Article Sidebar
Main Article Content
Abstract
Background. Postoperative delirium is common in intensive care and linked to adverse functional and cognitive outcomes. Valid risk prediction requires transparent predictor timing, outcome definition, validation design, and cohort independence.
Objective. To reproduce and extend a published postoperative delirium model, quantify record overlap between its public development and validation worksheets, and determine what level of validation the data support.
Methods. We analyzed a public prospective-cohort workbook (300 development and 100 designated-validation records). The published five-predictor equation, a refitted logistic model, and a ten-predictor penalized logistic model were compared by repeated nested 10-fold cross-validation. Discrimination, Brier score, and calibration were assessed with bootstrap intervals. Exact cross-sheet matches required agreement on ten predictors, the Intensive Care Delirium Screening Checklist score, and the delirium label. Designated-worksheet and non-overlap analyses were sensitivity analyses.
Results. Delirium occurred in 82/300 development (27.3%) and 29/100 designated-validation records (29.0%). The area under the receiver operating characteristic curve (AUROC) was 0.838 (95% bootstrap interval 0.782–0.886) for the refitted five-predictor model and 0.840 (0.785–0.888) for the penalized model; Brier scores were 0.132 and 0.134, and calibration slopes 0.751 and 1.061. Seventy-six of 100 validation records matched development records exactly. Designated-worksheet AUROC was high (0.957–0.961) but not independent. The 24-record non-overlap subset had 17 events and seven non-events, so estimates were imprecise.
Conclusion. The parsimonious model showed reproducible internal discrimination, and penalization added no material gain. Because record independence is not supported, the validation worksheet does not establish transportability. Prospective evaluation in an independent cohort is required before clinical use.
Keywords
Article Details
As our aim is to disseminate original research article, hence the publishing right is a necessary one. The publishing right is needed in order to reach the agreement between the author and publisher. As the journal is fully open access, the authors will sign an exclusive license agreement.
The authors have the right to:
- Share their article in the same ways permitted to third parties under the relevant user license.
- Retain copyright, patent, trademark and other intellectual property rights including research data.
- Proper attribution and credit for the published work.
For the open access article, the publisher is granted to the following right.
- The non-exclusive right to publish the article and grant right to others.
- For the published article, the publisher applied for the Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.
