Research in rehabilitation medicine: methodological challenges

Objective: To provide an overview of methodological issues specifically related to the evaluation of rehabilitation interventions studies.

Study design and setting: Narrative review covering studies evaluating interventions in rehabilitation medicine with methodological issues.

Results: Four main methodological issues could be identified. First, the inclusion of patients. Patients should be selected based on having the problem being addressed by the intervention and not based on other criteria such as disease diagnosis. Second, the description of the intervention. Rehabilitation is a problem-solving process undertaken by one group of people (therapists) with another group (patient and family). The specific intervention being studied is only one of many that may affect outcome. Describing the whole package is a real challenge. Third, the control group. A control intervention has to raise equal expectations in patients and therapists. Expectation bias is a very probable influence on measured outcomes. Fourth, the definition of an appropriate outcome. In contrast to traditional biomedical research, rehabilitation research should have at least one "process" (proximate) or intervening variable measure and several distal primary outcome measures.