Socioeconomic characteristics of individuals who were and were not enrolled in two cardiovascular registry-based randomised controlled trials in Sweden: a descriptive cross-sectional study

Objectives To describe socioeconomic differences between individuals who were and were not enrolled in two registry-nested trials, and to assess whether socioeconomic variables provide additional discriminatory value in predicting enrolment beyond routinely collected clinical data. Design Descriptive cross-sectional study in Sweden using linked socioeconomic data. Setting Two Swedish Web System for Enhancement and Development of Evidence-based Care in Heart Disease Evaluated According to Recommended Therapies (SWEDEHEART) registry-based cardiovascular randomised trials, Thrombus Aspiration in ST-Elevation myocardial infarction in Scandinavia (TASTE) (Swedish centres only) and Bivalirudin versus Heparin in ST-Segment and Non-ST-Segment Elevation Myocardial Infarction in Patients on Modern Antiplatelet Therapy (VALIDATE-SWEDEHEART). Participants All individuals who were and were not enrolled in TASTE and VALIDATE-SWEDEHEART trials. Primary outcome measure For both trials, baseline socioeconomic characteristics of individuals who were and were not enrolled were compared using standardised mean differences. Enrolment was then modelled using random forests based on (1) trial clinical variables, (2) socioeconomic variables and (3) both combined. The areas under the receiver operating characteristic curve (AUC) of the models were compared to assess the additional value of socioeconomic data in predicting enrolment. Results In TASTE, 6841 of 11 436 potentially eligible individuals were enrolled, and in VALIDATE-SWEDEHEART, 5947 of 12 561 individuals were enrolled. Across both trials, enrolled individuals were more likely to be born in Sweden, be employed, be in a relationship, have a higher income and have a longer education duration than individuals who were not enrolled. The AUCs for enrolment models based solely on socioeconomic variables were similar to those using clinical variables. Adding socioeconomic variables to clinical models resulted in little or no improvement in AUC. Conclusions Socioeconomic characteristics differed between individuals who were and were not enrolled in two registry-based randomised controlled trials. The limited improvement in the predictive performance after adding socioeconomic variables to clinical models likely reflects a strong correlation between socioeconomic status and clinical profile. Trial registration number ClinicalTrials.gov identifiers: TASTE - NCT01093404, VALIDATE NCT02311231.