Monday, June 7, 2021

Estimating Covid Wave Deaths II (update)


  With data available for the whole of May 2021, we revisit the issue of underestimation of Covid deaths at the start of the second wave, given the explosive spread of the virus, unprecedented rise in severe cases and the overwhelming of hospital facilities and the health infrastructure in many cities. Figure 2 shows that the anomaly of case curve (green, left scale) lying above the death curve (red, right scale) which prevailed during the rise in cases has now been corrected as number of new cases starts to fall.

    We divide the analysis into following periods: (1) The period between the first wave start to peak of active cases, (2) From this first wave peak to trough of active cases, (3) the period from first wave trough of active cases, to the second wave peak, (4) From the 2nd wave peak of active cases to the latest available data (June 3, 2021). A similar periodization is done using the rate of growth of cases to define peaks and troughs.  The difference in fatality rates between period one and period 3 is used to adjust for the deaths in the second wave. There is no anomaly between period (4) and period (2). On the contrary the fatality rate is higher in the latter.

Like the previous note on this subject it doesn’t address the issue of general underestimation, underestimation relative to developed countries or under estimation relative other lower middle income countries.  

The differential in fatality rates between the rising phase of the second and first waves is used estimate the underestimation during the second COVID wave relative to the first wave. The estimates are given in table 2. Columns 8, 9, 10 give the cumulative value of tested, confirmed cases and deaths (as per official numbers). Columns 4, 5 & 6 give the increase in tested, cases and deaths during the seven periods defined above(4 + 3 more in terms of case growth rates). Column 7 gives the adjusted death rates. The rows marked 3a and 3b are the ones in which our estimates of death are higher than the official numbers.

On a cumulative basis, we find that deaths have been underestimated by 17% during the second wave (last row, table 2)

Figure 2: India COVID cases and deaths



Table 2: Underestimation of deaths in second wave relative to first.



No comments: