P441 Conjoint analysis for quantification of inflammatory bowel disease patients’ preferences for outcome metrics
W. K. van Deen*1, D. Nguyen1, N. Duran1, E. Kane1, M. G. van Oijen2, D.W. Hommes1
1UCLA Centre for Inflammatory Bowel Diseases, Division of Digestive Diseases, University of California, Los Angeles, California, United States, 2Academic Medical Centre, Amsterdam, Department of Medical Oncology, Amsterdam, Netherlands
Incorporation of patient preferences in medical decision making, a process referred to as shared decision making, has been shown to improve patients’ satisfaction and patients’ knowledge. The importance of measuring and reporting outcomes of care delivery is increasingly recognised to be crucial for quality improvement. However, no method to quantify the incorporation of patient driven decisions in the care process is available. In this study we quantified inflammatory bowel disease (IBD) patients’ preferences for 3 disease outcomes: disease control (DC), quality of life (QoL), and productivity (Pr) using a choice based conjoint analysis (CBC).
IBD patients were recruited through e-mail and were asked to fill out an online CBC questionnaire. The questionnaire assessed current levels of DC, QoL, and Pr, and 10 CBC questions, in which patients were asked to choose 1 out of 2 scenarios with different levels of DC, QoL, and Pr. A hierarchical Bayes model was run to estimate the importance individual patients assigned to each of the 3 outcomes. A single patient-centred outcome metric was developed based on the individual weight assigned to the outcomes.
In total, 210 IBD patients were included with a median age of 40 years (range 20–83). Of these, 51% had Crohn’s disease; 46% ulcerative colitis; and 3% indeterminate colitis. Large variations in individual patients’ preferences were observed. On average, QoL was valued higher than DC and Pr, and for all 3 outcomes, increases from low to intermediate levels were felt to be more important compared with increases from intermediate to high levels. No clinical characteristics were shown to be associated with different preferences. Individual preference weighted scores were calculated and were shown to be significantly different from scores without individual weightings in patients with active disease.
We showed that CBC can be used to quantify individual patients’ preferences for different outcome metrics. These preferences can be used to quantify a single patient-centred outcome metric for IBD patients. Because measured outcomes are significantly different when weighted based on individual patients’ preferences, we propose that outcomes of care should be measured and rewarded accordingly.