External validation of two models to predict delirium in critically ill adults using either the confusion assessment method-ICU or the intensive care delirium screening checklist for delirium assessment
External validation of two models to predict delirium in critically ill adults using either the confusion assessment method-ICU or the intensive care delirium screening checklist for delirium assessment
Objectives: To externally validate two delirium prediction models (the E-PRE-DELIRIC and recalibrated PRE-DELIRIC) and compare the performance of each model usingboth ConfusionAssessment Method-ICU (CAM-ICU) and Intensive Care Delirium Screening Checklist (ICDSC)deliriumassessment scores.
Design: Prospective,multinational cohort study.
Setting: ElevenICUs from seven countries in 3 continents.
Patients: Consecutive, delirium-free adults admitted tothe ICU for ≥ 6 hours in whom delirium could be reliably assessed.
Interventions: None.
Measurements and Main Results: The predictors included in eachmodel were collected (E-PRE-DELIRIC: at the time of ICU admission; PRE-DELIRIC:within 24 hours of ICU admission). Delirium was assessed using the CAM-ICU or ICDSC. Discriminationand calibration of bothmodels was assessed for the total cohort and also for patients evaluated witheach delirium assessment tool. Discrimination was determined using the area under the receiveroperating characteristic curve (AUROC). The predictive performance using the CAM-ICU or ICDSC was compared and the performance of the total cohortwas compared with the original results for each model. Calibration was assessedgraphically. A total of 2,178 patients were included. Deliriumincidence was 21.4%. TheAUROC of the E-PRE-DELIRIC model was 0.68 (95%CI 0.66-0.71) versusprior reported AUROC of 0.75 (95%CI 0.71-0.79). TheAUROC of the recalibrated PRE-DELIRIC model was 0.74 (95%CI 0.71-0.76) versusprior AUROC of 0.77 (95%CI 0.74-0.79). Bothmodels were well calibrated. Use of CAM-ICU or the ICDSC did not significantly affect the AUROC for eitherthe E-PRE-DELIRIC [Z-score of -0.91 (p>0.1] or PRE-DELIRIC [Z-score of 1.50(p>0.1)] model.
Conclusions: Using alarge, heterogeneous, multinational cohort of critically ill adults, evaluatedwith either the CAM-ICU or ICDSC, we show that the E-PRE-DELIRIC and PRE-DELIRIC models both statistically perform at a moderately-to-goodlevel. These results allowfor generalization to ICUs around the world.
e827-e835
Wassenaar, Annelies
4a047852-d9cb-4760-879d-af4ec237204e
Schoonhoven, Lisette
46a2705b-c657-409b-b9da-329d5b1b02de
Devlin, John W.
3b973cd9-e462-4596-9674-fd9798b15c9a
van Haren, Frank M.P.
42e8c3f2-0dd5-4def-911e-67d961ec4bc1
Slooter, Arjen J.C.
b655a34a-8109-4baf-a463-189d403212fc
Jorens, Philippe G.
a53a9dea-9822-4b0a-8cd5-8386319fd40f
van der Jagt, Mathieu
db97d638-078b-402c-85a6-348bc0750ce6
Simons, Koen S.
e276578a-1001-4f9d-816f-c3c345cd4a09
Egerod, Ingrid
0b99d6f7-8163-40b0-9574-966904c41085
Burry, Lisa D.
ef8bf2fd-a823-4150-bce9-443bb8ea3263
Beishuizen, Albertus
0403a954-d9f1-4254-b35f-a50efb7345b7
Matos, Joaquim
e18c38e7-109a-4a00-9a9c-9c02f7cabfea
Donders, A. Rogier T.
dd864788-b402-459b-b4e7-ecd7ebaf1cad
Pickkers, Peter
516df191-7ae2-457e-a7f7-abd6ca935687
Van den Boogaard, Mark
4751824c-6a51-4bc8-8854-97f4579e045b
Wassenaar, Annelies
4a047852-d9cb-4760-879d-af4ec237204e
Schoonhoven, Lisette
46a2705b-c657-409b-b9da-329d5b1b02de
Devlin, John W.
3b973cd9-e462-4596-9674-fd9798b15c9a
van Haren, Frank M.P.
