Prevalence study of the randomized controlled trials in the Journal of Intellectual Disability Research: 1957-1994.

Systematic reviews of care are increasingly potent guides to clinical practice, and it is important that all relevant randomized controlled trials (RCTs) are identified by those within the speciality of learning disability who produce such works. All RCTs in the Journal of Intellectual Disability Research and its predecessor were identified by hand-searching (1957-1994), and the frequency, origin, intervention and quality of reporting of randomization were described. Electronic searches for the trials were undertaken for the years 1974-1994, and the quality of indexing was inspected and tested. These electronic searches were then compared to a 'gold standard' search. Fifty-six RCTs were identified. None contained the world 'randomized' in the title and only nine mentioned it in the abstract. Out of the 37 RCTs published between 1974 and 1994, 36 are in PsycLIT and 37 in MEDLINE. One MEDLINE record contained the wrong abstract. The methodological phrases used in the electronic records were poor, and thus, the precision of electronic searches, using both databases, was low. This international journal contains many relevant RCTs from around the world, involving several types of interventions. Unfortunately, these trials cannot be readily accessed electronically using methodological phrases designed to find RCTs. Improved quality of indexing would facilitate identification of RCTs and their dissemination.