Under-reporting does hurt the COVID fight

Without good data, accurate projections are impossible, making it difficult to gauge the true state of the pandemic

May 04, 2021 12:02 am | Updated July 06, 2022 12:19 pm IST

India, now home to the world’s worst ongoing coronavirus pandemic, is currently reporting nearly a million new cases and 10,000 deaths every three days, according to data released by the Ministry of Health and Family Welfare . The true extent of the second wave now ravaging India is likely much worse than official numbers suggest.

The reasons

Is it a problem noted only in India? Not capturing all COVID-19 cases and COVID-19-related deaths is not unique to India. Research on the behavioural dynamics of COVID-19 from a group at the Massachusetts Institute of Technology estimates the global under-reporting factor for cumulative cases around 7 and for deaths 1.4 as of December, 2020. Researchers further note that these factors vary substantially across nations. In India, this problem seems to be particularly acute during the second wave based on empirical evidence and epidemiological models.

Why is it hard to capture all COVID-19 infections and related deaths? There could be several reasons why we cannot capture all COVID-19 infections such as silent asymptomatic infections, barriers to testing due to cost and travel time, reluctance to get tested due to COVID-19 associated stigma, limited availability of tests, obtaining a false negative test (remember that diagnostic tests are not perfect) and alike. Deaths related to COVID-19 that are missed often consist of deaths that happened outside health-care facilities at home, and post-COVID-19 deaths where the cause of death is listed as a pre-existing comorbidity such as heart disease or kidney failure. India also has a poor and delayed infrastructure for reporting of deaths and certifying the cause of death in general, particularly in the rural areas. In a 2017 estimate, one out of five deaths was medically reported.

How do we estimate this under-reporting from epidemiologic models? For modelling growth of an epidemic, what we observe are deaths, cases or hospitalisations. However, what really defines an epidemic is not exactly the growth of these observed quantities but the infections, which in turn become these outcomes (deaths, cases, hospitalisations) with some delay, and not all infections get converted to these observed quantities. What proportion of people die from an infection is a very important quantity as it allows us to know how dangerous a disease is. Another important quantity is how many infections a health system identifies, that is, how many of these will actually end up being reported as cases. A high number of infections being caught by the health system shows a successful surveillance strategy.

 

On epidemiological models

In a recent study , epidemiological models attempt to capture covert infections by accounting for unreported, but infectious individuals. Expected deaths are then estimated from the estimated number of infections and assumed infection fatality rates based on historical data.  Such models indicate the under-reporting factor for cases between 10 and 20 and for deaths between two and five based on data from the first wave for India.

Can we validate what the models are saying? How do we validate the extent of unreported cases: We can cross-check under-reporting of infections directly with serosurveys carried out in India. The third serosurvey conducted by the Indian Council of Medical Research (from December 17, 2020 till January 8, 2021) reports that 21.5% of all Indians above the age of 18 have antibodies present that indicate SARS-CoV-2 infection in the past. Approximately 59% of India’s 1.36 billion citizens are above the age of 18. This implies nearly 173 million adults infected. Factoring in the nearly 11 million COVID-19 cases reported by January 8 (assuming most cases are adults), this points to an implied under-reporting factor of roughly 16 for infections. In other words, only 6% of India’s COVID-19 infections are reported. Hence, the question of “missing infections” in India is undeniable and not contingent on a belief in the legitimacy of mathematical models — it is evident based on figures released by the Indian government/bodies alone.

Mortality data

How to validate the extent of unreported deaths: During wave 1 , a group of volunteers collected every reported death from obituaries in newspapers and reported almost twice the number of deaths than officially reported. During this recent surge, a recent report in The New York Times  noted that authorities in Gujarat reported between 73 and 121 COVID-related deaths each day in mid-April. The report added that one of Gujarat’s leading newspapers, Sandesh , sent reporters to cremation and burial grounds across the State and reported that the number was several times higher, around 600 each day. The extent of this under-reporting is higher than our past model estimates. This current increased case-fatality is not only due to clinical lethality of the virus, but more patients are dying due to not receiving adequate medical care. A way to capture the holistic effect of COVID-19 (direct and indirect impact on mortality) will be to perform a proper excess death calculation where demographers can take the number of people who die from any cause in a given region and period and compare the same with a historical trend based on the past few years and come up with a difference of observed and expected number of deaths. This method of investigating excess deaths is something various nations have explored, such as the United States (22% excess deaths in 2020 and 72% of these attributed to COVID). India has not made historical mortality data and data from 2020 publicly available, making this calculation infeasible at this point.

Is India testing enough during wave 2? Lack of adequate testing could be one of the causes of missing infections.

Testing in India

To understand India’s scale of testing, let us look at the U.S. during its highest surge in January (November 1-February 15) and India during this recent surge (March 28-April 27) in terms of testing. The bar plots show that India has a maximum daily test positivity ratio (TPR) around 25% and the U.S. had a maximum daily TPR around 15%. More interesting is Figure 2 showing seven-day growth rate in cases versus testing. While testing and cases have grown at a comparable rate in the U.S., in India the growth in reported cases on an average has been nearly five times higher than the growth in testing. India is not testing enough.

How does under-reporting matter? Right now, the country is reeling from skyrocketing infection and death counts. This surge has thrown our health-care systems off balance. Crucial medical supplies run dangerously low and hospitals are forced to turn away patients. These forecasting models are used to predict the need for oxygen, hospital beds, intensive care unit care needs, the peak and duration of the pandemic. Without having more informative data, accurate projections are impossible. Knowing the truth is better for both public and policymakers to gauge the true state of the pandemic.

Finally, the exact extent of under-reporting is debatable, but we should never forget that these numbers represent people. The official system can fail to capture the diseased and the deceased, but families cannot. The tragedies that have unfolded in thousands of families in India, with an astounding number of people that are currently sick and grieving for the dead, can never be captured through the reported staggering numbers and the ones that were missed.

Bhramar Mukherjee is Chair of Biostatistics at the University of Michigan School of Public Health. Soumik Purkayastha is a doctoral student in Biostatistics at the University of Michigan School of Public Health. Maxwell Salvatore is a doctoral student in Epidemiology at the University of Michigan School of Public Health.

Swapnil Mishra is a research associate at the MRC Centre for Global Infectious Disease Analysis, School of Public Health, Imperial College London

0 / 0
Sign in to unlock member-only benefits!
  • Access 10 free stories every month
  • Save stories to read later
  • Access to comment on every story
  • Sign-up/manage your newsletter subscriptions with a single click
  • Get notified by email for early access to discounts & offers on our products
Sign in

Comments

Comments have to be in English, and in full sentences. They cannot be abusive or personal. Please abide by our community guidelines for posting your comments.

We have migrated to a new commenting platform. If you are already a registered user of The Hindu and logged in, you may continue to engage with our articles. If you do not have an account please register and login to post comments. Users can access their older comments by logging into their accounts on Vuukle.