An Introduction to Propensity Score Methods for Reducing the Effects of Confounding in Observational Studies

2011/05/31 by Peter C. Austin · 12,108 citations
Mathematics · #Advanced Causal Inference Techniques #Confounding #Covariate #Econometrics #Inverse probability weighting #Mathematics #Observational study #Propensity score matching #Statistical Methods and Bayesian Inference #Statistical Methods and Inference #Statistics

paper · pdf · doi:10.1080/00273171.2011.568786

published in Multivariate Behavioral Research 46(3), 399-424 (Taylor & Francis)

openalex publication_date 2011/05/31 · openalex created_date 2025/10/10 · openalex updated_date 2026/08/06

Abstract

The propensity score is the probability of treatment assignment conditional on observed baseline characteristics. The propensity score allows one to design and analyze an observational (nonrandomized) study so that it mimics some of the particular characteristics of a randomized controlled trial. In particular, the propensity score is a balancing score: conditional on the propensity score, the distribution of observed baseline covariates will be similar between treated and untreated subjects. I describe 4 different propensity score methods: matching on the propensity score, stratification on the propensity score, inverse probability of treatment weighting using the propensity score, and covariate adjustment using the propensity score. I describe balance diagnostics for examining whether the propensity score model has been adequately specified. Furthermore, I discuss differences between regression-based methods and propensity score-based methods for the analysis of observational data. I describe different causal average treatment effects and their relationship with propensity score analyses.

Citations

• Estimating causal effects of treatments in randomized and nonrandomized studies.
• The central role of the propensity score in observational studies for causal effects
• Variable Selection for Propensity Score Models

