It's been reported today (April 4) that

New York state has recorded 630 more coronavirus deaths, another daily record that takes its [death] toll to 3,565.

Most deaths have been in New York City. The state now has almost as many cases - over 113,000 - as the whole of Italy.

But clearly Italy has had more Covid-19 deaths, over 15,000 (15,362 on the same April 4 day), which puts the case fatality ratio (CFR) estimate substantially higher in Italy than in NY.

The CFR is sensitive to the number of asymptomatic cases, i.e. cases that should be included in the CFR denominator, but might not be so because not enough testing is done on the asymptomatic individuals. (There are also other factors that can drive up the CFR "for real", like the age structure of the population, and smoking history.)

But to stick to a public-health measures comparison question (as opposed to [all] other confounding factors), how do NY and Italy compare in terms of testing, per capita?

  • 3
    I think this question is on-topic here because Covid-19 testing is largely influenced by government measures, and in countries with socialized medicine (i.e. "universal healthcare"), it's practically exclusively done that way. Even in the US, the [private] testing was influenced by regulatory decisions like those of the FDA. I'll eventually ask on medical science SE if others really think this is off topic here, but I disagree this is the case.
    – Fizz
    Apr 4 '20 at 19:48
  • I won’t vote to close, but my worry with this question is that it’s a current events question, so the answer is going to change continuously over the coming days, weeks, and months.
    – divibisan
    Apr 4 '20 at 20:57
  • @divibisan: true, but the question is most relevant right now, when some stats (total cases) match between the two regions.
    – Fizz
    Apr 4 '20 at 23:53
  • The number of dead is lagging behind some days. Italy seems already to be slowing down with infections and especially death numbers. In New York death numbers are still rising very fast. Case fatality ration should only be computed for a time in the past when the currently dead people probably infected. Otherwise you will considerably underestimate the CFR during the exponential growth phase
    – Manziel
    Apr 6 '20 at 13:56

Interestingly enough, the number of tests per capita appears roughly the same between Italy as a whole and NY, according to Wikipedia, but in some regions of Italy like Veneto, the number of tests per capita was higher:

                     Tests      Positive        Tests per 1M pop       Positive tests / 1K tests
NY (3 Apr):          260,520    102,863         13,392                 395
Italy (4 Apr):       657,224    124,632         10,896                 190
Italy- Veneto:       133,289     10,824         27,169                  81
Italy- Lombardy      141,877     49,118         14,102                 346

In percentages, that's 1.34% of NY's population having been tested (assuming/approximating 1 test per person) vs 1.09% in Italy. Italy has even had some "denser" testing in some regions like in Veneto, were 27,169 per 1M pop (2.72%) were tested, according to the same source.

Italy has had only about half of the postive:total tests ratio of NY (190 vs 395). This is even lower in the Veneto region (81 / 1K tests, so about one fifth of NY's.) If we use these positive:total tests ratios to approximate the ratio of persons tested positive in the whole population, we get 5,290 per 1M pop (0.52%) in NY and 2,070 per 1M pop (0.2%) in Italy. In the Veneto region, that number is 2,200 per 1M pop (0.22%), so 2.5 times more testing in Veneto (per capita) found only about 10% more infected than in the rest of Italy.

In Lombardy, the hardest hit region of Italy, quite interestingly, the number of tests per capita is closest to NY's (14,102 vs 13,392 per 1M pop), as is the ratio of postive:total tests (395 vs 346), so that's approximately 0.49% of Lombardy's population being found infected (compared to 0.52% in NY).

Alas, that page doesn't have the number of tests in NYC proper, which might an interesting additional figure to look at. The testing "density" in NYC (i.e. per capita) might be higher than in NY state as a whole.

  • 2
    The reason why tests per capita are low in Lombardy is that the health system, and in particular the testing labs, have been overwhelmed by the disease (there are also accusations of bad management by the region's government -- I guess it's early to judge though), and they aren't able to test all people that ought to be tested (reference, in Italian) Apr 4 '20 at 20:44
  • @DenisNardin: well, NY might have been equally overwhelmed. This ratio of tests positive to total tests actually might be a good proxy for how overwhelmed the system is. Mike Pence said a few days ago that the US is looking at a scenario comparable to Italy. I guess we know a little better why now (as the number of deaths per se doesn't quite compare yet.)
    – Fizz
    Apr 4 '20 at 20:50

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .