Effects on preferences of violations of procedural invariance

Characterization of Physicians That Might Be Reluctant to Propose HIV Cure-Related Clinical Trials with Treatment Interruption to Their Patients? The ANRS-APSEC Study

HIV cure-related clinical trials (HCRCT) with analytical antiretroviral treatment interruptions (ATIs) have become unavoidable. However, the limited benefits for participants and the risk of HIV transmission during ATI might negatively impact physicians’ motivations to propose HCRCT to patients. Between October 2016 and March 2017, 164 French HIV physicians were asked about their level of agreement with four viewpoints regarding HCRCT. A reluctance score was derived from their answers and factors associated with reluctance identified. Results showed the highest reluctance to propose HCRCT was among physicians with a less research-orientated professional activity, those not informing themselves about cure trials through scientific literature, and those who participated in trials because their department head asked them. Physicians’ perceptions of the impact of HIV on their patients’ lives were also associated with their motivation to propose HCRCT: those who considered that living with HIV means living with a secret were more motivated, while those worrying about the negative impact on person living with HIV’s professional lives were more reluctant. Our study highlighted the need to design a HCRCT that minimizes constraints for participants and for continuous training programs to help physicians keep up-to-date with recent advances in HIV cure research.

HIV Cure Research: Risks Patients Expressed Willingness to Accept

Despite doing well on antiretroviral therapy, many people living with HIV have expressed a willingness to accept substantial risks for an HIV cure. To date, few studies have assessed the specific quantitative maximal risk that future participants might take; probed whether, according to future participants, the risk can be offset by the benefits; and examined whether taking substantial risk is a reasonable decision. In this qualitative study, we interviewed 22 people living with HIV and used standard gamble methodology to assess the maximum chance of death a person would risk for an HIV cure. We probed participants' reasoning behind their risk-taking responses. Conventional inductive content analysis was used to categorize key themes regarding decision-making. We found that some people would be willing to risk even death for an HIV cure, and some of their reasons were plausible and went far beyond the health-related utility of an HIV cure. We contend that people's expressed willingness to take substantial risk for an HIV cure should not be dismissed out of hand.

Anchoring bias in face-to-face Time-Trade-Off valuations of health states

OBJECTIVE

To test whether anchoring (a cognitive bias) occurs during face-to-face interviews to value health states by Time-Trade-Off.

METHODS

147 Colombian subjects (111 males, 36 females) valued five EQ-5D health states better than death during a face-to-face interview. Subjects were randomly assigned to two different starting points.

RESULTS

Shapiro-Wilk test discarded normality, while non-parametric tests, including Kolmogorov-Smirnov and Wilcoxon-Mann-Whitney, showed that anchoring was significant in four out of five health states. A higher starting point increased the elicited value by 15 %-188 %. The size of the anchoring effect was not uniform among health states.

CONCLUSION

Anchoring effects may bias face-to-face Time Trade-Off valuations. The size of the anchoring effect is relevant enough for health policy.

Influenced from the start: anchoring bias in time trade-off valuations

Purpose

The de facto standard method for valuing EQ-5D health states is the time trade-off (TTO), an iterative choice procedure. The TTO requires a starting point (SP), an initial offer of time in full health which is compared to a fixed offer of time in impaired health. From the SP, the time in full health is manipulated until preferential indifference. The SP is arbitrary, but may influence respondents, an effect known as anchoring bias. The aim of the study was to explore the potential anchoring effect and its magnitude in TTO experiments.

Methods

A total of 1249 respondents valued 8 EQ-5D health states in a Web study. We used the lead time TTO (LT-TTO) which allows eliciting negative and positive values with a uniform method. Respondents were randomized to 11 different SPs. Anchoring bias was assessed using OLS regression with SP as the independent variable. In a secondary experiment, we compared two different SPs in the UK EQ-5D valuation study TTO protocol.

Results

A 1-year increase in the SP, corresponding to an increase in TTO value of 0.1, resulted in 0.02 higher recorded LT-TTO value. SP had little impact on the relative distance and ordering of the eight health states. Results were similar to the secondary experiment.

Conclusion

The anchoring effect may bias TTO values. In this Web-based valuation study, the observed anchoring effect was substantial. Further studies are needed to determine whether the effect is present in face-to-face experiments.

The utility assessment method order influences measurement of parents' risk attitude

BACKGROUND

Standard gamble (SG) and time trade-off (TTO) are two methods used for obtaining health utility values (utilities). Whether the order in which the methods are applied alters the relative utilities obtained by each method is unknown.

