Volume 23, Issue 19
Research Article

Stratification and weighting via the propensity score in estimation of causal treatment effects: a comparative study

Jared K. Lunceford

Corresponding Author

E-mail address: jared_lunceford@merck.com

Merck Research Laboratories, RY34‐A316, P.O. Box 2000, Rahway, NJ 07065‐0900, U.S.A.

Merck Research Laboratories, RY34‐A316, P.O. Box 2000, Rahway, NJ 07065‐0900, U.S.A.Search for more papers by this author
Marie Davidian

Department of Statistics, North Carolina State University, Box 8203, Raleigh, NC 27695, U.S.A.

Search for more papers by this author
First published: 24 August 2004
Citations: 703

Abstract

Estimation of treatment effects with causal interpretation from observational data is complicated because exposure to treatment may be confounded with subject characteristics. The propensity score, the probability of treatment exposure conditional on covariates, is the basis for two approaches to adjusting for confounding: methods based on stratification of observations by quantiles of estimated propensity scores and methods based on weighting observations by the inverse of estimated propensity scores. We review popular versions of these approaches and related methods offering improved precision, describe theoretical properties and highlight their implications for practice, and present extensive comparisons of performance that provide guidance for practical use. Copyright © 2004 John Wiley & Sons, Ltd.

Number of times cited according to CrossRef: 703

  • Extending inferences from a randomized trial to a new target population, Statistics in Medicine, 10.1002/sim.8426, 39, 14, (1999-2014), (2020).
  • Toward Causally Interpretable Meta-analysis, Epidemiology, 10.1097/EDE.0000000000001177, 31, 3, (334-344), (2020).
  • G-computation, propensity score-based methods, and targeted maximum likelihood estimator for causal inference with different covariates sets: a comparative simulation study, Scientific Reports, 10.1038/s41598-020-65917-x, 10, 1, (2020).
  • Constructing inverse probability weights for institutional comparisons in healthcare, Statistics in Medicine, 10.1002/sim.8657, 39, 23, (3156-3172), (2020).
  • Understanding marginal structural models for time-varying exposures: pitfalls and tips, Journal of Epidemiology, 10.2188/jea.JE20200226, (2020).
  • Doubly Robust Estimation of Causal Effect, Circulation: Cardiovascular Quality and Outcomes, 10.1161/CIRCOUTCOMES.119.006065, 13, 1, (2020).
  • Effect of Low-Dose Nebivolol in Patients with Acute Myocardial Infarction: A Multi-Center Observational Study, Chonnam Medical Journal, 10.4068/cmj.2020.56.1.55, 56, 1, (55), (2020).
  • Improving external validity of epidemiologic cohort analyses: a kernel weighting approach, Journal of the Royal Statistical Society: Series A (Statistics in Society), 10.1111/rssa.12564, 183, 3, (1293-1311), (2020).
  • Quantifying the bias due to observed individual confounders in causal treatment effect estimates, Statistics in Medicine, 10.1002/sim.8549, 39, 18, (2447-2476), (2020).
  • Uncertainty in the design stage of two‐stage Bayesian propensity score analysis, Statistics in Medicine, 10.1002/sim.8486, 39, 17, (2265-2290), (2020).
  • Assessing exposure effects on gene expression, Genetic Epidemiology, 10.1002/gepi.22324, 44, 6, (601-610), (2020).
  • Doubly robust estimator of risk in the presence of censoring dependent on time-varying covariates: application to a primary prevention trial for coronary events with pravastatin, BMC Medical Research Methodology, 10.1186/s12874-020-01087-8, 20, 1, (2020).
  • Effect of adjuvant chemotherapy on survival benefit in stage III colon cancer patients stratified by age: a Japanese real-world cohort study, BMC Cancer, 10.1186/s12885-019-6508-1, 20, 1, (2020).
  • How Soon Should Patients With Colon Cancer Undergo Definitive Resection?, Diseases of the Colon & Rectum, 10.1097/DCR.0000000000001525, 63, 2, (172-182), (2020).
  • Expanding the phenotype of Bardet–Biedl syndrome: Newly diagnosed sibling cases, Pediatrics International, 10.1111/ped.14066, 62, 1, (101-103), (2020).
