75→77), Indonesia (66→79), South Africa (68→167), South Korea (333→374), China (9101→10534)

Compared to yesterday, here's how these regions are doing.

Much worse: Louisiana, France, Brazil, India, Japan, Malta

Worse: Denmark

The same: Ontario

Better: New York, Spain, Iran

Much better: U.S., Turkey, U.K., Germany, Canada, Italy, Australia, Norway, Toronto, Mexico, Thailand, Singapore, Malaysia, Indonesia, South Africa, South Korea, China

I enjoyed looking at all the new statistics the Ontario government is publishing. Then I had a brief scare when I saw Toronto's numbers were way up, before I realized that they have started publishing probable cases as well as confirmed ones. Then I had another brief scare because there were 165 probable cases. Then I calmed down, as much as one can.

I think my older son cited me in an online discussion in his Biology class.

Japan saw more new cases today than Canada, for the first time ever. The messaging I'm hearing in Japan is worrying: people being told that they should stop going out drinking at karaoke bars. That seems so March 2020.

Most interesting read in a day full of interesting reads:

http://www.simaonlus.it/wpsima/wp-content/uploads/2020/03/COVID_19_position-paper_ENG.pdf

- an unreviewed paper which reminds us that many respiratory viruses are transported on particulate pollution, so shutting down polluting activities might help slow down novel coronavirus transmission. I mentally file it with papers saying that hypertonic saline nasal irrigation delivers chloride ions to virally infected cells that convert it as a defence to hypochlorous acid (bleach): would be nice if true, and isn't going to be definitively tested on novel coronavirus anytime soon.

The U.S. looks like it's gotten past some sort of testing limit, and its numbers have risen again. After three days of 20,000 new cases identified daily, they're up to 23,000 today.

Spain continues on at just under 8,000 cases a day, for a week now. I don't think it's really doing better. I think they only have the capacity to identify 8,000 cases a day.

Stay safe.

2020-03-30

Today's numbers: New York State (7), Spain (16), Turkey (16), U.S.A. (16), Louisiana State (18), U.K. (21), Germany (24), Canada (26), France (26), Ontario (26), Italy (31), Toronto (31), Iran (35), Norway (35), Australia (39), Denmark (39), Mexico (45), Brazil (54), Thailand (65), Singapore (66), Indonesia (66), South Africa (68), Japan (72), India (73), Malaysia (75), Malta (91), South Korea (333), China (9,101).

Compared to yesterday, here's how these regions are doing.

Much worse: Ontario, South Korea

Same: Iran

Slightly better: New York State (again), Spain (again), Turkey (again), U.S. (again), U.K. (again), Germany (again), Canada, France (again), Toronto, Denmark, Thailand, Japan

Much better: Louisiana (again), Italy, Norway, Australia, Mexico, Brazil, Singapore (again), Indonesia, South Africa, India, Malaysia (again), Malta (again), China

Ontario shifted its reporting methods and timing, and it looks like there's a bump today because they added in a large part of another day's numbers into today's. We'll see if it settles back down tomorrow.

Spain's new case numbers dropped below 7000 for the first time in almost a week, but now they're back in the 7500-8000 range where they've been displaying suspiciously linear growth. Turkey too has been in the 1600-2000 range for four days now. Let's hope that the U.S.'s slight dip from yesterday is not part of a similar stalling of testing at the 16,000-21,000 case range where they've been for four days. The U.S. pattern at least looks more pointy and less linear.

Louisiana had a bad day, with almost 485 new cases; it looks better because four days ago they had a worse day at 510 new cases. It's perhaps not a good idea to look too carefully at the numbers if you want to remain optimistic.

Italy had a genuinely better day, with barely over 4,000 new cases, the first under-5,000 day in a week.

I am optimistic for the Maltese, who thanks to their being an island nation of relatively small population, that has now cut itself off from the rest of the world, are seeing only single-digit new case numbers.

Also, don't read too much into South Korea and China being much better or worse than yesterday; they both have a good handle on their situation, which makes their daily fluctuations seem bigger because their new case numbers are so low.

Stay safe.

2020-03-29

Today's numbers: New York State (7), Turkey (14), Louisiana State (15), Spain (15), U.S.A. (16), U.K. (20), France (24), Germany (24), Canada (26), Italy (27), Ontario (29), Norway (30), Toronto (31), Australia (33), Iran (36), Denmark (38), Mexico (39), South Africa (42), Brazil (44), Singapore (58), Malaysia (61), Indonesia (63), Thailand (65), India (67), Japan (71), Malta (89), South Korea (357), China (7615)

Compared to yesterday, here's how these regions are doing.

