Addressing Extreme Propensity Scores via the Overlap Weights
The DOI matches its Crossref record; 19 of the 19 references it cites were found in a registry or an index.
Abstract
The popular inverse probability weighting method in causal inference is often hampered by extreme propensity scores, resulting in biased estimates and excessive variance. A common remedy is to trim patients with extreme scores (i.e., remove them from the weighted analysis). However, such methods are often sensitive to the choice of cutoff points and discard a large proportion of the sample. The implications for bias and the precision of the treatment effect estimate are unclear. These problems are mitigated by a newly developed method, the overlap weighting method. Overlap weights emphasize the target population with the most overlap in observed characteristics between treatments, by continuously down-weighting the units in the tails of the propensity score distribution. Here we use simulations to compare overlap weights to standard inverse probability weighting with trimming, in terms of bias, variance, and 95% confidence interval coverage. A range of propensity score distributions are considered, including settings with substantial nonoverlap and extreme values. To facilitate practical implementation, we further provide a consistent estimator for the standard error of the treatment effect estimated using overlap weighting.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
In the citing paper
“Overlap weighting emphasizes patients for whom either strategy was clinically plausible and limits the influence of extreme weights (17, 18).”
From Day-4 antipseudomonal and anti-MRSA de-escalation was not associated with higher mortality in critically ill patients with sepsis (2026) · cites this paper as [17]
Citations
Cites
The central role of the propensity score in observational studies for causal effects
ROSENBAUM, P. R., & RUBIN, D. B. (1983) · Biometrika · doi:10.1093/biomet/70.1.41 · Cited by 31,241
Matching Methods for Causal Inference: A Review and a Look Forward
Stuart, E. A. (2010) · Statistical Science · doi:10.1214/09-sts313 · Cited by 5,478
Austin, P. C., & Stuart, E. A. (2015) · Statistics in Medicine · doi:10.1002/sim.6607 · Cited by 4,495
Constructing Inverse Probability Weights for Marginal Structural Models
Cole, S. R., & Hernan, M. A. (2008) · American Journal of Epidemiology · doi:10.1093/aje/kwn164 · Cited by 2,796
Lunceford, J. K., & Davidian, M. (2004) · Statistics in Medicine · doi:10.1002/sim.1903 · Cited by 1,674
Hirano, K., & Imbens, G. W. (2001) · Health Services and Outcomes Research Methodology · doi:10.1023/a:1020371312283 · Cited by 1,039
Dealing with limited overlap in estimation of average treatment effects
Crump, R. K., Hotz, V. J., Imbens, G. W., & Mitnik, O. A. (2009) · Biometrika · doi:10.1093/biomet/asn055 · Cited by 1,007
Balancing Covariates via Propensity Score Weighting
Li, F., Morgan, K. L., & Zaslavsky, A. M. (2018) · Journal of the American Statistical Association · doi:10.1080/01621459.2016.1260466 · Cited by 962
Self-citationRothman, K. J. (1988) · Cited by 690
Weight Trimming and Propensity Score Weighting
Lee, B. K., Lessler, J., & Stuart, E. A. (2011) · PLoS ONE · doi:10.1371/journal.pone.0018174 · Cited by 500
Cited by
An introduction to inverse probability of treatment weighting in observational research
Chesnaye, N. C., Stel, V. S., Tripepi, G., Dekker, F. W., Fu, E. L., Zoccali, C., & Jager, K. J. (2022) · Clinical Kidney Journal · doi:10.1093/ckj/sfab158 · Cited by 951
“Several weighting methods based on propensity scores are available, such as fine stratification weights [17], matching weights [18], overlap weights [19] and inverse probability of treatment weights—the focus of this article.”
MethodLong COVID after breakthrough SARS-CoV-2 infection
Al-Aly, Z., Bowe, B., & Xie, Y. (2022) · Nature Medicine · doi:10.1038/s41591-022-01840-0 · Cited by 866
Thomas (2020) · doi:10.1001/jama.2020.7819 · Cited by 835
Desai, R. J., & Franklin, J. M. (2019) · BMJ · doi:10.1136/bmj.l5657 · Cited by 657
Risks and burdens of incident diabetes in long COVID: a cohort study
Xie, Y., & Al-Aly, Z. (2022) · The Lancet Diabetes & Endocrinology · doi:10.1016/s2213-8587(22)00044-4 · Cited by 564
“33,34 Sixth, we applied the doubly robust adjustment method to further adjust for covariates after applying inverse probability weighting.”
MethodMehta, N., Kalra, A., Nowacki, A. S., Anjewierden, S., Han, Z., Bhat, P., Carmona-Rubio, A. E., Jacob, M., Procop, G. W., Harrington, S., et al. (2020) · JAMA Cardiology · doi:10.1001/jamacardio.2020.1855 · Cited by 400
Risks of mental health outcomes in people with covid-19: cohort study
Xie, Y., Xu, E., & Al-Aly, Z. (2022) · BMJ · doi:10.1136/bmj-2021-068993 · Cited by 323
“19 To test the robustness of our findings, we performed four sensitivity analyses.”
