Wearable-derived postdischarge mobility as a digital biomarker for 90-day readmission and mortality in metastatic cancer.

purposeThe postdischarge period after unplanned hospitalization is a vulnerable transition for patients with metastatic cancer, often accompanied by functional decline, symptom burden, psychological distress, early readmission, and death. However, short-term risk assessment during this period still relies largely on episodic clinician-rated measures. We evaluated whether wearable-derived postdischarge step counts are associated with 90-day readmission, mortality, and patient-reported outcomes in patients with metastatic cancer.

methodsIn this prospective, two-center cohort study, adults with stage IV solid tumors discharged after unplanned hospitalization from two tertiary oncology centers were followed for 90 days. Physical activity was assessed using a wrist-worn wearable device, and median daily step count during a prespecified 14-day postdischarge landmark period was used as the main exposure. The primary outcome was unplanned 90-day readmission, and the secondary outcome was 90-day all-cause mortality. Associations with health-related quality of life, sleep quality, anxiety, and depressive symptoms were also evaluated using validated patient-reported outcome measures.

resultsAmong 200 evaluable patients, 86 (43.0%) experienced unplanned readmission and 56 (28.0%) died within 90 days. Receiver operating characteristic analysis identified 3013 steps/day as the optimal cutoff for 90-day readmission. Patients with lower postdischarge activity (≤ 3013 steps/day) had substantially higher readmission (73.0% vs 13.0%) and mortality (47.0% vs 9.0%) rates than those with higher activity (both p < 0.001). In multivariable analysis, low step count remained independently associated with readmission (adjusted OR 22.9, 95% CI 9.3-56.6) and mortality (HR 5.46, 95% CI 2.54-11.74). Median daily step count demonstrated strong discrimination for 90-day readmission (AUC 0.86) and mortality (AUC 0.83). Higher postdischarge activity was also associated with better quality of life, better sleep, and lower anxiety and depressive symptom burden. Each 1000-step increase was associated with lower odds of poor sleep quality, clinically significant anxiety, and depressive symptoms.

conclusionWearable-derived postdischarge mobility was strongly associated with short-term clinical outcomes and supportive care-relevant patient-reported outcomes in metastatic cancer. These findings support further evaluation of objective activity monitoring as a scalable, patient-centered approach to postdischarge risk stratification. External validation and prospective interventional studies are needed before wearable-derived mobility measures can be used to guide supportive care strategies.

trial registrationClinicalTrials.gov Identifier: NCT06687330.