Cited by

• Long-term gastrointestinal outcomes of COVID-19
• LLM-Extracted Covariates for Clinical Causal Inference: Rethinking Integration Strategies
• A Peek into the Political Biases in Email Spam Filtering Algorithms During US Election 2020
• Knowledge-Guided Time-Varying Causal Inference for Arctic Sea Ice Dynamics
• Endogeneity and marketing strategy research: an overview
• Cell-type-directed network-correcting combination therapy for Alzheimer’s disease
• Two-year trajectory of cognitive decline and neurological sequelae in COVID-19 survivors with acute neurological symptoms
• Speed at the Cost of Quality
• Population-Level Effectiveness and Herd Protection 17 Years After HPV Vaccine Introduction
• Studying Philosophy Does Make People Better Thinkers
• Neurodegeneration and Stroke After Semaglutide and Tirzepatide in Patients With Diabetes and Obesity
• When deep learning meets causal inference: a computational framework for drug repurposing from real-world data
• Impact of tracheal extubation location after surgical procedures on peri‐operative times: a prospective dual‐centre observational study*
• proxymate: Diagnosis and Adjustment of Proxy Estimates for Reliable Inference
• A Survey on Causal Inference
• On Efficient Adjustment for Micro Causal Effects in Summary Causal Graphs
• The contribution of the color space in LSST-like photometry for the selection of extragalactic globular cluster candidates
• Stabilising Learner Trajectories: A Doubly Robust Evaluation of AI-Guided Student Support using Activity Theory
• Causal Explanations for Disparate Trends: Where and Why?
• Adversarial Counterfactual Learning and Evaluation for Recommender System
• Association of Advanced Airway Insertion Timing and Outcomes After Out-of-Hospital Cardiac Arrest
• Remote Patient Monitoring Program in Automated Peritoneal Dialysis: Impact on Hospitalizations
• Technique failure in remote patient monitoring program in patients undergoing automated peritoneal dialysis: A retrospective cohort study
• Sex-based differences in long-term respiratory and pulmonary vascular outcomes after acute pulmonary embolism: A propensity-matched analysis of 795,338 patients
• FedECA: Federated External Control Arms for Causal Inference with Time-To-Event Data in Distributed Settings
• Dissociation between resuscitation success and neurological recovery in nursing home cardiac arrest: A propensity-matched analysis
• Causal Inference About the Effects of Interventions From Observational Studies in Medical Journals
• Trends in drug–drug interaction prevalence and longitudinal associations with health outcomes in the older community‐dwelling population: Findings from The Irish LongituDinal study on Ageing
• Comparison of outcomes after living and deceased donor kidney transplantation: UK national cohort study
• Young People With and Without Intellectual Disability Accessing Mental Health Services: Evaluating Psychosocial Functioning Outcomes Using Electronic Health Records
• AI Diffusion in Low Resource Language Countries
• Beyond Single-Tokenomics: How Farcaster's Pluralistic Incentives Reshape Social Networking
• Association of adrenal insufficiency with patient-oriented health-care outcomes in adult medical inpatients
• Imitation-Regularized Offline Learning
• A Primer on Deep Learning for Causal Inference
• Regret Minimization for Causal Inference on Large Treatment Space
• Combining Offline Causal Inference and Online Bandit Learning for Data Driven Decision
• Vancomycin combined with piperacillin/tazobactam increases the risk of acute kidney injury compared with vancomycin plus other anti-pseudomonal beta-lactams: a systematic review and network meta-analysis
• Addressing the joint occurrence of self-selection and simultaneity biases in the estimation of program effects based on cross-sectional observational surveys: case study of travel behavior effects in carsharing
• The harms of hate
• The Causal Cookbook: Recipes for Propensity Scores, G-Computation, and Doubly Robust Standardization
• Intermittently scanned continuous glucose monitoring compared with blood glucose monitoring is associated with lower HbA1c and a reduced risk of hospitalisation for diabetes-related complications in adults with type 2 diabetes on insulin therapies
• Deciphering the influence of demographic factors on the treatment of pediatric patients in the emergency department
• How do dataset characteristics affect the performance of propensity score methods and regression for controlling confounding in observational studies? A simulation study
• What does a woman like me do in a network like this? Gendered self-regulation in professional networking
• Hypothermic vs Normothermic Machine Perfusion in Liver Transplantation in the US: A Cohort Study
