Developing and validating an interpretable machine learning model for predicting postoperative nausea and vomiting in pulmonary resection patients: a retrospective observational study

Postoperative nausea and vomiting (PONV) remains a prevalent complication hindering Enhanced Recovery After Surgery (ERAS) protocols, particularly after pulmonary resection. While established clinical risk factors exist, the Apfel score lacks precision for individualized prediction and ignores neuroanatomical biomarkers....

No actionable change for audiologists — this study concerns postoperative nausea and vomiting prediction in thoracic surgery patients and has no direct relevance to audiology practice.

While outside core audiology, PONV management is peripherally relevant in surgical settings where audiologists may co-manage patients (e.g., cochlear implant surgery), but this specific model is not yet validated for ENT/audiology surgical contexts.

1. 01Retrospective ML model developed to predict PONV in pulmonary resection patients.
2. 02Model is designed to be interpretable, addressing the 'black box' criticism of AI in medicine.
3. 03Study conducted within an Enhanced Recovery After Surgery (ERAS) protocol context.
4. 04Retrospective design limits causal inference; prospective validation would be needed before clinical adoption.
5. 05Minimal direct relevance to audiology clinical practice.

PMID

42393535

DOI

10.1186/s12871-026-03942-5.

Journal

BMC Anesthesiology

Publication type

research_article

Evidence level

4

Population

Patients undergoing pulmonary resection surgery

Intervention

Interpretable machine learning model for PONV prediction