OBJECTIVE

We sought to determine whether the order in which SG and TTO utilities were obtained affects the relative values of the utilities obtained by each technique.

METHODS

Utilities were assessed for 29 health states from 4016 parents by using SG and TTO. The assessment order was randomized by respondent. For analysis by health state, we calculated (SG - TTO) for each assessment and tested whether the SG - TTO difference was significantly different between the two groups (SG first and TTO first). For analysis by individual, we calculated a risk-posture coefficient, γ, defined by the utility curve, SG = TTO(γ). We predicted γ through regression analysis with the covariates: child age, child sex, birth order, respondent age, respondent education level, and assessment method order.

RESULTS

In 19 of 29 health states, the SG - TTO difference was significantly greater (more risk averse) when TTO was assessed first. In the regression analysis, "child age" and "assessment method order" were significant predictors of risk attitude. The risk posture coefficient γ was higher (more risk-seeking) with increasing child age and in the SG-first respondents.

CONCLUSION

The order in which the SG versus TTO method is used strongly influences the relative values of the utilities obtained.

Ordering errors, objections and invariance in utility survey responses: a framework for understanding who, why and what to do

BACKGROUND

Utilities are the quantification of the perceived quality of life associated with any health state. They are used to calculate QALYs, the outcome measure in cost-utility analysis. Generally measured through surveys of individuals, utilities often contain apparent or unapparent errors that can bias resulting values and QALYs calculated from these values.

OBJECTIVE

The aim of this study was to improve direct health utility elicitation methodology through the identification of the types of survey responses that indicate errors and objections, and the reasons underlying them.

METHODS

We conducted a systematic review of the medical (PubMed), economics (EconLit) and psychology (PsycINFO) literature from 1975 through June 2010 for articles describing the types and frequency of errors and objections in directly elicited utility survey responses, and strategies to address these responses. Primary data were collected through an internet-based utility survey (standard gamble) of community members to identify responses that indicate error or objections. A qualitative telephone survey was conducted among a subset of respondents with these types of responses using an open-ended protocol to elicit rationales for them.

RESULTS

A total of 11 papers specifically devoted to errors, objections and invariance in utility responses have been published since the mid-1990s. Error/objection responses can be broadly categorized into ordering errors (which include illogical and inconsistent responses) and objections/invariance (which include missing data, protest responses and refusals to trade time or risk in utility questions). Reported frequencies of respondents making ordering errors ranged from 5% to 100%, and up to 35% of respondents have been reported as objecting to the survey or task in some manner. Changes in the design, administration and analysis of surveys can address these potentially problematic responses. Survey data (n = 398) showed that individuals who provided invariant responses (n = 26) reported the lowest level of difficulty with the survey and often identified as religious (23% of invariant responders found the survey difficult vs 63% of all responders, and 77% of invariant responders identified as religious compared with 56% of entire sample; p < 0.05 for both). Respondents who provided illogical responses (n = 50) were less likely to be college educated (56% of illogical responders vs 73% of entire sample; p < 0.05), and less likely to be confident in their responses (62% vs 75% of entire sample; p < 0.05). Qualitative interviews (n = 42) following the survey revealed that the majority of ordering errors were a result of confusion, lack of attention or difficulty in responding to the survey on the part of the respondent, while invariant responses were often considered and thoughtful reactions to the premise of valuing health using the standard gamble task.

CONCLUSIONS

Rationales for error/objection responses include difficulty in articulating preferences or misunderstanding with a complex survey task, and also thoughtful and considered protestations to the task. Mechanisms to correct unintentional errors may be useful, but cannot address intentional responses to elements of the measurement task. Identification and analysis of the prevalence of errors and objections in responses in utility data sets are essential to understanding the accuracy and precision of utility estimates and analyses that depend thereon.

The effect of time of onset on community preferences for health states: an exploratory study

BACKGROUND

Health state descriptions used to describe hypothetical scenarios in community-perspective utility surveys commonly omit detail on the time of onset of a condition, despite our knowledge that among patients who have a condition, experience affects the value assigned to that condition. The debate regarding whose values to use in cost utility analysis is based in part on this observed difference between values depending on the perspective from which they are measured. This research explores the effect on community preferences for hypothetical health states of including the time of onset of a health condition in the health state description, to investigate whether this information induces community respondents to provide values closer to those of patients with experience with a condition. The goal of the research is to bridge the gap between patient and community preferences.

