Propensity score methods and their application in nephrology research
Lianne Barnieh, Matthew T. James, Jianguo Zhang, Brenda R. Hemmelgarn
Abstract
Lianne Barnieh, Matthew T. James, Jianguo Zhang, Brenda R. Hemmelgarn
Abstract
Propensity score methods are used to control for treatment-selection bias in observational studies. A propensity score reduces a collection of covariates into a single composite score. This score is the probability, or propensity, of receiving a specific treatment conditional on the observed covariates. A propensity score can be applied by either matching subjects on the score, stratification by the propensity score or including the propensity score as a predictor in a multivariable model. This paper focuses on propensity score-matched methods. There are 4 steps in a propensity score-matched analysis. The propensity score is derived, the propensity score-matched sample is constructed, the degree to which matching has balanced observed covariates is assessed and the effect of the treatment on the outcome is estimated. Propensity score methods are often used in observational studies in nephrology, thus understanding their appropriate implementation, strengths and limitations is important.
OpenAlex reports 19 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
Propensity score methods are used to control for treatment-selection bias in observational studies. A propensity score reduces a collection of covariates into a single composite score. This score is the probability, or propensity, of receiving a specific treatment conditional on the observed covariates. A propensity score can be applied by either matching subjects on the score, stratification by the propensity score or including the propensity score as a predictor in a multivariable model. This paper focuses on propensity score-matched methods. There are 4 steps in a propensity score-matched analysis. The propensity score is derived, the propensity score-matched sample is constructed, the degree to which matching has balanced observed covariates is assessed and the effect of the treatment on the outcome is estimated. Propensity score methods are often used in observational studies in nephrology, thus understanding their appropriate implementation, strengths and limitations is important.
Key concepts: Propensity score matching, Medicine, Covariate, Observational study, Internal medicine, Selection bias, Matching (statistics), Statistics