Severe Health States Research Articles

Assessing the external validity of algorithms to estimate EQ-5D-3L from the WOMAC.

BackgroundThe use of mapping algorithms have been suggested as a solution to predict health utilities when no preference-based measure is included in the study. However, validity and predictive performance of these algorithms are highly variable and hence assessing the accuracy and validity of algorithms before use them in a new setting is of importance. The aim of the current study was to assess the predictive accuracy of three mapping algorithms to estimate the EQ-5D-3L from the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) among Swedish people with knee disorders. Two of these algorithms developed using ordinary least squares (OLS) models and one developed using mixture model.MethodsThe data from 1078 subjects mean (SD) age 69.4 (7.2) years with frequent knee pain and/or knee osteoarthritis from the Malmö Osteoarthritis study in Sweden were used. The algorithms’ performance was assessed using mean error, mean absolute error, and root mean squared error. Two types of prediction were estimated for mixture model: weighted average (WA), and conditional on estimated component (CEC).ResultsThe overall mean was overpredicted by an OLS model and underpredicted by two other algorithms (P < 0.001). All predictions but the CEC predictions of mixture model had a narrower range than the observed scores (22 to 90 %). All algorithms suffered from overprediction for severe health states and underprediction for mild health states with lesser extent for mixture model. While the mixture model outperformed OLS models at the extremes of the EQ-5D-3D distribution, it underperformed around the center of the distribution.ConclusionsWhile algorithm based on mixture model reflected the distribution of EQ-5D-3L data more accurately compared with OLS models, all algorithms suffered from systematic bias. This calls for caution in applying these mapping algorithms in a new setting particularly in samples with milder knee problems than original sample. Assessing the impact of the choice of these algorithms on cost-effectiveness studies through sensitivity analysis is recommended.Electronic supplementary materialThe online version of this article (doi:10.1186/s12955-016-0547-y) contains supplementary material, which is available to authorized users.

A Stated Preference Approach to Assess whether Health Status Impacts on Marginal Utility of Consumption.

A sample of Danish citizens (n = 2000) was asked to participate in a hypothetical experiment in which they could self-insure against a certain loss in income across two periods of time (one in which the person is in good health and one in which the person is in poorer health). Our results suggest that reduced health impacts on the marginal utility of consumption, but not in a linear fashion. Amongst a majority of respondents inferior, health increases the marginal utility of consumption for intermediate health states, whereas this is not the case for less and more severe health states. Copyright © 2016 John Wiley & Sons, Ltd.

How important is severity for the evaluation of health services: new evidence using the relative social willingness to pay instrument.

The 'severity hypothesis' is that a health service which increases a patient's utility by a fixed amount will be valued more highly when the initial health state is more severe. Supporting studies have employed a limited range of analytical techniques and the objective of the present paper is to test the hypothesis using a new methodology, the Relative Social Willingness to Pay. Three subsidiary hypotheses are: (1) that the importance of the 'severity effect' varies with the type of medical problem; (2) that the relationship between value and utility varies with the severity of the initial health state; and (3) that there is a threshold beyond which severity effects are insignificant. For each of seven different health problems respondents to a web-based survey were asked to allocate a budget to five services which would, cumulatively, move a person from near death to full health. The time trade-off utilities of health states before and after the service were estimated. The social valuation of the service measured by the budget allocation was regressed upon the corresponding increase in utility and severity as measured by the pre-service health state utility. Results confirm the severity hypothesis and support the subsidiary hypotheses. However, the effects identified are quantitatively significant only for the most severe health states. This implies a relatively limited redistribution of resources from those with less severe to those with more severe health problems.

A comparison of EuroQol 5-Dimension health-related utilities using Italian, UK, and US preference weights in a patient sample.

