To Register      SMDM Homepage

Tuesday, 19 October 2004

This presentation is part of: Poster Session - Clinical Strategies; Judgment and Decison Making


Ahmed M. Bayoumi, MD, MSc, St. Michael's Hospital, University of Toronto, Inner City Health Research Unit, Toronto, ON, Canada and William Dale, MD, PhD, University of Chicago, Department of Medicine, Section of Geriatrics, Chicago, IL.

We have developed and implemented ProSPEQT (Program to Survey Preferences by Evaluating Quality of Life Tradeoffs), a computer program for computer-assisted utility elicitation interviews. ProSPEQT consists of extensible markup language (XML) documents which are transformed into hypertext markup language (HTML). Health state scenarios are kept in a distinct XML document, allowing for rapid customization. ProSPEQT has modules for the standard gamble (SG), time trade-off (TTO), and rating scale (RS) methods. Visual aids include representation of probabilities using a circle with two arcs or a set of faces (SG), representation of timelines using horizontal bars (TTO), and representation of a rating scale using a vertical “feeling thermometer” (RS). We have applied ProSPEQT to research assessing utilities relevant to Human Immunodeficiency Virus (HIV) infection and prostate cancer. We describe the program’s development and initial experience with 102 HIV-positive participants who evaluated 7 health states, each with 3 methods. ProSPEQT has several features including: the ability to incorporate multiple-choice questionnaires; random ordering of states and methods; storing of results as XML files; recording of each step of an individual’s deliberation and the associated time; allowing respondents to retract responses; and incorporation of audio files. Additionally, ProSPEQT allows for many options including: the method for seeking utilities (bisection; ping-pong, or titration); fixed or random values (with range constraints) for initial utility values; control over fidelity of utility range response and allowance for progression from coarse to fine fidelity during an elicitation; and customization of presented survival times to patient-specific age, race, and sex categories using United States life tables. To answer SG, RS, and FT elicitations, participants took a median of 30 (interquartile range 12 to 88), 15 (7 to 54), and 28 (12 to 87) seconds, respectively. The initial SG utility elicitation took 136 (40 to 233) seconds compared to the final SG elicitation, which took 17 (9 to 48) seconds. In 2.8% of responses, participants retracted a response. ProSPEQT is a web-based program for utility elicitations that has multiple features, is broadly customizable across a wide range of health conditions, allows users to select a variety of options when constructing utility elicitation interviews, and can provide detailed information about individual participants’ selections during the interview process. Such programs may help to standardize and de-bias utility elicitations.

See more of Poster Session - Clinical Strategies; Judgment and Decison Making
See more of The 26th Annual Meeting of the Society for Medical Decision Making (October 17-20, 2004)