Alexander Edwards, associate professor in biomedical technology at the University of Reading, says it "quickly became clear" that if the tests were used as instructed, "you don"t tend to get large numbers of false positives".