OBJETIVO: Comparar os desfechos de tratamento de tuberculose em pacientes hospitalizados e aqueles tratados exclusivamente na atenção primária na cidade de São Paulo (SP), bem como determinar as variáveis mais associadas à internação. MÉTODOS: Pesquisa prospectiva e longitudinal, realizada entre janeiro e dezembro de 2007 em dois hospitais de grande porte e em serviços de saúde em duas regiões na cidade de São Paulo. Os dados foram coletados através de um questionário estruturado, no caso dos pacientes internados, e no Banco de Dados de Tuberculose da Secretaria Estadual de Saúde de São Paulo. RESULTADOS: Dos 474 pacientes incluídos no estudo, 166 estavam hospitalizados, e 308 eram pacientes ambulatoriais. A análise multivariada mostrou associações entre internação por tuberculose e diagnóstico de tuberculose em hospital/pronto-socorro (OR = 55,42), coinfecção por HIV (OR = 18,57), retratamento (OR = 18,51) e procura por outro serviço anteriormente (OR = 12,32). Para os pacientes hospitalizados e ambulatoriais, as taxas gerais de cura foram de 41,6% e 78,3%, respectivamente, ao passo que, para aqueles coinfectados por HIV, essas foram de 30,4% e 58,5%, enquanto as taxas gerais de mortalidade foram de 29,5% e 2,6%, respectivamente, ao passo que, para aqueles coinfectados por HIV, essas foram de 45,7% e 9,8%. CONCLUSÕES: Este estudo evidenciou maior gravidade, maior dificuldade de diagnóstico, menor taxa de cura e maior taxa de mortalidade nos pacientes internados que naqueles ambulatoriais. Além disso, os desfechos foram piores no subgrupo de pacientes coinfectados com HIV.
OBJECTIVE: To compare inpatient and outpatient treatment of tuberculosis, in terms of outcomes, in the city of São Paulo, Brazil, as well to determine which variables are most frequently associated with hospitalization. METHODS: A prospective, longitudinal study carried out between January and December of 2007, at two large hospitals and at outpatient clinics, in two regions of the city of São Paulo. For inpatients, data were collected with a structured questionnaire. Additional data were obtained from the São Paulo State Department of Health Tuberculosis Database. RESULTS: Of the 474 patients included in the study, 166 were inpatients, and 308 were outpatients. The multivariate analysis showed that hospitalization for tuberculosis was associated with hospital/emergency room diagnosis of tuberculosis (OR = 55.42), with HIV co-infection (OR = 18.57), with retreatment (OR = 18.51), and with having previously sought treatment at another health care facility (OR = 12.32). For the inpatient and outpatient groups, the overall cure rates were 41.6% and 78.3%, respectively, compared with 30.4% and 58.5% for those who were co-infected with HIV, whereas the overall mortality rates were 29.5% and 2.6%, respectively, compared with 45.7% and 9.8% for those who were co-infected with HIV. CONCLUSIONS: Among inpatients, tuberculosis appears to be more severe and more difficult to diagnose, resulting in lower cure rates and higher mortality rates, than among outpatients. In addition, tuberculosis patients co-infected with HIV have less favorable outcomes.
