Test positivity rate (TPR) is a much-talked about term these days, based on which the government is now intensifying or relaxing COVID-19 containment steps. But public health professionals say that TPR need not be the most valid or sensitive indicator of pandemic control.
A senior Health Department official said that the current method of calculating the TPR, the percentage of people testing positive for the virus per 100 samples, is not based on a fixed denominator. “If 10 out of 100 people test positive in an area, the TPR is 10%. If 200 are tested and 10 people get infected, it will come down to 5%. If five out of 50 people test positive the next day, even if there is a decline in the number of new patients, the TPR remains 10%,” he said. If the number of infected persons rises to 10, the TPR also goes up to 20% in that place.
This means that the TPR is dependent only on the number of people getting tested, and not on the whole population in any area. “If tests are carried out in a nearby area which is more populated, the TPR will be similar. No matter how big the population is, the rate will not change,” he said.
The official said that if the TPR comparison between two places has to be fair, the number of people subjected to tests should be increased according to the population. Otherwise, a fixed number of tests should be mandatory based on the size of the population.
Better option
“A better epidemiological option will be to compare the tests per 1,000 of the population and the new cases per 1,000 of the population. The test rates and case rates per population can be determined thus during a particular period,” he added.
Another public health professional, who wished not to be quoted, said that the TPR depends on many factors. They include the sensitivity and specificity of the test kits, prevalence of the infection, and the probability of infection among people. If vulnerable people and those who have had contacts with already infected people are tested, they stand a chance of turning positive. If asymptomatic people and those who have had no contacts are tested, they will not turn positive as well.
“If the TPR is to be considered a valid indicator, all those who are vulnerable to get infected should be tested, and those who don’t stand a chance should be excluded. The rise and fall of TPR is determined based on the proportion between these two groups. The proportion also varies according to local conditions,” he said.
TPR can be subjected to a ‘selection bias’, which can be high here considering the fact that a majority of the infected are asymptomatic and highly unlikely to be tested, he added.
