Have a personal or library account? Click to login
A Bayesian Approach to Modeling Risk of Hospital Admissions Associated With Schizophrenia Accounting for Underdiagnosis of the Disorder in Administrative Records Cover

A Bayesian Approach to Modeling Risk of Hospital Admissions Associated With Schizophrenia Accounting for Underdiagnosis of the Disorder in Administrative Records

Open Access
|Feb 2018

Figures & Tables

Table 1.

Patient characteristics, overall and by SMI status (N = 87,806).

VariableTotal (N = 87,806)Schizophrenia (N = 114)Bipolar (N = 412)PTSD (N = 82)MDD (N = 81,405)Non-SMI (N = 5,793)Test statisticp Valuea
Age
Mean (SD):52.7 (18.9)52.7 (19.8)47.3 (17.4)47.0 (16.6)51.7 (18.1)52.8 (19.9)61.8 <0.01
Median (min–max):53 (18–103)55 (18–89)48 (18–89)48 (18–87)51 (18–98)53 (18–103)
Female (%)50,359 (57.4)63 (55.3)256 (62.1)47 (57.3)4,300 (74.2)45,693 (56.1)728.2 <0.01
Race (%)155.7 <0.01
 White62,394 (17.1)80 (70.2)319 (77.4)57 (69.5)4,493 (77.6)57,445 (70.6)136.9 <0.01
 Black5,816 (6.6)17 (14.9)13 (3.2)7 (8.5)285 (4.9)5,494 (6.7)50.4 <0.01
 Other19,596 (22.3)17 (14.9)80 (19.4)18 (22.0)1,015 (17.5)18,466 (22.7)88.8 <0.01
Hispanic (%)4,763 (5.4)2 (1.8)15 (3.6)5 (6.1)277 (4.8)4,464 (5.5)10.80.03
Selim
Mean (SD):1.7 (1.8)1.7 (1.8)1.7 (1.8)1.9 (2.1)2.0 (2.1)1.7 (1.8)122.3 <0.01
Median (min–max):1 (0–15)1 (0–9)1 (0–8)1 (0–14)1 (0–15)1 (0–15)
Admission (%)7,269 (8.3)16 (14.0)53 (12.9)10 (12.2)668 (11.5)6,522 (8.0)106.4 <0.01

[i] aPatient characteristics were compared across SMI groups by employing Chi-square analyses for categorical variables and the Kruskal-Wallis test for continuous variables, assuming a Type I error of α = 0.05.

Table 2.

Bayesian logistic regression model results for varying sensitivities, along with the classical logistic regression model approach

PredictorOR95% CIOR95% CIOR95% CIOR95% CI
Schizophrenia1.87[1.08, 3.23]1.83[1.02, 3.12]1.82[1.02, 3.09]1.82[1.02, 3.10]
Bipolar disorder1.73[1.28, 2.32]1.71[1.26, 2.29]1.71[1.26, 2.29]1.71[1.26, 2.29]
PTSD1.52[0.76, 3.05]1.46[0.69, 2.85]1.46[0.68, 2.86]1.46[0.69, 2.84]
MDD1.28[1.17, 1.40]1.28[1.17, 1.40]1.28[1.17, 1.40]1.28[1.17, 1.40]
Age (decade effect)1.03[1.01, 1.05]1.03[1.01, 1.05]1.03[1.01, 1.05]1.03[1.01, 1.05]
Female1.31[1.25, 1.38]1.31[1.25, 1.38]1.31[1.25, 1.38]1.31[1.25, 1.38]
Black race1.07[0.97, 1.18]1.07[0.97, 1.18]1.07[0.97, 1.18]1.07[0.97, 1.18]
Other race0.83[0.77, 0.89]0.83[0.77, 0.89]0.83[0.77, 0.89]0.83[0.77, 0.89]
Hispanic1.09[0.98, 1.22]1.09[0.97, 1.22]1.09[0.97, 1.22]1.09[0.97, 1.22]
Selim Physical1.35[1.34, 1.37]1.35[1.34, 1.37]1.35[1.34, 1.37]1.35[1.34, 1.37]

[i] Note. N = 87,806. CI = confidence interval (credible for Bayesian).

Language: English
Submitted on: Aug 8, 2016
Accepted on: Jun 26, 2017
Published on: Feb 1, 2018
Published by: MIT Press
In partnership with: Paradigm Publishing Services
Publication frequency: 1 issue per year

© 2018 Eileen M. Stock, James D. Stamey, John E. Zeber, Alexander W. Thompson, Laurel A. Copeland, published by MIT Press
This work is licensed under the Creative Commons Attribution 4.0 License.