Meta-analysis of clinical outcomes following surgical mitral valve repair or replacement.

Abstract

Although the use of mitral valve surgery has been successful at alleviating mitral valve disease, published studies on either replacement or repair have yielded mixed clinical outcomes regarding differences between repair and replacement. Meta-analysis of various outcomes from 29 published studies was conducted. Studies were separated into four groups by etiology of disease: ischemic; degenerative/myxomatous; rheumatic and mixed. The summary odds ratio for early mortality, comparing replacement to repair, was 2.24 (1.78-2.80), while the summary total survival hazard ratio was 1.58 (1.41-1.78), replacement compared to repair, indicating worse outcomes among those undergoing mitral valve replacement. The risk of thromboembolism was lower in the repair group (summary hazard ratio=1.86, replacement vs. repair), while there was no statistical difference in time to re-operation between the two treatment groups (hazard ratio=0.88 [95% confidence interval: 0.48, 1.62]). Analysis stratified by etiologic classification was able to detect strong evidence of differences in 30-day and total survival outcomes favoring repair for three disease groups (rheumatic, mixed and degenerative). Surgery for ischemic mitral valve had lower 30-day mortality for repair than replacement, but no statistically significant difference in the overall survival was detected. The reported information in the published studies used in the current work lacks sufficient detail to allow summary determination of outcomes by mitral valve repair techniques and by type of mitral valve replacement.

Cite this paper

@article{Shuhaiber2007MetaanalysisOC, title={Meta-analysis of clinical outcomes following surgical mitral valve repair or replacement.}, author={Jeffrey H. Shuhaiber and Robert J. Anderson}, journal={European journal of cardio-thoracic surgery : official journal of the European Association for Cardio-thoracic Surgery}, year={2007}, volume={31 2}, pages={267-75} }