• Comparison of Propensity Score Methods and Covariate Adjustment
• Why Adaptively Collected Data Have Negative Bias and How to Correct for It
• Deep Bayesian Estimation for Dynamic Treatment Regimes with a Long Follow-up Time
• Balance-Subsampled Stable Prediction
• Accelerating emergence: the causal (but contextual) effect of social impact accelerators on nascent for-profit social ventures
• Privacy-Preserving Causal Inference via Inverse Probability Weighting
• Novel Systemic Anticancer Treatments and Health Services Use at the End of Life Among Adults With Cancer
• A population-level analysis on the association of cannabis use and urologic cancers
• Causal inference with observational data: A tutorial on propensity score analysis
• Sample Debiasing in the Themis Open World Database System (Extended Version)
• Improving the Generation and Evaluation of Synthetic Data for Downstream Medical Causal Inference
• On Variance of the Treatment Effect in the Treated When Estimated by Inverse Probability Weighting
• Investigation of the effectiveness of a community-based parent education program to engage families in increasing language interactions with their children
• Coping styles and mental health in response to societal changes during the COVID-19 pandemic
• Racial/Ethnic Differences in Relationships Between Pregnancy Intentions and Maternal Outcomes
• Association of glucagon-like peptide-1 receptor agonists with serious liver events among patients with type 2 diabetes: A Scandinavian cohort study
• Missing Data Imputation in the Context of Propensity Score Analysis: A Systematic Review
• Impacts of glucagon‐like peptide‐1 receptor agonists on the risk of adverse liver outcomes in patients with metabolic dysfunction‐associated steatotic liver disease cirrhosis and type 2 diabetes
• Marginalization of Regression-Adjusted Treatment Effects in Indirect Comparisons with Limited Patient-Level Data
• The sociobiome–oral microbiome pathway in dental caries among Indigenous Australians
• Callous‐unemotional traits, cognitive functioning, and externalizing problems in a propensity‐matched sample from the ABCD study
• Mode of anaesthesia and persistent postoperative opioid use: A reply
• Philosophy programs in prison and rehabilitative outcomes – A scoping review
• Dynamic effects of COVID-19 vaccination on major acute cardiovascular events and mortality following SARS-CoV-2 infection in a target trial emulation study
• AI Companions Reduce Loneliness
• The Corporatization of Independent Hospitals
• Bridging ideal studies and real-world practice: the potential role of real-world evidence in reproductive medicine
• Acetaminophen and Clinical Outcomes in Sepsis
• Impact of Baseline Kidney Function on the Rate of Progressive Kidney Disease After Pregnancy: A Population-Based Cohort Study Research Protocol
• Does low self-esteem predict lower wellbeing following relationship dissolution?
• Childhood Factors and Positive Functioning Among Adults Born Very Preterm or With Very Low Birth Weight
• Home cooking, cooking skills and dementia requiring long-term care: a population-based cohort study in Japan
• Potassium binders and continuation of renin–angiotensin system inhibitors/mineralocorticoid receptor antagonist in chronic kidney disease and heart failure (the DEMONSTRATE database)
• Disability onset and labor market outcomes
• Outcomes in Early Adulthood for Individuals Born Very Preterm and/or with Very Low Birth Weight: Evidence from Multinational Cohorts
• Urban mobility and learning: analyzing the influence of commuting time on students' GPA at Politecnico di Milano
• Effect of focal adenomyosis on pregnancy outcomes: retrospective propensity score matching analysis of 13,186 IVF/ICSI cycles
• Propensity score matching and complex surveys
• Estimating the effect of treatment on binary outcomes using full matching on the propensity score
• Variasi Bentuk Penguatan pada Pembelajaran Bahasa Jawa Kelas VII di SMP Se-Kecamatan Winong Kabupaten Pati
• Double propensity-score adjustment: A solution to design bias or bias due to incomplete matching
• GLP-1 receptor agonist use and cancer risk in obese nondiabetic adults
• A Unified Three-State Model Framework for Analysis of Treatment Crossover in Survival Trials
• The Effect of an Emergency Savings Program on Employee Savings and Work Performance: A Two-Year Field Intervention
• Core concepts in statistics and research methods. Part 2: clinical research principles and observational studies
• Age at entry into the Dutch child protection system of children of parents with intellectual disability: A case-control study
• On variance estimation of the inverse probability‐of‐treatment weighting estimator: A tutorial for different types of propensity score weights