Weights associated with the EuroQol 5-Dimension 3-Level (EQ-5D-3L) instrument represent preferences for health states elicited from general population’s samples. Weights have not been calculated for every country; however, empirical research shows that cross-country differences exist. This empirical study aims at investigating the impact of recently developed Italian weights in comparison with UK and US scores on health-related utility calculation using a sample of patients with Crohn’s disease. The study is based on a survey on health-related quality of life in patients (n=552) affected by active Crohn’s disease conducted in Italy from 2012 to 2013. Utilities computed through the Italian algorithm (mean: 0.76; SD: 0.20; median: 0.81) are generally higher than US (mean: 0.69; SD: 0.22; median: 0.77) and UK (mean: 0.57; SD: 0.32; median: 0.69) utilities, except for extremely severe health states where US values outweigh the Italian ones. UK preference weights generate the highest number of negative results. All the three value distributions are left-skewed due to very low scores associated with the most serious health states (ie, three or four levels equal to 3). As expected, despite the tariff set considered, more severe disease (Harvey Bradshaw Index >16) reduces the mean conditional EQ-5D-3L index (P<0.0001). Kendall’s rank correlation between EQ Visual Analog Scale score and EQ-5D-3L index is positive (P<0.0001), even though patients tend to value their health-related quality of life more when responding to EQ-5D-3L questions than on EQ Visual Analog Scale. Regardless of the tariff set considered, ordinary least-square results highlight that more severe disease (Harvey Bradshaw Index >16) reduces the mean conditional EQ-5D-3L index (P<0.0001). Results reveal remarkable differences among the three national tariff sets and especially when severe health states occur, suggesting the need for country-specific preference weights when evaluating utilities, which can be problematic since they have not been calculated for every country yet.

Predicting the progression of Alzheimer's disease dementia: A multidomain health policy model

IntroductionWe develop a multidomain model to predict progression of Alzheimer's disease dementia (AD). MethodsData from the US National Alzheimer's Coordinating Center (n = 3009) are used to examine change in symptom status and to estimate transition probabilities between health states described using cognitive function, functional ability, and behavior. A model is used to predict progression and to assess a hypothetical treatment scenario that slows mild to moderate AD progression. ResultsMore than 70% of participants moved state over 12 months. The majority moved in domains other than cognitive function. Over 5 years, of those alive more than half are in severe AD health states. Assessing an intervention scenario, we see fewer years in more severe health states and a potential impact (life years saved) due to mortality improvements. DiscussionThe model developed is exploratory and has limitations but illustrates the importance of using a multidomain approach when assessing impacts of AD and interventions.

Can the general public use vignettes to discriminate between Alzheimer's disease health states?

BackgroundValid estimates of health-related quality-of-life (HRQoL) are often difficult to obtain from persons with Alzheimer’s disease (AD) and family caregiver proxies. To help assess whether the general public can serve as an alternate source of proxy HRQoL estimates in AD, we examined whether the general public can use vignettes to discriminate between AD health states.MethodsWe administered a telephone survey to randomly recruited participants from the general public who were aged 18 years or older. Interviewers read vignettes describing the mild, moderate, and severe AD health states to the participants, who answered the EQ-5D-5L and Quality of Life-Alzheimer’s Disease (QoL-AD) scales as if they had AD based on the vignette descriptions. Participants also answered the EQ-5D-5L for their current health states. We converted EQ-5D-5L responses into health utility scores using Canadian preference weights. We employed the Wilcoxon signed rank test to examine whether mean health utility scores and mean QoL-AD scores differed between health states. We used Pearson’s r to assess correlations between health utility and QoL-AD scores.ResultsForty-eight participants (median age = 53 years; 25 female) completed the telephone interview; health utility and QoL-AD scores decreased as AD severity increased (p <0.0001). Mean health utility scores were 0.65 (mild), 0.51 (moderate), and 0.25 (severe). Mean QoL-AD scores were 26.7 (mild), 23.0 (moderate), and 17.4 (severe). The correlations between health utility and QoL-AD scores were moderate to strong (r ≥ 0.62).ConclusionsUsing the vignettes, the general public provided HRQoL estimates that discriminated between the three AD health states. This finding suggests the general public may be a promising source of proxy HRQoL estimates in place of persons with AD.Electronic supplementary materialThe online version of this article (doi:10.1186/s12877-016-0207-4) contains supplementary material, which is available to authorized users.

