QOL-25. DIGITAL PHENOTYPING OF PERIOPERATIVE MOBILITY METRICS IN PATIENTS UNDERGOING SURGICAL RESECTION OF SKULL BASE TUMORS

Abstract Current methods to assess quality of life and recovery after surgery for skull base tumors are limited and often biased by subjective patient-reported assessments, while objective, in-situ assessments are lacking. Smartphone-based digital phenotyping has been increasingly studied across pathologies, utilizing built-in technologies to measure behavioral patterns pertaining to sleep, physical mobility, social interactions, and cognitive functioning, among others. Here, we report our initial experience with smartphone-based digital phenotyping in patients who underwent resection of skull base lesions. Patients seen at our multi-disciplinary skull base center were enrolled under an institutional board-approved protocol and installed the Beiwe application, which passively collected GPS data 20 days pre- and up to 140 days post-surgery. Variables measured included time spent at home, number of separate significant places visited outside home, daily walking time, and daily number of steps taken. 26 patients were enrolled, comprised of 14 pituitary adenomas, 8 meningiomas, and 4 vestibular schwannomas. All patients demonstrated increased time spent at home, fewer number of significant locations visited, decreased walking time, and decreased total steps in the immediate post-operative period. For pituitaries, mean return to baseline for these variables was 25.7, 24.3, 80.2, and 88.9 days, respectively. For meningiomas, return to baseline was 56.5, 214.4, 45.0, and 45.6 days, respectively. For vestibular schwannomas, return to baseline was 219, 39.9, 35.1, and 36.4 days, respectively. This study demonstrates initial efficacy of digital phenotyping for a cohort of patients undergoing surgical resection of skull base lesions, based on passively collected smartphone GPS data during the peri-operative period. This is part of a larger, ongoing study enrolling surgical patients as well as non-operative controls for comparison. Such peri-operative mobility metrics may potentially be refined to predict postoperative outcomes following surgery such as improvement of neurological symptoms, postoperative CSF leak, infection, dysnatremia, and hospital length of stay.