Methodological issues in evaluation of innovative training approaches to stroke rehabilitation
Eligibility for RR was articles that were systematic reviews of the literature. RESULTS: For the PS, a total of 1,674 citations were retrieved in the literature search. After screening, data were abstracted for 99 studies in six domains. For the RR, the initial literature search yielded a total of 9...
| Main Authors: | , , , |
|---|---|
| Corporate Authors: | , |
| Format: | eBook |
| Language: | English |
| Published: |
Rockville, Maryland
Agency for Healthcare Research and Quality
November 17, 2008, 2008
|
| Series: | Technology assessment report
|
| Subjects: | |
| Online Access: | |
| Collection: | National Center for Biotechnology Information - Collection details see MPG.ReNa |
| Summary: | Eligibility for RR was articles that were systematic reviews of the literature. RESULTS: For the PS, a total of 1,674 citations were retrieved in the literature search. After screening, data were abstracted for 99 studies in six domains. For the RR, the initial literature search yielded a total of 949 English-language citations. After screening, a final set of 38 systematic reviews were data abstracted. CONCLUSIONS: In the PS, major methodological problems involved sample size and the psychometric properties of outcome measurement instruments. Sample size was sometimes too small to have adequate power to detect meaningful effects. Many authors failed to show sample size calculations or report a minimum clinically important difference (MCID). For many of the instruments used to measure outcomes, the psychometric properties were not tested in the stroke population. Most systematic reviews were of good quality and presented the evidence for stroke rehabilitation adequately. Many of the reviews evaluated high level study designs (e.g., randomized trials). From a methods perspective, the majority of reviews evaluated randomization, blinding, and withdrawals/dropouts. Fewer reviews evaluated baseline comparability, adverse events, or co-intervention or contamination. Many reviews indicated that blinding of the patient and the provider was not possible in stroke rehabilitation and as such did not evaluate eligible studies for this criterion. These findings concur with those of the purposive sampling. Regarding outcome measures, the PR and RR found that no single stroke-related measure captures all relevant dimensions of important attributes of interest to patients and clinicians. This implies that multiple measures may need to be included in future studies to capture all relevant attributes OBJECTIVES: The assessment was undertaken to describe key methodological issues in studies designed to evaluate stroke rehabilitation therapies. DATA SOURCES: MEDLINE(r), CINAHL(r), PsycINFO(r), and the Cochrane Database of Systematic Reviews (CDSR). Search scope varied, but the widest range was from January 2000 through late-January 2008. REVIEW METHODS: Purposive sampling (PS) and a review of reviews (RR) were employed to describe study methodology. Eligibility criteria for PS were English-language, comparative studies with human subjects and a main focus on stroke (or cerebrovascular accident). Also, any type of rehabilitation therapy could be included, provided its effect was evaluated using an outcome in one of six domains of interest: ambulation, cognition, quality of life, daily activities, dysphagia and communication. We only included drug studies if the medications were used to treat cognitive impairment. |
|---|---|
| Item Description: | Title from PDF title page |
| Physical Description: | 1 PDF file (viii, 85, approximately 121 pages) illustrations |