Researchers conducted a single-center, cross-sectional study to examine the relationship between patient-reported quality-of-life (QOL), adverse events (AEs), and treatment characteristics (including tumor type, drug class, number of cycles, and treatment intent). The study’s findings were presented at the 2017 ASCO Annual Meeting.
Consecutive patients attending an outpatient chemotherapy unit completed two questionnaires: the European Organization for Research and Treatment of Cancer quality-of-life questionnaire and National Cancer Institute Patient-Reported Outcomes version of the Common Terminology Criteria for Adverse Events, per visit to identify factors associated with the lowest global QOL score (QL2) over a six-week period.
A total of 115 patients completed up to six sets of questionnaires with a median QL2 score of 67 (range = 50–83). No difference was found between QL2 and treatment characteristics, but QL2 was significantly associated with AEs related to gastrointestinal, respiratory, attention, pain, sleep/wake, and mood (see TABLE). Using AEs as covariates, support vector machine with radial basis kernel was the best at classifying patients into QOL groups (mean bootstrapped area under ROC curve = 0.812; 95% confidence interval = 0.700–0.925).
“Machine learning analysis suggests that a combined AE analysis may reliably characterize a patient’s QOL,” the authors concluded.