How Can Causal Relationships Be Measured in Observational Studies? Propensity Score Matching: A Tutorial Article
0
Citation
0
Reference
10
Related Paper
Abstract:
How Can Causal Relationships Be Measured in Observational Studies? Propensity Score Matching: A Tutorial ArticleIn using observational, nonrandomized data, there is often interest in studying the effect of a particular treatment on a specific outcome. However, the imbalance of potential confounding variables between the treatment groups can distort the relationship between treatment and outcome. Propensity score matching is one, increasingly utilized, method to help account for such imbalances, allowing for a more accurate estimation of the influence of treatment on outcome. In this paper, we provide the clinician with an overview of propensity score matching techniques and provide a practical example of how this has been used in clinical research relevant to spine surgery.
Cite
Citations (24)
Cite
Citations (0)
To stay competitive in the marketplace, healthcare programs must be capable of reporting the true savings to clients. This is a tall order considering most healthcare programs are setup to be available to the client’s entire population; thus, the program cannot be conducted as a randomized control trial. In order to evaluate the performance of the program for the client, we use an observational study design which has inherent selection bias due to its inability to randomly assign participants. To reduce the impact of bias, we apply propensity score matching to the analysis. This technique is beneficial to healthcare program evaluations because it helps reduce selection bias in the observational analysis and in turn provides a clearer view of the client’s savings. This paper will explore how to develop a propensity score, evaluate the use of inverse propensity weighting versus propensity matching, and determine the overall impact of the propensity score matching method on the observational study population. All results shown are drawn from a savings analysis using a participant (cases) versus non-participant (controls) observational study design for a healthcare decision support program aiming to reduce emergency room visits.
Cite
Citations (0)
Average treatment effect
Cite
Citations (6)
Observational studies are often used to investigate the effects of treatments on a specific
outcome. In many observational studies, the event of interest can be of recurrent
type, which means that subjects may experience the event of interest more than one
time during their follow-up. The lack of random allocation of treatments to subjects in
observational studies may induce the selection bias leading to systematic differences
in observed and unobserved baseline characteristics between treated and untreated
subjects. Propensity score matching is a popular technique to address this issue. It
is based on the estimation of conditional probability of treatment assignment given
the measured baseline characteristics. The use of the propensity score in the analysis
of observational studies with recurrent event outcomes has not been well developed.
In this study, we consider three matching methods called propensity score matching,
covariate matching and history matching, and compare the accuracy of them
to estimate the treatment effects in recurrent event rates through Monte Carlo simulation
studies. We consider various scenarios under the settings of time-fixed and
time-dependent treatment indicators. A synthetic data set is analyzed to illustrate
the methods discussed in the thesis.
Baseline (sea)
Cite
Citations (0)
In observational studies subjects are not randomly assigned to treatment and control groups. Therefore, both groups can be significantly different from each other and their differences after applying a treatment cannot be attributed to the treatment effect. Propensity scores methods offer a way to balance groups by matching treatment and control units based on a set of covariates. This paper explains how to use SAS® to match samples employing the most commonly used matching methods, such as nearest available neighbor, calipers and radius with- and without replacement. This paper also serves as an introduction to propensity score matching methods and illustrates both the relevance and the inherent problems of any method that attempts to select a valid comparison group.
Average treatment effect
Treatment effect
Cite
Citations (30)
Cardinality (data modeling)
Cite
Citations (0)
Cite
Citations (3)
Inferring causal effects from observational studies is desirable when the randomized experiments are not feasible or are unethical. Due to the lack of randomization in observational studies, treatment selection depends on various covariates that are associated with outcomes of interest. In multiple-time-point observational studies, in particular, temporal features may potentially affect subjects differently, resulting in time-varying confounders. Using conventional methods for cross-sectional observational studies is inappropriate, since they fail to compensate for the biases arising from the time component. This dissertation focuses on two types of multiple-timepoint study designs, and proposes a propensity score matching method for valid causal inference. Chapter 2 deals with the repeated cross-sectional observational study, where the intervention of primary interest is conducted repeatedly on different cohorts, and a second intervention occurs between two time points. However, the covariates may be imbalanced for subjects in different intervention groups and for subjects participating at different time points. We extend from Rubin’s causal model, and establish a potential outcomes framework for this study design. The assumptions for identification of various causal effects are discussed. We propose two matching algorithms for multigroup matching, and compare them via simulation studies. We further provide two
Cite
Citations (2)
In this article, we presented the rationale and calculation procedures of the propcnsity score matching (PSM), and its application in the designing stage of an cpidcrniological study. Based on existing observational data, PSM can be used to select one or more comparable controls for each subject in 'treatment' group according to the propensity scores estimated by 'treatment' variable and main covariates. The results of an example analysis showed that the bias for main confounders between the treated and control samples declined more than 55% after PMS. Conclusion: PSM can reduce most of the confounding bias of the observational study, and can obtain approximate study effect to the randomized controlled trials when used in the designing of thc cpidcmiological study.
Key words:
Propensity score matching; Confounding bias; Epidemiology
Cite
Citations (2)