A comparison of the abilities of nine scoring algorithms in predicting mortality

J. Wayne Meredith, Gregory Evans, Patrick D. Kilgo, Ellen MacKenzie, Turner Osler, Gerald McGwin, Stephen Cohn, Thomas Esposito, Thomas Gennarelli, Michael L Hawkins, Charles Lucas, Charles Mock, Michael Rotondo, Loring Rue

Research output: Contribution to journalArticle

143 Citations (Scopus)

Abstract

Objective The purpose of this study was to compare the abilities of nine Abbreviated Injury Scale (AIS)- and International Classification of Diseases, Ninth Revision (ICD-9)-based scoring algorithms in predicting mortality. Methods The scores collected on 76,871 incidents consist of four AIS-based algorithms (Injury Severity Score [ISS], New Injury Severity Score, Anatomic Profile Score [APS], and maximum AIS [maxAIS]), their four ICD to AIS mapped counterparts, and the ICD-9-based ISS (ICISS). A 10-fold cross-validation was performed and area under the receiver operating characteristic curve was used to determine algorithm discrimination. Hosmer-Lemeshow statistics were computed to gauge goodness-of-fit, and model refinement measured variance of predicted probabilities. Results Overall, the ICISS has the best discrimination and model refinement, whereas the APS has the best Hosmer-Lemeshow performance. ICD-9 to AIS mapped scores have worse discrimination than their AIS-based counterparts, but still show moderate performance. Conclusion Differences in performance were relatively small. Complex scores such as the ICISS and the APS provide improvement in discrimination relative to the maxAIS and the ISS. Trauma registries should move to include the ICISS and the APS. The ISS and maxAIS perform moderately well and have bedside benefits.

Original languageEnglish (US)
Pages (from-to)621-629
Number of pages9
JournalJournal of Trauma
Volume53
Issue number4
DOIs
StatePublished - Jan 1 2002

Fingerprint

Abbreviated Injury Scale
International Classification of Diseases
Injury Severity Score
Mortality
ROC Curve
Registries
Wounds and Injuries

Keywords

  • ICISS
  • Injury Severity Score
  • Injury scoring
  • Outcome prediction
  • Trauma

ASJC Scopus subject areas

  • Surgery
  • Critical Care and Intensive Care Medicine

Cite this

Meredith, J. W., Evans, G., Kilgo, P. D., MacKenzie, E., Osler, T., McGwin, G., ... Rue, L. (2002). A comparison of the abilities of nine scoring algorithms in predicting mortality. Journal of Trauma, 53(4), 621-629. https://doi.org/10.1097/00005373-200210000-00001

A comparison of the abilities of nine scoring algorithms in predicting mortality. / Meredith, J. Wayne; Evans, Gregory; Kilgo, Patrick D.; MacKenzie, Ellen; Osler, Turner; McGwin, Gerald; Cohn, Stephen; Esposito, Thomas; Gennarelli, Thomas; Hawkins, Michael L; Lucas, Charles; Mock, Charles; Rotondo, Michael; Rue, Loring.

In: Journal of Trauma, Vol. 53, No. 4, 01.01.2002, p. 621-629.

Research output: Contribution to journalArticle

Meredith, JW, Evans, G, Kilgo, PD, MacKenzie, E, Osler, T, McGwin, G, Cohn, S, Esposito, T, Gennarelli, T, Hawkins, ML, Lucas, C, Mock, C, Rotondo, M & Rue, L 2002, 'A comparison of the abilities of nine scoring algorithms in predicting mortality', Journal of Trauma, vol. 53, no. 4, pp. 621-629. https://doi.org/10.1097/00005373-200210000-00001
Meredith JW, Evans G, Kilgo PD, MacKenzie E, Osler T, McGwin G et al. A comparison of the abilities of nine scoring algorithms in predicting mortality. Journal of Trauma. 2002 Jan 1;53(4):621-629. https://doi.org/10.1097/00005373-200210000-00001
Meredith, J. Wayne ; Evans, Gregory ; Kilgo, Patrick D. ; MacKenzie, Ellen ; Osler, Turner ; McGwin, Gerald ; Cohn, Stephen ; Esposito, Thomas ; Gennarelli, Thomas ; Hawkins, Michael L ; Lucas, Charles ; Mock, Charles ; Rotondo, Michael ; Rue, Loring. / A comparison of the abilities of nine scoring algorithms in predicting mortality. In: Journal of Trauma. 2002 ; Vol. 53, No. 4. pp. 621-629.
@article{378693d577f94f96a672bfaa92eb4509,
title = "A comparison of the abilities of nine scoring algorithms in predicting mortality",
abstract = "Objective The purpose of this study was to compare the abilities of nine Abbreviated Injury Scale (AIS)- and International Classification of Diseases, Ninth Revision (ICD-9)-based scoring algorithms in predicting mortality. Methods The scores collected on 76,871 incidents consist of four AIS-based algorithms (Injury Severity Score [ISS], New Injury Severity Score, Anatomic Profile Score [APS], and maximum AIS [maxAIS]), their four ICD to AIS mapped counterparts, and the ICD-9-based ISS (ICISS). A 10-fold cross-validation was performed and area under the receiver operating characteristic curve was used to determine algorithm discrimination. Hosmer-Lemeshow statistics were computed to gauge goodness-of-fit, and model refinement measured variance of predicted probabilities. Results Overall, the ICISS has the best discrimination and model refinement, whereas the APS has the best Hosmer-Lemeshow performance. ICD-9 to AIS mapped scores have worse discrimination than their AIS-based counterparts, but still show moderate performance. Conclusion Differences in performance were relatively small. Complex scores such as the ICISS and the APS provide improvement in discrimination relative to the maxAIS and the ISS. Trauma registries should move to include the ICISS and the APS. The ISS and maxAIS perform moderately well and have bedside benefits.",
keywords = "ICISS, Injury Severity Score, Injury scoring, Outcome prediction, Trauma",
author = "Meredith, {J. Wayne} and Gregory Evans and Kilgo, {Patrick D.} and Ellen MacKenzie and Turner Osler and Gerald McGwin and Stephen Cohn and Thomas Esposito and Thomas Gennarelli and Hawkins, {Michael L} and Charles Lucas and Charles Mock and Michael Rotondo and Loring Rue",
year = "2002",
month = "1",
day = "1",
doi = "10.1097/00005373-200210000-00001",
language = "English (US)",
volume = "53",
pages = "621--629",
journal = "Journal of Trauma and Acute Care Surgery",
issn = "2163-0755",
publisher = "Lippincott Williams and Wilkins",
number = "4",

