Five studies (n=134) were included in the review. Three studies were RCTs (n=86), of which two used crossover designs, one study was a prospective cohort, (n=20) and one was a pre-test post-test design (n=28).
The quality of the studies was low, with PEDro scores ranging from 3 to 5 out of a possible 10.
The results of the studies were mixed. No differences in pain were found when a roll cervical pillow and a semi-customised cervical pillow were compared with usual pillow, but positive effects on pain were reported for a water-based pillow compared with usual pillow, for a cervical pillow compared with a travel pillow, and for a cervical pillow with no comparator.
Results for use of analgesics, sleep quality and duration, activity, drop-outs and satisfaction were also reported.