METHODS

A survey of community-perspective preferences for hypothetical health states was conducted among a convenience sample of healthy adults recruited from a hospital consortium's research volunteer pool. Standard gambles for three hypothetical health states of varying severity were compared across three frames describing time of onset: six months prior onset, current onset, and no onset specified in the description. Results were compared within health state across times of onset, controlling for respondent characteristics known to affect utility scores. Sub-analyses were conducted to confirm results on values meeting inclusion criteria indicating a minimum level of understanding and compliance with the valuation task.

RESULTS

Standard gamble scores from 368 completed surveys were not significantly different across times of onset described in the health state descriptions regardless of health condition severity and controlling for respondent characteristics. Similar results were found in the subset of 292 responses that excluded illogical and invariant responses.

CONCLUSIONS

The inclusion of information on the time of onset of a health condition in community-perspective utility survey health state descriptions may not be salient to or may not induce expression of preferences related to disease onset among respondents. Further research is required to understand community preferences regarding condition onset, and how such information might be integrated into health state descriptions to optimize the validity of utility data. Improved understanding of how the design and presentation of health state descriptions affect responses will be useful to eliciting valid preferences for incorporation into decision making.

Is the value of a life or life-year saved context specific? Further evidence from a discrete choice experiment

BACKGROUND

A number of recent findings imply that the value of a life saved, life-year (LY) saved or quality-adjusted life year (QALY) saved varies depending on the characteristics of the life, LY or QALY under consideration. Despite these findings, budget allocations continue to be made as if all healthy life-years are equivalent. This continued focus on simple health maximisation is partly attributable to gaps in the available evidence. The present study attempts to close some of these gaps.

METHODS

Discrete choice experiment to estimate the marginal rate of substitution between cost, effectiveness and various non-health arguments. Odds of selecting profile B over profile A estimated via binary logistic regression. Marginal rates of substitution between attributes (including cost) then derived from estimated regression coefficients.

RESULTS

Respondents were more likely to select less costly, more effective interventions with a strong evidence base where the beneficiary did not contribute to their illness. Results also suggest that respondents preferred prevention over cure. Interventions for young children were most preferred, followed by interventions for young adults, then interventions for working age adults and with interventions targeted at the elderly given lowest priority.

CONCLUSION

Results confirm that a trade-off exists between cost, effectiveness and non-health arguments when respondents prioritise health programs. That said, it is true that respondents were more likely to select less costly, more effective interventions - confirming that it is an adjustment to, rather than an outright rejection of, simple health maximisation that is required.

Can urban methadone patients complete health utility assessments?

OBJECTIVE

To assess the ability of methadone maintenance treatment (MMT) patients to use two standardized health assessment tools to value health states related to chronic hepatitis C virus (HCV) infection and HCV treatment-associated side effects. An estimated 65-90% of MMT patients are chronically infected with HCV.

METHODS

We employed qualitative methods to explore how patients completed computerized rating scale assessments and standard gamble utility assessments by (1) having them discuss their responses in a think-aloud interview immediately after each health state assessment, and (2) allowing them the opportunity to recalibrate prior responses after considering subsequent health states.

RESULTS

MMT patients used the rating scale boundaries appropriately and used the standard gamble to rank the health states in an a priori logical order. A guided assessment approach that allowed recalibration provided additional insight into values assigned to the health states presented.

CONCLUSION

MMT patients are able to perform the tasks associated with rating scale assessments and standard gamble utility assessments of HCV health states.

PRACTICE IMPLICATIONS

These assessment methods should be considered as a means to elicit MMT patients' values for HCV treatment, since the treatment outcome is uncertain but it is likely that side effects will adversely affect current health.

Quality-of-life tradeoffs for hepatitis C treatment: do patients and providers agree?

BACKGROUND

The authors investigated differences between how patients and providers evaluate the quality-of-life tradeoffs associated with HCV treatment in computer-assisted interviews. They interviewed 92 treatment-naive HCV patients at gastroenterology, methadone maintenance, and HIV clinics at 3 hospitals in New York City and 23 physicians or nurses experienced in treating HCV at other hospitals in New York City. Subjects completed rating scale and standard gamble evaluations of current health and hypothetical descriptions of HCV symptoms and treatment side effects on a scale from 0 (death or worse than death) to 1 (best possible health).

RESULTS

. Treatment side effects were rated worse by patients than providers using the rating scale (moderate side effects 0.42 v. 0.62; severe side effects 0.24 v. 0.40) and standard gamble (moderate side effects 0.61 v. 0.91; severe side effects 0.52 v. 0.75) (all P < or = 0.01). A year of severe side effects was equivalent to 4.1 years of mild HCV symptoms avoided for patients if they returned to their current health after treatment compared with 2.0 years avoided if they achieved average population health. For patients with depression symptoms, HCV treatment with severe side effects had lower value unless it would also improve their current health.

