VALUE IN HEALTH 14 (2011) 1153–1157
Available online at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/jval
Elicitation of Informed General Population Health State Utility Values: A
Review of the Literature
Helen McTaggart-Cowan, PhD*
Canadian Centre for Applied Research in Cancer and British Columbia Cancer Agency, Vancouver, BC, Canada
A B S T R A C T
Objective: Within a health-care decision-making context, whose
health state utility values (HSUVs) should be used is constantly debated. This discussion is important because patient and general population utilities can differ. These discrepancies may be due to the general population not being informed about the health states. This article
investigates approaches used to inform the general population about
the health states that they are valuing. Methods: Studies reporting
methods to obtain informed general population HSUVs were identified,
outlined, and critically appraised. Results: Fourteen studies were identified: seven used information sessions, two used simulation methods,
two used reflection and deliberation techniques, and three used adaptation exercises. Conclusions: This review demonstrated a range of
Introduction
Cost-utility analysis (CUA) is increasingly being used to guide decision makers regarding the allocation of health-care resources. In
CUA, health benefits are quantified using quality-adjusted lifeyears (QALYs) [1], a measure that combines both duration and
quality of life (QOL) into a single index. QOL information is obtainable from respondents in the form of health state utility values
(HSUVs). Whether patients or members of the general population
should provide HSUV to be incorporated into a CUA is currently
under debate [2].
Although patients may be better at valuing their own health [3],
most agencies that use QALYs have advocated that HSUVs should
be obtained from the general population [4 – 6]. This is because
general population utilities are better suited to inform policy decisions in publicly funded health-care systems. Rational citizens,
operating behind a “veil of ignorance,” provide HSUVs unbiased by
specific interests. General population respondents are asked to
envision life in the health state and then provide a utility value for
it. A drawback to using general population HSUVs, however, is that
health state descriptions are brief. As such, respondents may not
fully comprehend life with the impaired health state, leading to
decisions that do not maximize societal health benefits.
The debate on whose HSUVs to use is important because previous studies have shown significant discrepancies between patient and general population utilities [7,8]. Generally, patient utilities tend to be higher than those of the general population’s. This
approaches to elicit informed general population HSUVs. The majority of the studies indicated that informing the respondents significantly affected their opinions of the health states and hence their
HSUVs. This suggests that the utilities that are currently used to
guide health-care resource allocation decisions may not represent
the general population’s preferences for specific health states. This
could result in decisions that do not maximize societal health
benefits.
Keywords: adaptation, general population, health states, informed
utilities, review.
Copyright © 2011, International Society for Pharmacoeconomics and
Outcomes Research (ISPOR). Published by Elsevier Inc.
may be a result of misinterpreting the same health state or incorporating differing levels of disease adaptation into the valuations;
for a further discussion, refer to Ubel et al. [9], Brazier et al. [10],
and Stiggelbout and de Voel-Voogt [11]. This difference can have
significant ramifications when incorporated into a CUA [9].
To ensure that general population utilities are relevant, HSUVs
provided by respondents informed about the health condition
may be a better approach [4,10,12]. The informing process allows
the respondents to incorporate all available information regarding
the health condition into their own assessment. Thus, the current
challenge is to refine the health state elicitation method such that
respondents are informed and able to formulate their HSUVs. This
article synthesizes the results from studies that informed the general population when valuing health states.
Methods
PubMed, EMBASE, and PsycINFO databases were searched to identify empirical studies that informed the general population about
the health states that they are valuing. The search was restricted
to articles in English dated 2009 or earlier. The search keywords
were general population, public, health state$, inform$, and utility. The
references of retrieved articles were searched to trace potentially
relevant articles.
Peer-reviewed published studies were selected for review if
they used written or other audiovisual techniques to describe the
* Address correspondence to: Helen McTaggart-Cowan, Canadian Centre for Applied Research in Cancer Control, 675 West 10th Avenue,
Vancouver, BC Canada V5Z 1L3.
E-mail: hcowan@bccrc.ca.
1098-3015/$36.00 – see front matter Copyright © 2011, International Society for Pharmacoeconomics and Outcomes Research (ISPOR).