  • Substantial genetic divergence and lack of recent gene flow support cryptic speciation in a colour polymorphic bumble bee (Bombus bifarius) species complex, Systematic Entomology, 10.1111/syen.12419, 45, 3, (635-652), (2020).
  • Experimental characterization of liquid film behavior during droplets–polyethylene particle collision, AIChE Journal, 10.1002/aic.16909, 66, 5, (2020).
  • Association between surgical approach and survival following resection of abdominopelvic malignancies, Journal of Surgical Oncology, 10.1002/jso.25841, 121, 4, (620-629), (2020).
  • Discharge destination and patient‐reported outcomes after inpatient treatment for isolated lower limb fractures, Medical Journal of Australia, 10.5694/mja2.50485, 212, 6, (263-270), (2020).
  • Semiparametric inference with missing data: Robustness to outliers and model misspecification, Econometrics and Statistics, 10.1016/j.ecosta.2020.01.003, (2020).
  • Are multiple speed cameras more effective than a single one? Causal analysis of the safety impacts of multiple speed cameras, Accident Analysis & Prevention, 10.1016/j.aap.2020.105488, 139, (105488), (2020).
  • Flexible regression approach to propensity score analysis and its relationship with matching and weighting, Statistics in Medicine, 10.1002/sim.8526, 39, 15, (2017-2034), (2020).
  • Statistical Analysis and Evaluation of Macroeconomic Policies: A Selective Review, Applied Mathematics-A Journal of Chinese Universities, 10.1007/s11766-020-3775-1, 35, 1, (57-83), (2020).
  • An evaluation of inverse probability weighting using the propensity score for baseline covariate adjustment in smaller population randomised controlled trials with a continuous outcome, BMC Medical Research Methodology, 10.1186/s12874-020-00947-7, 20, 1, (2020).
  • Causal inference for left-truncated and right-censored data with covariate measurement error, Computational and Applied Mathematics, 10.1007/s40314-020-01152-4, 39, 2, (2020).
  • Impact evaluation of the free maternal healthcare policy on the risk of neonatal and infant deaths in four sub-Saharan African countries: a quasi-experimental design with propensity score Kernel matching and difference in differences analysis, BMJ Open, 10.1136/bmjopen-2019-033356, 10, 5, (e033356), (2020).
  • A review of the use of propensity score diagnostics in papers published in high-ranking medical journals, BMC Medical Research Methodology, 10.1186/s12874-020-00994-0, 20, 1, (2020).
  • Assessing the effects of maternal HIV infection on pregnancy outcomes using cross-sectional data in Malawi, BMC Public Health, 10.1186/s12889-020-09046-0, 20, 1, (2020).
  • Letter: gastrointestinal symptoms pre‐admission are associated with greater severity of COVID‐19, Alimentary Pharmacology & Therapeutics, 10.1111/apt.15985, 52, 7, (1229-1230), (2020).
  • Treatment effects may remain the same even when trial participants differed from the target population, Journal of Clinical Epidemiology, 10.1016/j.jclinepi.2020.05.001, 124, (126-138), (2020).
  • Multiple-Group Propensity Score Inverse Weight Trimming and Its Impact on Covariate Balance and Bias in Treatment Effect Estimation, Quantitative Psychology, 10.1007/978-3-030-43469-4_12, (147-159), (2020).
  • Repeated Endoscopic Submucosal Dissection for Esophageal Neoplasia Located Close to a Previous Endoscopic Submucosal Dissection Scar, Clinical and Translational Gastroenterology, 10.14309/ctg.0000000000000226, 11, 8, (e00226), (2020).
  • Recombinant Human Soluble Thrombomodulin Contributes to a Reduction In-Hospital Mortality of Acute Cholangitis with Disseminated Intravascular Coagulation: A Propensity Score Analyses of a Japanese Nationwide Database, The Tohoku Journal of Experimental Medicine, 10.1620/tjem.252.53, 252, 1, (53), (2020).
  • Evaluating the causal effects of cellphone distraction on crash risk using propensity score methods, Accident Analysis & Prevention, 10.1016/j.aap.2020.105579, 143, (105579), (2020).
  • Brief discussion on sampling variability in 1:1 propensity score matching without replacement, Pharmacoepidemiology and Drug Safety, 10.1002/pds.5094, 29, 9, (1194-1197), (2020).
