Differential Item Functioning of the Patient Health Questionnaire-9 in Decompensated Cirrhosis.
Academic Article
Overview
abstract
INTRODUCTION: We examined whether the symptom expression of depression as assessed using the Patient Health Questionnaire-9 (PHQ-9) depression screening tool differs between patients with decompensated cirrhosis (DC) compared to primary care patients. METHODS: Study included 218 DC patients (91% Child-Pugh Class B/C) recruited from a liver transplant center and a real-world cohort of 436 outpatients from four primary care clinics in a large tertiary academic health system who completed the PHQ-9. We calculated positive screening rates for depression (PHQ-9 cutoff score of 10) for both cohorts. We evaluated PHQ-9 items for differential item functioning (DIF) in both cohorts within an Item Response Theory (IRT) framework. We compared DIF-adjusted and unadjusted IRT scores to characterize the impact of DIF on PHQ-9 total scores. RESULTS: Positive screening rates using a PHQ-9 cutoff score of 10 were 39% and 30% for DC and primary care patients respectively. Three PHQ-9 somatic symptom items (sleep problems, low energy, psychomotor agitation or retardation) showed significant DIF, with DC more likely than primary care patients with similar levels of depression severity to endorse these symptoms. DIF-adjusted scores suggested a one-point increase (PHQ-9 cutoff score of 11) in the screening threshold for DC patients. CONCLUSIONS: Equating for depression severity, we found differences in the symptom expression of depression for DC patients relative to primary care patients. Our findings highlight the need for future clinical and basic research into the diagnostic performance of depression screening tests and the phenomenology of depression in patients with DC.