CONCLUSIONS

. Patients have more concerns about treatment side effects than providers. Further research is warranted to develop HCV decision aids that elicit patient preferences and to evaluate how improved communication of the risks and benefits of HCV treatment and more effective treatment of depression may alter these preferences.

Health state valuation in mild to moderate cognitive impairment: feasibility of computer-based, direct patient utility assessment

BACKGROUND

Most patients with dementia will, at some point, need a proxy health care decision maker. It is unknown whether persons with various degrees of cognitive impairment can reliably report their health-related preferences.

METHODS

The authors performed health state valuations (HSVs) of current and hypothetical future health states on 47 pairs of patients with mild to moderate cognitive impairment and their caregivers using computer-based standard gamble, time tradeoff, and rating scale techniques.

RESULTS

Patients' mean (SD) age was 74.6 (9.3) years. About half of the patients were women (48%), as were most caregivers (73%), who were on average younger (mean age= 66.2 years, SD= 12.2). Most participants were white (83%); 17% were African American. The mean (SD) Mini-Mental State Examination (MMSE) score of patients was 24.2 (4.6) of 30. All caregivers and 77% of patients (36/47) completed all 18 components of the HSV exercise. Patients who completed the HSV exercise were slightly younger (mean age [SD]= 74.1 [8.5] v. 75.9 [11.8]; P = 0.569) and had significantly higher MMSE scores (mean score [SD] = 25.0 [4.3] v. 21.4 [4.4]; P = 0.018). Although MMSE scores below 20 did not preclude the completion of all 18 HSV ratings, being classified as having moderate cognitive impairment was associated with a lower likelihood of completing all scenario ratings (44% v. 82%). Patient and caregiver responses showed good consistency across time and across techniques and were logically consistent.

CONCLUSION

Obtaining HSVs for current and hypothetical health states was feasible for most patients with mild cognitive impairment and many with moderate cognitive impairment. HSV assessments were consistent and reasonable.

Application of generalizability theory confirmed lower reliability of the standard gamble than the feeling thermometer

OBJECTIVES

Recent studies suggest that rating clinical marker states (CMS) does not improve the measurement properties of the standard gamble (SG) and only slightly improves those of the feeling thermometer (FT). The poor intrarater (test-retest) reliability of CMS may explain their meager performance. Further, lack of interrater reliability may compromise the use of CMS in interpreting health state ratings. The aim of this study was to assess the reliability of CMS ratings for the SG and the FT.

STUDY DESIGN AND SETTING

Two similar studies in patients with chronic obstructive pulmonary disease (COPD, n=91) and in patients with gastroesophageal reflux disease (GERD, n=112) provided data for this analysis. Patients rated three different CMS (mild, moderate, and severe disease) twice several weeks apart. We used generalizability theory to calculate reliability coefficients.

RESULTS

Test-retest reliability for CMS ratings was higher for the FT compared to the SG (COPD: 0.86 vs. 0.67; GERD: 0.86 vs. 0.67). Interrater reliability was much higher for the FT compared to the SG (COPD: 0.78 vs. 0.46; GERD: 0.71 vs. 0.26).

CONCLUSIONS

These results suggest that the markedly poorer reliability of CMS for the SG than the FT is driven largely by poor interrater reliability.

Invariance and inconsistency in utility ratings

PURPOSE

To assess utilities of composite health states for dependence in activities of daily living (ADLs) for invariance (i.e., when subjects provide a utility of 1 for all health states) and order inconsistency (i.e., when subjects order their utilities such that their utility for a combination of ADL dependencies is greater than their utility for any subset of the combination).

METHODS

Each of the 400 subjects, age 65 y and older, enrolled in one of several regional medical centers of the Kaiser Permanente Medical Care Program of Northern California and provided standard-gamble utilities for single ADL dependencies (e.g., bathing, dressing, continence) and for dependence in 8 other combinations of ADL dependencies. For order-inconsistent responses, the authors calculated the maximum magnitude of inconsistency as the maximum difference between the utility for the combined ADL dependence health state and that of its inconsistent subset.

