52. Optimizing Pain Management in Plastic Surgery: Insights From The Pain (Pain AI Intervention) Clinical Trial Using Wearable Sensors

Abstract

PURPOSE: The purpose of this study is to develop and evaluate a sensor-based pain management strategy using a wearable device that continuously monitors physiological and environmental biomarkers. Current pain assessment methods often fail to provide continuous monitoring and lack precision, while opioid therapy poses addiction risks. By using real-time data analysis and machine learning, this study aims to improve the accuracy of postoperative pain predictions and enhance pain management, reducing reliance on opioids. METHODS: In this single-arm feasibility clinical trial in the Division of Plastic Surgery at Mayo Clinic, Florida, we developed the OMNI-H555p, a wearable device positioned at the center of the chest to continuously record ECG, PPG, skin temperature, activity, and environmental waveforms over a 72-hour period. We recruited 100 patients randomly from the plastic surgery department and anonymized their data for assessment. Patients used a provided iPhone 13 to respond to pain scale questionnaires every 4 hours and completed feasibility (FIM), appropriateness (IAM), and acceptability (AIM) questionnaires regarding their experience with the device. The recorded physiological waveforms and questionnaire responses were analyzed using appropriate statistical methods, with P < 0.05 considered statistically significant. RESULTS: Out of 100 recruited patients, 72 participated in the study, and 67 successfully used the OMNI device postoperatively. Of these, 60 completed the pain questionnaires, resulting in 850 responses over an average of 73.34 hours (SD 24.95). The cohort had a mean age of 59.4 years (SD 15.6), was predominantly White (91.0%), and comprised 62.7% female and 37.3% male participants. Participants completed an average of 14.17 questionnaires (SD 4.53) over approximately three days, with an average of 4.64 questionnaires per day. Pain transitions were gradual, and extreme jumps in pain scores were rare. The median pain score was 2, with a median absolute deviation of 1, and a right-skewed normal distribution (mean = 1.0, SD = 2.2). Physiological and environmental measures correlated significantly with pain levels. Accelerometry data across all three axes showed higher activity in the low pain group (pain scores 0-5) compared to the high pain group (pain scores 6-10). For instance, on the x-axis, absolute energy was higher in the low pain group (mean 0.93 vs 0.85). Similar trends were observed in the y-axis (mean 0.0053 vs 0.47) and z-axis (mean 0.86 vs 0.87). ECG data revealed significant variations, with a higher RMS in the low pain group (mean 0.96 vs 0.94), indicating changes in heart function correlated with pain levels (P < 0.05). Differences in temperature, pressure, and humidity were also observed between the groups, with a significant difference in median temperature (P = 0.019), and humidity (P = 0.0029). Patient evaluations of the OMNI device were positive, with high levels of agreement on feasibility (62.5%), appropriateness (65.2%), and acceptability (60.93%), reflecting broad consensus on the device’s effectiveness. CONCLUSION: Our wearable device effectively enhances postoperative pain management by providing continuous physiological and environmental monitoring. The positive patient feedback and the precision of the predictive model support its potential for clinical adoption, offering a significant advancement in pain management strategies.

Ähnliche Arbeiten