Evaluation of the effectiveness of hospital treatment of acute stroke in National Health System: use of mortality as indicator of performance

OBJETIVO: Avaliar a efetividade do tratamento hospitalar do Acidente Vascular Cerebral Agudo no Sistema ??nico de Sa??de SUS, comparando a mortalidade hospitalar ajustada entre pacientes que realizaram ou n??o a tomografia computadorizada. M??TODO: A fonte de informa????o utilizada foi o Sistema de Informa????o Hospitalar do SUS (SIH-SUS). Foram selecionadas 328.087 interna????es ocorridas no SUS em todo o territ??rio nacional entre abril de 2006 e dezembro de 2007. As interna????es foram reunidas e estudadas em 4 grupos: Acidente Isqu??mico Transit??rio (CID-10: G459); Acidente Vascular Cerebral Hemorr??gico (CID-10: I60; I61 e I62); Acidente Vascular Isqu??mico (CID-10: I63) e Acidente Vascular Cerebral n??o especificado (CID-10: I64). Foram utilizadas as mortalidades hospitalares at?? o s??timo e at?? o trig??simo dias, como medidas de resultado para comparar pacientes que realizaram e n??o realizaram tomografia computadorizada. RESULTADOS: Em geral os pacientes que realizaram a tomografia computadorizada apresentaram menores taxas de mortalidade hospitalar em rela????o ??queles que n??o realizaram o exame, sendo essa diferen??a em favor da realiza????o do exame observada principalmente at?? o segundo dia de interna????o em todos os 4 grupos. A diferen??a entre os que realizaram e os que n??o realizaram o exame foi acentuada no grupo do Acidente Vascular Isqu??mico (OR: 0,325; p>0,000), sendo que no primeiro dia o odds ratio foi de 0,021(p>0,000), em favor dos que realizaram o exame. CONCLUS??ES: Os exames de tomografia computadorizada no SUS, em geral, s??o realizados mais tardiamente que o recomendado pela literatura. Apesar das limita????es ainda existentes na qualidade da informa????o diagn??stica dispon??vel no SIH-SUS que restringiram a estrat??gia de ajuste de risco empregada nesse estudo, sugere-se o uso da tomografia computadorizada, o mais cedo poss??vel, como tecnologia auxiliar no diagn??stico e tratamento do AVC. Al??m disso, sugere-se o emprego mais amplo de medidas de desempenho, tais como a mortalidade hospitalar aqui empregada, para o monitoramento da qualidade do cuidado prestado no ??mbito do SUS.OBJECTIVE: To evaluate the effectiveness of hospital care of the Stroke in the Brazilian Health System by comparing adjusted hospital mortality rate between patients who had done or not CT scanning. METHOD: Brazilian hospital information systems was the data source used. Three hundred twenty eight thousand and eighty seven inpatients were included in this study, covering all the Brazilian territory between April of 2006 and December of 2007. The inpatients had been grouped in 4 groups: Transient cerebral ischaemic attack, unspecified (ICD-10: G45.9); Haemorrhage Stroke (ICD -10: I60; I61 and I62); Cerebral infaction (ICD -10: I63) and Stroke not specified as haemorrhage or infarction (ICD -10: I64). Hospital mortality until seventh and the thirtieth day was used as a result measure to compare patients who had been submitted or not to Computerized tomography (CT) scanning. RESULTS: In general the patients who submitted to TC scanning presented lower hospital mortality rates in relation to those who had not done CT scanning, being this difference for the accomplishment of the examination observed until the second day of in-hospital all stroke group. The group of the ischemic stroke presented the higher difference among those who were submitted or not to Computerized tomography (CT) scanning (OR: 0.325; p>0.000). In the first in-hospital day for the stroke group the odds ratio 0.021 (p>0.000) in favor of the group who had done the CT. CONCLUSIONS: The TC scans in the Brazilian health system, in general, are used with a greater delay than the recommended in literature. This leads to a reduction of the benefits of the examination. Although the limitations in the data quality of Brazilian hospital, the use of the TC scanning, as soon as possible, is suggested as auxiliary technology in the diagnosis and treatment of the stroke. Furthermore, it is also suggested a more frequent employment of performance indicator, such as hospital mortality rate, to monitoring quality of care in Brazil