PSweight: An R Package for Propensity Score Weighting Analysis

Zhou, Tianhui; Tong, Guangyu; Thomas, Laine; Li, Fan

doi:10.32614/rj-2022-011

Cited by 40 publications

(48 citation statements)

References 31 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Analyses were performed using R statistical software version 3.5.1 (R Core Team) [40]. The following packages were combined in custom functions to provide a reproducible and configurable pipeline: MICE [41], glmnet [42], MatchIt [43], and PSWeight [44]. The code is available online for transparency [45].…”

Section: Methodsmentioning

confidence: 99%

Mining Electronic Health Records for Drugs Associated With 28-day Mortality in COVID-19: Pharmacopoeia-wide Association Study (PharmWAS) (Preprint)

Lerner¹,

Serret-Larmande²,

Rance³

et al. 2021

Preprint

View full text Add to dashboard Cite

BACKGROUND Patients hospitalized for a given condition may be receiving other treatments for other contemporary conditions or comorbidities. The use of such observational clinical data for pharmacological hypothesis generation is appealing in the context of an emerging disease but particularly challenging due to the presence of drug indication bias. OBJECTIVE With this study, our main objective was the development and validation of a fully data-driven pipeline that would address this challenge. Our secondary objective was to generate pharmacological hypotheses in patients with COVID-19 and demonstrate the clinical relevance of the pipeline. METHODS We developed a pharmacopeia-wide association study (PharmWAS) pipeline inspired from the PheWAS methodology, which systematically screens for associations between the whole pharmacopeia and a clinical phenotype. First, a fully data-driven procedure based on adaptive least absolute shrinkage and selection operator (LASSO) determined drug-specific adjustment sets. Second, we computed several measures of association, including robust methods based on propensity scores (PSs) to control indication bias. Finally, we applied the Benjamini and Hochberg procedure of the false discovery rate (FDR). We applied this method in a multicenter retrospective cohort study using electronic medical records from 16 university hospitals of the Greater Paris area. We included all adult patients between 18 and 95 years old hospitalized in conventional wards for COVID-19 between February 1, 2020, and June 15, 2021. We investigated the association between drug prescription within 48 hours from admission and 28-day mortality. We validated our data-driven pipeline against a knowledge-based pipeline on 3 treatments of reference, for which experts agreed on the expected association with mortality. We then demonstrated its clinical relevance by screening all drugs prescribed in more than 100 patients to generate pharmacological hypotheses. RESULTS A total of 5783 patients were included in the analysis. The median age at admission was 69.2 (IQR 56.7-81.1) years, and 3390 (58.62%) of the patients were male. The performance of our automated pipeline was comparable or better for controlling bias than the knowledge-based adjustment set for 3 reference drugs: dexamethasone, phloroglucinol, and paracetamol. After correction for multiple testing, 4 drugs were associated with increased in-hospital mortality. Among these, diazepam and tramadol were the only ones not discarded by automated diagnostics, with adjusted odds ratios of 2.51 (95% CI 1.52-4.16, <i>Q</i>=.1) and 1.94 (95% CI 1.32-2.85, <i>Q</i>=.02), respectively. CONCLUSIONS Our innovative approach proved useful in generating pharmacological hypotheses in an outbreak setting, without requiring a priori knowledge of the disease. Our systematic analysis of early prescribed treatments from patients hospitalized for COVID-19 showed that diazepam and tramadol are associated with increased 28-day mortality. Whether these drugs could worsen COVID-19 needs to be further assessed.

show abstract

Section: Methodsmentioning

confidence: 99%

Mining Electronic Health Records for Drugs Associated With 28-day Mortality in COVID-19: Pharmacopoeia-wide Association Study (PharmWAS) (Preprint)

Lerner¹,

Serret-Larmande²,

Rance³

et al. 2021

Preprint

View full text Add to dashboard Cite

show abstract

“…The asymptotic variance estimators used in the current study account for the uncertainty in estimating the propensity score model 14 . Had the asymptotic variance estimator not accounted for uncertainty in the estimate propensity score, one would anticipate that the estimated SEs would be too small, as all of the sources of variation had not been accounted for.…”

Section: Discussionmentioning

confidence: 99%

“…Within the PSweight function, we estimated the propensity score using logistic regression in which treatment status was regressed on the measured baseline covariates. In PSweight, the variance of the estimated treatment effect is obtained using the empirical sandwich variance for propensity score weighting estimators based on M‐estimation theory 14 . This variance estimator accounts for the uncertainty in estimating the propensity score.…”

Section: Monte Carlo Simulations–methodsmentioning

confidence: 99%

See 1 more Smart Citation

Bootstrap vs asymptotic variance estimation when using propensity score weighting with continuous and binary outcomes

