Received November 29, 2006
Accepted on February 26, 2007
Estimating Preference Scores in Conventional and Home Nocturnal Hemodialysis Patients
Philip A. McFarlane *1,
Andreas Pierratos 
,
Ahmed M. Bayoumi 
,
and
Donald A. Redelmeier 
*Home Dialysis and Department of Medicine, Inner City Health Research Unit, St. Michael’s Hospital, Home Dialysis, Humber River Regional Hospital, and Department of Clinical Epidemiology, Canada Research Chair in Medical Decision Sciences, Sunnybrook and Women’s College Health Sciences Centre, Toronto, Ontario, Canada
1 To whom correspondence should be addressed. E-mail: phil.mcfarlane{at}utoronto.ca.
 |
Abstract |
|---|
Previous studies have reported higher quality of life in patients who receive home nocturnal hemodialysis (HNHD) than conventional in-center hemodialysis (IHD). The optimal method for eliciting preferences from dialysis patients remains undefined, and there may be unique methodologic concerns in this population. Patients’ preferences for IHD (n = 20) and HNHD (n = 24) were studied using the standard gamble (SG), time trade-off (TTO), and modified willingness to pay (WTP) methods. This report describes experience with operationalizing these three techniques in this population. A higher preference for HNHD was found with all measures, with significant differences observed with the SG (HNHD: median 0.79 [interquartile range (IQR) 0.67 to 0.95]; IHD: median 0.60 [IQR 0.20 to 0.82]; P = 0.031) and WTP (HNHD: median 0.50 [IQR 0.40 to 0.68]; IHD: median 0.20 [IQR 0.20 to 0.38]; P < 0.001). SG and TTO scores were moderately correlated but not with WTP. In addition, qualitative issues arose during TTO and WTP interviews that seemed to influence the interpretation of these preference scores. In the TTO, time willing to trade became oriented toward the next pivotal life event, with a failure of the requirement for a constant proportional time trade-off. WTP preferences were oriented toward the smallest survival stipend. These issues represent range restriction biases. No significant issues arose during the SG interviews. HNHD patients expressed a greater preference for current health than IHD patients. The operational performance of SG was good in this study, whereas biases and methodologic concerns were identified with the TTO and WTP in this population.
