MP61-06 THE ROLE OF WEARABLE DEVICES AND CPET IN PREDICTING MAJOR COMPLICATIONS AFTER RADICAL CYSTECTOMY

摘要

INTRODUCTION AND OBJECTIVE: Wearable devices (WDs) are an inexpensive tool that may have utility in the collection of mobility data from patients outside the hospital environment. Cardiopulmonary exercise testing (CPET) is used routinely to pre-operatively assess patients undergoing major surgeries including RC for risk stratification. CPET is expensive, and its clinical utility for assessing pre-operative fitness is not validated. In this study, I compare WD-derived mobility data with CPET variables used for risk-stratification in RC, and compare the CPET and WD data in predicting 90-day major complications after RC. METHODS: As part of a multi-centre RCT comparing open and robotic radical cystectomy (RC), patients were consented to wear WD on their wrists for seven days for 7 consecutive days before RC. Total step-counts were collected from each device. Patients also had CPET, and anaerobic threshold (AT) <11 or VE/VCO2 ≥33 were used as used as predictors for high risk of complications. Patients were divided into two groups: Clavien-Dindo major (CD) ≥3 or minor (≤2) within 90 days. Spearman Rho was used to explore associations between non-parametrically distributed variables. Groups were compared using Mann-Whitney-U for non-parametrically distributed variables. A log transformation was performed prior to logistic regression analysis. RESULTS: 57 patients participated in the WD sub-study. 10 (17.5%) patients had major complications in the post-operative 90 days. Median step-counts of 8,626 (IQR: 6,561-12,358) were noted in patients pre-operatively. Step-counts at baseline correlated significantly with both AT and VE/VCO2 (p=0.005 and p=0.002). Using thresholds for AT and VE/VCO2 max, a significant difference in step-counts was identified in the high and low risk group (p<0.001). A statistically significant (p<0.05) correlation between step-counts and major complications was identified, but not for CPET and major complications. On ROC analysis, AUC (CI) for step-counts to predict CD≥3 complications was 0.719 (0.537-0.901). Logistic regression identified step-counts to be a statistically significant variable (p=0.04) in predicting major complications, but this was not noted for CPET. CONCLUSIONS: Step-counts measured from WDs correlate with CPET variables used for risk stratification, and step-counts are significantly lower in the CPET high-risk group. However, only step-counts from WD are statistically significant predictor of major complications. While the sample size is limited, this pilot data suggests that wearable devices may offer utility in assessing pre-operative fitness and risk-stratifying patients undergoing RC. Source of Funding: The Urology FoundationThe Champniss Foundation