Patient-reported outcomes collected in ambulatory oncology practices: Feasibility, patterns, and correlates.

Objective: To examine the feasibility of soliciting outcomes from adults who received chemotherapy treatment for cancer and describe the patterns and correlates of patient-reported toxicities.

Data Sources: Patient survey data from 29 Michigan ambulatory oncology practices collected in 2017.

Study Design: Secondary analysis of patient survey data. Descriptive statistics were generated at the patient and practice levels. Thematic analysis of open-text comments identified clusters of frequently reported toxicities.

Data Collection Methods: Patients completed 11 items from the Patient-Reported Outcomes version of the Common Terminology Criteria for Adverse Events. Using a 5-point Likert scale, patients rated the frequency of nausea, vomiting, diarrhea, and pain; the severity of nausea, vomiting, constipation, numbness/tingling, and pain; and how much numbness/tingling and pain interfered with usual or daily activities. Patients could also report two toxicities in open-text comments. Finally, patients reported unplanned health care service for toxicity or side effect management.

Principal Findings: Of 3565 eligible patients, 2245 participated (63%) and 457 (20%) rated at least one toxicity as severe/very severe. Across practices, the proportion of patients who reported at least one severe/very severe toxicity ranged from 8% to 50%. Troubling toxicities included pain frequency (mean 2.3, SD 1.3), pain severity (2.1, 1.1), and diarrhea frequency (1.9, 1.0). From completed assessments, 1653 (74%) reported at least one toxicity in open-text comments; fatigue (n = 182), stomach discomfort (n = 53), and skin/nail changes (n = 41) were most frequently reported. Regarding consequences, 156 patients (7%) reported unplanned health care service use: 41 (26%) visited an emergency department and 32 (21%) were admitted to a hospital.

Conclusions: Querying patients on chemotherapy treatment experiences and toxicities was feasible. Toxicity rates varied across practices, which informed quality improvement. Toxicity severity and service use incidence exceed previously published trial data, particularly for pain, fatigue, and gastrointestinal issues. Open-text questions enabled exploration with newer treatment regimens.