Tuberculosis in Tanzania: a follow-up of a national sampling survey of drug resistance and other factors.

A total of 1873 patients admitted to a random sampling survey in 15 of the 61 administrative districts in Tanzania in 1969 has been followed up at 1 year or later. The random sample included districts with a long-established tuberculosis service (A districts), those with a service of recent inception (B districts), and those with no specialized tuberculosis service (C districts). The main follow-up concerns 1607 patients with pulmonary tuberculosis of whom 693 had a positive culture at the initial survey, 557 a negative culture and 357 had not produced a specimen. At 1 year or later 12% of the 1607 patients were lost from observation, 10% were alive but with no specimen or no result, 60% were culture-negative, 5% were culture-positive, and 12% were known to be dead. The proportion of patients known to be dead was similar in the 3 types of service, but the proportion lost from observation was highest in the B districts, 24% compared with 7% in the A and 10% in the C districts. Most of the losses occurred early, 73% within the first 3 months. Of the 693 patients with positive culture initially 9% were culture-positive at 1 year, as were 1% of the 557 culture-negative initially. The estimated proportion culture-negative at 1 year for the patients culture-positive initially was highest in the A districts, 78%, and very similar in the B and C districts, 66% and 67%, respectively. The policies of therapy were studied in 1459 patients; 86% were treated in hospital initially for a mean duration of 63 days. The standard regimen of streptomycin, thiacetazone and isoniazid was prescribed in 78% of the patients initially, the proportions being 93% in the A, 48% in the B and 81% in the C districts. The proportion of patients who received or collected supplies of medicament for the full 12 months was only 35%, the proportions being 40% in the A, 20% in the B and 39% in the C districts.