Does promoting research advance planning in a general elderly population enhance completion of a research directive and proxies' predictive ability? a randomized controlled trial

ABSTRACTBackground: Communicating wishes regarding future research participation in the event of incapacity may help others make enrollment decisions that are more consistent with these wishes. We tested whether an advance planning intervention (1) motivated older adults to document their wishes regarding future research participation and (2) guided proxies in making hypothetical research decisions that agreed with those of the older adult. Methods: This randomized trial involved 235 dyads of older adults and proxies. The experimental intervention consisted of a group information session on advance directives, coupled with two encounters with a facilitator who helped older adults communicate their wishes to their proxy. Control participants attended three educational workshops on healthy behaviors. Short descriptions of hypothetical studies were used at baseline and twice post intervention to elicit older adults' and proxies' decisions regarding study enrollment. Answers were elicited under four health states of increasing cognitive impairment. Results: By the end of the trial, 80% of older adults allocated to the experimental group had recorded their preferences; 61% indicated a willingness to participate in clinical research should they be solicited after losing decisional capacity. At baseline, percentages of agreement on enrollment decisions ranged from 63% for a blood draw in the older adult's current health state to 24% for a dementia drug trial. From baseline to the second posttest, agreement increased in the experimental group while declining among controls. The positive effect of the intervention was restricted to the severe dementia health state, with agreement rising from 26 to 37%. Conclusions: The intervention motivated older adults to document their preferences regarding future research involvement. It had little effect on proxies' ability to make hypothetical research decisions that match those of the older adult. It remains to be seen whether documented preferences would facilitate substitute decision making in hypothetical and actual research situations.

Valuing QALYs in Relation to Equity Considerations Using a Discrete Choice Experiment.

BackgroundTo judge whether an intervention offers value for money, the incremental costs per gained quality-adjusted life-year (QALY) need to be compared with some relevant threshold, which ideally reflects the monetary value of health gains. Literature suggests that this value may depend on the equity context in which health gains are produced, but the value of a QALY in relation to equity considerations has remained largely unexplored.ObjectiveThe objective of this study was to estimate the social marginal willingness to pay (MWTP) for QALY gains in different equity subgroups, using a discrete choice experiment (DCE). Both severity of illness (operationalized as proportional shortfall) and fair innings (operationalized as age) were considered as grounds for differentiating the value of health gains.MethodsWe obtained a sample of 1205 respondents, representative of the adult population of the Netherlands. The data was analysed using panel mixed multinomial logit (MMNL) and latent class models.ResultsThe panel MMNL models showed counterintuitive results, with more severe health states reducing the probability of receiving treatment. The latent class models revealed distinct preference patterns in the data. MWTP per QALY was sensitive to severity of disease among a substantial proportion of the public, but not to the age of care recipients.ConclusionThese findings emphasize the importance of accounting for preference heterogeneity among the public on value-laden issues such as prioritizing health care, both in research and decision making. This study emphasises the need to further explore the monetary value of a QALY in relation to equity considerations.Electronic supplementary materialThe online version of this article (doi:10.1007/s40273-015-0311-x) contains supplementary material, which is available to authorized users.

Elicitation of health state utilities associated with varying severities of flares in Systemic Lupus Erythematosus.

