Incorporating process utility into quality adjusted life years: a systematic review of empirical studies.

OBJECTIVE: This review aimed to identify published studies that provide an empirical measure of process utility, which can be incorporated into estimates of QALY calculations.
METHODS: A literature search was conducted in PubMed to identify published studies of process utility. Articles were included if they were written in the English language and reported empirical measures of process utility that could be incorporated into the QALY calculation; those studies reporting utilities that were not anchored on a scale of 0 representing dead and 1 representing full health were excluded from the review.
RESULTS: Fifteen studies published between 1996 and 2012 were included. Studies included respondents from the USA, Australia, Scotland and the UK, Europe and Canada. Eight of the included studies explored process utility associated with treatments; six explored process utility associated with screening procedures or tests; and one was performed in preventative care. A variety of approaches were used to detect and measure process utility: four studies used standard gamble techniques; four studies used time trade-off (TTO); one study used conjoint analysis and one used a combination of conjoint analysis and TTO; one study used SF-36 data; one study used both TTO and EQ-5D; and three studies used wait trade-off techniques. Measures of process utility for different drug delivery methods ranged from 0.02 to 0.27. Utility estimates associated with different dosing strategies ranged from 0.005 to 0.09. Estimates for convenience (able to take on an empty stomach) ranged from 0.001 to 0.028. Estimates of process utility associated with screening and testing procedures ranged from 0.0005 to 0.031. Both of these estimates were obtained for management approaches to cervical cancer screening.
CONCLUSION: The identification of studies through conventional methods was difficult due to the lack of consistent indexing and terminology across studies; however, the evidence does support the existence of process utility in treatment, screening and preventative care settings. There was considerable variation between estimates. The range of methodological approaches used to identify and measure process utility, coupled with the need for further research into, for example, the application of estimates in economic models, means it is difficult to know whether these differences are a true reflection of the amount of process utility that enters into an individual's utility function, or whether they are associated with features of the studies' methodological design. Without further work, and a standardised approach to the methodology for the detection and measurement of process utility, comparisons between estimates are difficult. This literature review supports the existence of process utility and indicates that, despite the need for further research in the area, it could be an important component of an individual's utility function, which should at least be considered, if not incorporated, into cost-utility analyses.