Published by Elsevier Inc.
doi:10.1016/j.jval.2011.05.046
1154
Table 1 – Characteristics of the reviewed studies.
No.
Study
Design
Sample
Country of
study
Type
Size
Bramlett et al., 2006 [16]
SI
GP (r)
75
US
2
Briggs et al., 2008 [17]
SI
GP (c)
75
UK
3
Clarke et al., 1997 [18]
SI
GP (c)
39
Canada, US
4
Cunningham and Hunt,
2000 [19]
SI
GP (c)
57
UK
5
Happich and von
Lengerke, 2005 [20]
SI
GP (r)
210
6
Lee et al., 2000 [21]
SI
GP (r)
20
USA
7
Lenert et al., 2004 [22]
SI
GP (c)
620
USA
8
Aballéa and Tsuchiya,
2007 [23]
SI
GP (r)
38
Germany
UK
Pre- and
posttests
Health state valued
Valuation method
Information to enrich the health state descriptions
No
Mild stuttering
SG
TTO
Rating
Moderate stuttering
SG
TTO
Rating
Severe stuttering
SG
TTO
Rating
No
No
Stable schizophrenia
TTO
Relapsed schizophrenia
TTO
Yes (patient)
No
Gaucher’s disease (low blood
SG
counts)
TTO
Risk-risk trade-off
Gaucher ‘s disease (bone pain)
SG
TTO
Risk-risk trade-off
Gaucher ‘s disease (enlarged
SG
abdomen)
TTO
Risk-risk trade-off
Yes
No
Dentofacial deformities
SG
TTO
VAS
Yes (patient)
No
Tinnitus
SG
TTO
VAS
Yes (patient)
No
Mild schizophrenia
SG
VAS
Moderate schizophrenia
SG
VAS
No
No
Mild schizophrenia
SG
Moderate schizophrenia (type 1) SG
Moderate schizophrenia (type 2) SG
Severe schizophrenia (type 1)
SG
Severe schizophrenia (type 2)
SG
Severe schizophrenia (type 3)
SG
Severe schizophrenia (type 4)
SG
Extremely severe
SG
No
Simulation to reproduce symptoms of the health state descriptions
No
No
Mild visual impairment
TTO
Severe visual impairment
TTO
Postoperative vision
TTO
Mean informed HSUVs
Pretest
Posttest
Control
N/A
0.96
0.93
0.80
0.91
0.85
0.65
0.81
0.63
0.44
0.87
0.48
0.94
0.87
0.32
0.94
0.86
0.29
0.92
0.82
0.24
0.73
0.67
54.3
0.81
0.79
0.34
0.88
0.64
0.83
0.54
0.88
0.76
0.75
0.65
0.66
0.56
0.63
0.43
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
0.76
0.50
0.55
N/A
0.94
0.87
0.55
0.81
0.81
0.40
0.90
0.80
0.45
0.79
0.72
59.5
0.80
0.77
0.35
0.92
0.70
0.92
0.64
N/A
N/A
(continued on next page)
VALUE IN HEALTH 14 (2011) 1153–1157
1
Control
group
Table 1 (continued).
No.
