Person-reported outcomes in registered randomised diabetes trials: A mapping review of constructs

Recently, efforts have been made to use and report person-reported outcomes (PROs) in randomised clinical trials (RCTs). Here, we aim to (1) assess the status of inclusion of PROs in registered RCTs over 5 years in people with type 1 or 2 diabetes, and (2) map the PRO measures (PROMs) onto predefined domains.

The largest trial registries (Clinicatrials.gov, International Clinical Trial Platform and ISRCTN) were systematically searched for RCTs in people with type 1 and/or type 2 diabetes of all ages between 2018 and 2023. Coding of PROs comprised: (1) PRO measure(s) included yes or no; if yes: (2) PRO(s) as primary outcome yes or no; and (3) mapping PROMs onto predefined PRO domains and per type of intervention.

N = 1543 trials met our inclusion criteria, of which n = 673 (44%) included PROs, assessed by 545 different measures. Twenty per cent of drug trials (n = 112) and 71% of behavioural interventions (n = 405) included PROs. In 149 trials (9.6%), a PRO was the primary outcome. The psychological functioning domain was most often assessed across all trials (21.6%), specifically in behavioural (44.8%) and medical device interventions (29.7%). In drug trials, the physical functioning and functional health domain was most included (9%). Across all trials, the social and family functioning domain was least assessed (3%).

We noticed an increase in the inclusion of PROs in diabetes RCTs. However, PROs are rarely included as primary outcomes in the majority of studies, particularly in drug trials. The heterogeneity of PROMs used in RCTs underscores the need for standardisation of PROs.