MethodOrkaby, A. R., Driver, J. A., Ho, Y.-L., Lu, B., Costa, L., Honerlaw, J., Galloway, A., Vassy, J. L., Forman, D. E., Gaziano, J. M., et al. (2020) · JAMA · doi:10.1001/jama.2020.7848 · Cited by 260
Bréchot, N., Hajage, D., Kimmoun, A., Demiselle, J., Agerstrand, C., Montero, S., Schmidt, M., Luyt, C.-E., Lebreton, G., Hékimian, G., et al. (2020) · The Lancet · doi:10.1016/s0140-6736(20)30733-9 · Cited by 226
Eisemann, N., Bunk, S., Mukama, T., Baltus, H., Elsner, S. A., Gomille, T., Hecht, G., Heywang-Köbrunner, S., Rathmann, R., Siegmann-Luz, K., et al. (2025) · Nature Medicine · doi:10.1038/s41591-024-03408-6 · Cited by 217
“The empirical sandwich variance for propensity score weighting estimators was used 23,24 .”
Glucagon-Like Peptide-1 Receptor Agonist Use and Residual Gastric Content Before Anesthesia
Sen, S., Potnuru, P. P., Hernandez, N., Goehl, C., Praestholm, C., Sridhar, S., & Nwokolo, O. O. (2024) · JAMA Surgery · doi:10.1001/jamasurg.2024.0111 · Cited by 202
Karter, A. J., Parker, M. M., Moffet, H. H., Gilliam, L. K., & Dlott, R. (2021) · JAMA · doi:10.1001/jama.2021.6530 · Cited by 191
Using Propensity Score Methods to Create Target Populations in Observational Clinical Research
Thomas, L., Li, F., & Pencina, M. (2020) · JAMA · doi:10.1001/jama.2019.21558 · Cited by 174
Self-citationKuang, K., Li, L., Geng, Z., Xu, L., Zhang, K., Liao, B., Huang, H., Ding, P., Miao, W., & Jiang, Z. (2020) · Engineering · doi:10.1016/j.eng.2019.08.016 · Cited by 126
“Recently, related works have been proposed to address the above challenges, including continuous treatment [6], the interaction of treatments [7], unobserved confounders [8], and the limits on overlap [9,10].”
BackgroundPropensity score weighting for causal inference with multiple treatments
Li, F., & Li, F. (2019) · The Annals of Applied Statistics · doi:10.1214/19-aoas1282 · Cited by 168
Self-citationBuprenorphine versus Methadone for Opioid Use Disorder in Pregnancy
Suarez, E. A., Huybrechts, K. F., Straub, L., Hernández-Díaz, S., Jones, H. E., Connery, H. S., Davis, J. M., Gray, K. J., Lester, B., Terplan, M., et al. (2022) · New England Journal of Medicine · doi:10.1056/nejmoa2203318 · Cited by 162
Propensity score weighting under limited overlap and model misspecification
Zhou, Y., Matsouaka, R., & Thomas, L. E. (2020) · Statistical Methods in Medical Research · doi:10.1177/0962280220940334 · Cited by 161
“While there are many versions of IPW trimming [15, 17], Crump et al.”
