Recent coverage on COVID-19 has been celebratory of India’s 90%+ recovery rate. How is the recovery rate measured and does effectively convey fatality from COVID-19?

In the battle against COVID-19, recovery rates are recurrently used to convey that we are nearing the end of the pandemic. Recent coverage on COVID-19 has been dominated by India reaching a 90%+ rate of recovery. Does this mean that those infected are now at a lower risk of dying?

In its most simplistic form, the Recovery rate, or what is now being termed “Discharge Rate” on the Ministry of Health and Family Welfare’s website, is calculated by dividing the number of recoveries, discharges and migrations by the number of confirmed cases.

By this measure, India currently has a median Recovery Rate of 91.48%, with most states showing very high rates of recovery as shown in the map below.

Table of Contents

What do these rates signify?

When broken down into its elements, the recovery rate is determined by the number of people getting removed from the population that is confirmed as COVID-positive, and the number of confirmed cases.

In India, COVID-positive patients are considered removed or recovered, when they complete the 14-day isolation period without it culminating in hospital admission or when those admitted in a hospital get discharged. Those migrating to other states is presumably a smaller share of recoveries.

The Recovery Rate is typically used to measure fatalities for epidemics where the outcome is definitive. However, it can be misleading in the case of ongoing pandemics, where at any given point, the outcome of a share of the confirmed cases is unknown. By dividing the number of recoveries by the number of confirmed cases, we are assuming that those in the process of recovery do not recover from the illness, thereby underestimating the number of recoveries or the number of deaths should the outcome result in death.

In the case of COVID-19, the prevalence of asymptomatic patients and therefore, the underreporting of confirmed cases further affects the accuracy of this measure. Moreover, recovery rates are likely to increase over time, as the recovery rate for COVID-19 is 14 days. For instance, a recovery rate of 0 on Day 1 will increase over time, particularly around the 14th day which is the expected recovery period.

The projection of a simplistic recovery rate that does not account for these factors can be misleading.

A better indicator of fatalities is the Case Fatality Rate when it is adjusted for bias. As outlined in the World Health Organization’s Scientific brief, this rate is calculated by dividing the number of deaths by the number of confirmed recoveries and deaths. By not dividing deaths by confirmed cases, the adjusted rate only accounts for cases where the outcome is known.

As evident in the map below, the Case Fatality Rate (bias-adjusted) shows greater variation in fatalities across states.

India’s median Case Fatality Rate is currently 1.36% which must be interpreted as the percentage of deaths among COVID-positive patients in India whose outcome is known.

Maharashtra records the highest fatality rate, which is concurrent with the number of deaths that the state reports. Maharashtra has reported 43710 deaths, nearly 3 times the number of deaths in Karnataka. These variations are not captured in the simplistic Recovery Rate which projects high recovery rates across the country.