The yield and usefulness of PAIN+ and PubMed databases for accessing research evidence on pain management: a randomized crossover trial.

Only 77 participants were analysed, and the study was not powered to compare disciplines, so results may not apply equally to physical therapists, occupational therapists, physicians, nurses, or psychologists. Participants were already PAIN+ users with more than one year of exposure, so their ratings may reflect familiarity with that service. The text says 76 clinicians enrolled, while the tables report 77 participants in the analysed sample. There was no washout period, and the first search could have influenced the second search. Usefulness and quality were judged from titles and abstracts only, not full papers, and only the top three articles were rated. The first-page yield was capped at 20 articles, so total efficiency could not be fully measured. The databases were sorted by their default algorithms at the time, best match for PAIN+ and most recent for PubMed, which could affect perceived usefulness. The discipline-specific questions were different for each profession, which may add heterogeneity. Blinding was incomplete because the PAIN+ platform name was visible during the online interface. The study was completed between 2013 and 2015, so database content and sorting may have changed since then.

The article reports McMaster University as the funder and does not list other conflicts of interest.

Do not conclude that PAIN+ should replace PubMed for pain evidence searches. PAIN+ top three articles were rated more useful, but PubMed was preferred for multidisciplinary questions and its discipline-specific articles were rated more likely to change practice.