BackgroundMethodInfluentialSelf-citationA First Course in Causal Inference
Ding, P. (2024) · Chapman & Hall/CRC eBooks · doi:10.1201/9781003484080 · Cited by 157
Leadless vs. transvenous single-chamber ventricular pacing in the Micra CED study: 2-year follow-up
El‐Chami, M. F., Bockstedt, L. A., Longacre, C. F., Higuera, L., Stromberg, K. D., Crossley, G. H., Kowal, R. C., & Piccini, J. P. (2021) · European Heart Journal · doi:10.1093/eurheartj/ehab767 · Cited by 155
Piccini, J. P., El‐Chami, M. F., Wherry, K., Crossley, G. H., Kowal, R. C., Stromberg, K. D., Longacre, C. F., Hinnenthal, J. A., & Bockstedt, L. A. (2021) · JAMA Cardiology · doi:10.1001/jamacardio.2021.2621 · Cited by 152
Matching Methods for Confounder Adjustment: An Addition to the Epidemiologist’s Toolbox
Greifer, N., & Stuart, E. A. (2021) · Epidemiologic Reviews · doi:10.1093/epirev/mxab003 · Cited by 150
Stürmer, T., Webster‐Clark, M., Lund, J. L., Wyss, R., Ellis, A. R., Lunt, M., Rothman, K. J., & Glynn, R. J. (2021) · American Journal of Epidemiology · doi:10.1093/aje/kwab041 · Cited by 147
Risk of Autism after Prenatal Topiramate, Valproate, or Lamotrigine Exposure
Hernández–Díaz, S., Straub, L., Bateman, B. T., Zhu, Y., Mogun, H., Wisner, K. L., Gray, K. J., Lester, B. M., McDougle, C. J., DiCesare, E., et al. (2024) · New England Journal of Medicine · doi:10.1056/nejmoa2309359 · Cited by 132
Diagnosing Failures of Fairness Transfer Across Distribution Shift in Real-World Medical Settings
Schrouff, J., Harris, N., Koyejo, S., Alabdulmohsin, I., Schnider, E., Opsahl-Ong, K., Brown, A. P. Y., Roy, S., Mincu, D., Chen, C., et al. (2022) · neural information processing systems · doi:10.52202/068431-1403 · Cited by 35
Zhu, Y., Hubbard, R. A., Chubak, J., Roy, J., & Mitra, N. (2021) · Pharmacoepidemiology and Drug Safety · doi:10.1002/pds.5338 · Cited by 122
Gaudry, S., Grolleau, F., Barbar, S. D., Martin‐Lefèvre, L., Pons, B., Boulet, É., Boyer, A., Chevrel, G., Montini, F., Bohé, J., et al. (2022) · Critical Care · doi:10.1186/s13054-022-03955-9 · Cited by 121
Choosing the Causal Estimand for Propensity Score Analysis of Observational StudiesPreprint
Greifer, N., & Stuart, E. A. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2106.10577 · Cited by 123
“ATO Clinical equipoise Should those at clinical equipoise receive treatment? Is there an effect of the treatment for some patients? Caliper matching (10,12) Coarsened exact matching (18,19) Cardinality matching (20) Overlap weights (21) Matching weights (22) Weight trimming (23)”
BackgroundMethodMethodological considerations when analysing and interpreting real-world data
Stürmer, T., Wang, T., Golightly, Y. M., Keil, A. P., Lund, J. L., & Funk, M. J. (2019) · Rheumatology · doi:10.1093/rheumatology/kez320 · Cited by 116
Using propensity scores to estimate effects of treatment initiation decisions: State of the science
Webster‐Clark, M., Stürmer, T., Wang, T., Man, K. K. C., Marinac‐Dabic, D., Rothman, K. J., Ellis, A. R., Gokhale, M., Lunt, M., Girman, C. J., et al. (2020) · Statistics in Medicine · doi:10.1002/sim.8866 · Cited by 109
Estimating heterogeneous survival treatment effect in observational data using machine learning
Hu, L., Ji, J., & Li, F. (2021) · Statistics in Medicine · doi:10.1002/sim.9090 · Cited by 104
Self-citationSynthetic Control Estimator: A Tool for Comparative Case Studies in Economic History
Gilchrist, D. J., Emery, T., Garoupa, N. M., & Spruk, R. (2020) · Journal of Economic Surveys · doi:10.1111/joes.12493 · Cited by 99
Crossley, G. H., Piccini, J. P., Longacre, C. F., Higuera, L., Stromberg, K. D., & El‐Chami, M. F. (2023) · Journal of Cardiovascular Electrophysiology · doi:10.1111/jce.15863 · Cited by 94
Relative Effectiveness of Influenza Vaccines Among the United States Elderly, 2018–2019
Izurieta, H. S., Chillarige, Y., Kelman, J. A., Wei, Y., Lu, Y., Xu, W., Lu, M., Pratt, D., Wernecke, M., MaCurdy, T. E., et al. (2020) · The Journal of Infectious Diseases · doi:10.1093/infdis/jiaa080 · Cited by 93
“As sensitivity analyses, we ran a univariate Poisson model with no weight truncation, a doubly-robust model that implemented covariate adjustment along with IPTW, and a model that implemented overlap weights instead of IPTWs as an alternative method to truncation for addressing extreme propensity scores [26, 27].”
PSweight: An R Package for Propensity Score Weighting Analysis
Zhou, T. H., Tong, G., Li, F., Thomas, L. E., & Li, F. (2022) · The R Journal · doi:10.32614/rj-2022-011 · Cited by 92
“PSweight by default implements the empirical sandwich variance for propensity score weighting estimators (Lunceford and Davidian, 2004; Li et al., 2019; Mao et al., 2018) based on the M-estimation theory (Stefanski and Boos, 2002).”
BackgroundMethodSelf-citationRaad, H., Cornelius, V. R., Chan, S., Williamson, E., & Cro, S. (2020) · BMC Medical Research Methodology · doi:10.1186/s12874-020-00947-7 · Cited by 87
“This may result in an additional loss of precision in the small sample setting using such methods [29].”
