Charles M Court-Brown1, Margaret M McQueen. 1. Department of Orthopaedic Trauma, Royal Infirmary of Edinburgh, Edinburgh, United Kingdom. courtbrown@aol.com
Abstract
BACKGROUND: Meta-analysis has become accepted as a methodically rigorous research tool, and as a result, many meta-analyses have been undertaken in orthopedic trauma, the implication being that their conclusions have improved surgeons' knowledge and facilitated improved clinical care. There have been criticisms of the methodology used in a number of meta-analyses; however, there has been no previous study of the clinical usefulness of their conclusions. METHODS: A study of 60 orthopedic trauma meta-analyses published during or after 2000 was undertaken. We divided the conclusions of the meta-analyses into four types depending on their innovation and potential for improving clinical care. We also separated the meta-analyses according to whether they were undertaken by the Cochrane Collaboration or by independent surgeons. RESULTS: Only 4 (6.7%) of the meta-analyses contained new, clinically useful, information, and 22 (36.7%) had no conclusions at all. A further 25 (41.7%) contained conclusions that were already in the standard orthopedic trauma literature. Approximately 70% of the Cochrane Collaborations had no conclusions. CONCLUSIONS: We found considerable variability in the clinical usefulness of orthopedic trauma meta-analyses, and we question the clinical usefulness of this type of research.
BACKGROUND: Meta-analysis has become accepted as a methodically rigorous research tool, and as a result, many meta-analyses have been undertaken in orthopedic trauma, the implication being that their conclusions have improved surgeons' knowledge and facilitated improved clinical care. There have been criticisms of the methodology used in a number of meta-analyses; however, there has been no previous study of the clinical usefulness of their conclusions. METHODS: A study of 60 orthopedic trauma meta-analyses published during or after 2000 was undertaken. We divided the conclusions of the meta-analyses into four types depending on their innovation and potential for improving clinical care. We also separated the meta-analyses according to whether they were undertaken by the Cochrane Collaboration or by independent surgeons. RESULTS: Only 4 (6.7%) of the meta-analyses contained new, clinically useful, information, and 22 (36.7%) had no conclusions at all. A further 25 (41.7%) contained conclusions that were already in the standard orthopedic trauma literature. Approximately 70% of the Cochrane Collaborations had no conclusions. CONCLUSIONS: We found considerable variability in the clinical usefulness of orthopedic trauma meta-analyses, and we question the clinical usefulness of this type of research.