Accuracy is everything, When we do a diagnostic test, an incorrect result can usually lead to agony and faulty, if not harmful, treatments. Currently, the most reliable way to identify coronavirus infection is to use a polymerase chain reaction (P.C.R.) test: a swab, usually taken from the nasal passage, creates a sample that is then sent to a specialized laboratory. P.C.R. Tests that can detect tiny amounts of genetic material from the virus cost over $ 100. Under ideal circumstances, the analysis only takes a few hours. Due to high demand, supply shortages and other problems, many commercial laboratories take more than a week to process them. This means that a positive test often comes back too late to allow contact tracers to notify exposed people before they can infect others in turn. In these circumstances, the diagnosis is only useful in making personal health decisions and providing data on the infection rate in a community.
In a July 21 report in JAMA Internal Medicine, the C.D.C. for Covid-19, nine out of ten infections are not identified – and barriers to testing are likely main reasons. To capture more of these cases, many of which may not have obvious symptoms, we need to change our thinking, says Daniel Larremore, a computational biologist at the University of Colorado, Boulder. In particular, we need to move from prioritizing the accuracy of individual test results to prioritizing a test system's ability to reduce the virus rate in a given population – even if it leads to more misdiagnosis.
To see how this might work in practice, consider a strategy for increasing test capacity: merging samples for analysis. For example, let's say 1 in 100 people has the virus. Testers take and label each with a nasal swab; A portion of each sample is saved and the remainder is grouped with samples from nine other people. The laboratory then performs 10 analyzes, one for each group of 10 samples. Nine of these will give negative results, a decision given to all 90 members of these groups. The laboratory then retests each and every sample in the positive group to find the infected member. In total, the lab performed 20 analyzes instead of the 100 it would take to test each one individually.
At a certain threshold, diluting samples by combining them with so many others can make the virus harder to detect, but the technique has grown in batches of five for P.C.R. testing. Nebraska has been able to expand its supplies through pooling, except for populations with high infection rates, which result in more groups testing positive and therefore requiring more individual testing. "That can change from week to week and possibly from day to day," says Jonathan Kolstad, an economist at the University of California at Berkeley. “Florida, three months ago you could have made pretty big pools. Now you wouldn't want that. "However, he and his colleagues note in a July working paper published in the National Bureau of Economic Research that computer models could use factors such as a person's age, occupation, zip code, and social networks to classify people according to their risk of infection and to group their samples In theory, as more people are phased out with the virus, the rates of infection will decrease, and pools can be expanded, making testing more efficient, so economists' analysis found that daily tests would cost twice as much as monthly tests Testing.