Much worse: Japan (again)

Same: Denmark, Mexico, Ontario

Slightly better: Australia, Brazil (again), China, France, Germany, Iran, Italy (again), New York, Norway, Spain (again), Turkey, U.K. (again), U.S. (again)

Much better: Canada, Louisiana, Malaysia, Malta, Singapore, South Africa (again), Thailand (again), Toronto

Spain and Italy both saw several hundred fewer new cases today than yesterday, so maybe they are reaching their inflection point. I hope so: they're still seeing several thousand newly diagnosed cases daily.

In Ontario, we learned last night after we had done our grocery shopping for our weekly Sunday fictive family dinner that that dinner was now banned because it had more than five people. We cooked the food separately, exchanged some dishes with each other halfway between our houses, and had our dinner conversation over FaceTime. Thanks to earlier measures, Toronto's numbers have come down sharply in the last few days, and I'm hoping the new restrictions help confirm that we are past the worst of it. Fingers crossed.

I'm worried about Japan, which has not quite managed to bring their epidemic under control. People are generally very careful about respiratory disease transmission there, as they are in most Asian countries. It sounds like they're finding though that there are situations like that community centre yesterday where just one infected person can set back their efforts tremendously, and I'm not sure how they're going to control that.

Stay safe.

2020-03-28

Today's numbers: New York State (7), Louisiana State (12), Spain (13), Turkey (14), U.S.A. (16), U.K. (20), France (22), Germany (22), Canada (23), Italy (26), Toronto (27), Norway (29), Ontario (30), Australia (32), South Africa (33), Iran (36), Denmark (39), Mexico (40), Brazil (44), Malta (46), Singapore (47), Malaysia (55), Thailand (62), Japan (95), China (7,615).

Compared to yesterday, here's how these regions are doing.

Much worse: Japan

Worse: France, Malaysia

The same: New York State, Singapore, Turkey

Better: Brazil, Canada, Denmark, Italy, Malta, Ontario (again), Spain, Toronto, U.K., U.S. (again)

Much better: Australia (again), China (again), Norway, South Africa, Thailand (again)

As I posted elsewhere, I suspect Italy and Spain are not doing better; it just looks like they are because they've reached their testing capacity, so until they can increase it, they will keep seeing the same number of new cases each day, which will be a smaller percentage of a larger infected population.

Both Toronto and Ontario saw 12% increases today. That's good in terms of recent numbers, but the government is aware that here too we are restricting testing to critical cases, and as of today additional restrictions came into place banning social gatherings of more than five people. We need to get our spread much closer to zero, and now's the time to do it.

Japan reported a huge uptick today. It was announced that almost half of the new cases came from one super-spreader who worked in a community centre with a large number of disabled people. That's going to be a difficult lesson for Japan to learn. We've had similar cases across Canada too, and I worry about my aunts who are in a local nursing home.

The past history was starting to look cluttered, especially if you look at all the regions in my visualization tool, so I've added buttons that let you look only at past, present or future data. I've also added the American state of Louisiana, which is apparently being hit hard.

2020-03-27

Today's numbers: New York State (8), Spain (13), Turkey (15), U.S.A. (16), U.K. (19), Canada (21), Germany (22), Norway (23), South Africa (23), France (24), Italy (26), Toronto (26), Australia (28), Ontario (30), Iran (38), Denmark (39), Brazil (43), Mexico (45), Singapore (48), Malta (55), Thailand (56), Malaysia (57), Japan (120), China (7,543),

Boilerplate explanation paragraph: The number in question is my own way of boiling down your region or country's growth/doubling rate and number of people infected with COVID-19 into one number. It represents the number of days you have left at your current rate of infection (averaged over the last four days) before you reach 1%. It's not a forecast (your rate of change is likely to change before then, when public health authorities take positive or negative action), just one simple number that tells you how you're doing compared to others. I could have picked a bigger percentage than 1%, but you can't go too far before the exponential growth bends back down logistically, and it seemed as easy a place as any; most people can visualize the idea of 1 in 100.

Compared to yesterday, here's how these regions are doing.

Much worse: Denmark (again), Mexico, Norway, Toronto, U.K. (again)

A little better: Brazil, France, Germany, Japan, New York State, Ontario, South Africa, Spain, U.S.A.