• Mandatory Vaccination in Europe
• The long‐term effects of formal child support
• The Impact of Long-Term Finance on Job Quality, Investments and Firm Performance: Cross-Country Evidence
• Revisiting the impact of decoupled subsidies on farm performance: a counterfactual analysis using microdata
• Treatments with versus without medication for children with behavioural difficulties in clinical practice: an economic evaluation with observational data
• Young people's lived experience expertise: Insights from the DigiCAT project to develop a counterfactual analysis tool for mental health data
• A Novel Method for Generating Spatially Resolved Synthetic Populations for Health Impact Assessments in Vulnerable Populations
• Bridging genomics and pharmacoepidemiology to expand treatment options for alcohol use disorder
• Interpretable Clinical Classification with Kolmogorov-Arnold Networks
• contextual: Evaluating Contextual Multi-Armed Bandit Problems in R
• Moving towards best practice when using inverse probability of treatment weighting (IPTW) using the propensity score to estimate causal treatment effects in observational studies
• Learning Representations for Counterfactual Inference
• Stable Prediction across Unknown Environments
• Effects of registered nurse staffing level on hospital-acquired conditions in cardiac surgery patients: A propensity score matching analysis
• Effect Identification and Unit Categorization in the Multi-Score Regression Discontinuity Design with Application to LED Manufacturing
• Causal Intervention for Weakly-Supervised Semantic Segmentation
• Lower risks of cirrhosis and hepatocellular carcinoma with GLP‐1RAs in type 2 diabetes: A nationwide cohort study using target trial emulation framework
• Assessing the impact of day programs on individuals living with dementia and their family/friend caregivers (AIDA-DemCare): protocol of a prospective cohort study combined with a qualitative evaluation
• A hierarchical modelling approach for Bayesian Causal Forests on longitudinal data: A Case Study in Multiple Sclerosis Clinical Trials
• Style is NOT a single variable: Case Studies for Cross-Style Language Understanding
• Immunosuppressed patients with COVID-19 pneumonia in ICU: clinical characteristics and factors influencing outcomes
• Bridging Cultures in the Era of Big Data: A Cross-Language Equivalence Framework in Machine-Learning Research With Social Media Texts
• Impact of integrase strand transfer inhibitors on cardiovascular disease in people with HIV
• From exposure to preference: How solar electrification experiences shape long-term sustainability in rural Pakistan
• Socioeconomic Factors and Initiation of Semaglutide or Tirzepatide Among Medicare Beneficiaries With Type 2 Diabetes
• Learning Modular Safe Policies in the Bandit Setting with Application to Adaptive Clinical Trials
• DFW: A Novel Weighting Scheme for Covariate Balancing and Treatment Effect Estimation
• The impacts of widespread agricultural interventions on yields and food security in Ethiopia☆
• Probabilistic Matching: Causal Inference under Measurement Errors
• A Comparative Evaluation of Statistical Methods in Hybrid Controlled Trials
• Large-scale Causal Approaches to Debiasing Post-click Conversion Rate Estimation with Multi-task Learning
• A framework for massive scale personalized promotion
• Oncogenic stress alters predator-prey interactions between a globally invasive and a native planaria species
• Long-Term Effectiveness of Financial Incentives for Not Prescribing Unnecessary Antibiotics to Children With Acute Respiratory and Gastrointestinal Infections: Japan's Nationwide Quasi-Experimental Study
• Uplift Modeling from Separate Labels
• Long‐term efficacy of rituximab versus intravenous cyclophosphamide for severe ANCA‐associated vasculitis in multicenter REVEAL cohort study
• Effects of COVID‐19 economic impacts on alcohol use disorder symptoms are mediated by maximum and volume of alcohol intake: Data from the National Alcohol Survey's COVID Cohort
• Assessing the Impact of Covariate Distribution and Positivity Violation on Weighting-Based Indirect Comparisons: a Simulation Study
• Self-Employment and Eudaimonic Well-Being: The Mediating Role of Problem- and Emotion-Focused Coping
• Use of dipeptidyl peptidase‐4 inhibitors is associated with lower risk of severe renal outcomes in pre‐dialysis patients with Type 2 diabetes
• The landscape of hepatitis C virus infection in pediatric kidney transplantation
• A Propensity-score-based Fine Stratification Approach for Confounding Adjustment When Exposure Is Infrequent
• Effectiveness and safety of outpatient rivaroxaban versus warfarin for treatment of venous thromboembolism in patients with a known primary hypercoagulable state
