Publication

Assessing the prior event rate ratio method via probabilistic bias analysis on a Bayesian network

Thommes, E. W., Mahmud, S. M., Young-Xu, Y., Snider, J. T., van Aalst, R., Lee, J. K. H., Halchenko, Y., Russo, E. & Chit, A., 1-Dec-2019, In : Statistics in Medicine. 21 p.

Research output: Contribution to journalArticleAcademicpeer-review

APA

Thommes, E. W., Mahmud, S. M., Young-Xu, Y., Snider, J. T., van Aalst, R., Lee, J. K. H., ... Chit, A. (2019). Assessing the prior event rate ratio method via probabilistic bias analysis on a Bayesian network. Statistics in Medicine. https://doi.org/10.1002/sim.8435

Author

Thommes, Edward W. ; Mahmud, Salaheddin M. ; Young-Xu, Yinong ; Snider, Julia Thornton ; van Aalst, Robertus ; Lee, Jason K. H. ; Halchenko, Yuliya ; Russo, Ellyn ; Chit, Ayman. / Assessing the prior event rate ratio method via probabilistic bias analysis on a Bayesian network. In: Statistics in Medicine. 2019.

Harvard

Thommes, EW, Mahmud, SM, Young-Xu, Y, Snider, JT, van Aalst, R, Lee, JKH, Halchenko, Y, Russo, E & Chit, A 2019, 'Assessing the prior event rate ratio method via probabilistic bias analysis on a Bayesian network', Statistics in Medicine. https://doi.org/10.1002/sim.8435

Standard

Assessing the prior event rate ratio method via probabilistic bias analysis on a Bayesian network. / Thommes, Edward W.; Mahmud, Salaheddin M.; Young-Xu, Yinong; Snider, Julia Thornton; van Aalst, Robertus; Lee, Jason K. H.; Halchenko, Yuliya; Russo, Ellyn; Chit, Ayman.

In: Statistics in Medicine, 01.12.2019.

Research output: Contribution to journalArticleAcademicpeer-review

Vancouver

Thommes EW, Mahmud SM, Young-Xu Y, Snider JT, van Aalst R, Lee JKH et al. Assessing the prior event rate ratio method via probabilistic bias analysis on a Bayesian network. Statistics in Medicine. 2019 Dec 1. https://doi.org/10.1002/sim.8435


BibTeX

@article{9ed0c5fe59bc4529b6cc7d2c0c8ba669,
title = "Assessing the prior event rate ratio method via probabilistic bias analysis on a Bayesian network",
abstract = "Background: Unmeasured confounders are commonplace in observational studies conducted using real-world data. Prior event rate ratio (PERR) adjustment is a technique shown to perform well in addressing such confounding. However, it has been demonstrated that, in some circumstances, the PERR method actually increases rather than decreases bias. In this work, we seek to better understand the robustness of PERR adjustment. Methods: We begin with a Bayesian network representation of a generalized observational study, which is subject to unmeasured confounding. Previous work evaluating PERR performance used Monte Carlo simulation to calculate joint probabilities of interest within the study population. Here, we instead use a Bayesian networks framework. Results: Using this streamlined analytic approach, we are able to conduct probabilistic bias analysis (PBA) using large numbers of combinations of parameters and thus obtain a comprehensive picture of PERR performance. We apply our methodology to a recent study that used the PERR in evaluating elderly-specific high-dose (HD) influenza vaccine in the US Veterans Affairs population. That study obtained an HD relative effectiveness of 25{\%} (95{\%} CI: 2{\%}-43{\%}) against influenza- and pneumonia-associated hospitalization, relative to standard-dose influenza vaccine. In this instance, we find that the PERR-adjusted result is more like to underestimate rather than to overestimate the relative effectiveness of the intervention. Conclusions: Although the PERR is a powerful tool for mitigating the effects of unmeasured confounders, it is not infallible. Here, we develop some general guidance for when a PERR approach is appropriate and when PBA is a safer option.",
keywords = "Bayesian networks, observational studies, probabilistic bias analysis, prior event rate ratio (PERR), unmeasured confounders, DOSE INFLUENZA VACCINE, ADJUSTMENT, HOSPITALIZATIONS, EFFICACY, FRAILTY",
author = "Thommes, {Edward W.} and Mahmud, {Salaheddin M.} and Yinong Young-Xu and Snider, {Julia Thornton} and {van Aalst}, Robertus and Lee, {Jason K. H.} and Yuliya Halchenko and Ellyn Russo and Ayman Chit",
year = "2019",
month = "12",
day = "1",
doi = "10.1002/sim.8435",
language = "English",
journal = "Statistics in Medicine",
issn = "0277-6715",
publisher = "Wiley",

}

RIS

TY - JOUR

T1 - Assessing the prior event rate ratio method via probabilistic bias analysis on a Bayesian network

AU - Thommes, Edward W.

