Variance estimation when using propensity-score matching with replacement with survival or time-to-event outcomes

被引:50
作者
Austin, Peter C. [1 ,2 ,3 ]
Cafri, Guy [4 ]
机构
[1] ICES, G106,2075 Bayview Ave, Toronto, ON M4N 3M5, Canada
[2] Univ Toronto, Inst Hlth Management Policy & Evaluat, Toronto, ON, Canada
[3] Sunnybrook Res Inst, Toronto, ON, Canada
[4] Johnson & Johnson Med Devices, San Diego, CA USA
基金
加拿大健康研究院;
关键词
matching; Monte Carlo simulations; observational study; propensity score; survival analysis; PERFORMANCE; MODELS;
D O I
10.1002/sim.8502
中图分类号
Q [生物科学];
学科分类号
07 ; 0710 ; 09 ;
摘要
Propensity-score matching is a popular analytic method to estimate the effects of treatments when using observational data. Matching on the propensity score typically requires a pool of potential controls that is larger than the number of treated or exposed subjects. The most common approach to matching on the propensity score is matching without replacement, in which each control subject is matched to at most one treated subject. Failure to find a matched control for each treated subject can lead to "bias due to incomplete matching." To avoid this bias, it is important to identify a matched control subject for each treated subject. An alternative to matching without replacement is matching with replacement, in which control subjects are allowed to be matched to multiple treated subjects. A limitation to the use of matching with replacement is that variance estimation must account for both the matched nature of the sample and for some control subjects being included in multiple matched sets. While a variance estimator has been proposed for when outcomes are continuous, no such estimator has been proposed for use with time-to-event outcomes, which are common in medical and epidemiological research. We propose a variance estimator for the hazard ratio when matching with replacement. We conducted a series of Monte Carlo simulations to examine the performance of this estimator. We illustrate the utility of matching with replacement to estimate the effect of smoking cessation counseling on survival in smokers discharged from hospital with a heart attack.
引用
收藏
页码:1623 / 1640
页数:18
相关论文
共 43 条
[11]   The performance of different propensity score methods for estimating absolute effects of treatments on survival outcomes: A simulation study [J].
Austin, Peter C. ;
Schuster, Tibor .
STATISTICAL METHODS IN MEDICAL RESEARCH, 2016, 25 (05) :2214-2237
[12]   The performance of inverse probability of treatment weighting and full matching on the propensity score in the presence of model misspecification when estimating the effect of treatment on survival outcomes [J].
Austin, Peter C. ;
Stuart, Elizabeth A. .
STATISTICAL METHODS IN MEDICAL RESEARCH, 2017, 26 (04) :1654-1670
[13]   Optimal full matching for survival outcomes: a method that merits more widespread use [J].
Austin, Peter C. ;
Stuart, Elizabeth A. .
STATISTICS IN MEDICINE, 2015, 34 (30) :3949-3967
[14]   The use of bootstrapping when using propensity-score matching without replacement: a simulation study [J].
Austin, Peter C. ;
Small, Dylan S. .
STATISTICS IN MEDICINE, 2014, 33 (24) :4306-4319
[15]   A comparison of 12 algorithms for matching on the propensity score [J].
Austin, Peter C. .
STATISTICS IN MEDICINE, 2014, 33 (06) :1057-1069
[16]   The performance of different propensity score methods for estimating marginal hazard ratios [J].
Austin, Peter C. .
STATISTICS IN MEDICINE, 2013, 32 (16) :2837-2849
[17]   Comparing paired vs non-paired statistical methods of analyses when making inferences about absolute risk reductions in propensity-score matched samples [J].
Austin, Peter C. .
STATISTICS IN MEDICINE, 2011, 30 (11) :1292-1301
[18]   A Tutorial and Case Study in Propensity Score Analysis: An Application to Estimating the Effect of In-Hospital Smoking Cessation Counseling on Mortality [J].
Austin, Peter C. .
MULTIVARIATE BEHAVIORAL RESEARCH, 2011, 46 (01) :119-151
[19]   Optimal caliper widths for propensity-score matching when estimating differences in means and differences in proportions in observational studies [J].
Austin, Peter C. .
PHARMACEUTICAL STATISTICS, 2011, 10 (02) :150-161
[20]   The performance of different propensity-score methods for estimating differences in proportions (risk differences or absolute risk reductions) in observational studies [J].
Austin, Peter C. .
STATISTICS IN MEDICINE, 2010, 29 (20) :2137-2148