608P - Patient and physician preferences for metastatic colorectal cancer treatments: A discrete choice experiment

Date 29 September 2014
Event ESMO 2014
Session Poster Display session
Topics Bioethics, Legal, and Economic Issues
Colon Cancer
Rectal Cancer
Presenter Juan Marcos Gonzalez
Citation Annals of Oncology (2014) 25 (suppl_4): iv167-iv209. 10.1093/annonc/mdu333
Authors J.M. Gonzalez1, S. Ogale2, R. Morlock3, J. Posner1, B. Hauber1
  • 1Health Preference Assessment, RTI Health Solutions, 27709 - Research Triangle Park/US
  • 2Genentech, Genentech, San Francisco/US
  • 3Health Economics And Outcomes Research, Genentech, Inc., 94080 - San Francisco/US

Abstract

Aim

To quantify the relative importance of benefits and risks of metastatic colorectal cancer (mCRC) treatments among patients and physicians.

Methods

Patients with mCRC and physicians treating patients with mCRC in the United States completed an online discrete-choice experiment survey including questions with pairs of hypothetical mCRC treatments. Respondents were asked to choose their preferred treatment in each question. Treatments were described using 5 outcomes identified through a survey of 5 clinical experts (attributes) and the degree to which treatments induce those outcomes (levels). The attributes in the survey were: progression-free survival (PFS), chance of a severe skin rash, chance of serious bleeding, chance of cardiac arrest, and chance of gastrointestinal (GI) perforation. Chance of serious infusion reactions (SIR) was not included as a study attribute because it comprises multiple possible outcomes. Cardiac arrest was the most severe type of infusion reaction identified by clinical experts and was thus included instead of SIR. Using results from a main-effects random-parameters logit model, attribute importance was calculated as the expected change in treatment choice induced by varying outcomes between clinically relevant levels.

Results

127 patients and 150 physicians completed the survey. Given the clinically-relevant levels for each attribute (PFS: 6-10 months; chance of a severe skin rash: 0-20%; chance of serious bleeding: 0-5%; chance of cardiac arrest: 0-2%; chance of GI perforations: 0-2%) patients considered PFS to be most important, followed by chance of a severe skin rash, serious bleeding, cardiac arrest, and GI perforation. For physicians the most important attribute was PFS, followed by chance of cardiac arrest, severe skin rash, serious bleeding and GI perforation.

Conclusions

Patients and physicians agree that PFS is the most important attribute of treatment. Physicians and patients had different rankings for risks suggesting the potential need for additional communication between patients and physicians regarding treatment-related risks.

Disclosure

J.M. Gonzalez: Genentech Inc provided funding for this study to RTI-HS. Juan Marcos Gonzalez is an employee of RTI HS. The publication of study results was not contingent on Genentech's approval. RTI HS maintained independent scientific control of the study; S. Ogale: is a full time employee and stock owner of Roche/Genentech; R. Morlock: was paid by Genentech at the time of this work and is not a stock owner of Roche/Genentech; J. Posner: Genentech Inc provided funding for this study to RTI-HS. Joshua Posner is an employee of RTI HS. The publication of study results was not contingent on Genentech's approval. RTI HS maintained independent scientific control of the study; B. Hauber: Genentech Inc provided funding for this study to RTI-HS. B. Hauber is an employee of RTI HS- The publication of study results was not contingent on Genentech's approval. RTI HS maintained independent scientific control of the study.