Maintenance Notice

Due to necessary scheduled maintenance, the JMIR Publications website will be unavailable from Wednesday, July 01, 2020 at 8:00 PM to 10:00 PM EST. We apologize in advance for any inconvenience this may cause you.

Who will be affected?

Accepted for/Published in: Journal of Medical Internet Research

Date Submitted: Aug 30, 2022
Date Accepted: Oct 11, 2023

The final, peer-reviewed published version of this preprint can be found here:

Development and Validation of a Prognostic Classification Model Predicting Postoperative Adverse Outcomes in Older Surgical Patients Using a Machine Learning Algorithm: Retrospective Observational Network Study

Choi JY, Yoo S, Song W, Kim S, Baek H, Lee JS, Yoon YS, Yoon S, Lee HY, Kim Ki

Development and Validation of a Prognostic Classification Model Predicting Postoperative Adverse Outcomes in Older Surgical Patients Using a Machine Learning Algorithm: Retrospective Observational Network Study

J Med Internet Res 2023;25:e42259

DOI: 10.2196/42259

PMID: 37955965

PMCID: 10682929

Development and validation of a prognostic classification model predicting postoperative adverse outcomes in older surgical patients using a machine learning algorithm: a retrospective observational network study

  • Jung-Yeon Choi; 
  • Sooyoung Yoo; 
  • Wongeun Song; 
  • Seok Kim; 
  • Hyunyoung Baek; 
  • Jun Suh Lee; 
  • Yoo-Seok Yoon; 
  • Seonghae Yoon; 
  • Hae-Young Lee; 
  • Kwang-il Kim

ABSTRACT

Background:

Older adults are at an increased risk of postoperative morbidity. Numerous risk stratification tools exist, but effort and manpower are required.

Objective:

To develop a predictive model of general surgery postoperative adverse outcomes in older patients with an open-source patient-level prediction from the Observational Health Data Sciences and Informatics for internal and external validation.

Methods:

We used the Observational Medical Outcomes Partnership Common Data Model (CDM) and machine learning algorithms. The primary outcome was a composite of 90 days postoperative all-cause mortality and emergency department visits. Secondary outcomes were postoperative delirium and prolonged hospital stay (of ≥21 days). A 75% vs. 25% split of the Seoul National University Bundang Hospital (SNUBH) and Seoul National University Hospital (SNUH) CDM data was used for model training and testing vs. external validation. Model performance was evaluated using the area under the receiver operating characteristic curve (AUC) with a 95% confidence interval (CI).

Results:

Data from 25,332 (SNUBH) and 38,865 (SNUH) patients were analyzed. Compared to random forest, Adaboost, and decision tree, the least absolute shrinkage and selection operator logistic regression model showed good internal discriminative accuracy (internal AUC 0·728 [95% CI 0·712–0·745]) and transportability (external AUC 0·666 [95% CI 0·658–0·674]) for the primary outcome. The model also possessed good internal and external AUCs for postoperative delirium (internal AUC 0·791 [95% CI 0·762–0·819], external AUC 0·774 [95% CI 0·755–0·792]) and prolonged hospital stay (internal AUC 0·797 [95% CI 0·782–0·812], external AUC 0·732 [95% CI 0·724–0·739]). Compared with age or the Charlson comorbidity index, the model showed better prediction performance.

Conclusions:

The derived model shall assist clinicians and patients in understanding the individualized risks and benefits of surgery.


 Citation

Please cite as:

Choi JY, Yoo S, Song W, Kim S, Baek H, Lee JS, Yoon YS, Yoon S, Lee HY, Kim Ki

Development and Validation of a Prognostic Classification Model Predicting Postoperative Adverse Outcomes in Older Surgical Patients Using a Machine Learning Algorithm: Retrospective Observational Network Study

J Med Internet Res 2023;25:e42259

DOI: 10.2196/42259

PMID: 37955965

PMCID: 10682929

Download PDF


Request queued. Please wait while the file is being generated. It may take some time.

© The authors. All rights reserved. This is a privileged document currently under peer-review/community review (or an accepted/rejected manuscript). Authors have provided JMIR Publications with an exclusive license to publish this preprint on it's website for review and ahead-of-print citation purposes only. While the final peer-reviewed paper may be licensed under a cc-by license on publication, at this stage authors and publisher expressively prohibit redistribution of this draft paper other than for review purposes.