  • Machine learning outcome regression improves doubly robust estimation of average causal effects, Pharmacoepidemiology and Drug Safety, 10.1002/pds.5074, 29, 9, (1120-1133), (2020).
  • A Survey of Learning Causality with Data, ACM Computing Surveys, 10.1145/3397269, 53, 4, (1-37), (2020).
  • Determination of the optimal number of strata for propensity score subclassification, Statistics & Probability Letters, 10.1016/j.spl.2020.108951, (108951), (2020).
  • Information acquisition and the adoption of a new rice variety towards the development of sustainable agriculture in rural villages in Central Vietnam, World Development Perspectives, 10.1016/j.wdp.2020.100262, 20, (100262), (2020).
  • Efficient estimation of human immunodeficiency virus incidence rate using a pooled cross‐sectional cohort study design, Statistics in Medicine, 10.1002/sim.8661, 39, 24, (3255-3271), (2020).
  • Omission of cortical renorrhaphy during robotic partial nephrectomy: a Vattikuti Collective Quality Initiative (VCQI) database analysis, Urology, 10.1016/j.urology.2020.09.003, (2020).
  • The Blessings of Multiple Causes, Journal of the American Statistical Association, 10.1080/01621459.2019.1686987, 114, 528, (1574-1596), (2020).
  • Robust and efficient semi‐supervised estimation of average treatment effects with application to electronic health records data, Biometrics, 10.1111/biom.13298, 0, 0, (2020).
  • Evaluating the Health Outcomes of the Healthy Women Healthy Babies Program in Delaware, Maternal and Child Health Journal, 10.1007/s10995-020-02972-w, (2020).
  • The Causal Effects of Parental Divorce and Parental Temporary Separation on Children’s Cognitive Abilities and Psychological Well-being According to Parental Relationship Quality, Social Indicators Research, 10.1007/s11205-020-02428-2, (2020).
  • Random Forests Approach for Causal Inference with Clustered Observational Data, Multivariate Behavioral Research, 10.1080/00273171.2020.1808437, (1-24), (2020).
  • Evaluating multiple surrogate markers with censored data, Biometrics, 10.1111/biom.13370, 0, 0, (2020).
  • Effects of Surgery on Survival of Early-Stage Patients With SCLC: Propensity Score Analysis and Nomogram Construction in SEER Database, Frontiers in Oncology, 10.3389/fonc.2020.00626, 10, (2020).
  • Clinical efficacy of hydroxychloroquine in patients with covid-19 pneumonia who require oxygen: observational comparative study using routine care data, BMJ, 10.1136/bmj.m1844, (m1844), (2020).
  • Causal inference for recurrent event data using pseudo-observations, Biostatistics, 10.1093/biostatistics/kxaa020, (2020).
  • Health, Well-Being and Work History Patterns: Insight on Territorial Differences, Social Indicators Research, 10.1007/s11205-020-02393-w, (2020).
  • Does Knee Arthroscopy for Treatment of Meniscal Damage with Osteoarthritis Delay Knee Replacement Compared to Physical Therapy Alone?, Clinics in Orthopedic Surgery, 10.4055/cios19114, 12, (2020).
  • Accounting for matching structure in post-matching analysis of observational studies, Communications in Statistics - Simulation and Computation, 10.1080/03610918.2019.1708928, (1-19), (2020).
  • Causal Inference and Estimands in Clinical Trials, Statistics in Biopharmaceutical Research, 10.1080/19466315.2019.1697739, (1-14), (2020).
  • Effect of behavioral health services and neighborhood disadvantages on recidivism: a comparison of mental health court and traditional court participants, Journal of Experimental Criminology, 10.1007/s11292-019-09402-0, (2020).
  • Blinatumomab versus historical standard therapy in pediatric patients with relapsed/refractory Ph-negative B-cell precursor acute lymphoblastic leukemia, Leukemia, 10.1038/s41375-020-0770-8, (2020).
  • The burden of high workload on the health-related quality of life among home care workers in Northern Sweden, International Archives of Occupational and Environmental Health, 10.1007/s00420-020-01530-9, (2020).
  • Third-Line Antidiabetic Therapy Intensification Patterns and Glycaemic Control in Patients with Type 2 Diabetes in the USA: A Real-World Study, Drugs, 10.1007/s40265-020-01279-y, (2020).
  • Synthetic and External Controls in Clinical Trials – A Primer for Researchers

