OBJECTIVES : Core outcome sets (COS) represent a minimum set of outcomes to measure in a given condition to enable outcome comparison across studies. Patient reported-outcomes (PROs) are becoming increasingly common, together with instruments for measuring them (i.e. patient reported outcome measures, PROMs). This cross-sectional study aimed to map the PROMs recommended in existing COS, as well as to document the outcome domains covered and overlaps across disease areas.

METHODS The Core Outcome Measures in Effectiveness Trials (COMET) database was searched in July 2017 to identify a subset of COS development studies recommending PROMs. A pilot-tested template was used to collect standard information. First, we extracted recommended PROMs domains and matched these to a 38-domain outcomes taxonomy. Second, we extracted items and subscales from full-texts of recommended PROMs, and matched individual items with the same 38-outcome taxonomy. Third, we used cross-tabulation methods to investigate the adoption of individual PROMs and to summarize the domains targeted by the studies across different disease areas.

RESULTS We included 71 COS studies recommending 251 unique PROMs. The majority of the instruments (86.4%) were recommended in just one COS, 23 were recommended in 2 COS, 6 in 3 COS, 2 in 4 COS, and 3 in more than 4 COS (i.e. HAQ-DI, EQ-5D, and SF-36). 59.0% were full questionnaires, 18.7% were questions or subscales of existing instruments and 22.3% were stand-alone questions. Over 75% were disease-specific PROMs, and the rest were generic; 5 PROMs only (i.e. EQ-5D, HUI3, QWB, SF-6D, PORPUS) were preference-based. ‘Global quality of life’ and ‘physical functioning’ were the most common domains, covering, respectively, 24.9% and 21.6% of all target domains, and therefore those presenting most overlap across disease areas.

CONCLUSIONS : This study showed the highly fragmented status of PROMs recommendations in COS and unveiled the scope for harmonizing their use in clinical research and practice.