Psychometric performance of the PAncreatic CAncer disease impact (PACADI) score

Abstract

Background/objective: Pancreatic Cancer Disease Impact (PACADI) score measures the impact of pancreatic cancer (PC) on important health dimensions, selected by patients. The aim of this single center study was to test the psychometric performance of the Pancreatic Cancer Disease Impact (PACADI) score.

Methods: Patients with suspected pancreatic cancer (PC) completed PACADI, the EuroQol-5D (EQ-5D index) and Edmonton Symptom Assessment System (ESAS) in this longitudinal observational study. Measures were compared across patients with PC (n = 210), other malignant lesions (OML) (n = 109) and non-malignant lesions (NML) (n = 41). Associations, test-retest and internal consistency reliability, longitudinal changes, sensitivity to change and prediction of mortality during the first year were examined in patients with PC.

Results: The three measures discriminated between PC and OML. The PACADI score correlated strongly at baseline (n = 199)/after three months (n = 85) with the EQ-5D index and ESAS "sense of well-being" (0.64 and 0.66/0.73 and 0.69, p < 0.001, respectively), showed high test-retest reliability (ICC 0.84) and very good internal consistency reliability (Cronbach's alpha 0.81-0.85) across all visits. Scores improved over time at 3, 6, 9 and 12 months for survivors, and standardized response mean (SRM) for improvement between 2 and 3 months (n = 44) was 0.80 (PACADI), -0.59 (EQ-5D index) and 0.69 (ESAS "sense of well-being"). The PACADI score significantly predicted mortality within the first year (p = 0.02) in contrast to the EQ-5D index and ESAS "sense of well-being".

Conclusion: This study showed satisfactory psychometric performance of the PACADI score. The results support its use in clinical practice and intervention trials.

Keywords: Cancer of pancreas; Outcome measures; Patient preferences; Validation studies.