Some prognostic models for traumatic brain injury were not valid

Chantal W P M Hukkelhoven, Anneke J J Rampen, Andrew I R Maas, Elana Farace, J. Dik F Habbema, Anthony Marmarou, Lawrence F. Marshall, Gordon D. Murray, Ewout W. Steyerberg

Research output: Contribution to journalArticle

45 Citations (Scopus)

Abstract

Objective: Various prognostic models have been developed to predict outcome after traumatic brain injury (TBI). We aimed to determine the validity of six models that used baseline clinical and computed tomographic characteristics to predict mortality or unfavorable outcome at 6 months or later after severe or moderate TBI. Study Design and Setting: The validity was studied in two selected series of TBI patients enrolled in clinical trials (Tirilazad trials; n = 2,269; International Selfotel Trial; n = 409) and in two unselected series of patients consecutively admitted to participating centers (European Brain Injury Consortium [EBIC] survey; n = 796; Traumatic Coma Data Bank; n = 746). Validity was indicated by discriminative ability (AUC) and calibration (Hosmer-Lemeshow goodness-of-fit test). Results: The models varied in number of predictors (four to seven) and in development technique (two prediction trees and four logistic regression models). Discriminative ability varied widely (AUC: .61-.89), but calibration was poor for most models. Better discrimination was observed for logistic regression models compared with trees, and for models including more predictors. Further, discrimination was better when tested on unselected series that contained more heterogeneous populations. Conclusion: Our findings emphasize the need for external validation of prognostic models. The satisfactory discrimination indicates that logistic regression models, developed on large samples, can be used for classifying TBI patients according to prognostic risk.

Original languageEnglish (US)
Pages (from-to)132-143
Number of pages12
JournalJournal of Clinical Epidemiology
Volume59
Issue number2
DOIs
StatePublished - Feb 1 2006

Fingerprint

Logistic Models
selfotel
Calibration
Area Under Curve
Post-Head Injury Coma
Brain Injuries
Traumatic Brain Injury
Clinical Trials
Databases
Mortality
Population

All Science Journal Classification (ASJC) codes

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

Cite this

Hukkelhoven, C. W. P. M., Rampen, A. J. J., Maas, A. I. R., Farace, E., Habbema, J. D. F., Marmarou, A., ... Steyerberg, E. W. (2006). Some prognostic models for traumatic brain injury were not valid. Journal of Clinical Epidemiology, 59(2), 132-143. https://doi.org/10.1016/j.jclinepi.2005.06.009
Hukkelhoven, Chantal W P M ; Rampen, Anneke J J ; Maas, Andrew I R ; Farace, Elana ; Habbema, J. Dik F ; Marmarou, Anthony ; Marshall, Lawrence F. ; Murray, Gordon D. ; Steyerberg, Ewout W. / Some prognostic models for traumatic brain injury were not valid. In: Journal of Clinical Epidemiology. 2006 ; Vol. 59, No. 2. pp. 132-143.
@article{7e92355c8a7d48ad8ca4b6b3d3dfe504,
title = "Some prognostic models for traumatic brain injury were not valid",
abstract = "Objective: Various prognostic models have been developed to predict outcome after traumatic brain injury (TBI). We aimed to determine the validity of six models that used baseline clinical and computed tomographic characteristics to predict mortality or unfavorable outcome at 6 months or later after severe or moderate TBI. Study Design and Setting: The validity was studied in two selected series of TBI patients enrolled in clinical trials (Tirilazad trials; n = 2,269; International Selfotel Trial; n = 409) and in two unselected series of patients consecutively admitted to participating centers (European Brain Injury Consortium [EBIC] survey; n = 796; Traumatic Coma Data Bank; n = 746). Validity was indicated by discriminative ability (AUC) and calibration (Hosmer-Lemeshow goodness-of-fit test). Results: The models varied in number of predictors (four to seven) and in development technique (two prediction trees and four logistic regression models). Discriminative ability varied widely (AUC: .61-.89), but calibration was poor for most models. Better discrimination was observed for logistic regression models compared with trees, and for models including more predictors. Further, discrimination was better when tested on unselected series that contained more heterogeneous populations. Conclusion: Our findings emphasize the need for external validation of prognostic models. The satisfactory discrimination indicates that logistic regression models, developed on large samples, can be used for classifying TBI patients according to prognostic risk.",
author = "Hukkelhoven, {Chantal W P M} and Rampen, {Anneke J J} and Maas, {Andrew I R} and Elana Farace and Habbema, {J. Dik F} and Anthony Marmarou and Marshall, {Lawrence F.} and Murray, {Gordon D.} and Steyerberg, {Ewout W.}",
year = "2006",
month = "2",
day = "1",
doi = "10.1016/j.jclinepi.2005.06.009",
language = "English (US)",
volume = "59",
pages = "132--143",
journal = "Journal of Clinical Epidemiology",
issn = "0895-4356",
publisher = "Elsevier USA",
number = "2",

