How important is aspirin adherence when evaluating effectiveness of low-dose aspirin?

Low-dose aspirin (LDA) is advocated for women at high-risk of pre-eclampsia, providing a modest, 10%, reduction in risk. Cardiology meta-analyses demonstrate 18% reduction in serious vascular events with LDA. Non-responsiveness to aspirin (sometimes termed aspirin resistance) and variable clinical effectiveness are often attributed to suboptimal adherence. The aim of this review was to identify the scope of adherence assessments in RCTs evaluating aspirin effectiveness in cardiology and obstetrics and discuss the quality of information provided by current methods. We searched MEDLINE, EMBASE and the Cochrane Library, limited to humans and English language, for RCTs evaluating aspirin in cardiology; 14/03/13-13/03/16 and pregnancy 1957-13/03/16. Search terms; 'aspirin', 'acetylsalicylic acid' appearing adjacent to 'myocardial infarction' or 'pregnancy', 'pregnant', 'obstetric' were used. 38% (25/68) of obstetric and 32% (20/62) of cardiology RCTs assessed aspirin adherence and 24% (6/25) and 29% (6/21) of obstetric and cardiology RCTs, respectively, defined acceptable adherence. Semi-quantitative methods (pill counts, medication weighing) prevailed in obstetric RCTs (93%), qualitative methods (interviews, questionnaires) were more frequent in obstetrics (67%). Two obstetric RCTs quantified serum thromboxane B and salicylic acid, but no quantitative methods were used in cardiology Aspirin has proven efficacy, but suboptimal adherence is widespread and difficult to accurately quantify. Little is currently known about aspirin adherence in pregnancy. RCTs evaluating aspirin effectiveness show over-reliance on qualitative adherence assessments vulnerable to inherent inaccuracies. Reliable adherence data is important to assess and optimise the clinical effectiveness of LDA. We propose that adherence should be formally assessed in future trials and that development of quantitative assessments may prove valuable for trial protocols.