AU - Mahmud, Salaheddin M.

AU - Young-Xu, Yinong

AU - Snider, Julia Thornton

AU - van Aalst, Robertus

AU - Lee, Jason K. H.

AU - Halchenko, Yuliya

AU - Russo, Ellyn

AU - Chit, Ayman

PY - 2019/12/1

Y1 - 2019/12/1

N2 - Background: Unmeasured confounders are commonplace in observational studies conducted using real-world data. Prior event rate ratio (PERR) adjustment is a technique shown to perform well in addressing such confounding. However, it has been demonstrated that, in some circumstances, the PERR method actually increases rather than decreases bias. In this work, we seek to better understand the robustness of PERR adjustment. Methods: We begin with a Bayesian network representation of a generalized observational study, which is subject to unmeasured confounding. Previous work evaluating PERR performance used Monte Carlo simulation to calculate joint probabilities of interest within the study population. Here, we instead use a Bayesian networks framework. Results: Using this streamlined analytic approach, we are able to conduct probabilistic bias analysis (PBA) using large numbers of combinations of parameters and thus obtain a comprehensive picture of PERR performance. We apply our methodology to a recent study that used the PERR in evaluating elderly-specific high-dose (HD) influenza vaccine in the US Veterans Affairs population. That study obtained an HD relative effectiveness of 25% (95% CI: 2%-43%) against influenza- and pneumonia-associated hospitalization, relative to standard-dose influenza vaccine. In this instance, we find that the PERR-adjusted result is more like to underestimate rather than to overestimate the relative effectiveness of the intervention. Conclusions: Although the PERR is a powerful tool for mitigating the effects of unmeasured confounders, it is not infallible. Here, we develop some general guidance for when a PERR approach is appropriate and when PBA is a safer option.

AB - Background: Unmeasured confounders are commonplace in observational studies conducted using real-world data. Prior event rate ratio (PERR) adjustment is a technique shown to perform well in addressing such confounding. However, it has been demonstrated that, in some circumstances, the PERR method actually increases rather than decreases bias. In this work, we seek to better understand the robustness of PERR adjustment. Methods: We begin with a Bayesian network representation of a generalized observational study, which is subject to unmeasured confounding. Previous work evaluating PERR performance used Monte Carlo simulation to calculate joint probabilities of interest within the study population. Here, we instead use a Bayesian networks framework. Results: Using this streamlined analytic approach, we are able to conduct probabilistic bias analysis (PBA) using large numbers of combinations of parameters and thus obtain a comprehensive picture of PERR performance. We apply our methodology to a recent study that used the PERR in evaluating elderly-specific high-dose (HD) influenza vaccine in the US Veterans Affairs population. That study obtained an HD relative effectiveness of 25% (95% CI: 2%-43%) against influenza- and pneumonia-associated hospitalization, relative to standard-dose influenza vaccine. In this instance, we find that the PERR-adjusted result is more like to underestimate rather than to overestimate the relative effectiveness of the intervention. Conclusions: Although the PERR is a powerful tool for mitigating the effects of unmeasured confounders, it is not infallible. Here, we develop some general guidance for when a PERR approach is appropriate and when PBA is a safer option.

KW - Bayesian networks

KW - observational studies

KW - probabilistic bias analysis

KW - prior event rate ratio (PERR)

KW - unmeasured confounders

KW - DOSE INFLUENZA VACCINE

KW - ADJUSTMENT

KW - HOSPITALIZATIONS

KW - EFFICACY

KW - FRAILTY

U2 - 10.1002/sim.8435

DO - 10.1002/sim.8435

M3 - Article

JO - Statistics in Medicine

JF - Statistics in Medicine

SN - 0277-6715

ER -

ID: 109723219