Patient-reported and clinical outcomes among patients with pancreatic cancer.

Abstract

e23200 Background: Pancreatic cancer is associated with poor survival, high symptom burden, and psychological distress. Conventional assessments such as performance status (PS) have relied on provider-generated data to evaluate the selection of treatment and prognosis. Patient-reported outcome measures (PROMs) represent a patient-centric representation of the patient experience and are increasingly applied as tools to help improve outcomes and quality of life. However, little is known about the correlation of PROMs with ECOG PS and clinical outcomes in pancreatic cancer. Methods: We performed a retrospective analysis of patients with pancreatic cancer seen at the Henry Ford Health System between 09/2020 and7/2023, using ICD codes. The NIH’s validated and standardized Patient-Reported Outcomes Measurement Information System was used to capture 4 core domains: fatigue, pain interference, physical function, and depression. Patient-level variables and disease-specific variables were obtained by chart review from EHR. Kruskal-Wallis tests for continuous variables and or Fisher’s exact tests for categorial variables were used to compare the different ECOG scores and PROM scores and patient-level and disease-specific variables. Results: 176 patients were analyzed, with a median age of 65, 58% were male. Most patients had Stage 1 32.9% followed by Stage 4 25.9%, stage 3 22.9%, 1 32.9%. The majority had an ECOG score of 1, followed by, 0, 2, and only 10 had an ECOG PS of 3. There was no statistically significant difference in PS scores according to smoking status, race, or AJCC Stage but differed by age (P = 0.0007). PS score was not significantly associated with PROMscores on depression, fatigue, or pain interference. However, increasing PS scores were associated with a significant increase in low physical function PROM scores (P < 0.0001). Conclusions: Clinician-assessed PS is a single assessment of the patient’s tolerance to therapies subject to physician bias; our study provides encouraging data on the association between PS and patients’ reported physical function. The other PROM domains did not provide additional meaningful information on the patient’s function although are part of clinical decision-making. [Table: see text]