RESULTS

A total of 76 subjects (19%) gave a utility of 1.0 for all health states presented to them; 19 (5%) gave the same utility other than 1.0 for all health states; 130 (33%) gave at least 1 utility < 1.0 and had no order inconsistencies; and 175 (44%) had at least 1 order inconsistency. Invariance was associated with a Mini-Mental Status Examination score < 28.6 (P = 0.01), with education < 12 y (P = 0.004), with race/ethnicity other than non-Hispanic White/Caucasian (P = 0.001), and with shorter time spent on the utility elicitation task (P < 0.0001). Among the inconsistent subjects, 69% had a maximal magnitude of inconsistency that was within 1 standard deviation of the mean utilities. The maximal magnitude of inconsistency was associated with longer time spent on the elicitation task (P < 0.0001) and race/ethnicity other than non-Hispanic White/Caucasian (P = 0.005). The mean (s) utility for dependence in continence among consistent subjects who were not invariant (0.88 [0.24]) was higher than among inconsistent subjects (0.80 [0.27]; P = 0.01).

CONCLUSIONS

Invariance and order inconsistencies in utility ratings for complex health states occur frequently. Utilities of consistent subjects may differ from those of inconsistent subjects. Utility assessments should attempt to measure and report these patterns.

Utilities for prostate cancer health states in men aged 60 and older

PURPOSE

We sought to measure utilities for prostate cancer health states in older men.

METHODS

A total of 162 men aged 60 years or older (52% of whom had been diagnosed with prostate cancer) provided standard gamble utilities for 19 health states associated with prostate cancer or its treatment using an interactive, computer-based utility assessment program. Demographics and experience with specific health states were examined as predictors of ratings using ordinary least squares regression analysis.

RESULTS

Mean utilities ranged from 0.67 to 0.84 for living with symptom-free cancer under conservative management ("watchful waiting") and from 0.71 to 0.89 for symptoms occurring with treatment (prostatectomy, radiation, and hormone ablation). For long-term treatment complications, bowel problems (0.71) were rated as significantly worse than impotence (0.89), urinary difficulty (0.88), or urinary incontinence (0.83). Combinations of these conditions were rated as significantly worse than individual component states. Men who had experienced impotence or urinary incontinence rated these states as slightly better than men who had not experienced the specific problems.

CONCLUSIONS

Both "watchful waiting" and treatment complications from prostate cancer treatments can have large impacts on quality of life. Mean ratings are important for use in policy-making and cost-effectiveness analyses. Variation in ratings across patients suggests that mean scores do not reflect individual preferences and that shared decision-making may be best for clinical decisions.

The standard gamble demonstrated lower reliability than the feeling thermometer

BACKGROUND AND OBJECTIVE

Participants rated clinical marker states (CMS) to make respondents familiar with the task of preference instruments, ground their ratings in relation to other health states, and help investigators interpret patient ratings. The objective was to assess the reliability of CMS using appropriate reliability statistics.

STUDY DESIGN AND SETTING

Eighty-one patients rated CMSs for mild, moderate, and severe chronic respiratory disease using the feeling thermometer (FT) and the standard gamble (SG) before and after a 3-month respiratory rehabilitation program. To assess reliability we used (a) intraclass correlation coefficients (ICC) with the variance between CMSs as signal and the variance between raters, the variance within raters, and the signal as noise; (b) scatter plots; and (c) Bland-Altman plots.

RESULTS

ICCs were 0.47 for the FT and 0.37 for the SG. Scatter and Bland-Altman plots showed large between- and within-person variability; 64.2% and 11.3% of the CMSs ratings were in the correct order on both occasions on the FT and SG, respectively.

CONCLUSION

Our results suggest moderate reliability of CMSs ratings for the FT and poor reliability for the SG, which may explain their lack of improving the SG's measurement properties. Investigators should use appropriate reliability statistics when addressing related issues.

Use of willingness to pay to study values for pharmacotherapies for migraine headache

BACKGROUND

There is a wide range of therapeutic options for migraine headaches, yet little is known about how patients value different treatment alternatives nor how to tailor treatments to patients' preferences.

OBJECTIVE

Assess patients' preferences for outcomes of treatment of migraine headache based on their marginal willingness to pay (WTP) for treatment attributes.

MATERIALS AND METHODS

In an Internet-based study, we used computer software to measure participants' WTP for a hypothetical ideal drug (one that was 100% effective, worked quickly, and had no adverse effects) and for other hypothetical drugs, each with one "less than ideal" attribute of performance.

SUBJECTS

Two hundred fifty-seven self-identified migraineurs recruited via an Internet Web site.

RESULTS

A high proportion of participants in the study had symptoms consistent with migraine etiology of headache (99%). Median "out-of-pocket" monthly WTP cost for an ideal migraine therapy was 130 dollars. WTP was associated with participants' own migraine experience: severity of pain, frequency of headaches, and the types of medications used in the past. WTP was reduced when pharmaceuticals offered less benefit, ranging from a mean of 74% of ideal for treatments that failed to relieve nausea, to 43% of ideal in treatments associated with a 50% chance of a rebound headache.

