Patient Perception, Preference and ParticipationThe effect of information on preferences stated in a choice-based conjoint analysis
Introduction
In recent years, assessment of health care preferences has been promoted in health care decision making [1], [2], [3]. On a macro level, policy makers are interested in the values and preferences of the community to explain or predict the uptake of health care programs [4]. On a micro level the relevance of patient preferences in decision making is put forward in the models of shared and informed decision making [5], [6]. As a result the use and usability of preference elicitation techniques are becoming a domain of interest in health care.
A preference elicitation technique that is often used to evaluate the mode and effect of health care is a conjoint analysis (CA) [7]. A specific form of CA is a discrete choice experiment (DCE). In a DCE a subject is asked to choose the preferred health state, product or service from a set of two or more scenarios. The hypothetical scenarios are constructed from short statements (levels) on the key characteristics (attributes or criteria) of the health state, product or service. A subject is expected to weigh criteria importance and level attractiveness during the decision task. A set of part-worth utilities for the criteria levels is estimated from the observed choices of the subject. A part-worth utility is the value of a criterion level to the subject. More attractive levels have higher part-worth utility. With the part-worth utilities for all levels, the relative importance of decision criteria and the overall preference for treatment can be estimated [8], [9], [10].
In earlier studies some methodological issues were raised with regard to the application of discrete choice experiments [11], [12]. It is known that the framing of the scenarios can influence outcome [12], [13]. However, no previous studies were focused on the effect of a priori information on the outcome of a DCE. This is important, because although information is seen as a prerequisite for decision making, it is known that the order, type and framing of information can influence the way information is used to make real-life decisions [14], [15], [16]. Moreover, it was found that observed treatment preference is influenced by the information that is available to a patient [17]. It is unknown whether preferences elicited in a hypothetical situation, such as a DCE, are also influenced by the information that is available to a subject prior to partaking in the experiment. In DCEs, much attention is focused on the description of the scenarios by ensuring that relevant information is presented in a comprehensible way in the description of criteria levels [9]. It could be hypothesized that as much attention is required to determine how much and which information is presented prior to a DCE, if the outcome of such a study is influenced by the information which is available to a subject.
Therefore, the aim of the current study was to determine if informing subjects with the actual harms and benefits of treatment and the available treatment options in an informational brochure before participating in a DCE influences its outcome, i.e., the part-worth utilities of criteria levels, the importance of decision criteria, and the predicted preference for treatment.
The decision context in the study was the treatment of ankle and foot impairment in stroke. In stroke, a deviant position of the ankle and foot that hinders standing and walking is a common disability. Determining the best treatment in ankle and foot impairment is a value-based decision, as the evidence of the effect of the treatment alternatives on patient performance is limited [18], [19]. Surgical, technological and orthotic treatment alternatives are available, which differ widely in terms of impact of treatment to the patient, in comfort and cosmetics, and in the required use of walking aids or braces during and post-treatment. This makes the decision for treatment in ankle and foot impairment extremely suitable for a trading exercise such as a DCE.
Section snippets
Study design and procedure
The study was reviewed by the Human Subjects Ethics Review Board of the Roessingh Centre for Rehabilitation and was exempted from formal approval because it was a onetime experiment without emotional impact to subjects. A convenience sample of 80 bachelors and master students and 20 colleagues in the research department was approached for the study and agreed to participate. All subjects were familiar with health research, as they were involved in a health research project or in a health
Part-worth utilities
Some differences were found in the ranking of criteria and the order of criteria levels between the two groups (Table 1). The order of preference of the part-worth utilities in the criteria with a natural order was as expected for criterion 8 (“success rate”) in both groups. However, the expected order of preference was violated for criterion 1 in both groups and for criterion 5 in the minimally informed group (Table 1).
In the criteria with an expectation on order which was based on common
Discussion
In the current study significant differences in preference estimates were found between two groups of subjects. Preference differences were mainly observed in part-worth utilities. It seems that more extensive information resulted in a: (1) decrease in level order reversals in the criteria with a natural order; (2) higher acceptance of negative treatment aspects in favor of a more positive treatment result and (3) higher preference for ready-made shoes at the cost of ankle stability.
We
Acknowledgements
This research was supported by The Netherlands Organization for Health Research and Development ZonMw (grant number 143.50.026) and the VIDI-award to AM Stiggelbout of The Netherlands Organization for Scientific Research NWO Innovational Research Incentives (grant number 917.56.356). We would like to thank Rudolf de Ruiter for his contribution to this study and the reviewers of the manuscript for their valuable comments that helped to improve the quality of the manuscript.
References (25)
- et al.
Shared decision making in clinical medicine: past research and future directions
Am J Prev Med
(1999) - et al.
The use of conjoint analysis to elicit community preferences in public health research: a case study of hospital services in South Australia
Aust NZ J Public Health
(2000) - et al.
Using stated preference discrete choice modeling to evaluate health care programs
J Bus Res
(2004) - et al.
How do patients evaluate and make use of online health information?
Soc Sci Med
(2007) - et al.
Multimedia educational aids for improving consumer knowledge about illness management and treatment decisions: a review of randomized controlled trials
J Psychiatr Res
(2008) - et al.
The use of the analytic hierarchy process to aid decision making in acquired equinovarus deformity
Arch Phys Med Rehabil
(2008) - et al.
Enhancing informed consent for research and treatment
Neuropsychopharmacology
(2001) - et al.
Variability in patient preferences for participating in medical decision making: implication for the use of decision support tools
Qual Health Care
(2001) Partnerships with patients: the pros and cons of shared clinical decision-making
J Health Serv Res Policy
(1997)Health decision aids to facilitate shared decision making in office practice
Ann Intern Med
(2002)