Shah Baiju R, Laupacis Andreas, Hux Janet E, Austin Peter C
Institute for Clinical Evaluative Sciences, Toronto, Ontario, Canada.
J Clin Epidemiol. 2005 Jun;58(6):550-9. doi: 10.1016/j.jclinepi.2004.10.016. Epub 2005 Apr 19.
To determine whether adjusting for confounder bias in observational studies using propensity scores gives different results than using traditional regression modeling.
Medline and Embase were used to identify studies that described at least one association between an exposure and an outcome using both traditional regression and propensity score methods to control for confounding. From 43 studies, 78 exposure-outcome associations were found. Measures of the quality of propensity score implementation were determined. The statistical significance of each association using both analytical methods was compared. The odds or hazard ratios derived using both methods were compared quantitatively.
Statistical significance differed between regression and propensity score methods for only 8 of the associations (10%), kappa = 0.79 (95% CI = 0.65-0.92). In all cases, the regression method gave a statistically significant association not observed with the propensity score method. The odds or hazard ratio derived using propensity scores was, on average, 6.4% closer to unity than that derived using traditional regression.
Observational studies had similar results whether using traditional regression or propensity scores to adjust for confounding. Propensity scores gave slightly weaker associations; however, many of the reviewed studies did not implement propensity scores well.
确定在观察性研究中使用倾向得分调整混杂偏倚与使用传统回归模型是否会得出不同的结果。
利用医学文献数据库（Medline）和荷兰医学文摘数据库（Embase）来识别那些使用传统回归和倾向得分方法控制混杂因素、描述了至少一种暴露与结局之间关联的研究。从43项研究中，发现了78个暴露-结局关联。确定了倾向得分实施质量的衡量指标。比较了两种分析方法下各关联的统计学显著性。对两种方法得出的比值比或风险比进行了定量比较。
仅8个关联（10%）在回归和倾向得分方法之间的统计学显著性存在差异，kappa = 0.79（95%可信区间 = 0.65 - 0.92）。在所有情况下，回归方法得出了倾向得分方法未观察到的具有统计学显著性的关联。使用倾向得分得出的比值比或风险比平均比使用传统回归得出的更接近1，相差6.4%。
无论是使用传统回归还是倾向得分来调整混杂因素，观察性研究都有相似的结果。倾向得分得出的关联略弱；然而，许多纳入综述的研究对倾向得分的实施并不理想。
