Complications Following Facelift and Neck Lift: Implementation and Assessment of Large Language Model and Artificial Intelligence (ChatGPT) Performance Across 16 Simulated Patient Presentations

Abstract

Introduction

ChatGPT represents a potential resource for patient guidance and education, with the possibility for quality improvement in healthcare delivery. The present study evaluates the role of ChatGPT as an interactive patient resource, and assesses its performance in identifying, triaging, and guiding patients with concerns of postoperative complications following facelift and neck lift surgery.

Methods

Sixteen patient profiles were generated to simulate postoperative patient presentations, with complications of varying acuity and severity. ChatGPT was assessed for its accuracy in generating a differential diagnosis, soliciting a history, providing the most-likely diagnosis, the appropriate disposition, treatments/interventions to begin from home, and red-flag symptoms necessitating an urgent presentation to the emergency department.

Results

Overall accuracy in providing a complete differential diagnosis in response to simulated presentations was 85%, with an accuracy of 88% in identifying the most-likely diagnosis after history-taking. However, appropriate patient dispositions were suggested in only 56% of cases. Relevant home treatments/interventions were suggested with an 82% accuracy, and red-flag symptoms with a 73% accuracy. A detailed analysis, stratified according to latency of postoperative presentation (<48 h, 48 h–1 week, or >1 week), and according to acuity of complications, is presented herein.

Conclusions

ChatGPT overestimated the urgency of indicated patient dispositions in 44% of cases, concerning for potential unnecessary increase in healthcare resource utilization. Imperfect performance, and the tool’s tendency for overinclusion in its responses, risk increasing patient anxiety and straining physician-patient relationships. While artificial intelligence has great potential in triaging postoperative patient concerns, and improving efficiency and resource utilization, ChatGPT’s performance, in its current form, demonstrates a need for further refinement before its safe and effective implementation in facial aesthetic surgical practice.

Level of Evidence IV

This journal requires that authors assign a level of evidence to each article. For a full description of these Evidence-Based Medicine ratings, please refer to the Table of Contents or the online Instructions to Authors www.springer.com/00266.