OPIOID-INDUCED CONSTIPATION RATING SCALE (ORS)- A PSYCHOMETRIC REVIEW
Author(s)
Cheng R*1;Cole JC1;Chin K2;Li YP2, Hopkins A2 1Covance Market Access Services, Inc., San Diego, CA, USA, 2Theravance Inc., South San Francisco, CA, USA
OBJECTIVES: Existing measures of opioid-induced constipation (OIC) symptoms have not been developed in accordance with FDA’s 2009 Guidance for Patient-Reported Outcome (PRO) Measures. The OIC Rating Scale (ORS) was developed based on literature and comprehensive qualitative research. This study aimed to: 1) validate a conceptual framework of two domains; 2) determine the optimal scoring system; 3) examine item- and scale-level psychometrics; and 4) establish domain responder definitions (RD). METHODS: ORS was administered to 215 US adults with OIC in a Phase II study. Confirmatory factor analysis (CFA) was used to test the fit of the previously developed conceptual framework. The evaluation analyzed data from the 2-week baseline period and the 5-week treatment period. Item response theory (IRT) was used to determine an optimal scoring approach, followed by classical psychometrics including item- and scale-level reliability, and convergent validity. Distribution- and anchor-based approaches were used to determine the RD for each domain based on week 5 change from baseline. RESULTS: Based on descriptive statistics, item-level correlations, and CFA results (confirmatory fit index/non-normed fit index > 0.96, standardized root mean residual <0.04), a two-domain structure of the ORS was validated and carried forward in tests of reliability (αs > 0.85, test-retest ICCs > 0.60) and validity analyses (convergent rs 0.37-0.55), and to establish score interpretation. IRT results provided no compelling evidence requiring special item weights rather than a simple raw sum-score per domain. The two-item bowel movement symptom (BMS) domain RD = 2 and six-item gastrointestinal-related symptom (GRS) domain RD = 4. CONCLUSIONS: The ORS instrument demonstrated strong psychometric characteristics using analyses following the current FDA guidance for PRO measures. The two domains explained the correlations among the bowel movement and gastrointestinal-related symptoms in OIC. These results support further use of this instrument in studies of potential treatments for OIC.
Conference/Value in Health Info
2013-05, ISPOR 2013, New Orleans, LA, USA
Value in Health, Vol. 16, No. 3 (May 2013)
Code
PCN104
Topic
Patient-Centered Research
Topic Subcategory
Patient-reported Outcomes & Quality of Life Outcomes
Disease
Gastrointestinal Disorders, Oncology