A fourteen-year study to identify measles antigen in urine specimens by fluorescent-antibody methods.

Between 1965 and 1979 inclusive, 426 morning urine specimens from 410 patients admitted to hospital with the provisional diagnosis of measles were tested for measles antigen by fluorescent-antibody methods. One hundred and forty-two specimens (33.3 per cent) were positive. Of 74 patients who were clinically assessed not to have measles, 6/77 specimens (7.8 per cent) gave false positive results, and 71/77 (92.2 per cent) gave true negative results. Three hundred and forty-nine urine specimens were tested from 336 patients clinically judged to have had measles and 136 of these specimens (39 per cent) were positive. The positive rate was highest in a small number of specimens taken before the rash and next highest at the onset of the rash. The positive rate then fell; it rose again near the end of the first week, but the further increased positive rate after that time was associated with very small numbers of specimens. Of 12 child contacts of measles, 5/12 (41.67 per cent) had positive urine specimens and one of them developed measles two days after sampling. None of the seven patients with negative urine specimens developed measles. Measles antigen was detected in urine sediment by fluorescent-antibody tests in 39 per cent of the patients with measles studied.