Statistics in Medicine · 2013 · 1K citations · 18 references
EngineeringTreatment EffectQuasi-experimentCausal InferenceData ScienceData MiningSocial MatchingBiasRandomized Controlled TrialNearest Neighbor MatchingStatisticsHealth Services ResearchSelection BiasMatching TechniqueOutcomes ResearchRehabilitationPattern MatchingMatching MethodsPropensity ScoreNearest NeighborStatistical InferenceMedicine
Propensity‑score matching is increasingly employed to mitigate confounding in observational studies of treatment effects. The authors used Monte Carlo simulations to compare optimal matching, greedy nearest‑neighbor matching (with and without calipers and various selection orders), and matching with replacement for forming matched pairs. Results showed that nearest‑neighbor matching achieved balance comparable to optimal matching, while caliper matching produced at least as good balance, lower bias, and the best mean‑squared‑error performance, with only modest effects from selection order and no advantage from replacement.
Propensity-score matching is increasingly being used to reduce the confounding that can occur in observational studies examining the effects of treatments or interventions on outcomes. We used Monte Carlo simulations to examine the following algorithms for forming matched pairs of treated and untreated subjects: optimal matching, greedy nearest neighbor matching without replacement, and greedy nearest neighbor matching without replacement within specified caliper widths. For each of the latter two algorithms, we examined four different sub-algorithms defined by the order in which treated subjects were selected for matching to an untreated subject: lowest to highest propensity score, highest to lowest propensity score, best match first, and random order. We also examined matching with replacement. We found that (i) nearest neighbor matching induced the same balance in baseline covariates as did optimal matching; (ii) when at least some of the covariates were continuous, caliper matching tended to induce balance on baseline covariates that was at least as good as the other algorithms; (iii) caliper matching tended to result in estimates of treatment effect with less bias compared with optimal and nearest neighbor matching; (iv) optimal and nearest neighbor matching resulted in estimates of treatment effect with negligibly less variability than did caliper matching; (v) caliper matching had amongst the best performance when assessed using mean squared error; (vi) the order in which treated subjects were selected for matching had at most a modest effect on estimation; and (vii) matching with replacement did not have superior performance compared with caliper matching without replacement. We found that (i) nearest neighbor matching induced the same balance in baseline covariates as did optimal matching; (ii) when at least some of the covariates were continuous, caliper matching tended to induce balance on baseline covariates that was at least as good as the other algorithms; (iii) caliper matching tended to result in estimates of treatment effect with less bias compared with optimal and nearest neighbor matching; (iv) optimal and nearest neighbor matching resulted in estimates of treatment effect with negligibly less variability than did caliper matching; (v) caliper matching had amongst the best performance when assessed using mean squared error; (vi) the order in which treated subjects were selected for matching had at most a modest effect on estimation; and (vii) matching with replacement did not have superior performance compared with caliper matching without replacement.
18