Testing causal effects in observational survival data using propensity score matching design

Propensity score matching is widely used to determine the effects of treatments in observational studies. Competing risk survival data are common to medical research. However, there is a paucity of propensity score matching studies related to competing risk survival data with missing causes of failure. In this study, we provide guidelines for estimating the treatment effect on the cumulative incidence function when using propensity score matching on competing risk survival data with missing causes of failure. We examined the performances of different methods for imputing the data with missing causes. We then evaluated the gain from the missing cause imputation in an extensive simulation study and applied the proposed data imputation method to the data from a study on the risk of hepatocellular carcinoma in patients with chronic hepatitis B and chronic hepatitis C.

Download Full-text

Impact of morphine use on mortality in STEMI patients treated with primary PCI - preliminary data from a new registry

European Journal of Preventive Cardiology ◽

10.1093/eurjpc/zwab061.090 ◽

2021 ◽

Vol 28 (Supplement_1) ◽

Author(s):

D Szabo ◽

A Szabo ◽

IF Edes ◽

D Becker ◽

B Merkely ◽

...

Keyword(s):

Propensity Score ◽

Propensity Score Matching ◽

Survival Data ◽

Preliminary Data ◽

Laboratory Data ◽

Rank Test ◽

Primary Pci ◽

Kaplan Meier ◽

P2y12 Antagonists ◽

Current Guidelines

Abstract Funding Acknowledgements Type of funding sources: None. Background Opioids decrease the effect of P2Y12 receptor inhibitors in vitro and observational reports suggest that morphine use is associated with larger infarct size. Our research group presented previously, using a prospective single-center registry, that periprocedural morphine use may have no impact on long-term mortality in STEMI patients treated with primary PCI and clopidogrel. Purpose Our purpose is to check this interaction using a new registry of patients treated according to the current guidelines, including novel antiplatelet agents. Methods From May until November 2020, we collected 196 STEMI patients treated with primary PCI. 88 (44.9%) of them got morphine during the prehospital and periprocedural care. Baseline demographic, anamnestic, procedural, and laboratory data were collected. Survival data were analysed using Kaplan-Meier survival curves and the log-rank test. To adjust for confounding, a 1:1 propensity score-matching analysis was performed using 114 cases. Results An adequate balance on baseline covariates was achieved during propensity score-matching. Kaplan-Meier analysis showed no difference in 30-days mortality of the patients treated with or without morphine neither in the original nor in the propensity score-matched population (p = 0.094 and p = 0.309, respectively). Conclusion Our preliminary data suggest that morphine may have no impact on mortality in STEMI patients treated with primary PCI and medical therapy according to the current guidelines including novel P2Y12 antagonists. Abstract Figure. Kaplan-Meier curves

Download Full-text

Propensity score matching for treatment delay effects with observational survival data

Statistical Methods in Medical Research ◽

10.1177/0962280219877908 ◽

2019 ◽

Vol 29 (3) ◽

pp. 695-708 ◽

Cited By ~ 1

Author(s):

Erinn M Hade ◽

Giovanni Nattino ◽

Heather A Frey ◽

Bo Lu

Keyword(s):

Propensity Score ◽

Propensity Score Matching ◽

Survival Data ◽

Causal Effect ◽

Treatment Delay ◽

Cox Proportional Hazards Model ◽

Time Varying ◽

Delay Effect ◽

Delay Effects ◽

Time Varying Covariates

In observational studies with a survival outcome, treatment initiation may be time dependent, which is likely to be affected by both time-invariant and time-varying covariates. In situations where the treatment is necessary for the study population, all or most subjects may be exposed to the treatment sooner or later. In this scenario, the causal effect of interest is the delay in treatment reception. A simple comparison of those receiving treatment early vs. those receiving treatment late might not be appropriate, as the timing of the treatment reception is not randomized. Extending Lu’s matching design with time-varying covariates, we propose a propensity score matching strategy to estimate the treatment delay effect. The goal is to balance the covariate distribution between on-time treatment and delayed treatment groups at each time point using risk set matching. Our simulation study shows that, in the presence of treatment delay effects, the matching-based analyses clearly outperform the conventional regression analysis using the naive Cox proportional hazards model. We apply this method to study the treatment delay effect of 17 alpha-hydroxyprogesterone caproate (17P) for patients with recurrent preterm birth.