MethodReal-world evidence: the devil is in the detail
Gokhale, M., Stürmer, T., & Buse, J. B. (2020) · Diabetologia · doi:10.1007/s00125-020-05217-1 · Cited by 85
Survival outcomes of management in metastatic gastric adenocarcinoma patients
Hu, H., Tsai, H., Ku, H., Lo, S.-S., Shan, Y., Chang, H.-C., Chao, Y., Chen, J., Chen, S., Chiang, C., et al. (2021) · Scientific Reports · doi:10.1038/s41598-021-02391-z · Cited by 84
Cheng, C., Fan, L., Thomas, L. E., & Fan, L. (2022) · American Journal of Epidemiology · doi:10.1093/aje/kwac043 · Cited by 84
“Following the literature on PS trimming [9, 11], we consider α = 0.”
BackgroundMethodInfluentialSelf-citationWald, R., Gaudry, S., da Costa, B. R., Adhikari, N. K. J., Bellomo, R., Du, B., Gallagher, M. P., Hoste, E. A., Lamontagne, F., Joannidis, M., et al. (2023) · Intensive Care Medicine · doi:10.1007/s00134-023-07211-8 · Cited by 82
Propensity score weighting for causal subgroup analysis
Yang, S., Lorenzi, E., Papadogeorgou, G., WOJDYLA, D. M., Li, F., & Thomas, L. E. (2021) · Statistics in Medicine · doi:10.1002/sim.9029 · Cited by 79
Self-citationStraub, L., Hernández–Díaz, S., Bateman, B. T., Wisner, K. L., Gray, K. J., Pennell, P. B., Lester, B. M., McDougle, C. J., Suarez, E. A., Zhu, Y., et al. (2022) · JAMA Internal Medicine · doi:10.1001/jamainternmed.2022.0375 · Cited by 80
Propensity score methods for observational studies with clustered data: A review
Chang, T., & Stuart, E. A. (2022) · Statistics in Medicine · doi:10.1002/sim.9437 · Cited by 80
Costa, O. S., Connolly, S. J., Sharma, M., Beyer‐Westendorf, J., Christoph, M. J., Lovelace, B., & Coleman, C. I. (2022) · Critical Care · doi:10.1186/s13054-022-04043-8 · Cited by 78
“For hemostatic effectiveness, mortality, and thrombotic events, propensity score-overlap weighted binomial (logit) generalized estimating equations with a robust estimator were used to calculate odds ratios (ORs) with accompanying 95% confidence intervals (CIs) [17].”
BackgroundMethodSutton, J. D., Stevens, V. W., Chang, N.-C. N., Khader, K., Timbrook, T. T., & Spivak, E. S. (2020) · JAMA Network Open · doi:10.1001/jamanetworkopen.2020.20166 · Cited by 75
Propensity score weighting for covariate adjustment in randomized clinical trials
Zeng, S., Li, F., Wang, R., & Li, F. (2020) · Statistics in Medicine · doi:10.1002/sim.8805 · Cited by 74
“, 2018) as well as empirically reduce the variance of τh in finite samples(Li et al., 2019).”
BackgroundMethodSelf-citationAssociation of Uric Acid–Lowering Therapy With Incident Chronic Kidney Disease
Hassan, W., Shrestha, P., Sumida, K., Thomas, F., Sweeney, P. L., Potukuchi, P. K., Rhee, C. M., Streja, E., Kalantar‐Zadeh, K., & Kovesdy, C. P. (2022) · JAMA Network Open · doi:10.1001/jamanetworkopen.2022.15878 · Cited by 73
Heil, T. C., Verdaasdonk, E. G. G., Maas, H. A. A. M., Munster, B. C. V., Rikkert, M. G. M. O., de Wilt, J. H. W., & Melis, R. J. F. (2022) · Annals of Surgical Oncology · doi:10.1245/s10434-022-12623-9 · Cited by 73
“Additionally, instead of inverse probability weighting (IPW), which is often hampered by extreme propensity scores, this study used the overlap weighting (OW) method.(23) Compared with IPW combined with propensity score trimming, OW produces an unbiased treatment effect estimate with lower variance and good coverage.”
