Measuring Outcomes for Children with Cerebral Palsy Who Use Gait Trainers

The aim of this study was to describe systematically the best available intervention evidence for children with cerebral palsy (CP). This study was a systematic review of systematic reviews. The following databases were searched: CINAHL, Cochrane Library, DARE, EMBASE, Google Scholar MEDLINE, OTSeeker, PEDro, PsycBITE, PsycINFO, and speechBITE. Two independent reviewers determined whether studies met the inclusion criteria. These were that (1) the study was a systematic review or the next best available; (2) it was a medical/allied health intervention; and (3) that more than 25% of participants were children with CP. Interventions were coded using the Oxford Levels of Evidence; GRADE; Evidence Alert Traffic Light; and the International Classification of Function, Disability and Health. Overall, 166 articles met the inclusion criteria (74% systematic reviews) across 64 discrete interventions seeking 131 outcomes. Of the outcomes assessed, 16% (21 out of 131) were graded 'do it' (green go); 58% (76 out of 131) 'probably do it' (yellow measure); 20% (26 out of 131) 'probably do not do it' (yellow measure); and 6% (8 out of 131) 'do not do it' (red stop). Green interventions included anticonvulsants, bimanual training, botulinum toxin, bisphosphonates, casting, constraint-induced movement therapy, context-focused therapy, diazepam, fitness training, goal-directed training, hip surveillance, home programmes, occupational therapy after botulinum toxin, pressure care, and selective dorsal rhizotomy. Most (70%) evidence for intervention was lower level (yellow) while 6% was ineffective (red). Evidence supports 15 green light interventions. All yellow light interventions should be accompanied by a sensitive outcome measure to monitor progress and red light interventions should be discontinued since alternatives exist. © 2013 Mac Keith Press.

A systematic review of measurement properties of health-status instruments is a tool for evaluating the quality of instruments. Our aim was to appraise the quality of the review process, to describe how authors assess the methodological quality of primary studies of measurement properties, and to describe how authors evaluate results of the studies. Literature searches were performed in three databases. One hundred and forty-eight reviews were included. The purpose of included reviews was to identify health status instruments used in an evaluative application and to report on the measurement properties of these instruments. Two independent reviewers selected the articles and extracted the data. Reviews were often of low quality: 22% of the reviews used one database, the search strategy was often poorly described, and in many cases it was not reported whether article selection (75%) and data extraction (71%) was done by two independent reviewers. In 11 reviews the methodological quality of the primary studies was evaluated for all measurement properties, and of these 11 reviews only 7 evaluated the results. Methods to evaluate the quality of the primary studies and the results differed widely. The poor quality of reviews hampers evidence-based selection of instruments. Guidelines for conducting and reporting systematic reviews of measurement properties should be developed.