With seroprevalence surveys - done in Delhi and Mumbai - suggesting a much higher rate of COVID-19 infections than what were recorded in those cities so far, States have tried to ramp up testing levels to detect more infections.
But most new tests have been through Rapid Antigen Detection (RADT) method, which is not as reliable as RT-PCR tests because of the 'false negatives' problem. Some States have also limited Reverse Transcription-Polymerase Chain Reaction (RT-PCR) tests despite having the capacity to conduct more.
Actual vs detected
The serosurvey held during June 27-July 10 in Delhi found that 22.86% of Delhi's population had been exposed to the virus. But by July 10, COVID-19 tests had recorded only 1,09,410 infections, equivalent to 0.59% of the population.
Similar surveys in three Mumbai wards (slums - S, non-slums - NS) also found that much fewer cases had been recorded compared to the numbers who had been exposed to the virus.
Chart appears incomplete? Click to remove AMP mode
Two tests
States have ramped up RADTs instead of RT-PCR tests in order to capture the positive cases more quickly. RT-PCR tests are more accurate, while RADTs register a high rate of false negatives. Graphs show total-tests (bars, left-axis) and % of RADTs (line, right axis) in select States for whom data were available.
Accuracy vs speed
The RADTs can register as many as 50% 'false negatives' — i.e. samples of those who are actually infected by COVID-19 wrongly returning a negative result. RT-PCR is more accurate.
ICMR had directed that those tested as negative by RADT should be re-tested using RT-PCR. Data from Delhi show that very few have been re-tested.
Capacity vs actual
Instead of supplementing RT-PCR tests, States have replaced a chunk of them with RADTs. Delhi, for instance, has a RT-PCR capacity of 11,000 tests per day.
But on most days in July, fewer than half the capacity was used.
Source: State health ministries, ICMR, PIB, Delhi High Court