Abstract

SETTING:

Tribal villages in the jungles of the Jawadhu hills, South India.

OBJECTIVE:

To estimate the prevalence of tuberculosis (TB) infection and disease in a remote tribal population.

DESIGN:

A cross-sectional survey with two-stage screening for identification of cases. A stratified probability proportional sample with the hamlet as the unit.

METHODS:

Among 56 revenue divisions with a population of about 66,000, 24 revenue divisions were selected. Among 26,320 persons registered, children < 10 years were tuberculin tested and reactions were read after 72 hours. Those over 15 were X-rayed, and tuberculosis symptoms were investigated. Sputum was collected from those with abnormal X-ray or symptoms and examined for smear and culture positivity and sensitivity.

RESULTS:

Of the 6952 children tested and read, 5% had BCG scars and the prevalence of infection was 5%. The annual risk of infection was 1.1. Among adults, the prevalence of bacillary cases was 8/1000 and X-ray cases 29/1000. The prevalence of bacillary disease was higher among males, particularly with increasing age. Thirty symptomatic cases had normal X-rays and 63 X-ray cases had no symptoms. Thus prevalence would have been underestimated if either method had been used alone for screening. Isoniazid resistance was seen in 12% of patients, two of whom also had rifampicin resistance (2.6%).

CONCLUSIONS:

The prevalence and pattern of tuberculosis in this tribal group is similar to that observed in non-tribal areas.