BackgroundSystemic Lupus Erythematosus (SLE) is characterised by fluctuating periods of minimal disease activity and ‘flare’. Flare is an important outcome variable impacting the disease burden associated with SLE. The objective of this study was to obtain population-based utility values for varying severities of flare to measure the impact on health-related quality of life (HRQoL) in Australia, Canada, France, Japan, Spain and the UK.MethodsSix health states (HS) for varying severities of flare were developed based on literature, patient blogs, and interviews with patients (n = 12), rheumatologists (n = 7) and nurses (n = 2). HS were validated by independent clinical experts (n = 6) and pilot interviews (n = 10, UK). HS were evaluated using the time-trade-off (TTO) method during face-to-face interviews with a minimum representative sample (n = 100) of the general population, per-country. Visual Analog Scale (VAS) scores were obtained to validate TTO scores. TTO scores were converted into utility values.ResultsThe highest mean TTO utility scores were observed for the anchor HS (minimal disease activity) across all countries; means ranged from 0.66 in Japan to 0.82 in UK. All flare HS were associated with a disutility compared with the anchor HS (p < 0.001), means ranged across countries: mild flare HS: 0.55–0.71, moderate flare HS: 0.38–0.53, severe renal flare HS: 0.33–0.45, severe central nervous system (CNS) flare HS: 0.30–0.45 and severe generalised flare HS: 0.19–0.33. Mean VAS scores followed the same trend.ConclusionsThese results show increasing severity of flare has a detrimental impact on HRQoL. The severe generalised flare HS received the lowest mean utility score suggesting that the perceived day-to-day impact of a severe generalised flare was greater than a severe CNS or severe renal flare. This is, to the best of our knowledge, the first utility study to assess varying severities of flare in SLE across six different countries.

Utility estimation of hypothetical chronic obstructive pulmonary disease health states by the general population and health professionals.

PurposeThis study attempted to estimate the utility weights for hypothetical chronic obstructive pulmonary disease (COPD) health states, including the effect of exacerbation, and based on utilities elicited from a representative sample using the time trade-off (TTO).MethodsA total of 200 study subjects were recruited using a quota sampling method in Seoul, Korea. Ten COPD health profiles were described reflecting the severity of COPD and the extent of exacerbation. Respondents evaluated each health state using a visual analogue scale and TTO during a personal interview. TTO values were estimated using a linear mixed model, and the model performance was evaluated in terms of its predictive ability and goodness of fit.ResultsThe estimated TTO values were 0.824 in moderate, 0.646 in severe, and 0.305 in very severe COPD health states. The estimated utility decrements in TTO varied from 0.082 for a non-serious exacerbation to 0.228 for one non-serious plus one serious exacerbation per year. The mean absolute error of the TTO model was 0.008, and the generalized R2 was 0.86.ConclusionThe social preference of various COPD health states and the utility decrement due to exacerbation can be useful for the economic evaluation of COPD intervention in Korea.Electronic supplementary materialThe online version of this article (doi:10.1186/s12955-015-0228-2) contains supplementary material, which is available to authorized users.

Mapping the Clinical Chronic Obstructive Pulmonary Disease Questionnaire onto Generic Preference-Based EQ-5D Values

ObjectivesTo develop a model to predict EuroQol five-dimensional questionnaire (EQ-5D) values from clinical chronic obstructive pulmonary disease (COPD) questionnaire (CCQ) scores. MethodsWe used data from three clinical trials (the Randomized Clinical Trial on Effectiveness of Integrated COPD Management in Primary Care [RECODE], the Assessment Of Going Home Under Early Assisted Discharge [GO-AHEAD], and the Health Status Guided COPD Care [MARCH]). Data were randomly split into an estimation sample and a validation sample. The conceptual similarity between patient-reported CCQ and preference-based EQ-5D scores was assessed using correlation and principal-component analysis. Different types of models were estimated with increasing complexity. We selected the final models on the basis of mean absolute error and root mean square error when comparing predicted and observed values from the same population (internal validity) and from different trial populations (external validity). We also developed models for different country-specific EQ-5D value sets. ResultsThe principal-component analysis showed that the CCQ domains functional state and mental state are associated with four dimensions of the EQ-5D. The EQ-5D dimension pain/discomfort formed a separate construct on which no CCQ item loaded. The mean observed EQ-5D values were not significantly different from the mean predicted EQ-5D values in internal validation samples but did significantly differ in external validation samples. The models underestimated EQ-5D values in milder health states and overestimated them in more severe health states. The predictive ability of the models was similar across different EQ-5D value sets. ConclusionsThe models can predict mean EQ-5D values that are similar to observed mean values in a similar population. The overestimating/underestimating of the low/high EQ-5D values, however, limits its use in Markov models. Therefore, mapping should be used cautiously.