CONCLUSIONS

There is wide variability in patients' strength of preference for different attributes of migraine therapy. Choice of therapy for migraine headache should be individualized based on patients' preferences. WTP measures appear to be a valid and feasible metric for quantifying treatment preferences for migraine therapies.

Toward improved methods for measurement of utility: automated repair of errors in elicitations

This study examines the effects of use of an automated computer protocol to correct a common error in utility elicitations--namely, scoring of a health state with a greater impairment as being more desirable than one with lesser impairment. The authors studied the protocol in a sample of 563 members of the Internet-using public. Results revealed that errors were common (17% of ratings) but were typically successfully repaired (individuals who made only 1 or 2 errors had a 75% chance of repairing them). The values of individuals who repaired errors were similar to those without apparent error. In contrast, individuals who refused to repair errors had lower scores for the best health state in the series and higher ones for the worst health state. Results suggest that the repair procedures were successful and that inclusion of utility scores from individuals who fail to repair illogically ordered ratings may bias estimates of mean utilities.

iMPACT3: Internet-based development and administration of utility elicitation protocols

iMPACT3 (Internet Multimedia Preference Assessment Instrument Construction Tool, version 3) is a software development environment that helps researchers build Internet-capable multimedia utility elicitation software programs. The program is a free, openly accessible Web site (http.// preferences.ucsd.edu/impact3/asp). To develop a utility elicitation software program using iMPACT3, a researcher selects modular protocol components from a library and custom tailors the components to the details of his or her research protocol. iMPACT3 builds a Web site implementing the protocol and downloads it to the researcher's computer. In a study of 75 HIV-infected patients, an iMPACT3-generated protocol showed substantial evidence of construct validity and good internal consistency (logic error rates of 4% to 10% and procedural invariance error rates of 10% to 28%, depending on the elicitation method) but only fair 3- to 6-week test-retest reliability (intraclass correlation coefficient= 0.42 to 0.55). Further work may be needed on specific utility assessment procedures, but this study's results confirm iMPACT3's feasibility in facilitating the collection of health state utility data.

Assessing values for health: numeracy matters

BACKGROUND

Patients' values are fundamental to decision models, cost-effectiveness analyses, and pharmacoeconomic analyses. The standard methods used to assess how patients value different health states are inherently quantitative. People without strong quantitative skills (i.e., low numeracy) may not be able to complete these tasks in a meaningful way.

METHODS

To determine whether the validity of utility assessments depends on the respondent's level of numeracy, the authors conducted in-person interviews and written surveys and assessed utility for the current health for 96 women volunteers. Numeracy was measured using a previously validated 3-item scale. The authors examined the correlation between self-reported health and utility for current health (assessed using the standard gamble, time trade-off, and visual analog techniques) across levels of numeracy. For half of the women, the authors also assessed standard gamble utility for 3 imagined health states (breast cancer, heart disease, and osteoporosis) and asked how much the women feared each disease.

RESULTS

Respondent ages ranged from 50 to 79 years (mean = 63), all were high school graduates, and 52% had a college or postgraduate degree. Twenty-six percent answered 0 or only 1 of the numeracy questions correctly, 37% answered 2 correctly, and 37% answered all 3 correctly. Among women with the lowest level of numeracy, the correlation between utility for current health and self-reported health was in the wrong direction (i.e., worse health valued higher than better health): for standard gamble, Spearman r=-0.16, P = 0.44;for time trade-off, Spearman r=-0.13, P=0.54. Among the most numerate women, the authors observed a fair to moderate positive correlation with both standard gamble (Spearman r=0.22, P=0.19) and time trade-off (Spearman r=0.50, P=0.002). In contrast, using the visual analog scale, the authors observed a substantial correlation in the expected direction at all levels of numeracy (Spearman r= 0.82, 0.50, and 0.60 for women answering 0-1, 2, and 3 numeracy questions, respectively; all Ps < or = 0.003). With regard to the imagined health states, the most feared disease had the lowest utility for 35% of the women with the lowest numeracy compared to 76% of the women with the highest numeracy (P=0.03).

CONCLUSIONS

The validity of standard utility assessments is related to the subject's facility with numbers. Limited numeracy may be an important barrier to meaningfully assessing patients' values using the standard gamble and time trade-off techniques.

The reliability and internal consistency of an Internet-capable computer program for measuring utilities

OBJECTIVE

The purpose of this study was to assess the reliability and internal consistency of measurements of utilities performed with a computer program (iMPACT2) designed for Internet surveys and Internet patient decision-support systems.

