Published online before print January 13, 2005, doi: 10.1161/01.STR.0000152950.46598.f1
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
(Stroke. 2005;36:303.)
© 2005 American Heart Association, Inc.
Original Contributions |
Preference-Based Quality of Life in Patients With Cerebral Aneurysms
From the Section of Neurosurgery (J.T.K.), VA Connecticut Healthcare System, West Haven, Conn and the Department of Neurosurgery (J.T.K.), Yale University, New Haven, Conn; the Section of Outcomes Research (J.T.), Division of General Internal Medicine, Department of Internal Medicine, University of Cincinnati Medical Center, Cincinnati, Ohio, the Center for Clinical Effectiveness (J.T.), Institute for Health Policy and Health Services Research, Cincinnati, Ohio, and Veterans Affairs Medical Center (J.T.), Cincinnati, Ohio; the Center for Research on Health Care (M.S.R.), Section of Decision Sciences and Clinical Systems Modeling (M.S.R.), Division of General Internal Medicine, Department of Medicine, Division of General Internal Medicine (M.S.R.), University of Pittsburgh, Pittsburgh, Pa.
Correspondence to Dr Joseph T. King Jr, Section of Neurosurgery, VA Connecticut Healthcare System/112 950 Campbell Ave, West Haven, CT 06516. E-mail Joseph.KingJr{at}med.va.gov
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionConclusionsReferences |
|---|
Background and Purpose— Functional outcome scales are typically used to measure quality of life (QOL) and outcomes in patients with cerebral aneurysms; however, these instruments only examine a limited number of domains that contribute to QOL. An alternative are preference-based QOL methods, which integrate all factors contributing to QOL and provide a comprehensive individualized measure of how patients value their current health state. An additional advantage of preference-based QOL values is that they can be incorporated into decision analyses and cost-effectiveness analyses.
Methods— We used 4 preference-based QOL methods to measure QOL in 176 outpatients with cerebral aneurysms: (1) standard gamble; (2) time trade-off; (3) visual analogue scale; and (4) willingness to pay. We measured functional status with the Glasgow Outcome Scale (GOS), Rankin Scale, and Barthel Index. We then built multivariate linear regression models to examine the relationships between preference-based QOL, functional status, and patient characteristics.
Results— Preference-based QOL was moderately diminished in the aneurysm patients. Mean values were: standard gamble, 0.78; time trade-off, 0.79; visual analogue scale, 0.67; and willingness to pay, $121 000. Preference-based QOL was not well-explained by functional status or patient characteristics, as shown by regression models that accounted for <15% of the variation in preference-based QOL (R2<0.15).
Conclusions— Preference-based QOL instruments capture components of QOL in patients with cerebral aneurysms not assessed by functional status measures or patient characteristics. Studies of patients with cerebral aneurysms should consider incorporating preference-based QOL measures for a fuller evaluation of the impact of aneurysmal disease and its treatment on QOL.
Key Words: intracranial aneurysm • neurosurgery • outcome • quality of life • subarachnoid hemorrhage
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionsReferences |
|---|
Cerebral aneurysms can adversely affect quality of life (QOL) via subarachnoid hemorrhage (SAH), mass effect, or complications of treatment. QOL can be measured by several classes of instruments, including functional status instruments, health status instruments, and preference-based techniques.1 Functional status instruments usually assign patients a single score based on their ability to perform activities of daily living. Common functional scales used in patients with cerebral aneurysms include the Glasgow Outcome Scale (GOS),2 Rankin Scale,3 and the Barthel Index.4 Health status instruments (eg, Medical Outcomes Study Short Form-36 [SF-36]5) are based on a conceptual model that subdivides QOL into health domains and uses questionnaire responses or patient characteristics to assign a score for each domain. Preference-based QOL instruments, also known as utilities or health value measures, elicit patients’ valuations for their current health state expressed on a single 0-to-1 ratio scale. Preference-based measurements can differ markedly from functional status or health status assessments. For example, aneurysmal SAH may leave 2 patients with identical fine motor control deficits in their nondominant hand and similar functional status or health status. The deficit might have a minor effect on the preference-based QOL of a manual laborer, whereas it might be devastating to a professional musician. Preference-based QOL instruments such as the standard gamble,6 time trade-off,7 visual analogue scale,8 and willingness to pay9 would be sensitive to this difference.
