[Influenza vaccination by Lower Saxonian general practitioners - a secondary analysis of physicians' data from 1995/1996, 2002/2003, and 2005/2006].

Vaccination against infection with human influenza virus is considered to be one of the most effective preventive measures available, especially when complications such as hospitalisation or death and indirect costs from off-work are considered. General practice is the preferable place for annual influenza vaccination because here the elderly and those endangered from bad health conditions are cared for frequently and regularly. The aim of this study was to find out the frequencies of influenza vaccination by Lower Saxonian contract physicians during three time periods with special respect to patients of older age or at risk. Data from two sources of health-care service have been analysed, namely from the contract physicians' union of Lower Saxony with all physicians vaccinating against influenza in the winter seasons of 1995/1996, 2002/2003, and 2005/2006, and from direct access to the electronic practice record system of 79 general practices via the BDT software data interface. Contract physicians, of whom more than 90 % are general practitioners, from Lower Saxony, vaccinate patients of the statutory health insurance against influenza in markedly increasing numbers, since 1995/1996 and 2002/2003 up to 2005/2006. Those over sixty years old or at high risk from bad health conditions are vaccinated up to seven-fold more frequently, compared to other patients. Influenza vaccination coverage rates (VCRs) are significantly higher in small and in medium-sized practices, compared to those with many patients. Nevertheless, influenza VCRs in Lower Saxony are not yet as high as would be necessary or desirable when compared internationally. Secondary analysis of aggregated health service data revealed inconsistencies in the primary material on cross-checking and validating, probably being caused during the external process of data aggregation and anonymisation. Also major systematic obstacles were found in the subsequent process of analysis. Nevertheless, fundamental results have been produced and are valid for all statutory health insured Lower Saxonians. Data from direct access to electronic practice records allowed for a deeper and multi-faceted insight into 101 928 patients of the same population, limited by the possibility of selection bias ('convenience sample'). Secondary analysis of health service data from different sources and their cross-check comparison is possible and successful. It is important to inform and involve the holder of the primary data extensively, following the guide lines of "good practice secondary data analysis".