Study
Design
Sample
Country of
study
Control
group
Pre- and
posttests
No
Type
Size
9
Czoski-Murray et al.,
2009 [24]
SI
GP (c)
104
UK
No
10
Shiell et al., 2003 [25]
SI
GP (r)
233
Australia
Yes
Health state valued
Reading limit
Legal blindness
Untreated age-related macular
degeneration
Reflection and deliberation techniques
Yes
EQ-5D state 11222
EQ-5D state 11213
Stein et al. 2006, [26]
SI
GP (c)
15
12
Damschroder et al.,
2005 [27]
SQ
GP (c)
359
13
Damschroder et al.,
2008 [28]
SI
GP (r)
1117
UK
No
USA
Yes
USA
No
Yes
Multiple sclerosis state 1
Multiple sclerosis state 2
Gaucher’s disease
EQ-5D state 1
EQ-5D state 2
EQ-5D state 3
Heart failure state 1
Heart failure state 2
Moderate eczema
Severe eczema
Osteoarthritis state 1
Osteoarthritis state 2
Osteoarthritis state 3
Osteoarthritis state 4
Exercises to evoke adaptation to the health states
No
Preexisting paraplegia
Yes
SQ
GP (c)
852
698
USA
Yes
Yes
Control
0.71
0.59
0.36
N/A
SG
TTO
SG
TTO
SG
SG
SG
SG
SG
SG
SG
SG
SG
SG
SG
SG
SG
SG
0.71
0.75
0.58
0.55
0.75
0.94
0.88
0.97
0.30
0.75
0.84
0.51
0.782
0.70
0.96
0.83
0.96
0.98
0.68
0.72
0.53
0.53
0.83
0.93
0.88
0.98
0.30
0.75
0.84
0.50
0.786
0.69
0.96
0.83
0.97
0.98
0.66
0.61
0.57
0.52
N/A
PTO
N/A
Below-the-knee amputation
SG
TTO
SG
TTO
SG
TTO
SG
TTO
VAS
VAS
Paraplegia
Ubel et al., 2005 [29]
Posttest
N/A
PTO
Colostomy
14
Pretest
TTO
TTO
TTO
New-onset paraplegia
Severe back pain
Paraplegia
Paraplegia
Mean informed HSUVs
N/A
47.0
50.6
Saving the lives of 100
patients ⫽ saving
the lives of 100
healthy people
Saving the lives of 102
patients ⫽ saving
the lives of 100
healthy people
0.76
0.89
0.73
0.83
0.64
0.81
0.63
0.79
51.6
58.5
Saving the lives of 100
patients ⫽ saving the
lives of 102 healthy
people
Saving the lives of 1000
patients ⫽ saving the
lives of 100 healthy
people
0.76
0.90
0.72
0.86
0.67
0.81
0.67
0.80
62.2
52.7
VALUE IN HEALTH 14 (2011) 1153–1157
11
Valuation method
EQ-5D, EuroQoL; GP (c), general population (convenient sample); GP (r), general population (representative sample); HSUVs, health state utility values; N/A, not applicable; PTO, person trade-off; SG,
standard gamble; SI, structured interview; SQ, self-completed questionnaire; TTO, time trade-off; VAS, visual analog scale.
1155
1156
VALUE IN HEALTH 14 (2011) 1153–1157
health states to the respondents. Studies were excluded if the respondents did not value distinct health states. After a critical review of titles and abstracts, 14 full-text articles met the inclusion
criteria. Extracted information included design, type of sample
and size, country of study, presence of a control group, pre- and
post-tests used, health state labels, valuation techniques used,
and mean informed HSUVs.
Results
Table 1 summarizes the identified empirical studies. Interventions
used to inform the general population were categorized into information to enrich the health state descriptions (n ⫽ 7); simulation
to reproduce the symptoms of the health state (n ⫽ 2); opportunity
to reflect and deliberate on the health state descriptions (n ⫽ 2);
and exercises to evoke adaptation to the health state (n ⫽ 3). More
information about the reviewed studies is available in Appendix A
found at doi: 10.1016/j.jval.2011.05.046.
The majority of the identified studies provided respondents
with additional information through the use of audio recordings
and videos. The drawback of this approach is that the respondents
are not able to properly envision themselves in the health state
that they are being asked to value. One method to overcome this
limitation is to use devices to enhance the effect of the health state
(e.g., spectacles to replicate different severities of visual impairment). Such simulation techniques can help the respondents to
better imagine life in the health state; however, they are not applicable to all disease states. Deliberation and adaptation techniques explicitly encourage respondents to consider broader life
aspects in relation to the health state. The deliberation approach
may replicate the way in which “real-life” decisions are made,
whereas adaptation exercises inform the respondents about the
possibility of adapting to an impaired health state over time.
Discussion
This review aimed to gain an understanding of the methods currently being used to inform the general population when valuing
health states. Conventionally, health states in valuation studies
tend to be described as varying levels of a limited number of different domains. They may not, however, provide enough sensitivity for general population respondents to familiarize themselves
with the health state and provide a meaningful valuation. By providing additional information, alongside the descriptions, focusing illusion effect—an inattention to broad life domains and overestimation of the emotional impact of an event [13]—may be
reduced. Specifically, general population respondents may disproportionately focus on certain activities that could be negatively
affected by the impaired health state and ignore other aspects of
life (e.g., personal and spiritual relationships) that may be unaffected by a changed state of health [14]. These influences could
cause respondents to under- or overestimate their HSUVs, biasing
the outcome of a CUA. Although informing the respondents does
not ensure that all start at exactly the same knowledge point, it
should eliminate complete unawareness of the health state.