    ,
    Clinical Epidemiology, 10.2147/CLEP.S242097, Volume 12, (457-467), (2020).
  • Non-adherence in non-inferiority trials: pitfalls and recommendations, BMJ, 10.1136/bmj.m2215, (m2215), (2020).
  • Factors in Early Feeding Practices That May Influence Growth and the Challenges That Arise in Growth Outcomes Research, Nutrients, 10.3390/nu12071939, 12, 7, (1939), (2020).
  • Tourism development and residents’ well-being: Comparing two seaside destinations in Italy, Tourism Economics, 10.1177/1354816620916962, (135481662091696), (2020).
  • Association of Nonoperative Management Using Antibiotic Therapy vs Laparoscopic Appendectomy With Treatment Success and Disability Days in Children With Uncomplicated Appendicitis, JAMA, 10.1001/jama.2020.10888, (2020).
  • Variance estimation in inverse probability weighted Cox models, Biometrics, 10.1111/biom.13332, 0, 0, (2020).
  • Clopidogrel versus Aspirin after Dual Antiplatelet Therapy in Acute Myocardial Infarction Patients Undergoing Drug-Eluting Stenting, Korean Circulation Journal, 10.4070/kcj.2019.0166, 50, (2020).
  • The Effectiveness of Dance Therapy as an Adjunct to Rehabilitation of Adults With a Physical Disability, Frontiers in Psychology, 10.3389/fpsyg.2020.01963, 11, (2020).
  • Replication of randomized clinical trial results using real-world data: paving the way for effectiveness decisions, Journal of Comparative Effectiveness Research, 10.2217/cer-2020-0161, (2020).
  • Outcomes of outborn very-low-birth-weight infants in Japan, Archives of Disease in Childhood - Fetal and Neonatal Edition, 10.1136/archdischild-2019-318594, (fetalneonatal-2019-318594), (2020).
  • Bias and High-Dimensional Adjustment in Observational Studies of Peer Effects, Journal of the American Statistical Association, 10.1080/01621459.2020.1796393, (1-11), (2020).
  • Long-term abdominal wall benefits of the laparoscopic approach in liver left lateral sectionectomy: a multicenter comparative study, Surgical Endoscopy, 10.1007/s00464-020-07985-8, (2020).
  • Propensity score specification for optimal estimation of average treatment effect with binary response, Statistical Methods in Medical Research, 10.1177/0962280220934847, (096228022093484), (2020).
  • Propensity score weighting under limited overlap and model misspecification, Statistical Methods in Medical Research, 10.1177/0962280220940334, (096228022094033), (2020).
  • Exploring the alleged effect of lower academic achievement after the free semester in Korean Middle Schools, Asia Pacific Education Review, 10.1007/s12564-020-09641-1, (2020).
  • Further Evidence on the Effect of Clean Indoor Air Laws on Smoking: The Italian Case, Southern Economic Journal, 10.1002/soej.12409, 86, 3, (1110-1132), (2019).
  • Causal inference with noisy data: Bias analysis and estimation approaches to simultaneously addressing missingness and misclassification in binary outcomes, Statistics in Medicine, 10.1002/sim.8419, 39, 4, (456-468), (2019).
  • Organic farming for local markets in Kenya: Contribution of conversion and certification to environmental benefits, Canadian Journal of Agricultural Economics/Revue canadienne d'agroeconomie, 10.1111/cjag.12209, 68, 1, (83-105), (2019).
  • Regression‐with‐residuals estimation of marginal effects: a method of adjusting for treatment‐induced confounders that may also be effect modifiers, Journal of the Royal Statistical Society: Series A (Statistics in Society), 10.1111/rssa.12497, 183, 1, (311-332), (2019).
  • Estimating average treatment effects with a double‐index propensity score, Biometrics, 10.1111/biom.13195, 76, 3, (767-777), (2019).
  • Comparison of oncological outcomes after open and laparoscopic re‐resection of incidental gallbladder cancer, BJS (British Journal of Surgery), 10.1002/bjs.11379, 107, 3, (289-300), (2019).
  • Comparison of empirical Bayes and propensity score methods for road safety evaluation: A simulation study, Accident Analysis & Prevention, 10.1016/j.aap.2019.05.015, 129, (148-155), (2019).