Download Full-text

Within-center matching performed better when using propensity score matching to analyze multicenter survival data: empirical and Monte Carlo studies

Journal of Clinical Epidemiology ◽

10.1016/j.jclinepi.2013.03.018 ◽

2013 ◽

Vol 66 (9) ◽

pp. 1029-1037 ◽

Cited By ~ 8

Author(s):

Etienne Gayat ◽

Gabriel Thabut ◽

Jason D. Christie ◽

Alexandre Mebazaa ◽

Jean-Yves Mary ◽

...

Keyword(s):

Monte Carlo ◽

Propensity Score ◽

Propensity Score Matching ◽

Survival Data ◽

Monte Carlo Studies

Download Full-text

Propensity Score Matching Analysis for Causal Effects with MNAR Covariates

Statistica Sinica ◽

10.5705/ss.202016.0320 ◽

2018 ◽

Author(s):

Bo Lu ◽

Robert Ashmead

Keyword(s):

Propensity Score ◽

Propensity Score Matching ◽

Causal Effects ◽

Propensity Score Matching Analysis

Download Full-text

Using Propensity Score Matching to Improve Validity in Public Administration Research

10.4018/978-1-7998-8243-5.ch003 ◽

2022 ◽

pp. 45-60

Author(s):

Michael Howell-Moroney

Keyword(s):

Clinical Trials ◽

Propensity Score ◽

Public Administration ◽

Propensity Score Matching ◽

Observational Data ◽

Randomized Clinical Trials ◽

Causal Effect ◽

Social Science Research ◽

Causal Effects ◽

Government Employment

Randomized clinical trials have a longstanding status as the gold standard in detecting causal effects. In the social sciences, randomized clinical trials are rare because of their attendant logistical and cost burdens. Most social science research makes use of observational data. The empirical challenge posed by observational data is that treatment assignment is no longer random. This challenge continues to spur innovation across many disciplines toward more sophisticated techniques for estimating causal relationships. Scholars have developed a common theoretical framework for estimating causal effects, often called the potential outcomes or counterfactual framework. This chapter demonstrates the propensity score matching methodology as a way to estimate causal effects using observational data. Throughout, an example from public administration research, the effect of government employment on volunteerism, is used to illustrate the concepts. Empirical estimates of the treatment effects show that there may be a causal effect of government employment on volunteerism.

Download Full-text

Genetically adjusted propensity score matching: Closing the gap between non-experimental designs and true experiments in the social sciences.

10.31234/osf.io/kpbg6 ◽

2018 ◽

Author(s):

Ian A Silver

Keyword(s):

Social Sciences ◽

Propensity Score ◽

Propensity Score Matching ◽

Genetic Factors ◽

Difference Score ◽

Causal Effects ◽

Risk Scores ◽

Statistical Techniques ◽

Point Estimates ◽

The Social

Objectives: Social scientists employ various statistical techniques to approximate the causal association between two interrelated constructs. Although these methodologies have proven particularly useful for the advancement of knowledge, the limitations associated with preceding statistical techniques limit the ability of scholars to approximate causal associations within some conditions. As such, the current manuscript provides a new statistical technique designed to approximate causal associations.Methods: Genetically adjusted propensity score matching (GAPSM) represents an innovative iteration of propensity score matching (PSM) designed to integrate environmental and genetic factors into the matching process. As proposed, through the implementation of polygenic risk scores, future scholars can estimate genetically adjusted propensity scores (GAPS) through the implementation of two distinct statistical processes. To demonstrate the validity of the GAPSM approach, the current study employs simulation analyses to compare the point estimates derived from a post-GAPSM model, to the point estimates derived from a post-PSM model and a MZ difference score model. Results: The results of the simulation analyses demonstrated that when more robust environmental measures are introduced into the GAPSM approach and a larger amount of the variance in a treatment condition is explained by environmental factors, post-GAPSM models approach the true point estimate more closely than the point estimates derived from post-PSM models and MZ difference score models. Conclusions: Overall, the findings demonstrate that the GAPSM approach can prove useful when assessing the causal effects of treatment conditions on subsequent phenotypes by adjusting for observed environmental and genetic factors. Within the social sciences, this method could provide substantive advancements in our understanding of causal effects.

Download Full-text

Transarterial embolisation is associated with improved survival in patients with pelvic fracture: propensity score matching analyses

European Journal of Trauma and Emergency Surgery ◽

10.1007/s00068-020-01497-9 ◽

2020 ◽

Author(s):

Hohyun Kim ◽

Chang Ho Jeon ◽

Jae Hun Kim ◽

Hyun-Woo Sun ◽

Dongyeon Ryu ◽

...

