Value of drug-licensing documents in studying the effect of postmenopausalhormone therapy on cardiovascular disease

Background In a previous study of pooled data from published trials, we fou nd no evidence to support the claim that postmenopausal hormone therapy (PH T) is associated with a decrease in cardiovascular disease. The purpose of this study was to see whether reports of clinical trials attached to drug-l icensing applications in Finland could be obtained for scientific purposes, whether they are useful for studying cardiovascular events resulting from PHT, and if so, whether these unpublished reports corroborate the results o f published reports. Methods Since clinical trials in drug-licensing documents are confidential, we had to obtain special permission from the Ministry of Social Affairs an d Health to use the data for research purposes. After permission was grante d, we studied the clinical sections of licensing documents for PHT drugs se nt by drug companies to the Finnish Drug Agency. We aimed to identify trial s that compared PHT and a placebo (or no therapy, or vitamin-mineral drugs) , and that reported on cardiovascular and thromboembolic events or superfic ial phlebitis. New trials were identified and their data were pooled with t hose of published trials. Findings 17 licensing applications for drugs used as PHT were found. The nu mber, type, and quality of reporting of clinical trials varied widely betwe en applications. The trials and their reporting of unanticipated adverse ev ents were mostly inadequate. Six new trials tie, those fulfilling the inclu sion criteria and not included in our earlier report) were found. The new t rials added little to the conclusions of previously published studies: the calculated odds ratios of cardiovascular and thromboembolic events for wome n taking PHT versus those not taking it was 1.97 (95% CI 0.84-4.63), compar ed with 1.65 (0.65-4.21) in our previous study. Interpretation In this case, unpublished trials added only a little to the data available from published trials, mainly due to the type of clinical da ta used in the licensing applications. The new data did not change the prev ious conclusion that clinical trials do not support a beneficial effect of PHT on cardiovascular diseases.