Setting: A field project in Bangladesh.
Objective: To compare the effectiveness of commonly used carbolfuchsin staining variations.
Design: Routine hot Ziehl-Neelsen (ZN) 1% basic fuchsin staining for 15 min in 75 field clinics. Blind reading of duplicate smears stained by ZN 1% vs. 0.3% basic fuchsin applied for 5 min, or by ZN 1% 5 min vs. Kinyoun cold staining. Rechecking of discordant series.
Results: For comparable numbers of false positives, sensitivity was significantly lower with Kinyoun than with ZN 1% 5 min (85.6% vs. 93.0%, P < 0.001). Sensitivity with ZN 1% 5 min was not significantly higher than with 0.3% 5 min staining (89.9% vs. 86.5%). Routine examination using 1% 15 min ZN identified more positives than any of the study techniques.
Conclusions: Kinyoun cold staining sensitivity was unsatisfactory in field clinics. The sensitivity of the WHO/IUATLD recommended 0.3% fuchsin for 5 min was not significantly different from the original 1% ZN for 5 min, but 1% 15 min hot staining might be superior. A reduced fuchsin concentration together with a short staining time may leave too narrow a margin for error. TB programmes using hot ZN with a concentrated stain or longer staining time should not be urged to change.