Preference-based QOL measures are increasingly common in the medical literature and offer several advantages over functional status or health status measures. They incorporate individual attitudes toward functional status, pain, disability, and the like, and integrate these attitudes proportionate to their importance to each patient. Their valuations can be used in cost-effectiveness analyses and decision analyses.1 Despite these advantages, to date preference-based QOL instruments have received little attention in studies of patients with cerebral aneurysms. We assessed QOL in patients with cerebral aneurysms using 4 preference-based measures (standard gamble, time trade-off, visual analogue scale, and willingness to pay) and functional status using 3 functional scales (GOS, Rankin Scale, and Barthel Index). We then explored the relationships between preference-based QOL, functional status, and patient characteristics.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionsReferences |
|---|
Study Population
We studied patients with cerebral aneurysms at the University of Pittsburgh Medicial Center neurosurgery clinics recruited between June 2001 and February 2004. After obtaining informed consent, the patients underwent a structured interview and testing administered by a trained research assistant to obtain data on demographics, habits, comorbid diseases, functional status, and preference-based QOL. Additional data were abstracted from paper and electronic medical records. The protocol was approved by the institutional review boards of the University of Pittsburgh and Yale University. Patients received $25 after completing the interview.
Preference-Based QOL Assessment
We used the standard gamble,6 time trade-off,7 visual analogue scale,8 and willingness to pay9 to measure subject’s valuations of their current health. Health state valuations were obtained during interviews with a research assistant using a script and a portable computer. iMPACT310 interactive graphical software was used for standard gamble and time trade-off testing; a written survey was used for visual analogue scale testing; and willingness to pay was assessed with a custom Visual Basic program. Standard gamble, time trade-off, and visual analogue scale results are scored on a scale ranging from zero, corresponding to the value of death, to one, corresponding to the value of perfect health. In the standard gamble and time trade-off, patients make hypothetical choices involving a risk of death or a reduction in survival time, and QOL values are calculated from their responses. During standard gamble testing, patients choose between remaining in their current health state or accepting the results of a gamble with 2 possible outcomes—death or perfect health (Figure 1a). In the time trade-off, patients exchange a portion of their future survival time in exchange for perfect health during their shortened life span (Figure 1b). The visual analogue scale requires patients to place a mark on a line to rate their current health state (Figure 1c). Willingness to pay determines how much a patient would pay for a hypothetical cure for all of their health problems (see King et al11 for more details on the preference-based QOL measurement techniques).
|
Functional Status: GOS, Rankin Scale, and Barthel Index
Patient interview responses and a computerized scoring algorithm were used to classify patients on the GOS2 and a modification of the Rankin Scale,12 and to calculate the modified Barthel Index.13 Note that the study subjects were recruited from among patients well enough to travel to a neurosurgery clinic, thus patients with the lowest levels of functioning (GOS=2, persistent vegetative state; GOS=1, death; Rankin=5, severe disability, bedridden; Rankin=6, death) were not included in the analyses.
Data Analysis
Categorical variables were tabulated, and means, standard deviations, medians, and quartiles were calculated for continuous variables. Monetary values were converted to 2003 US dollars using the US Urban Consumer Price Index.14 Comorbid diseases were tabulated and counted. The relationships between preferences (standard gamble, time trade-off, visual analogue scale, and willingness to pay) versus patient characteristics and functional status (GOS, Rankin, and Barthel scales) were assessed with the Mann-Whitney U test for categorical variables, box plots and Cuzicks nonparametric test for trend15 for ordinal variables, and scatter plots and Spearman rank correlation for continuous variables. Associations among the standard gamble, time trade-off, visual analogue scale, and willingness to pay were examined with the Spearman rank correlation.
We used stepwise linear regression to model the association between preference-based QOL, functional status, and patient characteristics. Predictor variables assessed included age, sex, race, income (willingness to pay only), number of comorbid diseases, history of SAH, number of aneurysms, previous aneurysm treatment, presence of unsecured aneurysms, GOS score, Rankin score, and Barthel Index. Simple linear regression and a threshold P<0.20 were used to select candidate predictor variables for inclusion in the multivariate models. In the final multivariate models, P<0.05 were considered statistically significant.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionsReferences |
|---|
Study Population
Two hundred seventeen eligible patients consented to participate in the study, and 176 patients (81%) completed the standard gamble, time trade-off, visual analogue scale, willingness to pay, and enough questions to allow the determination of functional status measured by the GOS, Rankin Scale, and Barthel Index. Incomplete data collection was caused by errors in survey completion, research staffing issues, and patient time constraints. There were no significant differences between the study population and the 41 patients with incomplete data collection in their age, sex, race, education, history of SAH, number of aneurysms, presence of an unsecured aneurysm, or previous aneurysm treatments (for all, P
0.075). The mean±SD patient age was 54.1±12.8 years, 73% of the patients were women, 52% of patients had a SAH, and 29% had multiple aneurysms (Table 1).
