RESUMO
All 27 home-use tests sold in France in 1989 for the self-diagnosis of pregnancy were evaluated. The kits were first tested by qualified clinical chemistry technologists. Eleven kits with 100% specificity and 100% sensitivity were retained for the diagnostic study. Each of 638 laywomen was given a kit and asked to perform the assay with a coded urine specimen containing either no human chorionic gonadotropin (hCG) or an hCG concentration adjusted to the claimed detection limit (1 DL) or twice the detection limit (2 DL). After testing, each participant filled out a detailed questionnaire. The results showed a diagnostic specificity of 86-100% for 10 kits but a diagnostic sensitivity of 85-100% for only 5 kits at 2 DL and for only 2 at 1 DL. Among the 478 positive urine samples distributed, 230 were falsely interpreted as negative. The main explanation for such a high percentage of false-negative results was difficulty in understanding the explanatory leaflets accompanying the kits and hence in reading the results, regardless of the socioeconomic situation of the participant. We conclude that pregnancy home-use tests should be subjected to rigorous analytical controls and evaluated by a panel of potential users before being released on the market.