MethodInfluentialFuentes, A., Lüdtke, O., & Robitzsch, A. (2021) · Multivariate Behavioral Research · doi:10.1080/00273171.2021.1925521 · Cited by 66
Wallace, K. L., Zhang, S., Thomas, L. E., Stewart, E. A., Nicholson, W. K., Wegienka, G., Wise, L. A., Laughlin-Tommaso, S. K., Diamond, M. P., Marsh, E. E., et al. (2020) · Fertility and Sterility · doi:10.1016/j.fertnstert.2019.10.028 · Cited by 66
Self-citationCusimano, M. C., Chiu, M., Ferguson, S. E., Moineddin, R., Aktar, S., Liu, N., & Baxter, N. N. (2021) · BMJ · doi:10.1136/bmj-2021-067528 · Cited by 65
Effectiveness and Safety of Anticoagulation Therapy in Frail Patients With Atrial Fibrillation
Kim, D., Yang, P.-S., Sung, J.-H., Jang, E., Yu, H. T., Kim, T.-H., Uhm, J., Kim, J.-Y., Pak, H.-N., Lee, M.-H., et al. (2022) · Stroke · doi:10.1161/strokeaha.121.036757 · Cited by 64
Anchan, R. M., Spies, J. B., Zhang, S., WOJDYLA, D. M., Bortoletto, P. E., Terry, K. L., Disler, E. R., Milne, A., Gargiulo, A., Petrozza, J. C., et al. (2023) · American Journal of Obstetrics and Gynecology · doi:10.1016/j.ajog.2023.05.020 · Cited by 65
Self-citationFradley, M. G., Nguyen, N., Madnick, D. L., Chen, Y., DeMichele, A. M., Makhlin, I., Dent, S. F., Lefebvre, B., Carver, J. R., Upshaw, J. N., et al. (2023) · Journal of the American Heart Association · doi:10.1161/jaha.123.029361 · Cited by 63
Bovéda, S., Higuera, L., Longacre, C. F., Wolff, C., Wherry, K., Stromberg, K. D., & El‐Chami, M. F. (2023) · EP Europace · doi:10.1093/europace/euad016 · Cited by 62
“The overlap weights adjustment strategy uses the probability of receiving therapy with the opposing treatment based on characteristics used to construct the propensity score to place the most weight on patients considered the most exchangeable and the least emphasis on patients who are least likely to receive the oppo…”
MethodAustin, P. C. (2022) · Statistics in Medicine · doi:10.1002/sim.9519 · Cited by 61
Lawal, O. D., Aronow, H. D., Shobayo, F. O., Hume, A. L., Taveira, T. H., Matson, K. L., Zhang, Y., & Wen, X. (2023) · Circulation · doi:10.1161/circulationaha.122.060687 · Cited by 60
Eliminating Ambiguous Treatment Effects Using Estimands
Kahan, B. C., Cro, S., Li, F., & Harhay, M. O. (2023) · American Journal of Epidemiology · doi:10.1093/aje/kwad036 · Cited by 59
Self-citationLee, H., Lee, H. W., Kim, S. U., & Kim, H. C. (2022) · Clinical and Translational Gastroenterology · doi:10.14309/ctg.0000000000000435 · Cited by 58
“The association persisted when propensity score weighting was used to balance all covariates between the MAFLD and non-MAFLD groups (27).”
MethodTang, B., Sjölander, A., Wastesson, J. W., Maura, G., Blotiere, P.-O., Szilcz, M., Mak, J. K. L., Qin, C., Alvarsson, M., Religa, D., et al. (2024) · eClinicalMedicine · doi:10.1016/j.eclinm.2024.102689 · Cited by 57
“However, a popular science press release will be issued on the Karolinska Institutet website.”
Misra‐Hebert, A. D., Jehi, L. E., Ji, X., Nowacki, A. S., Gordon, S. M., Terpeluk, P., Chung, M. K., Mehra, R., Dell, K. M., Pennell, N. A., et al. (2020) · Journal of General Internal Medicine · doi:10.1007/s11606-020-06171-9 · Cited by 57
“The overlap propensity score weighting method was then applied where each patient’s statistical weight is the probability of that patient being assigned to the opposite group.(21)”
MethodEnriching single-arm clinical trials with external controls: possibilities and pitfalls
Lambert, J., Lengliné, É., Porcher, R., Thiébaut, R., Zohar, S., & Chevret, S. (2023) · Blood Advances · doi:10.1182/bloodadvances.2022009167 · Cited by 58
“Thus, balancing weights has been proposed as a simple way to de fi ne, based on speci fi c tilting functions, individual weights, and the resulting target population, 54 as it integrates most approaches, including PS matching.”
The Causal Cookbook: Recipes for Propensity Scores, G-Computation, and Doubly Robust Standardization
Chatton, A., & Rohrer, J. M. (2024) · Advances in Methods and Practices in Psychological Science · doi:10.1177/25152459241236149 · Cited by 58
“Alternatively, one may use overlap weights (see Table 2) that target causal effects in the overlapping population, that is, in the part of the population in which positivity holds (F. Li et al., 2019).”
BackgroundPallesen, A. V. J., Kristiansen, M., Westendorp, R. G. J., & Mortensen, L. H. (2022) · PLoS ONE · doi:10.1371/journal.pone.0264332 · Cited by 56
“The main aim of this method is to emulate randomised assignment to the treatment and control group [26].”
