COMPARISON OF MEASURES OF UTILITY IN PATIENTS UNDERGOING TOTAL SHOULDER ARTHROPLASTY

Abstract

Health utilities indicate the value of a given health state. They are essential components of decision analyses, and economic evaluations. In the area of total shoulder arthroplasty (TSA) we were unable to find literature estimating changes in utilities or the effect of method of obtaining utilities. The purpose of this pilot study was to describe the trajectory of utility scores before and after TSA using three approaches: the EQ-5D and the Health Utilities Index (HUI2 and HUI3) self-report format.

Twenty-four patients undergoing TSA at two teaching hospitals (Boston and Toronto) were assessed twice preoperatively, as well as at four and twelve weeks follow-up by self report mailed survey. At each survey all three utility estimates were obtained. Demographic and functional status was also gathered. The EQ-5D is a five item questionnaire which scores into a profile to which utility weights obtained from the developers were applied. The HUI self-report is a fifteen item scale obtaining a score on eight domains. A multiplicative formula is used to assign utility weights to these responses. Descriptive analysis of the sample, baseline characteristics and change in utility were completed. Intra-class correlation coefficients were used to calculate test-retest reliability between the two preoperative visits. Standardised response means (SRM) (mean change/SD of change) and relative efficiency (RE=ratio of SRM2) were calculated. Individual trajectories of change were graphed and examined for trends.

Twenty-four patients participated with average age of sixty-seven years, 58% were female and experienced large improvements in disability and pain (mean change DASH = 18.9/100, SPADI Pain = 30.3/100). Utility scores had low to moderate correlations with each other (0.26–0.68). Mean baseline scores were low (EQ5D=0.44, HUI2=0.68, HUI3 = 0.50). The average change in utility is shown in the following table along with effect size estimations and test-retest reliability.

Patients experience clinically important and statistically significant changes in their utility values even in the early stages of recovery after TSA (three months). The HUI3 and EQ-5D were most responsive to changes experienced in this sample.