Integration of patient-reported adverse events into a phase II chemotherapy treatment trial.

Abstract

6583 Background: It has been shown that patient reporting of symptomatic adverse events (AEs) is feasible during routine chemotherapy, and that patients report AEs that clinicians otherwise miss. It is not known if this approach is feasible in a treatment trial; whether clinicians will agree or disagree with patent-reported AE grades; or if clinicians will take actions based on patients' AE reports. Methods: Patients with advanced non-small cell lung cancer enrolling in a single-arm phase 2 trial of paclitaxel, pemetrexed, and bevacizumab at Memorial Sloan-Kettering Cancer Center (NCT00807573) reported 13 validated plain language questions adapted from the NCI's Common Terminology Criteria for Adverse Events (CTCAE), via iPads in waiting areas before each visit. Patients’ grades were transmitted to an investigator web page via which clinicians could agree or reassign each of the CTCAE grades, which were then transmitted to the institutional database as the official reported AEs for the trial. The proportion of patients and clinicians reporting AEs at each visit was tabulated with an a priori average compliance threshold set at 75% to conclude feasibility of the approach. The rate of clinician agreement with patient grades was computed. The frequency of clinical actions in response to patient-reported grade 3/4 AEs was calculated (ie, chemotherapy dose modification, ER referral, supportive medication). Results: 44 patients were enrolled between 1/09-9/11 (mean age 59; 50% female; 93% white), encompassing 721 total clinic visits (median 10; range 1-65). Patients reported AEs at visits 99.6% of the time and clinicians 98.7% of the time. Median time between patient self-reports and clinician reviews was 1 hour (range 0.2-18). Across symptoms, clinicians agreed with patients 93% of the time (weighted kappas .83 to .92), increasing the grade 5% of the time and decreasing it 2%. Clinical actions were taken in response to 67% of patient-reported grade 3/4 AEs. 71% of clinicians stated patient reporting informed their toxicity grading. Conclusions: Collection of patient-reported AEs at clinic visits via computer is feasible in a chemotherapy trial. Clinicians agree with patients’ self-reported AE grades and act on this information.