• A unified survey of treatment effect heterogeneity modeling and uplift modeling
• Deep Disentangled Representation Network for Treatment Effect Estimation
• Why the World Reads Wikipedia: Beyond English Speakers
• Applications and Modeling of Keystroke Logs in Writing Assessments
• Cardiovascular post-acute sequelae of SARS-CoV-2 in children and adolescents: cohort study using electronic health records
• The effect of expanded access to mental health care on economic status of households with a person with a mental disorder in rural Ethiopia: a controlled before-after study
• The effect of radical cystectomy on survival in patients with metastatic urothelial carcinoma of the urinary bladder
• A deterministic matching method for exact matchings to compare the outcome of different interventions
• Frontiers: The Impact of Ad-Blockers on Online Consumer Behavior
• The impact of extracurricular education on socioeconomic mobility in Japan: an application of causal machine learning
• The Effect of Morbid Obesity on Complications, Readmission, and Patient-Reported Outcomes Following Minimally Invasive Transforaminal Lumbar Interbody Fusion
• Effectiveness of Prehospital Critical Care Scene Response for Major Trauma: A Systematic Review
• Association of cannabis abuse/dependence on risks of erectile dysfunction and testosterone deficiency using a large claims database analysis
• Effectiveness of Track Cauterization in Reduction of Adverse Events for Lung Microwave Ablation
• Adverse effects of progestin-primed ovarian stimulation: combination of clinical study and single-cell analysis
• Learning Decomposed Representation for Counterfactual Inference
• Generalization Bounds and Representation Learning for Estimation of\n Potential Outcomes and Causal Effects
• Deep Counterfactual Networks with Propensity-Dropout
• Assessing the Effect of School‐Based Health Centers on Achievement of National Performance Measures
• A survey of using EHR as real-world evidence for discovering and validating new drug indications
• Evaluating Gender Wage Inequality in Academia using Causal Inference Methods for Observational Data
• Cross-Sectional Studies
• Multicalibrated Partitions for Importance Weights
• Counterfactual-based Incrementality Measurement in a Digital Ad-Buying Platform
• The use of propensity score methods with survival or time‐to‐event outcomes: reporting measures of effect similar to those used in randomized experiments
• A Distributionally Robust Framework for Nuisance in Causal Effect Estimation
• Framing Causal Questions in Sports Analytics: A Tutorial on Estimand Choice Illustrated Through Crossing in Soccer
• Characteristics of persons who die on their first suicide attempt: results from the National Violent Death Reporting System
• Evaluating sources of bias in observational studies of angiotensin-converting enzyme inhibitor/angiotensin II receptor blocker use during COVID-19: beyond confounding
• Vankeusaikaisten kuntoutusohjelmien vaikutusarvioinnit ja arvioitavuus
• The Efficacy of Propensity Score Matching for Separating Selection and Measurement Effects Across Different Survey Modes
• The Tradeoff of Temporariness: Economic and Social Impacts of H-2A Status on Mexican Migrant Men
• Non-TNFi biologic and targeted synthetic DMARDs in rheumatoid arthritis-associated interstitial lung disease: A propensity score-matched, active-comparator, new-user study
• Benchmarking Framework for Performance-Evaluation of Causal Inference Analysis
• Combating Summer Melt: The Impact of Near-Peer Mentor Matriculation Program in New York City
• Certificates without Electrons? Theory and Evidence on Impacts from AI-Driven Power Demand
• Estimating Covariate-balanced Survival Curve in Distributed Data Environment using Data Collaboration Quasi-Experiment
• Smooth Multi-Policy Causal Effect Estimation in Longitudinal Settings
• Gentrification, Geography, and the Declining Enrollment of Neighborhood Schools
• Entropy-balanced accruals
• Economic aspects of treating seizure clusters
• Safety and efficacy of digoxin: systematic review and meta-analysis of observational and controlled trial data
• Surgical outcomes in nonagenarian versus octogenarian patients: a propensity-matched analysis with implications for shared decision-making
• Creativity and Mood Disorder: A Systematic Review and Meta-Analysis
• Proton Pump Inhibitors and Risk of Incident CKD and Progression to ESRD
• The Estimation of Continual Causal Effect for Dataset Shifting Streams
• Balancing Method for High Dimensional Causal Inference
• Effect of tranexamic acid on mortality in patients with haemoptysis: a nationwide study
• Generalized propensity score weighting for functional causal inference framework
