Japan, masks, and flu
So, Japanese people wear masks a lot. Probably the main thing that got them through most of the pandemic, along with avoiding "the three Cs" and good contact tracing; Japan never had lockdowns. (And was worst in class for rich Asian countries for most of the pandemic, too.) But they already had the custom of wearing masks when sick, to protect others. So I wondered, did that make an observable difference in flu seasons, say?
I couldn't easily find comparative stats on flu cases. But this WHO site generates reports for a country "by the Global Influenza Surveillance and Response System (GISRS)". I'm lazy when it comes to uploading images, so just check this directory. And pay attention to the Y-axis numbers, because they are not all to the same scale! Probably what I want is the total area of a spike, but it's easier to look at the heights, and hell, the widths look similar. Anyway, we see US peaks around 5000 'specimens' in the mid 2010s, rising to 25,000 recently, while Japan has been peaking under 1000. And Japan has 36% of the US's population, so definitely has fewer specimens per capita. But what does that mean? What's the methodology? I have no idea. Canada has peaks of 4000, despite 10% the US's population, and Korea is a bit under 200, with 1/3 of Japan's population. And it's not like the US population went up by 5!
Flu *deaths* turn out to be more promising. Statista says "In 2019, the number of deaths from influenza in Japan increased to around 3.6 thousand from about 3.3 thousand in the previous year. The death rate from influenza amounted to 2.9 death cases per 100,000 inhabitants in 2019." Mainichi says "Annually, around 3,000 people in Japan die of influenza" (though 10,000 including indirect deaths from worsening other conditions.) And finally, Niid talks about 2018-19, with 3400 excess deaths, "similar to the average year". It also says 12 million flu patients visiting medical facilities.
The CDC has a nice explicit table listing stats for the past decade, and while deaths swing from 12,000 to 61,000, the average over 10 years is 35,900, or about 11 deaths per 100,000.
Japan: 3. USA: 11. That's *huge*. That's Japan having 72% fewer deaths, or the US having 3.67x more.
Though there may be complicated methodology, as the CDC tries to correct for underreporting, and may be trying to estimate the total burden, so maybe that 10,000 deaths number from Japan is the right one to use, which would be 8.3 deaths per 100,000 -- lower, but not vastly so.
The CDC gives other data, like medical visits, which aren't that different from Japan's 12 million! But, I know from other reading that Japanese people visit doctors a lot more (3x more?) than Americans, and spend more time in the hospital.
CDC also gives hospitalization -- in fact, that seems to be their primary source of actually measured data. The annual average is 442,000. Happily, Niid also gives hospitalization for that year in Japan: 20,389.
Japan: 17 hospitalizations per 100,000 people. US: 138. 8x more!
I also note that Japan has about 6 flu hospitalizations per death, while the US has 12. Is the US better at keeping hospitalized people from dying (more vaccination of the vulnerable?) Do Americans get hospitalized more easily for flu than Japanese people, contra the usual trend? Is the CDC "adjustment for underreporting" boosting the number? I don't know.
So what's the final verdict? It's *possible* that Japan has a much lower flu burden than the US: the strongest case for that is comparative hospitalizations per capita, with an 8-fold ratio. There's also a 4x ratio in deaths -- though that might be mostly canceled out if the US is estimating 'flu-associated' deaths (pulmonary and pneumonia and such, not just respiratory) and Japan isn't. OTOH medical visits don't seem dissimilar, given the greater likelihood of Japanese people to go see a doctor. But if flu transmission is similar, then Japan must have a much *lower* chance of being hospitalized for flu, which would be odd. Epidemiology is hard! Especially as a layman sticking my nose in and not knowing the messy details.
And of course even if there is a big difference in cases or deaths, I've done nothing to show that masking is responsible, though intuitively, sick people wearing masks when out of the house should help a lot. See the
DW comments at https://mindstalk.dreamwidth.org/591003.html#comments
I couldn't easily find comparative stats on flu cases. But this WHO site generates reports for a country "by the Global Influenza Surveillance and Response System (GISRS)". I'm lazy when it comes to uploading images, so just check this directory. And pay attention to the Y-axis numbers, because they are not all to the same scale! Probably what I want is the total area of a spike, but it's easier to look at the heights, and hell, the widths look similar. Anyway, we see US peaks around 5000 'specimens' in the mid 2010s, rising to 25,000 recently, while Japan has been peaking under 1000. And Japan has 36% of the US's population, so definitely has fewer specimens per capita. But what does that mean? What's the methodology? I have no idea. Canada has peaks of 4000, despite 10% the US's population, and Korea is a bit under 200, with 1/3 of Japan's population. And it's not like the US population went up by 5!
Flu *deaths* turn out to be more promising. Statista says "In 2019, the number of deaths from influenza in Japan increased to around 3.6 thousand from about 3.3 thousand in the previous year. The death rate from influenza amounted to 2.9 death cases per 100,000 inhabitants in 2019." Mainichi says "Annually, around 3,000 people in Japan die of influenza" (though 10,000 including indirect deaths from worsening other conditions.) And finally, Niid talks about 2018-19, with 3400 excess deaths, "similar to the average year". It also says 12 million flu patients visiting medical facilities.
The CDC has a nice explicit table listing stats for the past decade, and while deaths swing from 12,000 to 61,000, the average over 10 years is 35,900, or about 11 deaths per 100,000.
Japan: 3. USA: 11. That's *huge*. That's Japan having 72% fewer deaths, or the US having 3.67x more.
Though there may be complicated methodology, as the CDC tries to correct for underreporting, and may be trying to estimate the total burden, so maybe that 10,000 deaths number from Japan is the right one to use, which would be 8.3 deaths per 100,000 -- lower, but not vastly so.
The CDC gives other data, like medical visits, which aren't that different from Japan's 12 million! But, I know from other reading that Japanese people visit doctors a lot more (3x more?) than Americans, and spend more time in the hospital.
CDC also gives hospitalization -- in fact, that seems to be their primary source of actually measured data. The annual average is 442,000. Happily, Niid also gives hospitalization for that year in Japan: 20,389.
Japan: 17 hospitalizations per 100,000 people. US: 138. 8x more!
I also note that Japan has about 6 flu hospitalizations per death, while the US has 12. Is the US better at keeping hospitalized people from dying (more vaccination of the vulnerable?) Do Americans get hospitalized more easily for flu than Japanese people, contra the usual trend? Is the CDC "adjustment for underreporting" boosting the number? I don't know.
So what's the final verdict? It's *possible* that Japan has a much lower flu burden than the US: the strongest case for that is comparative hospitalizations per capita, with an 8-fold ratio. There's also a 4x ratio in deaths -- though that might be mostly canceled out if the US is estimating 'flu-associated' deaths (pulmonary and pneumonia and such, not just respiratory) and Japan isn't. OTOH medical visits don't seem dissimilar, given the greater likelihood of Japanese people to go see a doctor. But if flu transmission is similar, then Japan must have a much *lower* chance of being hospitalized for flu, which would be odd. Epidemiology is hard! Especially as a layman sticking my nose in and not knowing the messy details.
And of course even if there is a big difference in cases or deaths, I've done nothing to show that masking is responsible, though intuitively, sick people wearing masks when out of the house should help a lot. See the
DW comments at https://mindstalk.dreamwidth.org/591003.html#comments