BackgroundMethodInfluentialVaricella-zoster virus reactivation and the risk of dementia
Polisky, V., Littmann, M., Triastcyn, A., Horn, M., Georgiou, A., Widenmaier, R., Anspach, B., Tahrat, H., Kumar, S., Buser‐Doepner, C., et al. (2025) · Nature Medicine · doi:10.1038/s41591-025-03972-5 · Cited by 57
“We additionally tested alternative analysis approaches, including inverse probability of treatment weighting (IPTW) 58 and overlap weights (OW) 59 where samples are not matched but weighted based on their propensity scores.”
Subgroup balancing propensity score
Jing, D., Zhang, J. L., Zeng, S., & Li, F. (2019) · Statistical Methods in Medical Research · doi:10.1177/0962280219870836 · Cited by 56
“Note that the ATT weights is a special case of the class of balancing weights [8, 9]; other members include the inverse probability weights and the overlap weights, each corresponding to a different estimand.”
BackgroundSelf-citationLai, Q., Vitale, A., Iesari, S., Finkenstedt, A., Mennini, G., Onali, S., Hoppe‐Lotichius, M., Manzia, T. M., Nicolini, D., Avolio, A. W., et al. (2019) · Liver Transplantation · doi:10.1002/lt.25492 · Cited by 55
“The IPTW is prone to conceptual drawbacks, but this methodology represents the most rigorous way to re-equilibrate the sample to test.(12) Consistently, all the tests used to check the successful balancing of the 2 study groups confirmed the validity of our method.”
MethodXie, Y., Bowe, B., Gibson, A. K., McGill, J. B., Maddukuri, G., & Al-Aly, Z. (2021) · JAMA Internal Medicine · doi:10.1001/jamainternmed.2021.2488 · Cited by 48
Clinical evaluation of a machine learning–based early warning system for patient deterioration
Verma, A. A., Stukel, T. A., Colacci, M., Bell, S., Ailon, J., Friedrich, J. O., Murray, J., Kuzulugil, S. S., Yang, Z., Lee, Y., et al. (2024) · Canadian Medical Association Journal · doi:10.1503/cmaj.240132 · Cited by 53
“Each patient was weighted according to the overlap weight, 28 which is the probability of being assigned to the opposite exposure group based on propensity score.”
Quinn, K. L., Stukel, T. A., Huang, A., Abdel-Qadir, H. M., Altaf, A., Bell, C. M., Cheung, A. M., Detsky, A. S., Goulding, S., Herridge, M. S., et al. (2023) · JAMA Internal Medicine · doi:10.1001/jamainternmed.2023.2228 · Cited by 52
Saxena, F. E., Bierman, A. S., Glazier, R. H., Wang, X., Guan, J., Lee, D. S., & Stukel, T. A. (2022) · JAMA Network Open · doi:10.1001/jamanetworkopen.2022.22056 · Cited by 51
A Tutorial Introduction to Heterogeneous Treatment Effect Estimation with Meta-learners
Salditt, M., Eckes, T., & Nestler, S. (2023) · Administration and Policy in Mental Health and Mental Health Services Research · doi:10.1007/s10488-023-01303-9 · Cited by 51
Ohbe, H., Sasabuchi, Y., Yamana, H., Matsui, H., & Yasunaga, H. (2021) · The Lancet Regional Health - Western Pacific · doi:10.1016/j.lanwpc.2021.100185 · Cited by 52
The how and why of Bayesian nonparametric causal inference
Linero, A. R., & Antonelli, J. (2022) · Wiley Interdisciplinary Reviews Computational Statistics · doi:10.1002/wics.1583 · Cited by 50
Nethery, R. C., Mealli, F., & Dominici, F. (2019) · The Annals of Applied Statistics · doi:10.1214/18-aoas1231 · Cited by 51
Wei, J., Choi, H. K., Dalbeth, N., Li, X., Li, C., Zeng, C., Lei, G., & Zhang, Y. (2023) · JAMA Network Open · doi:10.1001/jamanetworkopen.2023.30885 · Cited by 50
Zhou, S., Chen, R., Liu, J., Guo, Z., Su, L., Li, Y., Zhang, X., Luo, F., Gao, Q., Lin, Y., et al. (2024) · Annals of Internal Medicine · doi:10.7326/m24-0178 · Cited by 50
Evaluating the causal effects of cellphone distraction on crash risk using propensity score methods
Lu, D., Guo, F., & Li, F. (2020) · Accident Analysis & Prevention · doi:10.1016/j.aap.2020.105579 · Cited by 48
Self-citationAbdel-Nabey, M., Chaba, A., Serre, J., Lengliné, É., Azoulay, É., Darmon, M., & Zafrani, L. (2022) · Annals of Intensive Care · doi:10.1186/s13613-022-00990-1 · Cited by 48
“This allows assigning a higher weight to patients with intermediate-risk and lower weight to outliers in both treatment groups, the analysis emphasizing the proportion of the population where the most treatment equipoise exists in clinical practice [25].”