|
Preference-Based QOL: Standard Gamble, Time Trade-Off, Visual Analogue Scale, Willingness to Pay
Preference-based QOL values measured with the time trade-off, standard gamble, visual analogue scale, and willingness to pay had distributions skewed toward the value of perfect health (Figure 2). The mean QOL values for the study population were: standard gamble, 0.78; time trade-off, 0.79; visual analogue scale, 0.67; and willingness to pay, $121 000 (Table 2). The willingness to pay value was the equivalent of 2.9-times the mean annual household income. There were significant associations between the standard gamble and the time trade-off, visual analogue scale, and willingness to pay (for all, P
0.042), and between the time trade-off and visual analogue scale (P=0.001) (Table 3).
|
|
|
Functional Status: Glasgow Outcome Scale, Rankin Scale, and Barthel Index
Patients were assigned scores on the GOS, Rankin Scale, and Barthel Index based on self-reported capabilities. All of the scales showed a marked "ceiling effect," whereby >70% of subjects were clustered at the top of the scales (Figure 3).
|
Multivariate Linear Regression Models of Preference-Based QOL
In general, multivariate analyses showed that patient characteristics, aneurysm characteristics, and functional status were able to explain only a small part of the variation in QOL measured with the standard gamble, time trade-off, visual analogue scale, and willingness to pay (Table 4). In fact, none of the variables was a significant predictor of standard gamble values. On the time trade-off, nonsmokers (P=0.031) had higher QOL, but the R2 of 0.03 indicates that smoking history scores explained only 3% of the variation in time trade-off QOL values. Fewer comorbid diseases (P=0.004) and a better Rankin score (P=0.004) were independently associated with better QOL measured with the visual analogue scale (R2=0.11; F <0.001). On the willingness to pay measure, higher income (P<0.001) was a significant independent predictor of higher willingness to pay values (R2=0.14; F <0.001). There were no independent associations between preference-based QOL and age, sex, race, education, history of SAH, number of aneurysms, previous aneurysm treatment, the presence of an unsecured aneurysm, GOS, or Barthel Index.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionsReferences |
|---|
We measured preference-based QOL in 176 patients with cerebral aneurysms using 4 different instruments: standard gamble, time trade-off, visual analogue scale, and willingness to pay. The mean QOL values were: standard gamble, 0.78; time trade-off, 0.79; visual analogue scale, 0.67; and willingness to pay, $121 000. To put these values in another perspective, our patients with aneurysms on average were willing to undertake a 22% risk of death to be restored to perfect health (standard gamble), were willing to exchange 21% of their remaining life expectancy to obtain perfect health (time trade-off), rated their current health at only 67% of perfect health (visual analogue scale), and were willing to pay 2.9-times their annual income to obtain perfect health (willingness to pay). These values are similar in magnitude to those measured in patients afflicted with neurologic or musculoskeletal diseases that affect functioning, such as minor stroke (standard gamble, 0.72)16 rheumatoid arthritis (time trade-off, 0.77),17 early amyotrophic lateral sclerosis (standard gamble, 0.79),18 or cervical spondylotic myelopathy (time trade-off, 0.75).11 The regression analyses showed only weak and inconsistent associations between preference-based QOL measures and patient characteristics or functional status. Thus, preference-based QOL methods measure unique aspects of QOL not captured by more conventional measures.