This review’s ability to determine the most effective method to
inform the general population is limited because most studies did
not include either independent control groups or pre- and posttest measurements. Without these specific components, the impact of the information on HSUVs cannot be directly assessed.
Only seven studies included controls; of these, three used patients
as “controls”. Of the seven, only studies 10 and 14 included preand posttest measurements (Table 1). As a result, the impact of
informing the general population on their HSUVs cannot be generalized in this review. Although both of these studies revealed
that the informed HSUVs were not significantly different from
those of the controls, these are too few from which to draw a
meaningful conclusion.
The majority of studies had general population respondents
value disease-specific health states rather than generic ones. The
use of disease-specific states may enable respondents to better
imagine the impact of the health state on their QOL; however, this
prevents comparison across diseases. As a result, it would be more
beneficial to understand how to best inform them about generic
health states [4 – 6]. However, there is concern that the value currently attached to the QOL of these generic states may be determined by unengaged and uninformed members of the general
population. Future work could involve the recalculation of tariffs
(e.g., EuroQoL 5D) that are currently being used to generate population HSUVs for use in economic evaluation with respondents
who are informed about the health states that they are valuing to
assess whether there was a difference between the uninformed
and informed HSUVs. By doing this, wider use of the QALY as a
health outcomes measure may be established [15]. Another way to
inform the general population is to provide them with patient
HSUVs; this was not explored in any of the reviewed studies. The
drawback of this method is that patient HSUVs needs to be collected at different time points to describe a series of events, including at disease onset, during adaptation, and after a period of
time living with the condition [10]. For this approach, however, the
QALY model will need to be reconsidered. By calculating individual QALYs for each time point, duration and QOL can no longer be
regarded as utility independent. This could significantly affect the
standard practice of using valuation sets in economic evaluation
of health-care technologies.
Despite the recommendation to use informed general population values in decision making [4 – 6], this review demonstrates
that relatively few studies have aimed at informing the general
population about the health states that they are being asked to
value; this may be a result of there being no guidance on how to
construct HSUVs of this type. Due to limitations within the designs
of the reviewed work (i.e., absence of independent control groups
or pre- and posttest measurements), there is no clear indication of
which methods for informing respondents are appropriate and
whether these methods would vary for different disease states.
The majority of the reports, however, indicated that informing the
respondents significantly affected their opinions of the health
states and hence their HSUVs. This suggests that the utilities that
are currently used to guide health-care resource allocation decisions may not represent the general populations’ preferences for
the health states. This could potentially result in decisions that do
not maximize societal health benefits. Given these important policy implications, further research to develop an appropriate
method for informing general population values is needed.
Acknowledgments
The author thanks the editor and the two anonymous reviewers
for their valuable suggestions, as well as Aki Tsuchiya and Stuart
Peacock for comments on earlier versions of this article. At the
time that this review was conducted, the author was a doctoral
candidate at the University of Sheffield and was supported by a
doctoral research award from the Canadian Institutes for Health
Research. The usual disclaimer applies.
Supplemental Materials
Supplemental material accompanying this article can be found in
the online version as a hyperlink at doi:10.1016/j.jval.2011.05.046,
or if hard copy of article, at www.valueinhealthjournal.com/issues
(select volume, issue, and article).
VALUE IN HEALTH 14 (2011) 1153–1157
REFERENCES
[1] National Institute for Health and Clinical Excellence. Guide to the
methods of technology appraisal. London: National Institute for
Health and Clinical Excellence, 2008.
[2] Brazier J, Ratcliffe J, Salomon J, Tsuchiya A. Measuring and Valuing
Health Benefits for Economic Evaluation. Oxford: Oxford University
Press, 2007.
[3] Nord E, Pinto JL, Richardson J, et al. Incorporating societal concerns for
fairness in numerical valuations of health programs. Health Econ
1999;8:25–39.