PSS32 - Non-Proliferative Diabetic Retinopathy: is It Cost-Effective to Treat Early?

Objectives Diabetic retinopathy (DR) is a leading cause of sight loss in people of working age. There have been recent developments in laser photocoagulation techniques, along with anti-vascular endothelial growth factor drugs for the treatment of severe non-proliferative and proliferative DR. The aim of the study was to evaluate the cost-effectiveness of panretinal photocoagulation administered at the severe non-proliferative diabetic retinopathy (NPDR) stage (early treatment), compared with waiting until high-risk proliferative (HR-PDR) characteristics (deferred treatment) developed. Methods A Markov model with a 30-year time horizon was developed, with clinical pathway options for patients presenting with moderate NPDR through to irreversible-severe vision loss and blindness (and to death). Once patients entered a post-treatment health state they can progress to more severe health states, regress back to earlier stages of the disease, persist where they are, or die. NHS and personal social services perspective was adopted. Transition probabilities were based mainly on data derived from the Early Treatment Diabetic Retinopathy Study. Health state utilities, costs and complications were based on information from the literature, supplemented by expert opinion. Costs and outcomes were discounted at 3.5%. Both deterministic and probabilistic sensitivity analyses were conducted. Results Administering panretinal photocoagulation at the severe NPDR stage was more effective and less costly than waiting until HR-PDR developed. Sensitivity analyses gave similar results, with early treatment continuing to dominate deferred treatment. The probabilistic sensitivity analysis suggests that at willingness-to-pay threshold of £20-£30,000 per quality-adjusted life year, the probability of early treatment being cost-effective is 60%. Conclusions Panretinal photocoagulation administered at the severe NPDR stage is likely to be cost-effective. However, given the limitations of the evidence on current treatments, these results to be interpreted with caution. A trial of early versus deferred laser therapy is needed to provide better data based on modern treatments.

Cost-Utility Analysis Of Fidaxomicin Compared To Vancomycin In The Management Of Severe Clostridium Difficile Infection In Poland.

In recent years a number of infections caused by Clostridium difficile has been significantly increasing. In Poland oral metronidazole constitutes the therapy of choice of non-severe infection and first-recurrence, while oral vancomycin is recommended to be given in case of severe disease and subsequent recurrences. Fidaxomicin is a novel treatment for Clostridium difficile infections (CDI). The aim of this study was to perform a cost-utility analysis of fidaxomicin for the treatment of severe CDI compared to vancomycin. A meta-analysis of two randomized clinical trials phase III comparing oral fidaxomicin and oral vancomycin in CDI was conducted. A Markov model was used to determine the cost-utility of fidaxomicin in patients with severe CDI. The cycle length was 10 days and the time horizon was 1 year. The patient entered the model in the severe CDI health state and was given either fidaxomycin or vancomycin for 10 days. The analysis was performed from the third-party payer perspective – the Polish National Health Fund. Only direct health care costs (drug costs, hospitalization) were included. Given the lack of formal utility measures for CDI, the utilities for the alternative health states described in the literature were adapted. In the base case, fidaxomicin was dominant compared to vancomycin, resulting in cost savings of PLN 905 and an incremental QALY gain of 0.015. Fidaxomicin was associated with higher cost savings (PLN 30,883) assuming that patients with severe CDI would be hospitalized at intensive care unit. One-way sensitivity analyses revealed that fidaxomicin remained dominant even if considering marginal values of both antibiotics’ acquisition cost. Fidaxomicin was dominant compared to vancomycin, generating additional QALYs with cost-savings in severe CDI patients in Poland.

