Symptoms and Needs Monitoring in Diverse Ambulatory Oncology Patients: Usage Characteristics and Impact on Emergency Room Visits and Hospitalization.

Abstract

Symptoms and needs monitoring using patient-reported outcomes (PRO) is associated with improved clinical outcomes in cancer care. However, these improvements have been observed predominantly in non-Hispanic White patients using English assessments with high completion rates. The documented impact of such monitoring on system-level outcomes including emergency room (ER) visits and hospitalizations remains limited. We explored factors affecting the completion of PRO measures and evaluated clinical outcomes in an ambulatory oncology setting with a diverse racial, ethnic, and linguistic population. A retrospective analysis (October 2019-February 2022) was performed for patients with cancer assigned to My Wellness Check (MWC), a patient-portal-administered and electronic health record-based PRO assessment that generates automated alerts to oncology providers. Patient demographics, clinical characteristics, and clinical outcomes were collected. Logistic regression models examined factors affecting the completion of MWC questionnaires. Cumulative incidence of ER visits and hospitalization were assessed by Cox proportional hazards regression models adjusting for demographics. We identified 9,553 patients; 43.1% (n = 4,117) answered one or more questions. Patients age 65 years or older (adjusted odds ratio [aOR], 0.77; P < .0001), male (aOR, 0.81; P < .0001), Hispanic/Latino ethnicity (aOR, 0.70; P < .0001), living without partners (aOR, 0.75; P < .0001), or receiving no treatment (aOR, 0.76; P < .0001) were less likely to answer MWC questionnaires. Patients who completed the entire MWC questionnaires had a reduced risk of an ER visit (adjusted hazard ratio, 0.78; P < .0001) and hospitalization (adjusted hazard ratio, 0.80; P = .0007) relative to patients who did not. Completing electronic health record-based PRO assessments was associated with significantly better clinical outcomes in a diverse cancer population. Specific patient groups were less likely to participate. Further research is needed to identify barriers to completing PRO measures and the long-term benefits of such programs.