}

Hukkelhoven, CWPM, Rampen, AJJ, Maas, AIR, Farace, E, Habbema, JDF, Marmarou, A, Marshall, LF, Murray, GD & Steyerberg, EW 2006, 'Some prognostic models for traumatic brain injury were not valid', Journal of Clinical Epidemiology, vol. 59, no. 2, pp. 132-143. https://doi.org/10.1016/j.jclinepi.2005.06.009

Some prognostic models for traumatic brain injury were not valid. / Hukkelhoven, Chantal W P M; Rampen, Anneke J J; Maas, Andrew I R; Farace, Elana; Habbema, J. Dik F; Marmarou, Anthony; Marshall, Lawrence F.; Murray, Gordon D.; Steyerberg, Ewout W.

In: Journal of Clinical Epidemiology, Vol. 59, No. 2, 01.02.2006, p. 132-143.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Some prognostic models for traumatic brain injury were not valid

AU - Hukkelhoven, Chantal W P M

AU - Rampen, Anneke J J

AU - Maas, Andrew I R

AU - Farace, Elana

AU - Habbema, J. Dik F

AU - Marmarou, Anthony

AU - Marshall, Lawrence F.

AU - Murray, Gordon D.

AU - Steyerberg, Ewout W.

PY - 2006/2/1

Y1 - 2006/2/1

N2 - Objective: Various prognostic models have been developed to predict outcome after traumatic brain injury (TBI). We aimed to determine the validity of six models that used baseline clinical and computed tomographic characteristics to predict mortality or unfavorable outcome at 6 months or later after severe or moderate TBI. Study Design and Setting: The validity was studied in two selected series of TBI patients enrolled in clinical trials (Tirilazad trials; n = 2,269; International Selfotel Trial; n = 409) and in two unselected series of patients consecutively admitted to participating centers (European Brain Injury Consortium [EBIC] survey; n = 796; Traumatic Coma Data Bank; n = 746). Validity was indicated by discriminative ability (AUC) and calibration (Hosmer-Lemeshow goodness-of-fit test). Results: The models varied in number of predictors (four to seven) and in development technique (two prediction trees and four logistic regression models). Discriminative ability varied widely (AUC: .61-.89), but calibration was poor for most models. Better discrimination was observed for logistic regression models compared with trees, and for models including more predictors. Further, discrimination was better when tested on unselected series that contained more heterogeneous populations. Conclusion: Our findings emphasize the need for external validation of prognostic models. The satisfactory discrimination indicates that logistic regression models, developed on large samples, can be used for classifying TBI patients according to prognostic risk.

AB - Objective: Various prognostic models have been developed to predict outcome after traumatic brain injury (TBI). We aimed to determine the validity of six models that used baseline clinical and computed tomographic characteristics to predict mortality or unfavorable outcome at 6 months or later after severe or moderate TBI. Study Design and Setting: The validity was studied in two selected series of TBI patients enrolled in clinical trials (Tirilazad trials; n = 2,269; International Selfotel Trial; n = 409) and in two unselected series of patients consecutively admitted to participating centers (European Brain Injury Consortium [EBIC] survey; n = 796; Traumatic Coma Data Bank; n = 746). Validity was indicated by discriminative ability (AUC) and calibration (Hosmer-Lemeshow goodness-of-fit test). Results: The models varied in number of predictors (four to seven) and in development technique (two prediction trees and four logistic regression models). Discriminative ability varied widely (AUC: .61-.89), but calibration was poor for most models. Better discrimination was observed for logistic regression models compared with trees, and for models including more predictors. Further, discrimination was better when tested on unselected series that contained more heterogeneous populations. Conclusion: Our findings emphasize the need for external validation of prognostic models. The satisfactory discrimination indicates that logistic regression models, developed on large samples, can be used for classifying TBI patients according to prognostic risk.

UR - http://www.scopus.com/inward/record.url?scp=30944455273&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=30944455273&partnerID=8YFLogxK

U2 - 10.1016/j.jclinepi.2005.06.009

DO - 10.1016/j.jclinepi.2005.06.009

M3 - Article

C2 - 16426948

AN - SCOPUS:30944455273

VL - 59

SP - 132

EP - 143

JO - Journal of Clinical Epidemiology

JF - Journal of Clinical Epidemiology

SN - 0895-4356

IS - 2

ER -