3.1 Sex differences in generalizability of HFrEF trials 99 comparison to males in the RCT-eligible population, with cardiovascular mortality risk 43% higher in the RCTs than expected in the registry. The higher percentage of CV death in the RCTs is consistent with use of enrichment strategies in inclusion/exclusion criteria. However, despite the same inclusion/exclusion criteria for males and females, females in trials showed no evidence of enrichment. On the contrary, there seemed a trend towards lower-than-expected CV mortality risk for females enrolled in trials compared to eligible females from the registries. Enrichment strategies are often used in RCTs to identify patients who will experience CV events sooner than non-CV events in order to decrease time to target endpoint and improve efficiency of RCTs.45 It is unclear what could explain this opposing response to enrichment. One explanation could be that there are some sex-specific factors affecting patient selection and willingness to participate. These are numerous reports that point out that females can be underrepresented due to significant patient-oriented biopsychosocial barriers which results in the exclusion of females who are elderly, obese, depressed, nonwhite, with greater comorbidity, and who have less social support.7,36–38 This could hold true for the studied population here, as baseline characteristic differences between the RCT group and registry groups were larger for females than males. In addition, although females and males in the RCT were prescribed medication similarly, females in the registries were less often prescribed anticoagulants and ACEI or ARBs, which is consistent with previous literature.2,4,15 This is concerning because the use of ACEI or ARBs was a significant driver for RCT-ineligibility in registry females and is possibly an additional barrier for female recruitment in RCTs. Patients in RCTs are also known to receive better care, and gender-related differences in clinical management has been shown to negatively affect females in the real world.42,44 Taken together, these barriers could lead to a healthier female RCT population that is less representative of their real-world counterparts, especially in comparison to males. Lastly, it is also conceivable that risk factors used to calculate the standardized mortality ratios have sex-specific impact. For example, diabetes, hypertension, and smoking have long been recognized as important risk factors in HF development, with evidence of a greater effect in females due to earlier onset adverse LV remodeling with increased wall thickness.38,40,46,47 Although the MAGGIC risk model was chosen for this study due to its validity in predicting mortality for both sexes,28,47 there are valid arguments for testing the interaction with sex in the
RkJQdWJsaXNoZXIy MjY0ODMw