Avaliação de uma intervenção com codificadores hospitalares para melhoria da informação sobre internações por causas externas

AUTOR(ES)
FONTE

IBICT - Instituto Brasileiro de Informação em Ciência e Tecnologia

DATA DE PUBLICAÇÃO

30/06/2010

RESUMO

The Hospital Information System of the Brazilian Unified Health System (SIH-SUS) isan important national database on hospital morbidity. However, the underreporting ofsome types of causes has become a problem in terms of data reliability. Theobjective of this study is to characterize the process of morbidity coding in hospitaladmissions by external causes carried out by the SUS, and to evaluate the impact ofan intervention with hospital data coders on the quality of this information. The studywas conducted in Londrina (PR), in three phases. The first phase was undertook tocharacterize hospital data coders according to personal characteristics, training incoding hospital causes of admissions and work’s characteristics, as well as to identifythe difficulties related to the process of coding external causes of morbiditymentioned by these professionals. Data were collected using a structured formapplied in their working sites. The second phase compared the quality of the codingbefore and after an intervention that consisting of a training session which includedthree lectures on the subject and the application of an instrument, before and afterthese explanatory lectures, with admissions due to external causes to be coded. Thethird phase included an evaluation of the SIH-SUS records regarding causes ofadmission coding. In this phase, physician’s reports on hospital admissions (AIHs),from three general hospitals, in two distinct periods (sixty days before and sixty daysafter the training sessions) were revised manually. These reports were transcribedand coded by two specially trained persons. The SIH databank was obtained fromthe DATASUS site, in the same selected periods. Using the AIH number as theidentifying field between the two banks, it was possible to compare the data fromthese two sources before and after the training, using the Epi Info program, version6.04d. As for the codifiers profile, there was a predominance of women between 40and 50 years of age; 85.7% of the participants said they had learned to code in theirown work, without any previous training. Difficulties with the process of coding mostlyreported were: lack of information about the external cause on the physician’s reportand the illegibility of the information. In the evaluation carried out immediately afterthe training session, an increase in the number of right answers was verified for mostcodes of causes categorized in three characters levels, with a rate of right answersfor after/before of 1.22 (p=0.03). In the third phase, an improvement in the quality ofinformation on external causes in the SIH was observed in two hospitals. Hospital 1showed the best quality of information on admissions due to external causes, beforeand after the training sessions (Kappa 0.85 before and 0.90 after). In hospital 2 therewas a reduction in all agreement indicators after the training and, in hospital 3, allagreement indicators, according to SIH information and the research, increased inthe post-intervention period. Training sessions may be strategies which cancontribute to the improvement of the quality of the information; however, to beeffective, they must be done periodically, involving all the people responsible forgenerating this type of information, and they should also be connected with thehealth services transformation proposals.

ASSUNTO(S)

sistemas de recuperação da informação - hospitais recuperação da informação - confiabilidade (probabilidade) hospitais - administração administração e saúde pública violência - mortalidade acidentes - mortalidade saúde pública information storage and retrieval systems hospitals hospitals administration violence mortality accidents mortality

ACESSO AO ARTIGO

Documentos Relacionados