Thresholds in PROMIS Scores Anchored to Subsequent Unscheduled Health Service Use Among People Diagnosed With Cancer

Abstract

PURPOSE To establish thresholds in the Patient-Reported Outcomes Measurement Information System (PROMIS) pain interference, physical function, fatigue, and depression scores on the basis of their association with subsequent use of the emergency department (ED) or urgent care by people diagnosed with cancer. METHODS Retrospective data from 952 people seen at Henry Ford Cancer and insured through the Health Alliance Plan were analyzed using generalized linear mixed-effects models. The log odds of ED or urgent care use during 14 or 30 days after each patient-reported outcome (PRO) assessment were related to PRO scores, while adjusting for comorbidity, sociodemographic, and tumor characteristics. RESULTS Pain interference and physical function were associated with subsequent ED or urgent care visits, but fatigue and depression were not, and the results for 14- and 30-day visits were similar. Thresholds anchored in the likelihood of these visits differed according to cancer stage. For people with advanced cancer, a pain interference score of 60 or higher (odds ratio [OR] 3.75, [95% CI, 1.53 to 7.87]) and a physical function score lower than 40 (OR 2.94, [95% CI, 1.22 to 7.06]) produced the largest ORs with narrowest CIs for 30-day visits. For people with nonadvanced cancer, the thresholds of 65 for pain interference (OR 2.64, [95% CI, 1.40 to 5.01]) and 35 for physical function (OR 1.87, [95% CI, 1.01 to 3.45]) produced largest ORs with narrowest CIs for 30-day visits. CONCLUSION These anchor-based thresholds in PROMIS scores can inform clinicians' actions with the goal of preventing ED or urgent care visits.