CHAPTER 2.2 72 including any dose background therapy of ACEI/ARB and BB and found between 10 and 30% absolute decrease in eligibility, which seems a fair trade-off particularly for evaluating incremental benefit of add-on therapies. However, stepping up required background therapy to include MRA substantially lowers eligibility by two-thirds, highlighting the need to base decisions for selection criteria not only by guidelinedirected medical therapy (GDMT) recommendations but on the actual use of these GDMTs. Instead of mandating specific drug classes, an alternative involves utilising a score to summarise type and intensity of background GDMT as basis to compare within and between trials.40 Among the strengths of this study is extensive analysis of eligibility criteria for trials from ClinicalTrials.gov , which is among the most complete trial register on drugs and devices by major pharmaceutical companies . 14 As therapeutics are eventually aimed at global markets, assessing eligibility using multinational registries from Asia and Europe enables testing the hypothesis for equal eligibility across patient profiles. There are also several limitations in this study. Information on trial phase was not available for 54% of studies labelled as interventional. Natriuretic peptide levels as a criterion could not be compared here, due to incomplete data from ASIAN-HF registry. That said, this diagnostic and prognostic criterion is infrequently measured in limited resource settings and selecting patients by natriuretic peptides is known to affect distribution of trial patient characteristics2, raising further questions on generalizability. Next, eligibility criteria recorded in ClinicalTrials.gov represent only part of the full list, albeit the most important ones. Thus, the proportion of eligibility criteria here is likely underestimated. Similarly, because only a subset of criteria could be accounted for when calculating eligibility scores, these would be overestimated compared to actual eligibility. As the definitions for HF subtypes by LVEF evolves with time, HFmrEF subtype is more likely covered within the HFrEF trials with some minimal overlap with HFpEF trials. It is necessary to acknowledge that both ASIAN-HF and BIOSTAT-CHF cohorts apply selection criteria and therefore have narrower spectrum of real-world patients than those found within electronic medical records (EMR). Nonetheless, present challenges such inherent lack of clarity in analogue clinical text, unstructured data formats and restriction to single centres or payer41 preclude the use of EMRs for large scale comparisons. For these reasons, HF registries represent the next best data source, given that they are specifically designed for the disease and have benefits of
RkJQdWJsaXNoZXIy MjY0ODMw