Risk of tuberculosis among contacts of isoniazid-resistant and isoniazid-susceptible cases.

OBJECTIVE: To compare the risk to household contacts of isoniazid (INH) susceptible and INH-resistant cases of tuberculosis (TB) in a rural community in South India.

METHODS: In all, 5562 contacts of INH-susceptible and 779 contacts of INH-resistant patients and 246,845 persons with no TB case in the home were followed for 15 years, with surveys every 2.5 years comprising radiographic and sputum examination, selective follow-up of high-risk individuals and passive surveillance. If a new case developed, the household members were assigned to the 'INH-susceptible' (n = 7088) or 'INH-resistant' series (n = 526), whichever appropriate. Logistic regression and Cox's proportional hazards model were employed.

RESULTS: The baseline prevalence of tuberculous infection was respectively 70% and 56% in contacts of INH-resistant and INH-susceptible patients (P < 0.001), compared to 46% in non-contacts. The incidence of culture-positive TB was respectively 295 and 311 per 100,000, compared to 162/100,000 in non-contacts. The adjusted hazard ratios were 2.4 and 2.0 for contacts of INH-resistant and INH-susceptible patients.

CONCLUSION: The baseline prevalence of tuberculous infection was substantially higher in contacts of INH-resistant than INH-susceptible patients, but the incidence of tuberculous disease over a 15-year follow-up was similar in the two series, and twice as high as in non-contacts.