How is the most severe health state being valued by the general population?

BackgroundIt has been reported that valuation of health states that are close to death, such as the most severe health state, can be affected by health state valuation procedure, and their utility values are difficult to predict. We examined how the most severe health states of Short Form-6 dimension (SF-6D) and EuroQoL-5 dimension-3 level (EQ-5D-3L) were valued by the Singapore general population.MethodsOverall, 249 SF-6D and 42 EQ-5D-3L states were valued by two separate samples from the Singapore general population using the visual analogue scale (VAS) method. Ordinary least-square regression model was employed to explain deficit in the valuation of the most severe state using the health state descriptors.ResultsA total of 1021 participants from the SF-6D sample and 1015 participants from the EQ-5D-3L sample were included in the analysis. We observed that 67% of the SF-6D participants and 74% of the EQ-5D-3L participants considered the most severe state worse than dead. The most severe state had mean VAS valuation scores more than 20–25 points lower than the adjacent states that are better by only one level in only one dimension. SF-6D VAS valuation score for the most severe state was 27 points and 12 points lower than expected according to the health state descriptors among the participants who considered the most severe state worse than dead and better than dead, respectively. Similar results were found for the EQ-5D-3L valuation.ConclusionsThe most severe health state was valued lower than expected according to its descriptors.

Quantifying Long-Term Care Preferences

. Current policies redirecting long-term care (LTC) delivery away from institutional care to home- and community-based services are being made in the absence of crucial evidence on preferences. Studies indicate that the shift to home care is generally not cost-saving; thus, an empirical assessment of effectiveness is needed to evaluate policies incenting home care investment. This study quantifies LTC preferences between different delivery modes. . This study extended the time tradeoff method to elicit utilities and LTC preferences associated with the receipt of different modes of LTC services, conditional on health states defined by varying levels of functional and cognitive impairment. Users' LTC preferences are measured as differential utilities between alternative LTC options for each health state. . For the same health state, respondents (n = 81) significantly preferred home care over institutional care, except for the most impaired health state. The preference for home care over institutional care is quantified as 0.30 quality-of-life (QOL) weight when people need help with only 1 activity of daily living (ADL). The preference for home care depends significantly on levels of disability and was weaker once the need for help became greater. Under the most severe health state of having moderate to severe dementia and needing help with 6 ADLs, the quantified home care preference was only 0.03 QOL weight and was not statistically significant. . Because the sample is mostly composed of African Americans, the results may not be generalizable to other racial and ethnic groups. . People do not always strongly prefer home care over institutional care, as is often assumed. The costs of expanding home- and community-based care should be weighed against these preferences.

Differences in EQ-5D-3L health state valuations among patients with musculoskeletal diseases, health care professionals and healthy volunteers.

Results from existing studies indicate that different respondent groups' health state valuations in cost-utility analyses are not equivalent. The objectives in our study were to analyse differences in health state valuations among three respondent groups in the context of medical rehabilitation in Germany. Using the time trade-off (TTO) technique, valuations of EQ-5D-3L health states were obtained from patients with musculoskeletal diseases, healthy volunteers and health care professionals. We used linear mixed models to predict TTO utilities and specified and tested interaction effects. We identified statistically significant (p < 0.05) differences among the three groups in six out of 42 health states. On average, patients' TTO values were somewhat higher compared with other respondent groups. Most of these differences occurred in severe health states. Mean differences and mean absolute differences were 0.02 and 0.14 for patients vs healthy volunteers and 0.06 and 0.14 for patients vs health care professionals. Furthermore, significant effects among respondents were observed for seven of the 22 possible interactions describing differences between respondent groups. Coefficients associated with significant interaction effects ranged from 0.08 to 0.18 (absolute values). The results of our study suggest that TTO valuations of health states differ depending on the specific respondent group from which valuations are obtained. On average, these differences were small. However, researchers and decision makers should remain aware of these differences when interpreting incremental cost-utility assessments.

