Since 2005, German hospitals are required by law to publish structured quality reports (QRs). Because of the detailed data basis, the QRs are being increasingly used for secondary data analyses in health services research. Up until now, methodological difficulties that can cause distorted results of the analyses have essentially been overlooked. The aim of this study is to systematically list the methodological problems associated with using QR and to suggest solution strategies. The QRs from 2006-2012 form the basis of the analyses and were aggregated in a database using an individualized data linkage procedure. Thereafter, a correlation analysis between a quality indicator and the staffing of hospitals was conducted, serving as an example for both cross-sectional as well as longitudinal studies. The resulting methodological problems are described qualitatively and quantitatively, and potential solutions are derived from the statistical literature. In each reporting year, 2-15% of the hospitals delivered no QR. In 2-16% of the QRs, it is not recognizable whether a report belongs to a hospital network or a single location. In addition, 6-66% of the location reports falsely contain data from the hospital network. 10% of the hospitals changed their institution code (IC), in 5% of the cases, the same "IC-location-number-combination" was used for different hospitals over the years. Therefore, 10-20% of the QRs cannot be linked with the IC as key variable. As a remedy for the linking of QR, the combination of the IC, the address and the number of beds represents a suitable solution. Using this solution, hospital network reports, location reports and missing reports can be identified and considered in an analysis. Secondary data analyses with quality reports provide a high potential for error due to the inconsistent data base and the problems of the data linkage procedure. These can distort calculated parameters and limit the validity of results. Only the unequivocal identification of the reporting hospitals guarantees meaningful results.