Incidence of in-hospital adverse events in the State of Rio de Janeiro, Brazil: Evaluation of patient medical record

OBJETIVO: Avaliar a qualidade da informa????o dos prontu??rios de tr??s hospitais de ensino do Estado do Rio de Janeiro, participantes do estudo de base para a estimativa da incid??ncia de eventos adversos (EA). MATERIAL E M??TODOS: Estudo descritivo, baseado em informa????es coletadas na revis??o de prontu??rios do estudo de base. Foi aplicado escore de completitude, medido pela propor????o de informa????o ignorada, composto pelos graus de avalia????o: excelente (menor que 5 por cento), bom (5 por cento - 10 por cento), regular (11 por cento - 20 por cento), ruim (21 por cento - 50 por cento) e muito ruim (mais de 50 por cento). Foram calculados propor????es e intervalos de confian??a de 95 por cento, para cada informa????o do prontu??rio. A an??lise foi realizada para o conjunto dos pacientes, para os tr??s hospitais e para pacientes com e sem EA. Foram calculadas m??dias na an??lise do conjunto de vari??veis e, para fins de compara????o, foi realizado o teste t de Student. Foi aplicado o teste qui-quadrado e a estat??stica de Fisher na an??lise comparativa entre pacientes com e sem EA. RESULTADOS: A qualidade dos prontu??rios foi considerada ruim, no conjunto dos pacientes. As vari??veis que apresentaram maior propor????o de aus??ncia de informa????o foram: "Avalia????o inicial da enfermagem" (63,9 por cento) e "Avalia????o do paciente pelo assistente social" (80 por cento). O hospital 3 apresentou melhor qualidade dos prontu??rios e o hospital 1 apresentou o pior resultado. Os pacientes com EA apresentaram melhor qualidade dos prontu??rios do que aqueles sem EA. CONCLUS??ES: Informa????es indispens??veis ao cuidado apresentaram baixo registro. Ressalta-se a import??ncia da elabora????o de medidas que visem melhorias na qualidade do prontu??rio, que ir??o refletir na qualidade da assist??ncia ao paciente.To evaluate the quality of information obtained from medical records of three teaching hospitals in the State of Rio de Janeiro, Brazil, which participated in a previous study on the incidence of adverse events (AE). METHODS: Descriptive study based on the information collected from medical record review during the incidence study. An evaluation score of completeness, measuring the proportion of absent information was applied, based on the following items: excellent (less than 5 percent), good (5 percent - 10 percent), moderate (11 percent - 20 percent), bad (21 percent - 50 percent), and very bad (more than 50 percent). Proportions and 95 percentconfidence intervals were calculated for each variable obtained from medical records. Analysis was developed for the whole group of patients, by hospital groups, and by patients with and without AE. For analyzing the group of variables, means were estimated and the t-student's test was applied for comparisons. The Chi-squared test and Fisher's statistics were applied for comparing patients with and without AE. RESULTS: The quality of information obtained from patient medical records was considered bad in the group of patients as a whole. Variables with the highest proportion of absent information were: "Nurse's first evaluation of the patient" (63.9 percent) and "Social worker's evaluation" (80 percent). Hospital 3 had the best results in terms of quality of medical records and hospital 1 had the worst results. The quality of medical records of patients with AE was better than the quality of medical records of patients that had no AE. CONCLUSIONS: Variables considered essential for patient's healthcare were not well documented. Measures to develop quality of medical records are necessary to improve patient's healthcare