Austin

2022

Statistics in Medicine

View full text Add to dashboard Cite

We used Monte Carlo simulations to compare the performance of asymptotic variance estimators to that of the bootstrap when estimating standard errors of differences in means, risk differences, and relative risks using propensity score weighting. We considered four different sets of weights: conventional inverse probability of treatment weights with the average treatment effect (ATE) as the target estimand, weights for estimating the average treatment effect in the treated (ATT), matching weights, and overlap weights. We considered sample sizes ranging from 250 to 10 000 and allowed the prevalence of treatment to range from 0.1 to 0.9. We found that, when using ATE weights and sample sizes were ≤ 1000, then the use of the bootstrap resulted in estimates of SE that were more accurate than the asymptotic estimates. A similar finding was observed when using ATT weights and sample sizes were ≤ 1000 and the prevalence of treatment was moderate to high. When using matching weights and overlap weights, both the asymptotic estimator and the bootstrap resulted in accurate estimates of SE across all sample sizes and prevalences of treatment. Even when using the bootstrap with ATE weights, empirical coverage rates of confidence intervals were suboptimal when sample sizes were low to moderate and the prevalence of treatment was either very low or very high. A similar finding was observed when using the bootstrap with ATT weights when sample sizes were low to moderate and the prevalence of treatment was very high. K E Y W O R D Sbootstrap, inverse probability of treatment weighting, propensity score, variance estimation INTRODUCTIONPropensity score weighting is a popular causal inference method for estimating the effects of treatments, exposures, and interventions when using observational data. The propensity score is the probability of a subject being treated (vs receiving the control) based on the subject's observed baseline covariates. 1 Propensity score weighting entails constructing weights that are derived from the propensity score. The first set of propensity-score based weights, referred to as inverse probability of treatment weights (IPTW), weight each subject by the inverse of the probability of receiving the treatment that subjectThis is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.

show abstract

“…Because the application of TTM can be affected by the patient’s clinical condition and because the clinical parameters of the two groups were expected to be considerably different, the overlap propensity score weighting method was applied. 17 , 18 A propensity score for receiving TTM was calculated from a multivariable logistic regression analysis incorporating the following variables: age, sex, baseline CPC, diabetes, hypertension, witnessed arrest, bystander cardiopulmonary resuscitation, initial shockable rhythm, prehospital ROSC, no flow time, low flow time, presence of ST-elevation myocardial infarction, emergency percutaneous coronary intervention, Sequential Organ Failure Assessment score, Acute Physiology And Chronic Health Evaluation II score, and the cause of cardiac arrest. Multiple imputations by chained equations were applied for missing values.…”

Section: Methodsmentioning

confidence: 99%

Target temperature management versus normothermia without temperature feedback systems for out-of-hospital cardiac arrest survivors

et al. 2022

View full text Add to dashboard Cite

Objective The clinical benefit of automatic temperature control devices remains unclear. We investigated the outcomes of out-of-hospital cardiac arrest (OHCA) survivors who had undergone either target temperature management (TTM) with a temperature feedback system (TFS) or maintenance of normothermia without a TFS during post-resuscitation care. Methods This study was a retrospective analysis of a multicenter prospective cohort of OHCA survivors who had received postcardiac arrest care from August 2014 to December 2018. The overlap propensity score weighting method was applied for adjustment between groups. Results A total of 405 OHCA survivors were included. TTM with a TFS and normothermia without a TFS were applied to 318 and 87 patients, respectively. Fever events were more common in patients with normothermia without a TFS. After propensity score matching, no statistically significant differences were observed in the 1-month good neurologic outcome (odds ratio 0.99, 95% confidence interval [CI] 0.56–1.25) or survival rate (odds ratio 1.25, 95% CI 0.88–1.78). Conclusion No significant differences in the 1-month neurologic outcome were observed between patients receiving TTM with a TFS and those undergoing normothermia without a TFS.

show abstract

PSweight: An R Package for Propensity Score Weighting Analysis

Cited by 40 publications

References 31 publications

Mining Electronic Health Records for Drugs Associated With 28-day Mortality in COVID-19: Pharmacopoeia-wide Association Study (PharmWAS) (Preprint)

Mining Electronic Health Records for Drugs Associated With 28-day Mortality in COVID-19: Pharmacopoeia-wide Association Study (PharmWAS) (Preprint)

Bootstrap vs asymptotic variance estimation when using propensity score weighting with continuous and binary outcomes

Target temperature management versus normothermia without temperature feedback systems for out-of-hospital cardiac arrest survivors

Contact Info

Product

Resources

About