Only 1 in 200 of those who tested negative in an antigen test in Delhi to detect possible coronavirus cases were re-tested, a fraction that epidemiologists say is too low given what is known about the disease. From June 25, daily cases appear to have peaked at 3,390 and steadily declined to 1,790 as of 15 July.
Delhi formally began using antigen testing kits on June 18 and has accelerated their use since July. While these tests dominate daily testing, using them widely without following up with adequate retests contradicts Indian Council of Medical Research (ICMR) guidelines on use of the test, experts told The Hindu.
Also read | Nearly 3 lakh rapid antigen tests in Delhi
From June 15- July 16, Delhi has conducted 305,820 Rapid Antigen Detection Tests (RADT) — a quicker complement to the standard RT-PCR (reverse transcription polymerase chain reaction) test that while most accurate takes as many as 24-48 hours to deliver a verdict — to check for the COVID-19 virus. Of these, 285,225 tests came ‘negative’. Of them, 1,670 (or about 0.5%) were chosen for re-test by RT-PCR and 252 of these were confirmed positive, according to the most updated figures as on July 16 confirmed by The Hindu.
The limitation of the RADT is that they may miss as many as half of those who are actually positive, which is why the ICMR recommends that these tests be administered in regions where the background prevalence of the disease is known to be high and those negatives “suspected” to be harbouring the virus be re-tested by the standard RT-PCR test.
When approving the RADT by SD Biosensor, a South Korean company, the ICMR underlined that the test ought to be used only in containment zones, hotspots, hospital settings and laboratories among those who manifested one or other symptoms of the disease, influenza-like illnesses, and people with co-morbidities who were asymptomatic direct and high-risk contacts of those confirmed positive. Under these settings, those who tested ‘negative’ and whom clinicians “suspected” to be harbouring the disease ought to be “definitely tested sequentially by RT-PCR to rule out infection”.
Those who test positive don’t need a re-test and must be considered positive, the guidelines note.
On July 9, a Delhi government order as part of its “Test Track Treat” strategy expanded the ambit of testing beyond containment zones. Symptomatic individuals would be identified via surveys and they would all need to be tested with RADT kits. “We consistently need to test at least 20,000 people each day. Since the contacts can be in any part of the city, depending on the movement of the patient, the testing facility will be made available at all identified dispensaries and polyclinics,” the order said.
This initiative also underlines that “high-risk” persons who test negative in the RADT will have to undergo RT-PCR.
However, the low number of re-tests doesn’t square with the ICMR recommendation regarding re-tests nor with previous research. ICMR scientists, in a national survey published in May in the Indian Journal of Medical Research found 10% positives among symptomatic contacts of laboratory confirmed cases; 6.1% among hospitalised patients with Severe Acute Respiratory Illness (SARI); and 5.1% among asymptomatic family members of laboratory confirmed cases. This was in April and cases have only exponentially increased everywhere, including Delhi, since.
“A low level of re-testing with RT-PCR in persons who are testing antigen negative will underestimate the cases and make the tracking of the decline imprecise. The ability to detect true positive cases must be maintained at a high level, even as testing rates are being increased,” said Dr. Srinath Reddy, President, Public Health Foundation of India, and a member of ICMR’s national COVID-19 task force.
An epidemiologist who has been part of the ICMR surveys but declined to be identified said, “If the overall positivity is 15%-20%, it’s very odd if only 0.5% samples are chosen for retesting.”
Officials told The Hindu that in Delhi the tests were being deployed widely and there were camps being set up outside containment zones in various districts to facilitate testing.
A modified ‘Revised COVID Response Plan’ was released by Delhi on July 8. Some districts officials started holding camps for testing ‘Special Surveillance Groups’, including domestic workers, drivers and electricians in areas where they work and also stay, using RADTs.
“We retest a person using RT-PCR if the doctor on the ground feels that the person has ILI (influenza like illness) symptoms. But not every negative case is re-tested using RT-PCR,” a Delhi government Health Department official said, adding, “If we re-test everyone, then the purpose of having another [rapid antigen] test is defeated.”
Of those re-tested with RT-PCR, around 15% test positive, which is higher than the antigen test positivity of 6%.
According to sources in the Ministry of Home Affairs, from April 1-June 14, Delhi averaged about 2,800 tests per day, and from June 15-July 8 these increased nearly tenfold to about 20,000 a day (including RT-PCR and RDAT). A Delhi government affidavit to the High Court says that from July 11-July 14 there were on average every day two-three RADTs for every RT-PCR test — or 52,000 RADTs for about 20,000 RT-PCR tests.
The consequence of indiscriminately deploying antigen tests would mean expanding the number of tests and presenting a lower positivity rate while not necessarily being able to reliably establish the extent of the spread of the coronavirus in the population.
Experts say that when deploying RADTs in the field, there has to be an appropriate strategy in place. In hospitals or laboratory settings, where kits are analysed to determine their ‘sensitivity’ and ‘specificity’ (metrics to gauge the accuracy of the test), there would be higher prevalence of positives, said Dr. Giridara Gopal Parameswaran, an epidemiologist and research scientist at the All India Institute of Medical Sciences, Delhi.
“If many of those who tested negative were regular people [who weren’t tested on the basis of specific criteria], you have diluted the purpose of the entire exercise. However, we need to know the sensitivity and specificity of these kits in the field, and the basis for deciding to appropriately test. If the prevalence is low, it will affect these parameters,” he told The Hindu.