3.42 General discussion 195 increased participation of Asian individuals from 1% in 2001-2004 to 20% in 20132016 though trial participation from Black and Hispanic patients from other regions either remained stagnant or declined.15 In this aspect, expansion of sites to SubSaharan Africa, Latin America and other Asia Pacific regions would contribute to racially/ethnically more generalizable HF trials. Assessing generalizability based on race/ethnicity is also complicated by variation in nomenclature and classification of racial/ethnic groups between populations. For instance, the Malaysian population is broadly classified as Asians or South-East Asians in clinical studies43, but in itself can be subclassified to its Malay ethnic majority (51%) and Chinese, Indian, Indigenous group minorities and noncitizens.44 Within what is regarded a homogenous population of South-East Asians, we observed inter-ethnic differences in incident heart failure hospitalizations, whereby Indians had a 20% higher risk compared to the nation’s major ethnic group. Similarly for HF outcomes, we report disparities between ethnicities within a setting that has tax-funded universal health access. Others, consisting mainly Indigenous groups, experience markedly higher inpatient mortality (1.9-fold higher) and 30-day mortality (1.3-fold higher) compared to Malays. On the contrary, Indians had 20% and 13% lower risk of 30-day and 1-year mortality. These observed ethnic disparities for HF remained after adjustment for age, sex and year of admission, highlighting that racial/ethnic information in trials need to go beyond broad racial groups such as Asian/non-Asian or Hispanic/non-Hispanic42 and include specific countries, origins and ethnicities for disentangling heterogeneity of risks and therapeutic responses. Qualitative representation on HF severity, prevalent comorbidities and background heart failure therapy By estimating eligibility on inclusion and exclusion criteria, we showed that enrolment for HFrEF trials have become stricter by more than two-fold in the past 20 years. One of the reasons to this change is growing numbers of exclusion criteria per trial as part of strategies to maximize validity of causal estimates, termed broadly as practical (efficacy) enrichment.45–47 Although opinions on the value of representation in study samples is non-unanimous,48 it is clear that extensive lists of exclusion cause greater therapeutic uncertainty for underrepresented patients which becomes more problematic when it concerns a significant fraction of the treated population.
RkJQdWJsaXNoZXIy MjY0ODMw