Psychomtetric Evaluation of the Pediatric Constipation Assessment Scale (P-CAS) in Children with Cancer or Hematologic Disorders

Abstract

Background: Constipation is an under-recognized symptom in children receiving opioids and/or cancer treatments whose management requires accurate assessment yet few self-report measures exist. The purpose of this three-phase instrument testing study, guided by the Model for Constipation Across the Lifespan, was to revise and evaluate the psychometric properties of the Pediatric Constipation Assessment Scale (P-CAS), a self-report measure in children aged 8 to 17 years with cancer or hematologic disorders. The aims were to: (Phase 1) estimate content validity using an expert panel of healthcare professionals (HCPs) and children; (Phase 2) administer the P-CAS to children followed by cognitive debriefing interviews to evaluate responses and assist in revising the scale; and (Phase 3) administer the P-CAS to two groups of children with cancer or hematologic disorders at low and high risk for constipation to evaluate reliability and validity. Methods: In Phase 1, 12 HCPs and 12 children rated the relevancy of the P-CAS items and content validity indices were estimated at item- (I-CVI) and scale-level (S-CVI). In Phase 2, 12 children completed the P-CAS followed by recorded cognitive interviews which were analyzed using content analysis. In Phase 3, 25 children completed the P-CAS and analysis addressed reliability and validity. In all phases, descriptive profiles were compiled for participants. Results: I-CVI was .25 -1.0 for HCPs, .58-1.0 for children, and .50-1.0 combined. S-CVI was acceptable and with six items deleted, increased to .83 (HCPs), .91 (children), and .87 (combined). Cognitive interviews revealed potential sources of measurement error including variable time-frame recall and problems with item interpretation and summated rating selections. Alpha coefficients were acceptable in Phases 2 (.84) and 3 (.80). Nine items in Phase 2 and 8 in Phase 3 had item-total correlations of .30-.71. Conclusions: Results demonstrated evidence of content validity and internal consistency reliability. Variability in I-CVI across items demonstrated the importance of evaluating each item even though S-CVI was acceptable. Pre-testing accompanied by cognitive interviews was beneficial in identifying difficulties not recognized during content validation. Findings provide a foundation for further psychometric testing that will expand the ability to assess constipation in children with cancer/hematologic disorders.