METHODS

We implemented the Internet Multimedia Preference Assessor Construction Tool, version 2 (iMPACT2) program using the combination of a web server, HTML files, and a web-accessible database. The program randomized subjects, screened their responses for missing data and failures of internal consistency, assisted patients with resolving certain inconsistencies, and, upon a subject's completion of the protocol, provided a report of results to the research assistant administering the program. To validate the iMPACT2 program, we recruited 60 healthy community volunteers and elicited preferences in a research-lab setting using a visual analog scale (VAS) and the standard gamble (SG) for subject's current health and three hypothetical states. For purposes of comparison, we also administered a Short Form-12 (SF-12) health-assessment questionnaire. Subjects used the computer software on two occasions separated by 2-4 weeks of time.

RESULTS

Visual analog scale and standard gamble ratings for subjects' current health were reliable (intraclass correlation coefficient (ICC) of 0.82 and 0.84 (two outliers excluded -0.60 without exclusions), respectively) were comparable with the reliability of the Physical and Mental Component scales of the SF-12 (ICCs of 0.84 and 0.75, respectively). Subjects could easily discriminate between hypothetical states (D scores 0.74 for SG and 0.90 for VAS), and 94% had a completely internally consistent ordering of preference ratings for states.

CONCLUSIONS

iMPACT2 produces measurements of standard gamble utilities that are reliable and have a high degree of internal consistency. Procedures for assessment of utilities developed for desktop computer programs can be translated to software designed for the Internet, facilitating the use of utilities and endpoints in clinical trials and development of web-based decision-support applications for patients. However, further testing, including direct comparisons with traditional interviewer administered utility elicitation protocols, is needed.

Preference for fractures and other glucocorticoid-associated adverse effects among rheumatoid arthritis patients

OBJECTIVE

The objective of this study was to determine rheumatoid arthritis (RA) patients' preferences for validated health state scenarios depicting glucocorticoid adverse events, predictors of these preferences, and psychometric properties of different preference techniques in this population.

METHODS

Preferences were elicited by rating scale and time trade-off methods. Time trade-offs included trading current health for either time spent alive in an adverse health state for chronic conditions (time trade-off) or time spent in a sleeplike state for acute conditions (sleep trade-off).

RESULTS

A total of 107 subjects with long-standing RA participated in the preference interviews. Mean preference values (rating scale/trade-off) were lowest for serious fracture adverse events, including hip fracture requiring a nursing home stay (0.55+/-0.22/0.76+/-0.36) and vertebral fracture with chronic pain (0.59+/-0.23/0.67+/-0.35), and highest for cataracts (0.84 + 0.17/0.96 0.09) and wrist fracture (0.82+/-0.18/0.81+/-0.29). Rating scales had a stronger correlation (r= 0.88) with physician ranking of scenarios than trade-off methods (r = 0.31). All methods were feasible and demonstrated good reliability, while rating scale method showed better construct validity than trade-off techniques.

CONCLUSION

Relative to their current health, RA patients assigned low preference values to many glucocorticoid adverse events, particularly those associated with chronic fracture outcomes. Results varied with the preference measure used, indicating that methodological attributes of preference determinations must be considered in clinical decision making.

Differences in health values among patients, family members, and providers for outcomes in schizophrenia

OBJECTIVE

The objectives of this study were to determine whether there are important differences in how patients, family members, and health care providers (HCPs) value health outcomes in schizophrenia and to assess the degree to which such differences, if they exist, could adversely affect clinical and policy decision making.

METHODS

Participants viewed videotaped depictions of simulated patients with mild and moderate symptoms of schizophrenia, with and without a common adverse drug effect (pseudoparkinsonism), and then provided standard gamble and visual analog scale ratings of desirability of these states.

SUBJECTS

A convenience sample of unrelated patients (n = 148), family members of patients (n = 91), and HCPs (nurses, psychologists, doctors of pharmacy, and doctors of medicine; n = 99) was drawn from geographically and clinically diverse environments.

RESULTS

Patients' and family members' utilities for health states averaged 0.1 to 0.15 units higher than those of HCPs (P <0.002 for differences between groups, ANOVA for multiple observations). The disutility of adverse drug effects was less for health professionals than patients and family members (P = 0.008). Health professionals tended to prefer states with mild symptoms with extrapyramidal side effects to states with moderate symptoms. Patients and family members found these states equally preferable (P <0.007 for differences between groups).

CONCLUSIONS

There are systematic differences in values for health outcomes between patients and HCPs with regard to states with adverse effects of antipsychotic drugs. Family members of patients in general had values that were more similar to those of patients than were those of health professionals. The results emphasize the importance of participation by patients (or family member proxies) in clinical decision making and guideline development.