}

TY - JOUR

T1 - A comparison of the abilities of nine scoring algorithms in predicting mortality

AU - Meredith, J. Wayne

AU - Evans, Gregory

AU - Kilgo, Patrick D.

AU - MacKenzie, Ellen

AU - Osler, Turner

AU - McGwin, Gerald

AU - Cohn, Stephen

AU - Esposito, Thomas

AU - Gennarelli, Thomas

AU - Hawkins, Michael L

AU - Lucas, Charles

AU - Mock, Charles

AU - Rotondo, Michael

AU - Rue, Loring

PY - 2002/1/1

Y1 - 2002/1/1

N2 - Objective The purpose of this study was to compare the abilities of nine Abbreviated Injury Scale (AIS)- and International Classification of Diseases, Ninth Revision (ICD-9)-based scoring algorithms in predicting mortality. Methods The scores collected on 76,871 incidents consist of four AIS-based algorithms (Injury Severity Score [ISS], New Injury Severity Score, Anatomic Profile Score [APS], and maximum AIS [maxAIS]), their four ICD to AIS mapped counterparts, and the ICD-9-based ISS (ICISS). A 10-fold cross-validation was performed and area under the receiver operating characteristic curve was used to determine algorithm discrimination. Hosmer-Lemeshow statistics were computed to gauge goodness-of-fit, and model refinement measured variance of predicted probabilities. Results Overall, the ICISS has the best discrimination and model refinement, whereas the APS has the best Hosmer-Lemeshow performance. ICD-9 to AIS mapped scores have worse discrimination than their AIS-based counterparts, but still show moderate performance. Conclusion Differences in performance were relatively small. Complex scores such as the ICISS and the APS provide improvement in discrimination relative to the maxAIS and the ISS. Trauma registries should move to include the ICISS and the APS. The ISS and maxAIS perform moderately well and have bedside benefits.

AB - Objective The purpose of this study was to compare the abilities of nine Abbreviated Injury Scale (AIS)- and International Classification of Diseases, Ninth Revision (ICD-9)-based scoring algorithms in predicting mortality. Methods The scores collected on 76,871 incidents consist of four AIS-based algorithms (Injury Severity Score [ISS], New Injury Severity Score, Anatomic Profile Score [APS], and maximum AIS [maxAIS]), their four ICD to AIS mapped counterparts, and the ICD-9-based ISS (ICISS). A 10-fold cross-validation was performed and area under the receiver operating characteristic curve was used to determine algorithm discrimination. Hosmer-Lemeshow statistics were computed to gauge goodness-of-fit, and model refinement measured variance of predicted probabilities. Results Overall, the ICISS has the best discrimination and model refinement, whereas the APS has the best Hosmer-Lemeshow performance. ICD-9 to AIS mapped scores have worse discrimination than their AIS-based counterparts, but still show moderate performance. Conclusion Differences in performance were relatively small. Complex scores such as the ICISS and the APS provide improvement in discrimination relative to the maxAIS and the ISS. Trauma registries should move to include the ICISS and the APS. The ISS and maxAIS perform moderately well and have bedside benefits.

KW - ICISS

KW - Injury Severity Score

KW - Injury scoring

KW - Outcome prediction

KW - Trauma

UR - http://www.scopus.com/inward/record.url?scp=0036797168&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0036797168&partnerID=8YFLogxK

U2 - 10.1097/00005373-200210000-00001

DO - 10.1097/00005373-200210000-00001

M3 - Article

C2 - 12394857

AN - SCOPUS:0036797168

VL - 53

SP - 621

EP - 629

JO - Journal of Trauma and Acute Care Surgery

JF - Journal of Trauma and Acute Care Surgery

SN - 2163-0755

IS - 4

ER -