Safety and efficacy of cardiovascular medication in women and men; the added value of clinical care data

Women and men are different. This may seem obvious, but cardiovascular research too often treats women and men as interchangeable. Due to this, there is a lack of sex-specific knowledge about optimal treatment for patients suffering from cardiovascular diseases such as heart failure. Sex differences in pharmacokinetics and – dynamics can affect how women react to cardiovascular medication compared with men, supporting the idea of sex-specific approaches to treatment. However, women remain underrepresented in clinical heart failure trials and thus data on safety and efficacy of heart failure medication in women is scarce. This thesis explores whether clinical care data can close this evidence gap. Clinical care data comprise electronic patient records, which better reflect the patient population seen in daily care than for example clinical trials. This means that women are properly represented in these data, creating opportunities for sex differences research. Using a large clinical care dataset, we show that women more often stop their cardiovascular medication due to adverse drug reactions than men. In addition, women report different adverse drug reactions than men for the same medication types. We also show that the optimal dosage of heart failure medication may be lower for women compared with men. Women treated with ≤50% of the guideline-recommended dose had a lower mortality risk than those treated with higher doses, an effect that was not seen in men. Lastly, we show that primary prevention statins reduce mortality risk more strongly in women than in men, taking away remaining doubts surrounding the effectiveness of statins in women free of cardiovascular disease. However, it remains debatable whether clinical care data alone are sufficient to completely close existing evidence gaps as these data have their own inherent limitations. The last chapter of this thesis discusses how these limitations can affect the quality of clinical care data-based research, and how researchers can mitigate these effects via statistical and methodological techniques. We conclude that clinical care data are and will remain an important source of information for sex differences research, as current and future efforts aimed at quality improvements will only reduce the size and effect of any inherent limitations that these data have