Keyword(s):

Propensity Score ◽

Propensity Score Matching ◽

Pelvic Fracture ◽

Survival Data ◽

Trauma Patients ◽

Multivariate Logistic Regression ◽

Severely Injured Patients ◽

Transarterial Embolisation ◽

Arterial Haemorrhage ◽

Injured Patients

Abstract Introduction Transarterial embolisation (TAE) is an effective intervention for management of arterial haemorrhage associated with pelvic fracture. However, its effects on survival and clinical outcomes are unclear. Methods Trauma patients with survival data between November 2015 and December 2019 were identified using a trauma database. Patients were divided between TAE and non-TAE groups, and a propensity score was developed using multivariate logistic regression. Survival at 28 days was compared between the groups after propensity score matching. Results Among 881 patients included in this study, 308 (35.0%) were treated with TAE. After propensity score matching, 130 pairs were selected. Survival at 28 days was significantly higher among patients treated with TAE than among those treated without TAE [122 (93.9%) vs. 112 (86.2%); odds ratio = 2.45; 95% CI 1.02–5.86; p = 0.039]. Conclusions TAE use was associated with improved survival at 28 days in patients with pelvic fracture and should therefore be considered in the management of severely injured patients with pelvic fracture.

Download Full-text

Is marijuana really a gateway drug? A nationally representative test of the marijuana gateway hypothesis using a propensity score matching design

Journal of Experimental Criminology ◽

10.1007/s11292-021-09464-z ◽

2021 ◽

Author(s):

Cody Jorgensen ◽

Jessica Wells

Keyword(s):

Propensity Score ◽

Propensity Score Matching ◽

Gateway Hypothesis ◽

Matching Design ◽

Nationally Representative

Download Full-text

Besondere Förderung von Kern- kompetenzen an Spezialgymnasien?

Zeitschrift für Pädagogische Psychologie ◽

10.1024/1010-0652/a000001 ◽

2010 ◽

Vol 24 (1) ◽

pp. 5-22 ◽

Cited By ~ 7

Author(s):

Jürgen Baumert ◽

Michael Becker ◽

Marko Neumann ◽

Roumiana Nikolova

Keyword(s):

Propensity Score ◽

Propensity Score Matching

Der vorliegende Beitrag geht der Frage nach, ob Schülerinnen und Schüler, die nach der vierten Klasse in Berlin in ein grundständiges Gymnasium wechseln, in Abhängigkeit vom Profil des besuchten Gymnasiums im Vergleich zu Grundschülern mit vergleichbaren Lernvoraussetzungen unterschiedliche Lernzuwächse im Leseverständnis, in Mathematik und Englisch erreichen. Auf der Datengrundlage der ELEMENT-Studie wurde die Leistungsentwicklung von Schülerinnen und Schülern an grundständigen Gymnasien (N = 1758) und Grundschulen (N = 3169) während der 5. und 6. Jahrgangsstufe mithilfe von Propensity Score Matching-Analysen (PSM) modelliert. Nach Kontrolle von leistungsrelevanten Unterschieden zwischen den Schülergruppen am Ende der 4. Jahrgangsstufe zeigten sich für das Leseverständnis am Ende der 6. Klasse keine statistisch signifikanten Unterschiede. Für die Mathematikleistung ließen sich Unterschiede lediglich zugunsten eines grundständigen Gymnasiums, das zum Untersuchungszeitpunkt noch kein spezifisches Profil entwickelt hatte, nachweisen. In der Domäne Englisch, in der die curricularen Unterschiede zwischen den Schulzweigen stärker akzentuiert sind, wurden positive Ergebnisse im Vergleich zu den Grundschulen für die so genannten Schnellläuferzüge, die englisch-bilingualen Klassen und das grundständige Gymnasium ohne spezifisches Profil ermittelt. Die Lernstände am Ende der 6. Klasse in den altsprachlichen Gymnasien fielen dagegen im Vergleich zu den Grundschulen geringer aus. Die Befunde widersprechen der Annahme, dass mit dem frühen Übergang auf ein grundständiges Gymnasium automatisch eine besondere Förderung der Lesefähigkeit und des mathematischen Verständnisses besonders leistungsfähiger Schülerinnen und Schüler erreicht werde. Die Ergebnisse zu den Englischleistungen weisen hingegen darauf hin, dass Unterschiede in der Leistungsentwicklung auftreten können, sofern die Aufteilung auf Schulen mit unterschiedlichen Bildungsprogrammen mit curricularen Unterschieden im Unterricht einhergeht. Methodische und inhaltliche Implikationen der Befunde und Grenzen ihrer Generalisierbarkeit werden diskutiert.

Download Full-text