Major reasons for the considerable heterogeneity among published results of heart valve surgery are inconsistency in follow up techniques, reporting systems and classification of adverse events. The present recommendations are intended to harmonize the presentation of clinical material in order to improve comparison of data from different sources for the analysis of pooled data. The quality of an observational study is largely, if not entirely, due to the follow up technique, which may be graded according to six categories: Self-reporting of adverse events/well-being by the patients may be classified 'excellent'; if the information is gathered and re-checked at short-term intervals. Data obtained from in-hospital or outpatient examinations by qualified examiners at least twice a year or other personal contact through qualified examiners may be regarded as 'sufficient', if the results are re-checked by contacting the treating home physician. All other follow up techniques may be regarded as inappropriate. Consequences of complications are entirely dependent on severity and possible sequelae. It is therefore recommended to grade any reported complication according to its severity by utilizing a score system. Embolisms are best categorized by utilizing the performance status scale. Bleeding events may be categorized according to severity as fatal, major (requiring hospital transmission with transfusion, surgery or with permanently increased disability) or minor (not requiring hospital admission, surgery or transfusion). In some cases it will remain unclear whether an event was primarily embolic or hemorrhagic. These complications should be summarized as 'not categorized'. The reporting of morbid events due to thrombosis, embolism and bleeding should go along with information regarding the quality of antithrombotic management.