Date of Graduation


Document Type


Degree Type



School of Pharmacy


Pharmaceutical Systems and Policy

Committee Chair

Michael J Smith


Preference-Based Measures of Health (PBMH) provide 'preference' or 'utility' weights that enable the calculation of quality-adjusted life years for the economic evaluations of interventions. The Diabetes Utility Index (DUI) was developed as a two-page, self-administered diabetes-specific PBMH that can replace expensive time-consuming interviews with patients to estimate their health state utilities. Inputs from theory, an existing diabetes-specific measure of quality of life, and statistical analyses were submitted to a clinical expert panel. After three rounds of pilot surveys (n1=52, n2=65, n3=111) at primary care clinics in Morgantown, WV, five attributes and severity categories for each attribute were finalized on the basis of the results of Rasch Analysis and consultations with the panel. The final attributes were: 'Physical Ability & Energy', 'Relationships', 'Mood & Feelings', 'Enjoyment of Diet', and 'Satisfaction with Management of diabetes'. The next step involved obtaining preferences for health states based on combinations of DUI attributes and severity levels from 100 individuals with diabetes, recruited from primary care and community settings in and around Morgantown, WV, in hour-long one-on-one interviews. These health states were anchor states, single-attribute level states including corner states, and marker states. The interviews provided data to calculate a Multi-Attribute Utility Function (MAUF) that calculates utilities for any of the 768 health states that can be defined by the DUI, on a scale where 1.00=Perfect Health and 0.00=the all worse 'Pits' state, from respondents' answers to its five questions. In addition to an overall index score, attribute-level preference scores were also calculable by the function. Finally, a validation survey was conducted in collaboration with the West Virginia University (WVU) Diabetes Institute. For concurrent and construct validation purposes, the DUI was mailed to individuals with diabetes along with generic PBMH like the EuroQol EQ-5D, the SF-6D and other patient-reported outcomes measures like the Diabetes Symptoms Checklist-Revised, the Short Form 12 (SF-12) and the Well-Being Questionnaire (W-BQ12), and their surveys responses (n=396) were merged with a clinical database consisting of ICD-9 diagnosis codes. The DUI utilities were found to be largely free of socio-demographic effects and its scores were well distributed between 0.00 and 1.00. The DUI moderately correlated with generic PBMH and distinguished between severity groups based on diabetes symptoms and complications. The scoring function of the DUI calculated utilities favorably compared against cardinal Standard Gamble utilities obtained directly from patients for three DUI health states. These results show evidence of the feasibility and validity of the DUI. Further research is suggested to demonstrate the generalizability of these findings, to study the responsiveness of the DUI, and to examine the clinical meaningfulness of the DUI change scores.