Psychometric evaluation of a patient-reported outcome measure in pancreatic exocrine insufficiency (PEI)

Abstract

Background/objectivesPancreatic exocrine insufficiency (PEI) is commonly caused by chronic pancreatitis (CP) or cystic fibrosis (CF). There are no PEI-specific patient-reported assessments of symptoms and impacts. The PEI Questionnaire (PEI-Q) was developed through qualitative research with PEI patients and expert clinical input. This study evaluated the psychometric properties of the PEI-Q. Methods162 PEI patients (CF = 71 and CP = 91), 62 diarrhoea-specific irritable bowel syndrome (IBS-D) patients and 60 healthy controls completed the 26-item PEI-Q and the Gastrointestinal Quality of Life Index (GIQLI) at baseline. PEI patients completed the measures again two weeks later to assess the test-retest reliability of the PEI-Q. Analyses supported item reduction and scoring algorithm development, followed by psychometric evaluation. ResultsOver 90% of PEI patients completed at least 23 of the 26 items at baseline. Item responses and clinical relevance supported retention of 18 items. Factor analysis supported a three-factor solution (abdominal symptoms, bowel movements, impacts) with adequate model fit. PEI-Q scores had good internal consistency (Cronbach's alpha: 0.77–0.82) and test-retest reliability (ICC: 0.73–0.87). Correlations between PEI-Q and GIQLI supported convergent validity. Known-groups and receiver operating characteristic analyses demonstrated that PEI-Q scores discriminated (p < 0.001) between differing PEI severities, and PEI patients and controls. ConclusionsThe PEI-Q has good validity and reliability. Results indicate that the PEI-Q could be used to aid identification and diagnosis of PEI, assist in the management of patients already diagnosed with PEI, ensuring correct and optimum treatment as well as enhance patient-clinician communication.