Sero studies, also known as sero surveys, use serology tests to check for the presence of antibodies among the population.
Professor Mohanan said that while people often think that large samples are needed for an unbiased estimate, it is not true.
COVID-19 sero survey: Soon after the coronavirus pandemic struck India in March, various sero surveys started being conducted to determine the proportion of population that has been exposed to the virus. But what exactly are sero surveys, how do they work, why are their results so high as compared to the RT-PCR tests that are conducted and how are they supposed to be read? Associate Professor at Duke Unversity’s Sanford School of Public Policy Manoj Mohanan addressed these questions in an article for IE.
Seroprevalence studies explained
Sero studies, also known as sero surveys, use serology tests to check for the presence of antibodies among the population. If the tests detect sufficiently high quantities of a specific antibody in a person, it suggests that the person has been infected in the past. Usually, these tests are conducted on a sample of the population which has been selected based sampling techniques to ensure that the results can be scaled up to the entire population. Using these techniques, the sero survey does not need to be conducted on everyone or even a majority of the population. The only thing that needs to be kept in mind is that the people who agree to participate in the survey truly represent the population.
Professor Mohanan said that while people often think that large samples are needed for an unbiased estimate, it is not true, and a large sample is needed only to achieve precision. He added that precision can help in determining whether the estimates and findings of two or more studies overlap with each other. However, precision from a large number of observations does not necessarily rule out bias, he cautioned.
Why are serosurvey results so different from RT-PCR numbers?
The professor said that the national number of cases and the results from sero survey are so different partly because of the fact that most cases of the COVID-19 cases that the country has seen are asymptomatic. Apart from that, there are significant variations in symptoms among the symptomatic cases. That, along with the stigma and fear of being quarantined, has kept a lot of symptomatic people from getting tested, which is why the number of current infections detected by RT-PCR remains much too low as compared to the findings of the sero survey.
Learning from overall testing in India
Studies conducted in large parts of India, including Pune, Mumbai, Delhi and Hyderabad, have estimated that a large portion of the population in these places have antibodies and have therefore been exposed to the virus. He added that a recent study conducted by his team in Karnataka found 44% of the rural population having antibodies. Moreover, with the spread of the virus, the population having antibodies is expected to rise.
This rise depends on the way people interact, the precautions they take as well as the number of people infected presently.
Lower numbers in second sero survey rounds explained
A second round of sero survey among the same population could produce lower numbers for several reasons. The first reason could be that the people sampled in the first round might not want to participate again and therefore, the survey would be carried out from among people who had not been tested earlier. The deviation could be written up to the change in the sample.
Another reason is the decline in antibodies over time, which is normal considering once the infection passes, the body has no reason to produce antibodies. This means that a person might have antibodies, just not enough for it to be a ‘positive’ in a lab test, the professor said. He also assured that a decline in the number of antibodies also does not mean that the body could be re-infected immediately, adding that several studies are being conducted on whether the human immune system is generating any form of long-term immunity from the infection.
Difference in numbers in different studies
Professor Mohanan said that different sero studies being conducted in the same city could still show different results due to difference in sampling and testing methods. The difference in the sensitivity of the test kit can lead to different results unless they have been adjusted for sampling methods and test accuracy.
Apart from that, studies can have varied timeframes. He said that in case of a rapidly evolving epidemic, the exposed population can vary greatly within a matter of weeks, which can lead to different results from different studies.
He also said that the infection spreads differently in different parts of a city or a state, leading to variation in results of studies conducted in the same city or state.
Seroprevalence and herd immunity
The professor said the studies have so far cleared three things – a large part, if not majority, of the Indian population has contracted the virus, the rural area has also been infected equally due to large migration of people from urban to rural areas in the lockdown, and even with about 50% seroprevalence in some parts, it can’t be guaranteed that the remaining population would be protected by herd immunity.
He added it is a big concern that if people drop their guard on the assumption of herd immunity, a lot of people in the country could get very sick very fast. Therefore, it is important to continue to take precautions like wearing masks, washing hands, and maintaining physical distancing as much as possible.