Russia’s Covid fatality rate

From today’s issue of The Telegraph:

Russia boasts one of the world’s lowest Covid-19 mortality rates, but new figures released by Moscow authorities suggest that hundreds of coronavirus deaths could have gone unreported.

Scientists and doctors say the low death count could be explained by Russia’s conservative protocol of attributing fatalities to Covid-19, as well as suspected under-reporting in the regions.

Three days earlier, The New York TImes quoted a Russian researcher:

“[All-cause m]ortality figures in Moscow seem to be much higher than average for Aprils over the last decade… The number of Covid-19 victims is possibly almost three times higher than the official toll.”

Regrettably, the NYT opened the piece with this an irrelevancy: “researchers have been puzzled by Russia’s mortality rate of only about 13 deaths per million, far below the world average of 36 in a country…”

The number of deaths divided by the total population isn’t what we’re looking for. We’re interested in the fatality rate among the infected and/or the sick, currently 0.9%. That’s strikingly low compared with 6% so far for the US, 7% for Brazil and over 10% for several European countries.

According to the FT:

Russia’s national death toll from coronavirus could be 70 per cent higher than the government’s official data show… The Financial Times’ analysis of all-cause mortality data in Moscow, the capital, and St Petersburg, Russia’s second-largest city, found 2,073 more deaths in April relative to the historical average of the previous five years. Official Covid deaths in the two cities came to just 629 for the same period, leaving 1,444 deaths in excess of normal mortality levels unaccounted for.

According to The Guardian, “Moscow has ascribed its surprisingly low death toll… to a mass testing regime and rapid response to confirmed cases.” Yes, the mass testing could be a valid explanation for a relatively low fatality rate, but hardly for sub-1% vs. 10% observed in the EU.

A more plausible scenario would be this. If, as NYT and the FT suggest, the actual death rate is roughly three times as high as the reported rate, it should be close to 3%. If the scale of systematic underreporting of Covid-19 deaths outside of Russia’s two largest cities is even greater, perhaps the actual fatality rate should be adjusted towards 4%. If the difference were 4% (rather than 1%) for Russia against 7% in Brazil, “we’re simply testing more people” would sound a lot more convincing.