Abstract

Analysis of eight groups of data collected at varying intervals during a period of seven years showed fluctuations in the sensitivity of tests to diagnose trichomoniasis in women. The best results were obtained from fresh, correctly prepared Diamond's medium, Feinberg-Whittington's medium, and modified Squires and McFadzean's medium (which showed 82% to 94% relative sensitivity). Poor results were identified consistently in commercially prepared Bushby medium from one source (40% relative sensitivity) and in a batch of commercially prepared Squires and McFadzean's medium from which chloramphenicol had been omitted (23% relative sensitivity). Examination of wet film, culture, and exfoliative cytology stained by Papanicolaou's method were shown to be indispensable for auditing the performance of each test and to maintain the quality of a diagnostic service.