Confirmatory factor analysis of the Patient Assessment of Constipation-Symptoms (PAC-SYM) among patients with chronic constipation

Abstract

Background and aim: PAC-SYM is widely adopted to asses constipation severity. However, it has been validated in a small sample, few items have been included based on expert opinion and not on empirical grounds, and its factor structure has never been replicated. We aimed at evaluating the psychometric properties of PAC-SYM in patients with chronic constipation. Methods: We enrolled 2,203 outpatients with chronic constipation in two waves. We used wave I sample to test the psychometric properties of the PAC-SYM and wave II sample to cross-validate its factor structure, to assess criterion validity, responsiveness to clinical change, and its minimal clinically important difference. Results: Only a minority of patients reported any rectal tearing (38 %). Deletion of such item leads to a 11-item version (M:PAC-SYM). The remaining items in the rectal domain were moderately correlated with the stool domain. Exploratory factor analysis and confirmatory factor analysis revealed a bifactor structure with two subscales (stool and abdominal symptoms) and a general severity factor. The M:PAC-SYM demonstrated excellent reliability, moderate correlation with SF-12 and treatment satisfaction (r = 0.28-0.45), discrimination across Rome III criteria for functional constipation and abdominal pain, and responsiveness to clinical change (β = -0.49; ω2 = 0.25). M:PAC-SYM minimal clinically important difference was 0.24. Conclusion: Our analysis shows that the rectal domain may not represent a relevant cluster of symptoms for patients with chronic constipation. We developed a modified version of the PAC-SYM which might better represent symptom severity of most patients seeking care in gastroenterology referral centers.