Epidemiological analysis of tuberculosis treatment outcome as a tool for changing TB control policy in Israel

Sensing an inadequacy of tuberculosis control due to an influx of TB associated with immigration, we analyzed TB treatment outcome in Israel by population groups.

To provide an epidemiological basis necessary for any new national TB control policy, and to bring it to the attention of the medical profession in Israel and abroad since its results led to a change in Israel's TB control policy.

We reviewed all TB cases notified during the period 1990 to September 1992. "New cases" (820 cases, 93.5%) and "re-treatment cases" (57 cases, 6.5%) were analyzed according to three mutually exclusive groups: "successful outcome," "death," and "potentially unsatisfactory outcome" (according to WHO/IUATLD definitions).

Of 820 "new cases," 26.6% had a "satisfactory outcome," 68.5% had a "potentially unsatisfactory outcome" and 4.9% died; compared to 47.4%, 45.6% and 7% among 57 "re-treatment cases," respectively. Using logistic regression analysis, outcome was associated with the district health office (P < 0.0001), the TB "experience" of the notifying clinic (P < 0.0001), and the form of TB (P = 0.02). No significant relationships were obtained for population groups, gender and age, interval between arrival in Israel and TB notification, and bacteriological results.

Non-supervised TB treatment resulted in poor outcomes regardless of population groups. Better outcomes occurred in the larger TB clinics. Therefore, in addition to measures such as adequate drug supplies, reorganization of TB laboratories and training of TB personnel, we recommend the "directly observed treatment short-course" for all cases as well as reducing the number of treatment centers thereby increasing their case load.