The Positive Test Paradox

We’re all probably overcome by covid news fatigue at this point and I can imagine many who would rather avoid all mentions of it like the plague (pun intended). However, there was one report today on the undercounting of Indian COVID cases by a factor of 90!!!! that really caught my eye. On the face of it, you expect India to miss cases which seems perfectly reasonable. This begs the question, what’s the reason for the low test positivity rate — the number of positive cases per 1000 tests — in India? India has lower tests per capita as compared to richer nations. Naturally, you’d then expect only those with symptoms or an actual risk of infection to get tested which would increase the positivity rate. However, counter-intuitively, we notice that the positivity rate is lower in India and presumably, other countries where there are lower tests per capita. This can possibly be explained by the incentives for getting tests done.

Before I speak more about COVID, I’ll speak a bit of competitive exams. People who appear in competitive exams usually spend a certain amount of time preparing for an exam before applying. Further, students who feel their preparation is inadequate choose not to appear in the exams. Naturally, the scores in the exams is not representative of what the general population would have scored. The students who appeared for the exam were the subsection of the population that was more suited to pass the exam. As a result, the percentage of people who fail the exam would be lower than the percentage of people from the general population who would have failed the exam. This seems to apply to COVID tests too. A lot of people who are required to get a covid test as a pass to do something ensure they are shielded and in peak health by the time they take their COVID tests. This is certainly true for me as I’m completely isolating myself for 2 weeks prior to my COVID test so that I can fly to India. When the potential downsides of not getting a test are adverse in ~1% of all cases, it’s quite likely that people would only take the test if really necessary — poor health or mandated for an activity.

This does explain why positivity rates are often quite low but this doesn’t still explain why a country like India which has a low positivity rate also has a significantly higher undercounting. This can be explained by 3 crucial factors — demographics, culture and economics. A lot of the developing world has a younger population. During a pandemic where a higher proportion of youth is expected to be asymptomatic, it’s only natural that you’d see a lot of young people with infections not getting tested in countries like India. Also, people in developing nations have a higher propensity to self-medicate or shrug off infections. This leads to fewer people with actual symptoms visiting a hospital. Another crucial factor is the economics of visiting a healthcare centre, getting tested, and potentially getting a positive result. Each of these carry costs which are less significant in more developed nations. This means that in a country like India, the ones who get tested are more likely to be ones that don’t need it and that creates a situation where low positivity in fact is sees a large number of actual cases outside thus leading to the positivity test paradox