  • Misuse of Regression Adjustment for Additional Confounders Following Insufficient Propensity Score Balancing, Epidemiology, 10.1097/EDE.0000000000001023, 30, 4, (541-548), (2019).
  • Targeting poverty under complementarities: Evidence from Indonesia's unified targeting system, Journal of Development Economics, 10.1016/j.jdeveco.2019.06.002, (2019).
  • Does unemployment contribute to poorer health-related quality of life among Swedish adults?, BMC Public Health, 10.1186/s12889-019-6825-y, 19, 1, (2019).
  • Globalization and Working Environment Nexus: Evidence From Pakistan, SAGE Open, 10.1177/2158244019852474, 9, 2, (215824401985247), (2019).
  • Does outpatient cardiac rehabilitation help patients with acute myocardial infarction quit smoking?, Preventive Medicine, 10.1016/j.ypmed.2018.10.010, 118, (51-58), (2019).
  • Assessing the effectiveness of Japan's community-based direct payment scheme for hilly and mountainous areas, Ecological Economics, 10.1016/j.ecolecon.2019.01.036, 160, (62-75), (2019).
  • Leveraging the entire cohort in drug safety monitoring: part 1 methods for sequential surveillance that use regression adjustment or weighting to control confounding in a multisite, rare event, distributed data setting, Journal of Clinical Epidemiology, 10.1016/j.jclinepi.2019.04.012, 112, (77-86), (2019).
  • Applying sequential surveillance methods that use regression adjustment or weighting to control confounding in a multisite, rare-event, distributed setting: Part 2 in-depth example of a reanalysis of the measles-mumps-rubella-varicella combination vaccine and seizure risk, Journal of Clinical Epidemiology, 10.1016/j.jclinepi.2019.04.019, 113, (114-122), (2019).
  • Is enrollment in a Medicaid health maintenance organization associated with less preventable hospitalizations?, Preventive Medicine Reports, 10.1016/j.pmedr.2019.100964, (100964), (2019).
  • Adjustments of multi-sample -statistics to right censored data and confounding covariates , Computational Statistics & Data Analysis, 10.1016/j.csda.2019.01.012, (2019).
  • Impact of radiotherapy administered simultaneously with systemic treatment in patients with melanoma brain metastases within MelBase, a French multicentric prospective cohort, European Journal of Cancer, 10.1016/j.ejca.2019.02.009, 112, (38-46), (2019).
  • Trends in Regionalization of Care and Mortality For Patients Treated With Radical Cystectomy, Medical Care, 10.1097/MLR.0000000000001143, 57, 9, (728-733), (2019).
  • Investigating the impact of the economic crisis on children's wellbeing in four European countries, Social Science Research, 10.1016/j.ssresearch.2019.06.013, (2019).
  • College attendance type and subsequent alcohol and marijuana use in the U.S., Drug and Alcohol Dependence, 10.1016/j.drugalcdep.2019.107580, (107580), (2019).
  • On the Relation Between G-formula and Inverse Probability Weighting Estimators for Generalizing Trial Results, Epidemiology, 10.1097/EDE.0000000000001097, 30, 6, (807-812), (2019).
  • Deep Representation Learning for Individualized Treatment Effect Estimation using Electronic Health Records, Journal of Biomedical Informatics, 10.1016/j.jbi.2019.103303, (103303), (2019).
  • Multi-institutional trial of non-operative management and surgery for uncomplicated appendicitis in children: Design and rationale, Contemporary Clinical Trials, 10.1016/j.cct.2019.06.013, 83, (10-17), (2019).
  • undefined, Proceedings of the Twelfth ACM International Conference on Web Search and Data Mining - WSDM '19, 10.1145/3289600.3291021, (618-626), (2019).
  • Chemotherapy, Radiation, or Combination Therapy for Stage III Uterine Cancer, Obstetrics & Gynecology, 10.1097/AOG.0000000000003287, 134, 1, (17-29), (2019).
  • Sensitivity analysis for inverse probability weighting estimators via the percentile bootstrap, Journal of the Royal Statistical Society: Series B (Statistical Methodology), 10.1111/rssb.12327, 81, 4, (735-761), (2019).
  • See more

The full text of this article hosted at iucr.org is unavailable due to technical difficulties.