Validity and interpretation of preference-based measures of health-related quality of life

Utilities are numeric measurements that reflect an individual's beliefs about the desirableness of a health condition, willingness to take risks to gain health benefits, and preferences for time. This report discusses the approaches to assess and compare the validity of methods used to assign utilities for cost-utility analysis. Threats to validity include construct underrepresentation and construct-irrelevant variance. Construct underrepresentation occurs when a stimulus presented to a judge fails to fully represent the depth and complexity of information required in actual judgments. Construct-irrelevant variation occurs when factors irrelevant to preferences influence measurements of utilities. Among several factors that cause construct-irrelevant variation are cognitive abilities, numeracy skills, emotions and prejudices, and the elicitation procedure. Commonly used elicitation methods (visual-analog scales, time tradeoff, and standard gamble) capture different facets of utilities (desirableness of states, time preferences, and risk attitude) to different degrees. The validity of an elicitation protocol depends (1) on the degree to which its scaling method captures the relevant facets of utility and (2) on the degree to which measurements are influenced by construct-irrelevant variation. Discrete-state health index models provide an alternative to direct elicitation of utilities and work by attaching fixed preference weights to observable health states. The creation of discrete-state models with current technologies requires the adoption of strong assumptions about the scaling properties of utilities. Future research must refine methods of eliciting utilities and identify sources of construct-irrelevant variability that reduce the validity of utility assessments. Because of the impact of variation in techniques on measurements, we do not recommend the combination of utilities elicited with different protocols in cost-utility analysis and do not recommend the display of cost-utility ratios from different studies in comparison or "league" tables.

Estimation of utilities for the effects of depression from the SF-12

BACKGROUND

Utilities for health conditions, including major depressive disorder, have a theoretical relationship to health-related quality of life (HRQOL). Because of the complexity of utility measurement and the existence of large numbers of completed studies with HRQOL data but not utility data, it would be desirable to be able to estimate utilities from measurements of HRQOL.

OBJECTIVE

The objective of this study was to estimate utility for remission in major depression by use of information on associated variation in Short Form 12 (SF-12) scores.

DESIGN

A mapping function for SF-12 scores (based on a 6-health-state model with patient-weighted preferences) was applied to longitudinal data from a large naturalistic study to estimate changes in utilities.

SUBJECTS

Preference ratings for states were performed in a convenience sample of depressed primary care patients (n = 140). Outcomes were evaluated in patients in the Course of Depression Study (n = 295) with a DSM III diagnosis of depression at the onset of the study.

MEASURES

From clinical interview data, differences in utilities and global physical and mental health-related quality of life at 1- and 2-year follow-up were compared for patients who did and did not experience remission as determined by the Course of Depression Interview.

RESULTS

Remission of depression resulted in health status improvement, as measured by the SF-12, equivalent to a gain of 0.11 quality-adjusted life-years over 2 years.

CONCLUSIONS

Utilities for changes in health status, associated with a clinical change in depression, can be modeled from the SF-12 scales, which results in utilities within the range of estimates described in the literature.

Comparison of preferences for health outcomes in schizophrenia among stakeholder groups

BACKGROUND

To determine the effectiveness of psychiatric interventions for use in cost-effectiveness analysis, we assessed the feasibility of using a multimedia computer survey to study preferences (utilities) for health outcomes among persons with schizophrenia, family members of persons with schizophrenia, health professionals, and the public.

METHODS

We developed videos depicting two patterns of mental health impairment in schizophrenia, both with and without pseudo-parkinsonism side-effects. These descriptions were integrated into a computer program that measured preferences using two psychometric methods: (1) standard gamble and (2) a visual analog scale. This program was used to compare preferences among potential stakeholder groups.

RESULTS

20 persons with schizophrenia, 11 family members, 20 healthy volunteers and 14 health professionals participated in the computerized interview. All but one subject completed the survey. The correlation among ratings of various states was high (r=0.7-0.95) and ratings were internally consistent in 89% of participants. There were significant differences in values between groups for health states (p=0.024) and in values for the effects of pseudo-parkinsonism on quality of life (p<0.001). Persons with schizophrenia valued the disease states more highly and placed more significance than did other groups on the effects of pseudo-parkinsonism on quality of life.

CONCLUSIONS

Computer-based multimedia techniques can offer a feasible and valid approach to measure preferences for outcomes in schizophrenia. The study found significant differences in preferences among stakeholder groups for schizophrenia outcomes. Further work is needed to clarify how these differences affect clinical decision-making and policies for health resource allocation.