Comparison of logistic regression and neural network analysis applied to predicting living setting after hip fracture

Kenneth Ottenbacher, Richard T. Linn, Pamela M. Smith, Sandra B. Illig, Melodee Mancuso, Carl V. Granger

Research output: Contribution to journalArticle

35 Citations (Scopus)

Abstract

Purpose Describe and compare the characteristics of artificial neural networks and logistic regression to develop prediction models in epidemiological research. Methods The sample included 3708 persons with hip fracture from 46 different states included in the Uniform Data System for Medical Rehabilitation. Mean age was 75.5 years (sd=14.2), 73.7% of patients were female, and 82% were non-Hispanic white. Average length of stay was 17.0 days (sd=10.6). The primary outcome measure was living setting (at home vs. not at home) at 80 to 180 days after discharge. Results Statistically significant variables (p < .05) in the logistic model included follow-up therapy, sphincter control, self-care ability, marital status, age, and length of stay. Areas under the receiver operating characteristic curves were 0.67 for logistic regression and 0.73 for neural network analysis. Calibration curves indicated a slightly better fit for the neural network model. Conclusions Follow-up therapy and independent bowel and/or bladder function were strong predictors of living at home up to 6 months after hospitalization for hip fracture. No practical differences were found between the predictive ability of logistic regression and neural network analysis in this sample.

Original languageEnglish (US)
Pages (from-to)551-559
Number of pages9
JournalAnnals of Epidemiology
Volume14
Issue number8
DOIs
StatePublished - Sep 2004

Fingerprint

Hip Fractures
Logistic Models
Aptitude
Length of Stay
Neural Networks (Computer)
Marital Status
Self Care
Information Systems
ROC Curve
Calibration
Urinary Bladder
Hospitalization
Rehabilitation
Outcome Assessment (Health Care)
Therapeutics
Research

Keywords

  • Joint Replacement
  • Rehabilitation Outcomes

ASJC Scopus subject areas

  • Medicine(all)
  • Public Health, Environmental and Occupational Health
  • Epidemiology

Cite this

Comparison of logistic regression and neural network analysis applied to predicting living setting after hip fracture. / Ottenbacher, Kenneth; Linn, Richard T.; Smith, Pamela M.; Illig, Sandra B.; Mancuso, Melodee; Granger, Carl V.

In: Annals of Epidemiology, Vol. 14, No. 8, 09.2004, p. 551-559.

Research output: Contribution to journalArticle

Ottenbacher, Kenneth ; Linn, Richard T. ; Smith, Pamela M. ; Illig, Sandra B. ; Mancuso, Melodee ; Granger, Carl V. / Comparison of logistic regression and neural network analysis applied to predicting living setting after hip fracture. In: Annals of Epidemiology. 2004 ; Vol. 14, No. 8. pp. 551-559.
@article{82e4b15900424a1aa07672ab626f4f16,
title = "Comparison of logistic regression and neural network analysis applied to predicting living setting after hip fracture",
abstract = "Purpose Describe and compare the characteristics of artificial neural networks and logistic regression to develop prediction models in epidemiological research. Methods The sample included 3708 persons with hip fracture from 46 different states included in the Uniform Data System for Medical Rehabilitation. Mean age was 75.5 years (sd=14.2), 73.7{\%} of patients were female, and 82{\%} were non-Hispanic white. Average length of stay was 17.0 days (sd=10.6). The primary outcome measure was living setting (at home vs. not at home) at 80 to 180 days after discharge. Results Statistically significant variables (p < .05) in the logistic model included follow-up therapy, sphincter control, self-care ability, marital status, age, and length of stay. Areas under the receiver operating characteristic curves were 0.67 for logistic regression and 0.73 for neural network analysis. Calibration curves indicated a slightly better fit for the neural network model. Conclusions Follow-up therapy and independent bowel and/or bladder function were strong predictors of living at home up to 6 months after hospitalization for hip fracture. No practical differences were found between the predictive ability of logistic regression and neural network analysis in this sample.",
keywords = "Joint Replacement, Rehabilitation Outcomes",
author = "Kenneth Ottenbacher and Linn, {Richard T.} and Smith, {Pamela M.} and Illig, {Sandra B.} and Melodee Mancuso and Granger, {Carl V.}",
year = "2004",
month = "9",
doi = "10.1016/j.annepidem.2003.10.005",
language = "English (US)",
volume = "14",
pages = "551--559",
journal = "Annals of Epidemiology",
issn = "1047-2797",
publisher = "Elsevier Inc.",
number = "8",

}

TY - JOUR

T1 - Comparison of logistic regression and neural network analysis applied to predicting living setting after hip fracture

AU - Ottenbacher, Kenneth

AU - Linn, Richard T.

AU - Smith, Pamela M.

AU - Illig, Sandra B.

AU - Mancuso, Melodee

AU - Granger, Carl V.

PY - 2004/9

Y1 - 2004/9

N2 - Purpose Describe and compare the characteristics of artificial neural networks and logistic regression to develop prediction models in epidemiological research. Methods The sample included 3708 persons with hip fracture from 46 different states included in the Uniform Data System for Medical Rehabilitation. Mean age was 75.5 years (sd=14.2), 73.7% of patients were female, and 82% were non-Hispanic white. Average length of stay was 17.0 days (sd=10.6). The primary outcome measure was living setting (at home vs. not at home) at 80 to 180 days after discharge. Results Statistically significant variables (p < .05) in the logistic model included follow-up therapy, sphincter control, self-care ability, marital status, age, and length of stay. Areas under the receiver operating characteristic curves were 0.67 for logistic regression and 0.73 for neural network analysis. Calibration curves indicated a slightly better fit for the neural network model. Conclusions Follow-up therapy and independent bowel and/or bladder function were strong predictors of living at home up to 6 months after hospitalization for hip fracture. No practical differences were found between the predictive ability of logistic regression and neural network analysis in this sample.

AB - Purpose Describe and compare the characteristics of artificial neural networks and logistic regression to develop prediction models in epidemiological research. Methods The sample included 3708 persons with hip fracture from 46 different states included in the Uniform Data System for Medical Rehabilitation. Mean age was 75.5 years (sd=14.2), 73.7% of patients were female, and 82% were non-Hispanic white. Average length of stay was 17.0 days (sd=10.6). The primary outcome measure was living setting (at home vs. not at home) at 80 to 180 days after discharge. Results Statistically significant variables (p < .05) in the logistic model included follow-up therapy, sphincter control, self-care ability, marital status, age, and length of stay. Areas under the receiver operating characteristic curves were 0.67 for logistic regression and 0.73 for neural network analysis. Calibration curves indicated a slightly better fit for the neural network model. Conclusions Follow-up therapy and independent bowel and/or bladder function were strong predictors of living at home up to 6 months after hospitalization for hip fracture. No practical differences were found between the predictive ability of logistic regression and neural network analysis in this sample.

KW - Joint Replacement

KW - Rehabilitation Outcomes

UR - http://www.scopus.com/inward/record.url?scp=4544258100&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=4544258100&partnerID=8YFLogxK

U2 - 10.1016/j.annepidem.2003.10.005

DO - 10.1016/j.annepidem.2003.10.005

M3 - Article

C2 - 15350954

AN - SCOPUS:4544258100

VL - 14

SP - 551

EP - 559

JO - Annals of Epidemiology

JF - Annals of Epidemiology

SN - 1047-2797

IS - 8

ER -