[4] Gold MR, Siegal JE, Russell LB, Weinstein MC, eds. Cost-effectiveness
in Health and Medicine. Oxford: Oxford University Press, 1996.
[5] National Institute for Health and Clinical Excellence. Guide to the
Methods of Technology Appraisal. London: National Institute for
Clinical Excellence, 2008.
[6] Canadian Agency for Drugs and Technologies in Health. Guidelines for
the Economic Evaluation of Health Technologies: Canada. Ottawa:
Canadian Agency for Drugs and Technologies in Health, 2006.
[7] Sackett DL, Torrance GW. The utility of different health states as
perceived by the general public. J Chron Dis 1978;31:697–704.
[8] Peeters Y, Stiggelbout AM. Health state valuations of patients and the
general public analytically compared: a meta-analytical comparison of
patient and population health state utilities. Value Health
2010;13:306 –9.
[9] Ubel PA, Loewenstein G, Jepson C. Whose quality of life? A
commentary exploring discrepancies between health state evaluation
of patients and the general public. Qual Life Res 2003;12:599 – 607.
[10] Brazier J, Akehurst R, Brennan A, et al. Should patients have a greater role in
valuing health states? Appl Health Econ Health Policy 2005;4:201–8.
[11] Stiggelbout AM, de Voel-Voogt E. Health state utilities: a framework
for studying the gap between the imagined and the real. Value Health
2008;11:76 – 87.
[12] Fryback DG. Whose quality of life? Or whose decision? Qual Life Res
2003;12:609 –10.
[13] Schkade DA, Kahneman D. Does living in California make people
happy? A focusing illusion in judgements of life satisfaction. Psychol
Sci 1998;9:340 – 6.
1157
[14] Dolan P, Kahneman D. Interpretations of utility and their implications
for the valuation of health. Econ J 2008;118:215–34.
[15] Kind P, Elston Lafata J, Matuszewski K et al. The use of QALYs in
clinical and patient decisionmaking: issues and prospects. Value
Health 2009;12(Suppl. 1):S27–30.
[16] Bramlett RE, Bothe AK, Franic DM. Using preference-based measures
to assess quality of life in stuttering. J Speech Lang Hearing Res 2006;
49:381–94.
[17] Briggs A, Wild D, Lees M, et al. Impact of schizophrenia and
schizophrenia treatment-related adverse events on quality of life:
directly utility experience. Health Qual Life Outcomes 2008;6:105.
[18] Clarke AE, Goldstein MK, Michelson D, et al. The effect of assessment
method and respondent population on utilities elicited for Gaucher
disease. Qual Life Res 1997;6:169 – 84.
[19] Cunningham SJ, Hunt NP. A comparison of health state utilities for
dentofacial deformity as derived from patients and members of the
general population. Eur J Orthodon 2000;22:335– 42.
[20] Happich M, von Legerke T. Valuing the health state ‘tinnitus’:
differences between patients and the general public. Hearing Res 2005;
207:50 – 8.
[21] Lee TL, Ziegler JK, Sommi R, et al. Comparison of preferences for
health outcomes in schizophrenia among stake holder groups.
J Psychiat Res 2000;34:201–10.
[22] Lenert LA, Sturley AP, Rapaport MH, et al. Public preferences for
health states with schizophrenia and a mapping function to estimate
utilities from positive and negative symptom scale scores. Schizo Res
2004;71:155– 65.
[23] Aballéa S, Tsuchiya A. Seeing for yourself: feasibility study towards
valuing visual impairment using simulation spectacles. Health Econ
2007;16:537– 43.
[24] Czoski-Murray C, Carlton J, Brazier J, et al. Valuing condition-specific
health states using simulation contact lenses. Value Health 2009;12:
793–9.
[25] Shiell A, Hawe P, Fletcher M. Reliability of health utility measures and
a test of values clarification. Soc Sci Med 2003;56:1531– 41.
[26] Stein K, Ratcliffe J, Round A, et al. Impact of discussion on preferences
elicited in a group setting. Health Qual Life Outcomes 2006;4:22.
[27] Damschroder LJ, Zikmund-Fisher BJ, Ubel PA. The impact of
considering adaptation in health state valuation. Soc Sci Med 2005;61:
267–77.