Objective: To evaluate the association of the characteristics of the family system in adherence observed treatment short in patients with pulmonary tuberculosis in three cities of the south-southeast of Mexico.
Material and methods:Cases of pulmonary tuberculosis were studied, who started treatment at first-level units. The follow-up was 6 months or until the patient lost his grip, emigrated to another city, died, or decided not to continue the study. Intrafamilial relationships were assessed, family functioning (FACES III) and social network (DUKE UNC-11); adherence was assessed with test-Greene Morinsky. We calculated incidence of non-adherence and relative risk factors studied. Variables with significant differences in the bivariate analysis were subjected to the proportional hazards model of Cox.Results: Two hundred and thirty four patients were included; the total track joined 36,937 days, with a median of 175 days. Patients with dysfunctional family have RR=8.95 (95% CI=4.51-17.76, p<0.001) compared with those with functional family, and patients with non-functional network showed RR=2.22 (95% CI=1.13-4.35, p 0.002) compared with those with functional social network. In the Cox regression model adjusted for education, statistical significance was maintained for family functionality. Family functioning influences the cohorts studied in treatment adherence