Statistical profiling methods with hierarchical logistic regression for healthcare providers with binary outcomes

Xiaowei Yang, Bin Peng, Rongqi Chen, Qian Zhang, Dianwen Zhu, Qing J. Zhang, Fuzhong Xue, Lihong Qi

Research output: Contribution to journalArticle

5 Citations (Scopus)

Abstract

Within the context of California's public report of coronary artery bypass graft (CABG) surgery outcomes, we first thoroughly review popular statistical methods for profiling healthcare providers. Extensive simulation studies are then conducted to compare profiling schemes based on hierarchical logistic regression (LR) modeling under various conditions. Both Bayesian and frequentist's methods are evaluated in classifying hospitals into 'better', 'normal' or 'worse' service providers. The simulation results suggest that no single method would dominate others on all accounts. Traditional schemes based on LR tend to identify too many false outliers, while those based on hierarchical modeling are relatively conservative. The issue of over shrinkage in hierarchical modeling is also investigated using the 2005-2006 California CABG data set. The article provides theoretical and empirical evidence in choosing the right methodology for provider profiling.

Original languageEnglish (US)
Pages (from-to)46-59
Number of pages14
JournalJournal of Applied Statistics
Volume41
Issue number1
DOIs
StatePublished - Jan 2014

Fingerprint

Binary Outcomes
Logistic Regression
Profiling
Healthcare
Hierarchical Modeling
Coronary Artery
Shrinkage
Statistical method
Surgery
Outlier
Simulation Study
Tend
Methodology
Modeling
Logistic regression
Simulation

Keywords

  • Bayesian mixed-effects models
  • hierarchical logistic regression models
  • provider profiling
  • quality of care
  • risk-adjusted mortality

ASJC Scopus subject areas

  • Statistics and Probability
  • Statistics, Probability and Uncertainty

Cite this

Statistical profiling methods with hierarchical logistic regression for healthcare providers with binary outcomes. / Yang, Xiaowei; Peng, Bin; Chen, Rongqi; Zhang, Qian; Zhu, Dianwen; Zhang, Qing J.; Xue, Fuzhong; Qi, Lihong.

In: Journal of Applied Statistics, Vol. 41, No. 1, 01.2014, p. 46-59.

Research output: Contribution to journalArticle

Yang, Xiaowei ; Peng, Bin ; Chen, Rongqi ; Zhang, Qian ; Zhu, Dianwen ; Zhang, Qing J. ; Xue, Fuzhong ; Qi, Lihong. / Statistical profiling methods with hierarchical logistic regression for healthcare providers with binary outcomes. In: Journal of Applied Statistics. 2014 ; Vol. 41, No. 1. pp. 46-59.
@article{7594f5c771bc493285eca861b7645b5d,
title = "Statistical profiling methods with hierarchical logistic regression for healthcare providers with binary outcomes",
abstract = "Within the context of California's public report of coronary artery bypass graft (CABG) surgery outcomes, we first thoroughly review popular statistical methods for profiling healthcare providers. Extensive simulation studies are then conducted to compare profiling schemes based on hierarchical logistic regression (LR) modeling under various conditions. Both Bayesian and frequentist's methods are evaluated in classifying hospitals into 'better', 'normal' or 'worse' service providers. The simulation results suggest that no single method would dominate others on all accounts. Traditional schemes based on LR tend to identify too many false outliers, while those based on hierarchical modeling are relatively conservative. The issue of over shrinkage in hierarchical modeling is also investigated using the 2005-2006 California CABG data set. The article provides theoretical and empirical evidence in choosing the right methodology for provider profiling.",
keywords = "Bayesian mixed-effects models, hierarchical logistic regression models, provider profiling, quality of care, risk-adjusted mortality",
author = "Xiaowei Yang and Bin Peng and Rongqi Chen and Qian Zhang and Dianwen Zhu and Zhang, {Qing J.} and Fuzhong Xue and Lihong Qi",
year = "2014",
month = "1",
doi = "10.1080/02664763.2013.830086",
language = "English (US)",
volume = "41",
pages = "46--59",
journal = "Journal of Applied Statistics",
issn = "0266-4763",
publisher = "Routledge",
number = "1",

}

TY - JOUR

T1 - Statistical profiling methods with hierarchical logistic regression for healthcare providers with binary outcomes

AU - Yang, Xiaowei

AU - Peng, Bin

AU - Chen, Rongqi

AU - Zhang, Qian

AU - Zhu, Dianwen

AU - Zhang, Qing J.

AU - Xue, Fuzhong

AU - Qi, Lihong

PY - 2014/1

Y1 - 2014/1

N2 - Within the context of California's public report of coronary artery bypass graft (CABG) surgery outcomes, we first thoroughly review popular statistical methods for profiling healthcare providers. Extensive simulation studies are then conducted to compare profiling schemes based on hierarchical logistic regression (LR) modeling under various conditions. Both Bayesian and frequentist's methods are evaluated in classifying hospitals into 'better', 'normal' or 'worse' service providers. The simulation results suggest that no single method would dominate others on all accounts. Traditional schemes based on LR tend to identify too many false outliers, while those based on hierarchical modeling are relatively conservative. The issue of over shrinkage in hierarchical modeling is also investigated using the 2005-2006 California CABG data set. The article provides theoretical and empirical evidence in choosing the right methodology for provider profiling.

AB - Within the context of California's public report of coronary artery bypass graft (CABG) surgery outcomes, we first thoroughly review popular statistical methods for profiling healthcare providers. Extensive simulation studies are then conducted to compare profiling schemes based on hierarchical logistic regression (LR) modeling under various conditions. Both Bayesian and frequentist's methods are evaluated in classifying hospitals into 'better', 'normal' or 'worse' service providers. The simulation results suggest that no single method would dominate others on all accounts. Traditional schemes based on LR tend to identify too many false outliers, while those based on hierarchical modeling are relatively conservative. The issue of over shrinkage in hierarchical modeling is also investigated using the 2005-2006 California CABG data set. The article provides theoretical and empirical evidence in choosing the right methodology for provider profiling.

KW - Bayesian mixed-effects models

KW - hierarchical logistic regression models

KW - provider profiling

KW - quality of care

KW - risk-adjusted mortality

UR - http://www.scopus.com/inward/record.url?scp=84889097831&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84889097831&partnerID=8YFLogxK

U2 - 10.1080/02664763.2013.830086

DO - 10.1080/02664763.2013.830086

M3 - Article

AN - SCOPUS:84889097831

VL - 41

SP - 46

EP - 59

JO - Journal of Applied Statistics

JF - Journal of Applied Statistics

SN - 0266-4763

IS - 1

ER -