• Earlier Surgery Reduces Complications in Acute Traumatic Thoracolumbar Spinal Cord Injury: Analysis of a Multi-Center Cohort of 4108 Patients
• Reality of Accomplishing Surgery within 24 Hours for Complete Cervical Spinal Cord Injury: Clinical Practices and Safety
• The impact of 3D printing implementation on stock returns
• Estimating Heterogeneity in Travel Mode Choice Shifts with Causal Forests
• Methods for population adjustment with limited access to individual patient data: A review and simulation study
• On variance estimation for generalizing from a trial to a target population
• Trajectories of coparenting quality across ethnically diverse and interethnic parents
• Learning Weighted Representations for Generalization Across Designs
• Sarcopenia using muscle mass prediction model and cognitive impairment: A longitudinal analysis from the English longitudinal study on ageing
• Breath analysis for non-invasive detection of breast cancer: A prospective observational cohort study
• Potential benefit and lack of serious risk from corticosteroids in drug‐induced liver injury: An international, multicentre, propensity score‐matched analysis
• Performance of Prehospital Antibiotic Administration and Blood Culture Collection in a Physician-Staffed Mobile Unit: A Retrospective Cohort Study
• The use of propensity score methods with survival or time-to-event outcomes: reporting measures of effect similar to those used in randomized experiments. [europepmc]
• Moving towards best practice when using inverse probability of treatment weighting (IPTW) using the propensity score to estimate causal treatment effects in observational studies. [europepmc]
• The REporting of studies Conducted using Observational Routinely-collected health Data (RECORD) statement. [europepmc]
• The Propensity Score. [europepmc]
• Proton Pump Inhibitors and Risk of Incident CKD and Progression to ESRD. [europepmc]
• A Propensity-score-based Fine Stratification Approach for Confounding Adjustment When Exposure Is Infrequent. [europepmc]
• Natural Experiments: An Overview of Methods, Approaches, and Contributions to Public Health Intervention Research. [europepmc]
• Environmental Risk Factors for Developing Type 2 Diabetes Mellitus: A Systematic Review. [europepmc]
• Sex and Gender Differences Research Design for Basic, Clinical, and Population Studies: Essentials for Investigators. [europepmc]
• Coronary Atherosclerotic Precursors of Acute Coronary Syndromes. [europepmc]
• Treatment correlates of successful outcomes in pulmonary multidrug-resistant tuberculosis: an individual patient data meta-analysis. [europepmc]
• Balance diagnostics after propensity score matching. [europepmc]
• COSMOS-E: Guidance on conducting systematic reviews and meta-analyses of observational studies of etiology. [europepmc]
• Association of Patient Priorities-Aligned Decision-Making With Patient Outcomes and Ambulatory Health Care Burden Among Older Adults With Multiple Chronic Conditions: A Nonrandomized Clinical Trial. [europepmc]
• Diabetes medications and risk of Parkinson's disease: a cohort study of patients with diabetes. [europepmc]
• Bidirectional associations between COVID-19 and psychiatric disorder: retrospective cohort studies of 62 354 COVID-19 cases in the USA. [europepmc]
• 6-month neurological and psychiatric outcomes in 236 379 survivors of COVID-19: a retrospective cohort study using electronic health records. [europepmc]
• Interim findings from first-dose mass COVID-19 vaccination roll-out and COVID-19 hospital admissions in Scotland: a national prospective cohort study. [europepmc]
• Risk of clinical sequelae after the acute phase of SARS-CoV-2 infection: retrospective cohort study. [europepmc]
• An introduction to inverse probability of treatment weighting in observational research. [europepmc]
• Long-term cardiovascular outcomes of COVID-19. [europepmc]
• Risks of mental health outcomes in people with covid-19: cohort study. [europepmc]
• Risks and burdens of incident diabetes in long COVID: a cohort study. [europepmc]
• Six-month sequelae of post-vaccination SARS-CoV-2 infection: A retrospective cohort study of 10,024 breakthrough infections. [europepmc]
• Postacute sequelae of COVID-19 at 2 years. [europepmc]

Related

• Invited Commentary: Propensity Scores
• The central role of the propensity score in observational studies for causal effects
• Propensity Score Analysis with Matching Weights
• Moving towards best practice when using inverse probability of treatment weighting (IPTW) using the propensity score to estimate causal treatment effects in observational studies
• Balance diagnostics for comparing the distribution of baseline covariates between treatment groups in propensity‐score matched samples