Classification of leprosy cases under field conditions in Bangladesh. II. Reliability of clinical criteria.

In 2 non-governmental organization projects 244 new leprosy patients in Bangladesh were classified in the field according to clinical criteria i.e. number of skin lesions and number of enlarged nerves. Comparison of these classification results with the results of skin smears and biopsies yielded a sensitivity (for detection of a MB case) of 92.1%, but the 'unconfirmed MB rate' amounted to 52.6%. In order to improve the reliability of the operational classification, several additional clinical criteria were investigated. It was found that neither the presence of anaesthesia in the skin lesions nor the presence of grade 2 disabilities or peripheral anaesthesia or voluntary muscle testing (VMT) impairment contributed to an improved classification. Counting the number of body areas showing signs of leprosy, which had proven very useful in other programmes, did not result in a more reliable classification in the 2 projects in Bangladesh. The presence of clinical signs of lepromatous leprosy, more specifically nodules or diffuse infiltration, could be a useful addition to the classification criteria. If the sensitivity must remain higher than 90%, the lowest 'unconfirmed MB rate' obtainable in Bangladesh, using clinical criteria only, is 46.4%, for a sensitivity of 91.0%. However, the inclusion of skin-smear results in the classification criteria could improve the sensitivity to 96.6% and lower the 'unconfirmed MB rate' to 40.3%. A reduction in MB overclassification will result in more efficient and more cost-effective leprosy control programmes.