BackgroundPropensity score methods in observational research: brief review and guide for authors
Andrew, B. Y., Brookhart, M. A., Pearse, R. M., Raghunathan, K., & Krishnamoorthy, V. (2023) · British Journal of Anaesthesia · doi:10.1016/j.bja.2023.06.054 · Cited by 48
Multinomial Extension of Propensity Score Trimming Methods: A Simulation Study
Yoshida, K., Solomon, D. H., Haneuse, S., Kim, S. C., Patorno, E., Tedeschi, S. K., Lyu, H., Franklin, J. M., Stürmer, T., Hernández–Díaz, S., et al. (2018) · American Journal of Epidemiology · doi:10.1093/aje/kwy263 · Cited by 48
Mlcoch, T., Hrnčiarová, T., Tužil, J., Zadak, J., Marian, M., & Doležal, T. (2019) · Value in Health · doi:10.1016/j.jval.2019.06.010 · Cited by 46
Use of an AI scribe and electronic health record efficiency
Pearlman, K., Wan, W., Shah, S., & Laiteerapong, N. (2025) · JAMA Network Open · doi:10.1001/jamanetworkopen.2025.37000 · Cited by 45
Cipolletta, E., Nakafero, G., McCormick, N., Yokose, C., Avery, A. J., Mamas, M. A., Choi, H. K., Tata, L. J., & Abhishek, A. (2024) · The Lancet Rheumatology · doi:10.1016/s2665-9913(24)00248-0 · Cited by 45
Wee, L. E., Tay, A. T., Chiew, C. J., Young, B. E., Wong, B., Lim, R., Lee, C. L., Tan, J. Y. S., Vasoo, S. S., Lye, D. C., et al. (2023) · Clinical Microbiology and Infection · doi:10.1016/j.cmi.2023.06.016 · Cited by 44
“To address the possibility 191 that extreme propensity scores encountered using IPTW could bias the estimator and induce 192 excessive variance, we utilised overlap weighting as an additional comparison.[24] To evaluate if an 193”
Basile, C., Lindberg, F., Benson, L., Guidetti, F., Dahlström, U., Piepoli, M. F., Mol, P. G. M., Scorza, R., Maggioni, A. P., Lund, L. H., et al. (2025) · Circulation · doi:10.1161/circulationaha.124.072855 · Cited by 42
“14,15 Average treatment effect in the overlap population was chosen as estimand because of its better ability to achieve covariate balance and robustness to unmeasured confounding compared with conventional weighting estimands.”
Massicotte‐Azarniouch, D., Petrcich, W., Walsh, M. W., Canney, M., Hundemer, G. L., Milman, N., Hladunewich, M., Fairhead, T., & Sood, M. M. (2021) · Clinical Kidney Journal · doi:10.1093/ckj/sfab229 · Cited by 42
Sahai, A., Bhandari, R., Koupenova, M., Freedman, J. E., Godwin, M., McIntyre, T. M., Chung, M. K., Iskandar, J., Kamran, H., Aggarwal, A., et al. (2020) · doi:10.21203/rs.3.rs-119031/v1 · Cited by 41
Hager, K., Shi, P., Li, Z., Chui, K. K. H., Berkowitz, S. A., Mozaffarian, D., Chhabra, J., Wilken, J., Vergara, C., Becker, E. C., et al. (2023) · Diabetes Care · doi:10.2337/dc22-1645 · Cited by 40
“Statistical Methods We used an overlap weight propensity score approach to weight the control and treatment groups (29).”
MethodPrescription Stimulant Use During Pregnancy and Risk of Neurodevelopmental Disorders in Children
Suarez, E. A., Bateman, B. T., Hernández–Díaz, S., Straub, L., McDougle, C. J., Wisner, K. L., Gray, K. J., Pennell, P. B., Lester, B. M., Zhu, Y., et al. (2024) · JAMA Psychiatry · doi:10.1001/jamapsychiatry.2023.5073 · Cited by 41
McCormick, N., Yokose, C., Wei, J., Lu, N., Wexler, D. J., Aviña-Zubieta, J. A., Vera, M. A. D., Zhang, Y., & Choi, H. K. (2023) · Annals of Internal Medicine · doi:10.7326/m23-0724 · Cited by 39
Jouffroy, R., Darmon, M., Isnard, F., Géri, G., Beurton, A., Fartoukh, M. S., Tudesq, J., Nemlaghi, S., Demoule, A., Azoulay, É., et al. (2021) · Journal of Critical Care · doi:10.1016/j.jcrc.2021.04.014 · Cited by 39
“This allows higher weight to be assigned to patients with intermediate risk and lower weight to outliers in both treatment groups, the analysis emphasizing the proportion of the population where the most treatment equipoise exists in clinical practice [19].”
