Understand how we got here — and how to move forward.


October 5, 2020

The 5 biggest surprises in the largest-ever study of Covid-19

Daily Briefing

    In what is the largest epidemiological study of Covid-19 conducted so far, researchers studied 85,000 of the more than six million confirmed cases of the novel coronavirus in India, as well as nearly 600,000 of their contacts---and their extensive research produced some surprising findings.

    Ready-to-use slides: 11 ways health care is changing globally

    Study details

    For the study, Ramanan Laxminarayan of the Centers for Disease Dynamics, Economics and Policy in New Delhi, India; Joseph Lewnard, an epidemiologist at the University of California-Berkeley; and their colleagues traced more than three million contacts of 435,539 confirmed cases of Covid-19, the disease caused by the coronavirus, and then focused their efforts on the 575,071 contacts—of a cumulative 84,965 confirmed cases—for whom testing information was available.

    The researchers centered their work on two states in southern India, Andhra Pradesh and Tamil Nadu. According to the New York Times, these states are two of the five states in the country with the highest number of Covid-19 cases. However, the states also possess some of India's most sophisticated health care systems, and they have the highest levels of public health spending in the country.

    To conduct the level of contact tracing necessary for the study, the researchers were able to leverage disease-surveillance networks in both states that had originally been established during the AIDS epidemic. When someone in those states tested positive for the novel coronavirus, a public health worker would call them by phone to ask where they had been and whom they'd been in contact with for the previous two weeks. All of the immediate family members of those with confirmed cases were tested and isolated quickly, while other contacts were located via phone calls and text messages.

    Overall, according to Laxminarayan, the research comprises "the largest epidemiological study anywhere on Covid[-19] by far."

    Key findings

    Mortality risk declined after 65

    Among several surprising findings, the study identified an overall case-mortality rate of 2%, a rate that increased predictably with age until age 65 when—unlike other nations—it started to decline.  

    According to Lewnard and his colleagues, the overall life expectancy in India is 69. As a result, those who make it to age 65 or older in general may already be at an advantage in terms of their health or access to health care than their counterparts, which in turn could make them more likely to survive the virus.  

    Children transmit the virus

    In addition, the study found that while children who are younger than 17 were the least likely to die from Covid-19, they transmitted the coronavirus at rates similar to their older counterparts. Specifically, the researchers found that children between the ages of five and 17 transmitted the virus to 18% of close contacts their own age.

    More broadly speaking, the researchers found that infected people tended to transmit the virus to others of similar ages. For instance, the researchers found that more than 5,300 school age children included in the study had transmitted the virus to 2,508 contacts, but were more likely to spread the virus to children of similar ages.

    According to Lewnard, the findings contradict "claims that children have no role in the infection process." He added, "There's, granted, not an enormous number of kids in the contact tracing data, but those who are in it are certainly transmitting."

    Super-spreading is 'the rule'

    The study also demonstrated that "[s]uper-spreading events are the rule rather than the exception," Laxminarayan said. Specifically, the researchers found that just 5% of people with confirmed cases accounted for 80% of new infections identified via contact tracing—while about 71% of people with confirmed cases were linked to no new cases at all.

    But according to the researchers, while the study did not assess whether some infected people spread the virus more effectively than others because of biological factors, it did suggest that super-spreading events are affected by various behaviors, such as physical distance, length of contact, and ambient conditions. For instance, the researchers found that 80% of people exposed to someone with a confirmed case of the virus in a high-risk environment (in this case, sitting within three rows of an infected person on a bus or train for at least six hours) contracted the virus. By comparison, just 1.6% of people exposed to an infected person in a lower-risk environment—such as being in the same room, but being more than three feet away—became ill.

    Hospital stay before death is just 5 days

    The study also found that the median hospital stay before death from Covid-19 was five days—significantly shorter than the two week median for patients in the United States. According to Lewnard, patients in India may deteriorate more quickly because of underlying medical conditions, such as high blood pressure or diabetes.

    Initially infected people tended to be male, older

    According to the study, the people most likely to be infected first tended to be male and older than their contacts. Lewnard said this trend could stem from several potential causes, such as men being more likely to be in situations with a risk of infection, more likely to experiences symptoms of infection, more likely to get tested if they feel unwell, or even more likely to answer the phone if a contact tracer calls them. 


    The study is key in part because it is one of the first to focus on a lower-income nation, Lewnard noted. He explained that until this study was published, most of the data on the epidemiology of Covid-19 had come from higher-income nations, even though the majority of cases were occurring in lower- or middle-income countries.

    Jeffrey Shaman, an epidemiologist at Columbia University, echoed that thought. "To see it in the Indian milieu is very important," he said. Noting that "India has been the nexus of the most cases recorded for the last three, four weeks," he added, "We can't just study it in a few countries and then walk away."

    However, Ashish Jha, dean of Brown University's School of Public Health, who was not involved in the study, said while he appreciated the study overall, he would warn against extrapolating its findings too far.

    He explained that the two states included in the study are among the best prepared to deal with an outbreak in India. "It is really important to understand this is not the experience of Bihar [a rural, lower-income state in India], this is not the experience of [the Democratic Republic of the Congo]," Jha said. "This is a much rosier picture than what you are likely to see in those places" (Mandavilli, New York Times, 9/30; Bengali, Los Angeles Times, 9/30).

    Have a Question?


    Ask our experts a question on any topic in health care by visiting our member portal, AskAdvisory.