Juni P, Tallon D, Egger M; International Society of Technology Assessment in Health Care. Meeting.
Annu Meet Int Soc Technol Assess Health Care Int Soc Technol Assess Health Care Meet. 2000; 16: 199.
MRC Health Services Research Collaboration, Department of Social Medicine, University of Bristol, Canynge Hall, Whiteladies Road, Bristol BS8 2PR, UK
Introduction: The assessment of the methodological quality of primary studies is an important element of meta-analysis but there is debate on the most appropriate approach. We examined whether and how quality assessment was performed in meta-analyses published in leading English-language journals. Methods: A handsearch 1993-1997 of four general medicine journals (Ann Intern Med, BMJ, Lancet, JAMA) and four specialist journals (Am J Cardiol, Cancer, Circulation, Obstet Gynaecol) identified 133 meta-analyses. We used a standardised questionnaire to extract relevant information and logistic regression for analysis. Results: The quality of primary studies was assessed in 54 (41%) meta-analyses, 31 (23%) reported on concealment of treatment allocation and blinding and 25 (19%) performed sensitivity analyses according to quality. Over 40 different approaches were identified, with checklists and quality scales used in similar proportions. In multivariable analysis quality assessment was less likely to be reported in specialist journals (or 0.32 compared to general medicine journals, 95% CI 0.12-0.87) and journals published in the United Kingdom (or 0.37 compared to US journals, 95% CI 0.16-0.88). Affiliation with the Cochrane Collaboration predicted assessment of quality (or 6.30, 95% CI 1.94-20.4). Conclusions: The quality assessment of primary studies is relatively uncommon and inconsistent in meta-analyses published in leading medical journals. These results indicate uncertainty about the importance of quality assessment and the appropriate approach to quality assessment, with some evidence for an Atlantic divide. Methodological research and the development of guidelines are warranted.
Publication Types:
Keywords:
- Biomedical Research
- Double-Blind Method
- Evaluation Studies
- Great Britain
- Language
- Meta-Analysis
- Periodicals
- Physical Examination
- Publishing
- methods
- hsrmtgs
Other ID:
UI: 102271850
From Meeting Abstracts