Systematic restaining of sputum smears for quality control is useful in Burundi

Routine tuberculosis control services in Burundi.

To determine whether systematic restaining of sputum smears for acid-fast bacilli (AFB) prior to rechecking quality assessment is necessary.

Blinded rechecking of peripheral routine smears, including a second control of discordants, before and after restaining.

Without restaining, 10/825 (1.2%) negative, and 59/189 (31.2%) positive results were declared false. After restaining, there were 34 (4.1%) false negatives and 13 (6.9%) false positives, both highly significant changes. Before restaining, quantification of positive smears was usually considered too high, while after restaining 41 out of 42 positives were found to have too low readings.

Despite mild climatic conditions in Burundi, restaining all slides before rechecking revealed an unrecognised, widespread problem of false negatives, rather than false positives. This indicated the need for critical re-appraisal of the standard procedure using cold staining, rather than re-training, as would have been inferred from results without restaining. Systematic restaining of all slides prior to rechecking may be more widely needed in National Tuberculosis Programmes to cover all possible serious causes of error. Cold staining should be avoided in field programmes since its performance is easily affected by frequently encountered adverse factors.