Objectives: This study describes the epidemiological profile of cases of repeated treatment of tuberculosis occurring among residents of the city of Recife in the year of 1997. Material and methods: A cross-sectional study was performed and 240 cases of repeated treatment were studied, representing 16.2% of all cases of tuberculosis. Results: Interruption of previous treatment was the principal reason for repeated treatments (55.8%), followed by cases of recurrence of the disease (39.2%). Patients undergoing repeated treatment for tuberculosis are characterized as predominantly male, ages between 30 and 49, with little or no school education, with a significant predominance of pulmonary tuberculosis. It is striking that 42.7% of the cases of repeated treatment had a history of two or more previous courses of treatment. With regard to the outcome of the repeated treatment, 52.5% of the cases showed unfavorable results. Among these, 44.2% of the patients gave up the course of treatment and 39.6% followed it for the first month. The cases that returned for treatment following recurrence showed a higher percentage of successful outcomes (56.7%), in comparison to those that returned after interrupting previous treatment (40.2%), this difference being statistically significant (χ2 = 19,55; p < 0,01). It was found that, among the cases of pulmonary tuberculosis, 27.1% resumed chemotherapy without undergoing the Koch bacillus catarrh test and only 27.1% carried out BK culture with sensitivity test. Conclusions: The results show the need to enhance tuberculosis treatment and follow-up procedures within the sphere of the health care service. An adequate treatment is the cheapest method to avoid drug resistance and should be understood as a strategy to improve the tuberculosis surveillance system.

Keywords: Tuberculosis. Pulmonary tuberculosis. Repeated treatment.