Preference-based QOL measures have seen little use in the study of cerebral aneurysms despite some unique advantages. First, preference-based instruments integrate all aspects of QOL, because subjects can incorporate all health domains that contribute to their QOL. Health status instruments are confined to specific activities or health domains, and functional status instruments focus on physical functioning to the exclusion of other domains. Second, preference-based instruments allow for individual differences in priorities of numerous factors that contribute to quality of life, whereas health status and functional status measures use standardized formulae to calculate scores. Third, the values from 2 of the preference-based instruments—the standard gamble and the time trade-off—can be incorporated into decision analyses and combined with economic data in cost-effectiveness analyses. Because they do not involve explicit trade-offs, visual analogue scale scores cannot be used interchangeably with health values derived from the standard gamble or time trade-off, and thus cannot be incorporated into decision analyses and cost-effectiveness analyses.19 Decision analyses and cost-effectiveness analyses use mathematical models of disease and treatment that incorporate clinical outcomes data and preference-based QOL values to compare treatment strategies. The outputs from such analyses can inform clinicians, patients, and policy-makers during decision-making.
One of the logistical challenges of collecting preference-based QOL data such as the standard gamble and time trade-off has been the need for a face-to-face interview to gather the data. The development of interactive software permitting computerized self-administration has simplified data collection.10 Telephone administration of the standard gamble and time trade-off during follow-up testing once individuals have had an initial face-to-face interview is also a reliable technique.20 Recent work with the "paper gamble," a self-administered paper survey version of the standard gamble, has shown a high correlation with computer-administered standard gamble results21 and has demonstrated excellent test-retest reliability.22 Hopefully, these developments and others will increase the feasibility of preference-based QOL assessment.21 Collection of visual analogue scale and willingness to pay data are much simpler—visual analogue scale data can be collected via a written survey, and willingness to pay data can be collected via telephone interview or written survey.
In contrast, functional status measures are much easier to use than preference-based QOL instruments. Data can be conveniently and inexpensively collected via face-to-face interview, telephone interview, postal survey, or from proxy respondents (ie, spouses, family members, or caretakers). It may be possible to extract functional status data from medical records, allowing retrospective research studies based on existing medical records. These advantages are offset by a major drawback, although functional status instruments are user-friendly and investigator-friendly, the regression models show that these measures fail to explain most of the variation in preference-based QOL in patients with cerebral aneurysms. This finding makes sense, because many factors besides functional status contribute to QOL. Studies that use functional status measures to assess the impact of disease or treatment on QOL may not be measuring the factors that are most important to patients. These unmeasured factors can be captured by preference-based QOL instruments.
Functional status and patient characteristics were only able to explain a small amount of the variation in preference-based QOL values. Sporadic predictors of QOL included functional status, comorbid disease, cigarette smoking, and income. The relationship between decreased functional status, comorbid disease, and worse QOL is fairly intuitive. More vexing is trying to understand the relative contributions of the condition of interest (ie, cerebral aneurysms) and other comorbid diseases to global QOL. Several methods have been proposed for determining the condition-specific contributions to QOL;23,24 however, much work remains to be performed on this important and challenging issue. The association between cigarette smoking and lower QOL measured with the time trade-off may be because of the protean detrimental effects of cigarette smoke. Finally, it is not surprising that individuals with more economic resources are willing to pay more for a cure.11
|
Conclusions |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionsReferences |
|---|
Preference-based QOL instruments such as the standard gamble, time trade-off, visual analogue scale, and willingness to pay allow investigators to assess how patients with cerebral aneurysms value their current health. These valuations document the impact of cerebral aneurysms on QOL from the perspective of the patient. The QOL values from preference-based instruments are not well-predicted by patient characteristics or functional status. Preference-based QOL instruments provide unique insights into QOL in patients with cerebral aneurysms not captured by functional status measures. Studies of patients with cerebral aneurysms should consider incorporating preference-based QOL measures for a fuller understanding of the impact of aneurysms and their treatment on QOL.
Received October 7, 2004; accepted November 3, 2004.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionsReferences |
|---|
1. McDowell I, Newel C. General health status and quality of life. Measuring Health. A Guide to Rating Scales and Questionnaires, 2nd ed. New York: Oxford University Press; 1996: 380–492.
2. Jennett B, Bond M. Assessment of outcome after severe brain damage. A practical scale. Lancet. 1975; 1: 480–484.[Medline] [Order article via Infotrieve]
3. Rankin J. Cerebrovascular accidents in patients over the age of 60: II. Prognosis. Scot Med J. 1957; 2: 200–215.[Medline] [Order article via Infotrieve]
4. Mahoney FI, Barthel DW. Functional evaluation: the Barthel Index. Md Med J. 1965; 14: 61–65.
5. Ware JE, Sherbourne CD. A 36-item short-form health survey (SF-36): conceptual framework and item selection. Med Care. 1992; 30: 473–483.[Medline] [Order article via Infotrieve]
6. von Neumann J, Morgenstern O. Theory of Games and Economic Behavior. New York: Wiley; 1953.
