Read on Twitter
Some general thoughts (in random order) on COVID-19 death undercounting in India following the recent article in The Hindu ( https://www.thehindu.com/opinion/op-ed/covid-19-deaths-are-not-as-low-as-reported/article32310814.ece) by @giridar100 and collaborator, and the response from @oommen ( https://www.orfonline.org/expert-speak/from-covid19-pandemic-to-epidemic-of-spurious-analysis/). 1/
The documented instances of COVID-19 death undercounting are too many and too major in scale to be treated as aberrations. Many have followed similar patterns - for example omitting deaths from "comorbidities" and suspected deaths. https://twitter.com/muradbanaji/status/1284481215563575296
The "urban areas have high MCCD coverage" argument is something of a red herring. Delhi, with its high MCCD coverage, saw *huge* undercounting before subjected to pressure. Mumbai too added in lots of missed COVID-19 deaths in June (~1700). https://twitter.com/muradbanaji/status/1284893376567885824
Most of the documented cases of undercounting have emerged as a consequence of media scrutiny and external pressure, rather than as a result of some sudden change of heart from officials. Clearly scrutiny is vital, and arguing for less scrutiny is unethical...
...especially since the dominant narrative about COVID has shifted from "there's not much disease" to "not many are dying". This shift was already becoming clear by June. In this context there needs to be extra vigilance around fatality underreporting. https://science.thewire.in/health/covid-19-india-case-fatality-rate-confusion-propaganda/
Where there have been "reconciliations", the explanations have been unclear, and the reconciled deaths have rarely been dated, making it very difficult to judge if the corrections were complete. This lack of transparency seems to be a choice, and not about "weak systems".
Of course there is a lot of variation by region. But with so many instances, as epidemics pick up in new areas, assuming little or no death undercounting is as much conjecture, as assuming undercounting is occurring.
Aside from systematic undercounting, routine/inadvertent undercounting could increase as spread picks up in rural areas with poorer death registration and MCCD coverage. The article in The Hindu should sensitise us to this possibility.
With limited data, estimating the scale of undercounting locally or nationally is difficult, leaving only a few options. E.g., examining funeral or hospital data; using excess mortality data where available; tracking data for unusual patterns; and using modelling...
...e.g. modelling can tell us about apparent changes in infection fatality rate (IFR) over time - as in Gujarat recently. These could be interpreted as genuine drops in IFR or in terms of undercounting. Such changes should spur further investigation. https://twitter.com/muradbanaji/status/1291065734181711872
So, there is necessarily an element of speculation in attempts to quantify the problem. But while people may disagree on the scale, the key is acknowledging undercounting, and examining *how* it happens, as @PriyankaPulla did recently in this piece. n/n https://science.thewire.in/health/india-mccd-comorbidities-covid-19-deaths-undercounting/