Much better: Australia, Canada, China, Iran, Italy, Malaysia, Malta, Singapore, Thailand (again)

I understand the issue in Toronto is still returning travellers being diagnosed, and officials have warned that our numbers could increase substantially for a few days. At the other end of the country, officials in British Columbia were guardedly optimistic, saying that social distancing seemed in general to reduce transmission rates by a factor of two, and that there was a reasonable chance that everyone who needs care will receive it.

Denmark's "much worse" is still starting at a fairly good baseline, so not too much to worry about. They're drifting up toward 10% daily from 5%. Norway had a somewhat worse day than usual, at 12% rather than their recent 9%. Not good trends, but much better than other places.

Mexico was down to 10% one day recently, but is now consistently back in the 20% daily range.

2020-03-26

New York State (7), Spain (13), U.S.A. (16), Canada (18), Germany (21), South Africa (21), France (22), U.K. (23), Australia (25), Italy (26), Ontario (29), Norway (33), Malta (36), Brazil (41), Toronto (41), Singapore (43), Iran (44), Denmark (46), Malaysia (46), Thailand (47), Mexico (50), Japan (120), China (6,805),

Boilerplate explanation paragraph: The number in question is my own way of boiling down your region or country's growth/doubling rate and number of people infected with COVID-19 into one number. It represents the number of days you have left at your current rate of infection (averaged over the last four days) before you reach 1%. It's not a forecast (your rate of change is likely to change before then, when public health authorities take positive or negative action), just one simple number that tells you how you're doing compared to others. I could have picked a bigger percentage than 1%, but you can't go too far before the exponential growth bends back down logistically, and it seemed as easy a place as any; most people can visualize the idea of 1 in 100.

Compared to yesterday, here's how these regions are doing.

Much worse: Denmark, Germany, Iran (again), Japan, Malaysia, South Africa, U.K.

A little worse: China, France, Ontario, Singapore

The same: Canada, Spain

A little better: Australia, Italy, Mexico, New York State, Norway, U.S.

Much better: Brazil, Thailand (again), Toronto

Toronto saw a rise of only 6% today. 5% if you leave out returning travellers, who are becoming scarcer and contributing less to the new case numbers. That's the third time since the epidemic hit in full force that we've had new case numbers below 10%. It's not cause for complacency though; local public health staff are restricting access to testing, at least until the current test backlog is cleared. So some unknown part of the decline is reduced testing. The rule now is that if you have severe symptoms, or if you are in an at risk population and have good reason to think you're infected, you can get tested. Otherwise, just isolate and protect yourself from others and vice versa.

Canada's daily rate has dropped in the past four days 44%, 33%, 23%, 19%. Again, this might be due to undertesting, but let's hope not.

New York and the rest of the U.S. are gradually bringing their outbreak under control, with new daily reported cases down into the 20%s.

Australia has had two days in a row of increases down in the 10%s for the first time.

Italy, Japan and Denmark are now consistently below 10%. This is good news for Italy, whose overall numbers are so large that even small percentages affect large absolute numbers. It's less good news for Japan, which has shown difficulty in getting its rate down another order of magnitude to Chinese levels.

Stay safe, everybody.

2020-03-25

Today's numbers: New York State (6), Spain (14), U.S.A. (16), Canada (19), Australia (23), France (24), Germany (24), South Africa (25), Italy (26), Malta (26), U.K. (26), Norway (31), Thailand (32), Brazil (35), Ontario (36), Toronto (36), Singapore (47), Denmark (50), Iran (50), Malaysia (50), Mexico (50), Japan (164), China (7,997).

Compared to yesterday, here's how these regions are doing.

Much worse: Iran.

Worse: South Africa, Ontario, Singapore, Denmark.

A little worse: Canada, Malta, Japan, China.

Same: New York State, Germany, Norway.

A little better: Spain, U.S.A, Australia, France, U.K., Brazil, Toronto.

Better: Italy, Thailand, Malaysia, Mexico.

I know locally that Ontario has been seeing a steady influx of returning residents from out of the country, who as of this morning constituted half of the known new cases. Not much we can do about this, other than stay away from high concentrations of travellers, and ask the travellers to obey the requirements of the Quarantine Act.

Iran is much worse in the sense that they had 2200 or 9% new cases today when they had gotten down to below 5% daily a few days ago. News reports say the government is concerned that a second wave is building up, and that they are signalling that there will be tighter controls to come.

On the bright side, Italy has really turned things around, going from being the worst case in Europe to having the lowest new case rate outside of Asia. Their problem continues to be that even 7.5% of 70,000 people is a huge number of new cases.

Stay safe.