Karacabeyli, D., Lacaille, D. V., Lu, N., McCormick, N., Xie, H. W., Choi, H. K., & Aviña-Zubieta, J. A. (2024) · PLoS ONE · doi:10.1371/journal.pone.0308533 · Cited by 39
Effect of aspirin on short-term outcomes in hospitalized patients with COVID-19
Sahai, A., Bhandari, R., Godwin, M., McIntyre, T. M., Chung, M. K., Iskandar, J., Kamran, H., Hariri, E. H., Aggarwal, A., Burton, R., et al. (2021) · Vascular Medicine · doi:10.1177/1358863x211012754 · Cited by 38
Ren, J., Cislo, P. R., Cappelleri, J. C., Hlavacek, P., & DiBonaventura, M. (2023) · BMC Medical Research Methodology · doi:10.1186/s12874-023-01835-6 · Cited by 37
“Both IPTW and SMR use the propensity score as a part of denominator for calculating the weight, but OW does not.”
MethodMontagne, F., Chaari, Z., Bottet, B., Sarsam, M., Mbadinga, F., Selim, J., Guisier, F., Gillibert, A., & Baste, J. (2022) · Cancers · doi:10.3390/cancers14112611 · Cited by 36
“The difference of 5-year DFS was calculated from 5-year Kaplan–Meier estimates with propensity-score weighting, using overlap weights [38], combined with multiple imputation by fully conditional specification and non-parametric percentile bootstrap in a Boot-MI sequence [39].”
MethodCausal Inference with Generative Artificial Intelligence: Application to Texts as Treatments
Imai, K., & Nakamura, K. (2026) · Journal of the American Statistical Association · doi:10.1080/01621459.2026.2689629 · Cited by 35
“If some estimated scores are too close to 0 or 1, one may trim them (Crump et al., 2009), clip them (Dorn, 2025), or use overlap weights (Li et al., 2019).”
Xu, X., Zhang, Y., Gan, J., Ye, X., Yu, X., & Huang, Y. (2021) · British Journal of Anaesthesia · doi:10.1016/j.bja.2021.05.031 · Cited by 35
Dexamethasone for Inpatients With COVID-19 in a National Cohort
Mourad, A., Thibault, D. P., Holland, T. L., Yang, S., Young, A. R., Egloff, S. A. A., & Thomas, L. E. (2023) · JAMA Network Open · doi:10.1001/jamanetworkopen.2023.8516 · Cited by 35
Self-citationvan der Laan, L. W. P., Carone, M., & Luedtke, A. R. (2024) · arXiv (Cornell University) · doi:10.48550/arxiv.2402.01972 · Cited by 35
Causal inference in the absence of positivity: The role of overlap weights
Matsouaka, R., & Zhou, Y. (2024) · Biometrical Journal · doi:10.1002/bimj.202300156 · Cited by 34
“…to both propensity score misspecification and extrapolation of outcome regression models matter most in the area of limited covariate distributions overlap (Canavire-Bacarreza et al., 2021; Ju et al., 2018; J. Kang et al., 2016; Khan & Tamer, 2010; Li, Thomas, et al., 2018 1 − ˜𝑒(𝑋) showcases…”
BackgroundMethodResultInfluential
Metadata
- Authors
- Li, F., & Thomas, L. E.
- Affiliations
- Duke University (US)
- Corresponding author
- Li, Fan
- Venue
- American Journal of Epidemiology
- Publisher
- Oxford University Press (OUP)
- Published
- 2018-09-05
- Volume
- 188
- Issue
- 1
- Pages
- 189-190
- ISSN
- 0002-9262 (ISSN-L)
- 1476-6256
- OpenAlex
- W2907932858
- Cited by
- 694
- References
- 19
- Citations by year
- 2026: 97 · 2025: 124 · 2024: 145 · 2023: 119 · 2022: 83 · 2021: 73 · 2020: 35 · 2019: 15 · 2018: 1
- FWCI
- 25.73
- Type
- journal article
- Language
- en
- Open access
- closed
- Licence
- academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model (am)
- Locations
- American Journal of Epidemiology (publishedVersion)
- PubMed (publishedVersion)
- Funders
- Agency for Healthcare Research and Quality (SES-1424688)
- National Science Foundation
- Topics
- Physical Sciences › Mathematics › Statistics and Probability › Advanced Causal Inference Techniques
- Physical Sciences › Mathematics › Statistics and Probability › Statistical Methods in Clinical Trials
- Social Sciences › Decision Sciences › Statistics, Probability and Uncertainty › Meta-analysis and systematic reviews
- Keywords
- inverse probability weighting, treatment effect estimation, bias and variance, confidence interval coverage, standard error estimation, causal inference
- MeSH terms
- Epidemiologic Methods, Humans, Models, Statistical, Bias, Causality, Propensity Score
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.