7. Torrance GW, Thomas WH, Sackett DL. A utility maximization model for evaluation of health care programs. Health Serv Res. 1972; 7: 118–133.[Medline] [Order article via Infotrieve]
8. Streiner DL, Norman GR. Health Measurement Scales. A practical guide to their development and use. New York: Oxford University Press; 1989.
9. Diener A, O’Brien B, Gafni A. Health care contingent valuation studies: a review and classification of the literature. Health Econ. 1998; 7: 313–326.[CrossRef][Medline] [Order article via Infotrieve]
10. Lenert LA, Sturley A, Watson ME. iMPACT3: internet-based development and administration of utility elicitation protocols. Med Decis Making. 2002; 22: 464–474.
11. King JT, Jr, Tsevat J, Moossy JJ, Roberts MS. Preference-based quality of life measurement in patients with cervical spondylotic myelopathy. Spine. 2004; 29: 1271–1280.[CrossRef][Medline] [Order article via Infotrieve]
12. The Dutch TIA trial: protective effects of low-dose aspirin and atenolol in patients with transient ischemic attacks or nondisabling stroke. The Dutch TIA Study Group. Stroke. 1988; 19: 512–517.
13. Collin C, Wade DT, Davies S, Horne V. The Barthel ADL index: a reliability study. Int Disabil Studies. 1988; 10: 61–63.[Medline] [Order article via Infotrieve]
14. Consumer Price Index—All Urban Consumers. U S Department of Labor, Bureau of Labor Statistics Web site, 2003. Available at: URL: http://data.bls.gov/cgi-bin/surveymost. Accessed May 5, 2004.
15. Cuzick J. A Wilcoxon-type test for trend. Stat Med. 1985; 4: 87–90.[Medline] [Order article via Infotrieve]
16. Post PN, Stiggelbout AM, Wakker PP. The utility of health states after stroke: a systematic review of the literature. Stroke. 2001; 32: 1425–1429.
17. Tijhuis GJ, Jansen SJ, Stiggelbout AM, Zwinderman AH, Hazes JM, Vlieland TP. Value of the time trade off method for measuring utilities in patients with rheumatoid arthritis. Ann Rheum Dis. 2000; 59: 892–897.
18. Kiebert GM, Green C, Murphy C, Mitchell JD, O’Brien M, Burrell A, Leigh PN. Patients’ health-related quality of life and utilities associated with different stages of amyotrophic lateral sclerosis. J Neurol Sci. 2001; 191: 87–93.[CrossRef][Medline] [Order article via Infotrieve]
19. Gold MR, Siegel JE, Russell LB, Weinstein MC. Cost-effectiveness in Health and Medicine. New York: Oxford University Press; 1996.
20. van Wijck EE, Bosch JL, Hunink MG. Time-tradeoff values and standard-gamble utilities assessed during telephone interviews versus face-to-face interviews. Med Decis Making. 1998; 18: 400–405.
21. Ross PL, Littenberg B, Fearn P, Scardino PT, Karakiewicz PI, Kattan MW. Paper standard gamble: a paper-based measure of standard gamble utility for current health. Int J Technol Assess Health Care. 2003; 19: 135–147.[CrossRef][Medline] [Order article via Infotrieve]
22. Littenberg B, Partilo S, Licata A, Kattan MW. Paper Standard Gamble: the reliability of a paper questionnaire to assess utility. Med Decis Making. 2003; 23: 480–488.
23. Fryback DG, Lawrence WFJ. Dollars may not buy as many QALYs as we think: a problem with defining quality-of-life adjustments. Med Decis Making. 1997; 17: 276–284.
24. King JT, Jr., Styn MA, Tsevat J, Roberts MS. "Perfect health" versus "disease free": the impact of anchor point choice on the measurement of preferences and the calculation of disease-specific disutilities. Med Decis Making. 2003; 23: 212–225.
This article has been cited by other articles:
 |
|
 |
|
  J. T. King Jr., J. Tsevat, J. R. Lave, and M. S. Roberts Willingness to Pay for a Quality-Adjusted Life Year: Implications for Societal Health Care Resource Allocation Med Decis Making, November 1, 2005; 25(6): 667 - 677. [Abstract] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||