42e8c3f2-0dd5-4def-911e-67d961ec4bc1
Slooter, Arjen J.C.
b655a34a-8109-4baf-a463-189d403212fc
Jorens, Philippe G.
a53a9dea-9822-4b0a-8cd5-8386319fd40f
van der Jagt, Mathieu
db97d638-078b-402c-85a6-348bc0750ce6
Simons, Koen S.
e276578a-1001-4f9d-816f-c3c345cd4a09
Egerod, Ingrid
0b99d6f7-8163-40b0-9574-966904c41085
Burry, Lisa D.
ef8bf2fd-a823-4150-bce9-443bb8ea3263
Beishuizen, Albertus
0403a954-d9f1-4254-b35f-a50efb7345b7
Matos, Joaquim
e18c38e7-109a-4a00-9a9c-9c02f7cabfea
Donders, A. Rogier T.
dd864788-b402-459b-b4e7-ecd7ebaf1cad
Pickkers, Peter
516df191-7ae2-457e-a7f7-abd6ca935687
Van den Boogaard, Mark
4751824c-6a51-4bc8-8854-97f4579e045b
Wassenaar, Annelies, Schoonhoven, Lisette, Devlin, John W., van Haren, Frank M.P., Slooter, Arjen J.C., Jorens, Philippe G., van der Jagt, Mathieu, Simons, Koen S., Egerod, Ingrid, Burry, Lisa D., Beishuizen, Albertus, Matos, Joaquim, Donders, A. Rogier T., Pickkers, Peter and Van den Boogaard, Mark
(2019)
External validation of two models to predict delirium in critically ill adults using either the confusion assessment method-ICU or the intensive care delirium screening checklist for delirium assessment.
Critical Care Medicine, 47 (10), .
(doi:10.1097/CCM.0000000000003911).
Abstract
Objectives: To externally validate two delirium prediction models (the E-PRE-DELIRIC and recalibrated PRE-DELIRIC) and compare the performance of each model usingboth ConfusionAssessment Method-ICU (CAM-ICU) and Intensive Care Delirium Screening Checklist (ICDSC)deliriumassessment scores.
Design: Prospective,multinational cohort study.
Setting: ElevenICUs from seven countries in 3 continents.
Patients: Consecutive, delirium-free adults admitted tothe ICU for ≥ 6 hours in whom delirium could be reliably assessed.
Interventions: None.
Measurements and Main Results: The predictors included in eachmodel were collected (E-PRE-DELIRIC: at the time of ICU admission; PRE-DELIRIC:within 24 hours of ICU admission). Delirium was assessed using the CAM-ICU or ICDSC. Discriminationand calibration of bothmodels was assessed for the total cohort and also for patients evaluated witheach delirium assessment tool. Discrimination was determined using the area under the receiveroperating characteristic curve (AUROC). The predictive performance using the CAM-ICU or ICDSC was compared and the performance of the total cohortwas compared with the original results for each model. Calibration was assessedgraphically. A total of 2,178 patients were included. Deliriumincidence was 21.4%. TheAUROC of the E-PRE-DELIRIC model was 0.68 (95%CI 0.66-0.71) versusprior reported AUROC of 0.75 (95%CI 0.71-0.79). TheAUROC of the recalibrated PRE-DELIRIC model was 0.74 (95%CI 0.71-0.76) versusprior AUROC of 0.77 (95%CI 0.74-0.79). Bothmodels were well calibrated. Use of CAM-ICU or the ICDSC did not significantly affect the AUROC for eitherthe E-PRE-DELIRIC [Z-score of -0.91 (p>0.1] or PRE-DELIRIC [Z-score of 1.50(p>0.1)] model.
Conclusions: Using alarge, heterogeneous, multinational cohort of critically ill adults, evaluatedwith either the CAM-ICU or ICDSC, we show that the E-PRE-DELIRIC and PRE-DELIRIC models both statistically perform at a moderately-to-goodlevel. These results allowfor generalization to ICUs around the world.
Text
External validation of two models to predict delirium in critically ill adults
- Accepted Manuscript
More information
Accepted/In Press date: 7 May 2018
e-pub ahead of print date: 1 October 2019
Identifiers
Local EPrints ID: 428091
URI: http://eprints.soton.ac.uk/id/eprint/428091
ISSN: 0090-3493
PURE UUID: 549e86a1-8753-45f5-a128-34f598ed7522
Catalogue record
Date deposited: 08 Feb 2019 17:30
Last modified: 16 Apr 2024 04:02
Export record
Altmetrics
Contributors
Author:
Annelies Wassenaar
Author:
John W. Devlin
Author:
Frank M.P. van Haren
Author:
Arjen J.C. Slooter
Author:
Philippe G. Jorens
Author:
Mathieu van der Jagt
Author:
Koen S. Simons
Author:
Ingrid Egerod
Author:
Lisa D. Burry
Author:
Albertus Beishuizen
Author:
Joaquim Matos
Author:
A. Rogier T. Donders
Author:
Peter Pickkers
Author:
Mark Van den Boogaard
Download statistics
Downloads from ePrints over the past year. Other digital versions may also be available to download e.g. from the publisher's website.
View more statistics