2020-03-24

Today's numbers: New York State (7), Spain (14), U.S.A. (15), Australia (21), Canada (21), Italy (22), France (24), Germany (25), U.K. (25), Malta (28), South Africa (28), Thailand (29), Norway (32), Brazil (33), Toronto (34), Ontario (40), Malaysia (46), Mexico (47), Singapore (50), Denmark (55), Iran (64), Japan (164), China (8,228).

Boilerplate explanation paragraph: The number in question is my own way of boiling down your region or country's growth/doubling rate and number of people infected with COVID-19 into one number. It represents the number of days you have left at your current rate of infection (averaged over the last four days) before you reach 1%. It's not a forecast (your rate of change is likely to change before then, when public health authorities take positive or negative action), just one simple number that tells you how you're doing compared to others. I could have picked a bigger percentage than 1%, but you can't go too far before the exponential growth bends back down logistically, and it seemed as easy a place as any; most people can visualize the idea of 1 in 100.

I added NYS today, because I'm worried about friends and family in New York. New York and Italy are the only regions on the list that have more than 1 person infected in a thousand, and New York is growing at 32% daily compared to Italy's 10%. That's the difference between growing tenfold in 8 days or 24.

I added the city of Toronto as well, because I live here. Apart from that, it's uninteresting: it has a higher density of infected people, but its growth rate is the same as the province of Ontario (15%), because it accounts for half the cases.

Some countries are doing a tiny bit worse: Spain, Canada, South Africa, Denmark, Iran, Japan. China has been having about 40 new cases out of 80,000 for the last few days, and suddenly has 80; I hope it's just a blip. Two countries are on the same track as yesterday: Australia, U.K. This is not a good thing, but not as bad as doing worse. The rest of these regions have shown at least a little improvement, thanks to tightening controls on the pandemic: U.S.A., Italy, France, Germany, Malta, Thailand, Norway, Brazil, Ontario, Malaysia, Mexico, Singapore.

Stay safe, everyone.

2020-03-23

Today's numbers: U.S.A. (14), Spain (16), Italy (20), Germany (21), Australia (22), Malta (22), Canada (23), France (24), Brazil (25), U.K. (26), Thailand (28), Norway (30), South Africa (30), Ontario (34), Malaysia (41), Mexico (41), Singapore (48), Denmark (55), Iran (64), Japan (164), China (10,168).

The U.S. reported new case rate looks like it has actually been decreasing over the past four days from the 40%s down to the 30%s. Today's is actually 27% as reported now, but there are usually additional reports that come in overnight, so that number is likely to go up. But still, it looks like measures to prevent spread are starting to have an effect, taking the doubling rate from once every two days to once very 2.5~3 days. In the current climate, you could call this good news.

There's a big blip in the Canadian numbers, reportedly because the province of Quebec has started accepting test result numbers from hospitals where they were previously only counting tests processed centrally. So that means that the situation wasn't as good as we thought before. Things like this happen all the time though, which is one reason I average the rate over the last four days.

In the province of Ontario, the daily rate has been remarkably steady at about 18% for three days. This is good, in that it's not going up, but it’s still a doubling rate of four days. Until today, we still had a lot of unnecessary activity in the province that was resulting in contagion. I'm glad that the province's announcement today has persuaded one of my friend's employers to finally shut down for the duration.

The rest of the countries I'm monitoring haven't had any major changes in the past day. That means that if they were doing a good job of bringing the epidemic under control (China, Japan), they are still doing so; and if they aren't (everyone else), they should be doing more now to do so.

2020-03-22

Today's numbers: U.S.A. (13), Spain (17), Italy (18), Germany (20), Brazil (22), Australia (24), Malta (25), U.K. (25), France (27), Thailand (27), Canada (29), Norway (29), Mexico (34), Ontario (34), South Africa (36), Malaysia (44), Denmark (47), Singapore (51), Iran (66), Japan (164), China (10,482).

The U.S. daily new case rate (or doubling time) remains more or less the same at 43% (or doubling faster than every two days). I've added Brazil, so that there's at least one other country in the same ballpark (38%). Spain, Italy, Germany are actually doing okay on the new case rate with their last few days all being in the teens (%) or lower; their day numbers are high because they all have a bigger proportion of their population infected already (1 in 1,000 in Italy), so they don't have as far to go. Thailand is in the opposite situation: they just had a big jump in their reported cases, but they were starting off with so few cases that their situation doesn't look as serious.

Canada and the province of Ontario have had their new case rates drop significantly over each of the last three days, as social distancing measures take effect. The daily rates are down in the 10-15% range now, so there is still work to do.