Evaluating the Use of a Novel Patient-Reported Outcomes Measure in Cancer Care: A Pilot Study in Patients Receiving Radiation Therapy

Abstract

Patient safety is a cornerstone of high quality cancer care. However, current efforts rely on clinician-centric methods. In this IRB-approved pilot study, we evaluated the feasibility of an innovative approach to prospectively capture system safety data using patient-reported non-routine events (PNREs), defined as events that deviate from expected/optimal care. We postulated that safety-related PNREs will identify potential flaws in cancer care pathways that increase the likelihood of unplanned treatment events/errors (UTEs) and lower standardized experience measures. Eligibility included ECOG 0-2 adult patients with newly diagnosed head/neck & lung cancer scheduled to receive curative radiation (RT). PNREs, using our previously validated Patient-reported Comprehensive Open-ended Non-routine Event Survey (PCONES) tool, and patient-reported experience measures including: the EORTC-QLQ-30 (to compute a Global Health Status (GHS) on a 0-100 scale), the NCCN Distress Thermometer (DT, 0-10 scale), and 4 core ambulatory Consumer Assessment of Healthcare Providers and Systems (CAHPS) survey questions, were collected at each on-treatment visit. Additionally, physicians completed the NCI CTCAE v4.0 toxicity score and collected UTEs from the patients’ medical records. Propensity score weighting and regression methods were used to quantify associations between PNREs and patient experience. 20/24 patients (pts) who enrolled with 2 months f/u during RT comprise our study cohort. Mean age was 64.5±2.1 year (SEM); 60% male. Overall, 80 encounters occurred, median of 4 (range: 3-5). In 39% of encounters, at least one PNRE was reported yielding 40 total PNREs from 75% of the pts. 32.5% of PNREs were safety related while the remaining were related to pts’ care experience. In encounters when pts did not report PNREs, the GHS and DT measurements were 70.8±3.0 and 2.5±0.4, respectively. In cases where pts reported at least one PNRE, GHS decreased to 63.9±4.4 (p=0.182) and distress increased to 3.7±0.6 (p=0.088). The GHS was even lower (56.4±6.9) and distress higher (4.6±0.9) in pts who experienced safety-related PNREs. Three study pts each had one UTE (15% of participants), including a hospital readmission for surgical complications. 15.4% of the adult safety-related PNREs related to these UTEs. The 5 pts who never reported a PNRE experienced no UTEs. Encounters in which PNREs were reported were associated with lower CAHPS composite scores – no PNRE: 58% top-box ratings; ≥1 PNRE: 19% top-box ratings (p<0.001). This pilot study demonstrates the feasibility of collecting PNREs to identify potential systems safety risks in a radiation oncology clinic. Future efforts will include a large phase II trial to validate these findings, and to assess the value of PNREs in improving cancer care processes and patients’ outcomes.