[Methods for assessing the completeness of data in health information systems in Brazil: a systematic review]

The scope of this study was to review the literature on the methods used to assess the completeness of the data in health information systems. Searches were conducted in the Medline, SciELO and Lilacs databases using the key words "quality," "data quality," "quality of information" and "information systems." It included articles, dissertations and theses that assessed the fulfillment of variables. In the 19 studies reviewed, the epidemiological rationality systems were the most evaluated (90.5%). Studies on completeness are still scarce (one per year on average). Methods vary from isolated analysis to the evaluation of more than one dimension of quality. The term completeness is the word most used to characterize the fulfillment of variables. The Romero and Cunha score is the most frequent. The SIS databases and Datasus site were the sources and means of access to the most common data. Studies with data from the Southeast of Brazil were predominant. The indicators that measure the completeness vary from simple relative frequency to indices and synthesis-indicators. The heterogeneity of methods lacks standardization and dissemination of information to enable comparisons between the studies.