Valuations of epilepsy-specific health states: a comparison of patients with epilepsy and the general population

AimsUtility values that can be used in the economic evaluation of treatments for epilepsy can be elicited from the general population and the patient population, but it is unclear how the health state values differ. The aim of this study is to compare the preferences of the general population and a sample of people with epilepsy for health states described by the NEWQOL-6D QALY measure. MethodsThe Time Trade Off preference elicitation technique was used to value eight NEWQOL-6D health states. The general population sample was recruited and interviewed in their homes, and the sample with epilepsy was recruited and interviewed in an epilepsy service in North West England. Descriptive analysis and regression modeling were used to compare health state values across the populations. ResultsA sample of 70 people with epilepsy and a sample of 60 members of the general population were included. The populations differed across a range of background characteristics, but there were limited differences between the health state values. Patients provided significantly higher (better) values for the most severe health state described by the NEWQOL-6D (p<0.01) and nonsignificant higher values for states with intermediate severity. The general population health state value was only higher for the best health state described by the NEWQOL-6D. ConclusionsThe similarities in the patient and general population values for NEWQOL-6D health states suggest that the use of the general population utility weights for the estimation of QALYs in the economic evaluation of epilepsy interventions is appropriate and largely representative of patient preferences.

Does diabetes have an impact on health-state utility? a study of Asians in Singapore.

Our objective was to compare the time trade-off (TTO) values of EQ-5D-3L health states elicited from Singaporeans with and without type 2 diabetes mellitus (T2DM) and T2DM patients with and without complications. The TTO values of ten EQ-5D-3L health states were elicited from a consecutive sample of T2DM patients and a general Singaporean population sample using similar valuation protocols. In face-to-face interviews, T2DM patients and members of the general population were asked to value five and ten health states, respectively. The difference in TTO values between the two samples and between T2DM patients with and without complications was examined using multiple linear regression models. A total of 109 T2DM patients and 46 individuals without T2DM provided data. All ten health states considered, the mean TTO value was -0.02 for the general population sample and -0.04 for T2DM patients, with the unadjusted and adjusted difference being -0.06 (95 % confidence interval [CI] -0.16, 0.03) and 0.02 (95 % CI -0.12, 0.15). The general population sample had systematically lower TTO values for mild health states, with the adjusted difference being -0.13 (95% CI -0.25, -0.02); while the two samples had similar TTO values for severe health states, with the adjusted difference being 0.02 (95% CI -0.16, 0.19). T2DM patients without complications had systematically lower TTO values than those with complications, with the adjusted difference being -0.10 (95% CI -0.23, 0.03). It appears that diabetes and its complications affect patients' valuation of health states. Hence, the EQ-5D-3L health-state values based on the general population may underestimate the utility of health interventions for T2DM.

A Comparison of Direct and Indirect Methods for the Estimation of Health Utilities from Clinical Outcomes

Analysts frequently estimate health state utility values from other outcomes. Utility values like EQ-5D have characteristics that make standard statistical methods inappropriate. We have developed a bespoke, mixture model approach to directly estimate EQ-5D. An indirect method, "response mapping," first estimates the level on each of the 5 dimensions of the EQ-5D and then calculates the expected tariff score. These methods have never previously been compared. We use a large observational database from patients with rheumatoid arthritis (N = 100,398). Direct estimation of UK EQ-5D scores as a function of the Health Assessment Questionnaire (HAQ), pain, and age was performed with a limited dependent variable mixture model. Indirect modeling was undertaken with a set of generalized ordered probit models with expected tariff scores calculated mathematically. Linear regression was reported for comparison purposes. Impact on cost-effectiveness was demonstrated with an existing model. The linear model fits poorly, particularly at the extremes of the distribution. The bespoke mixture model and the indirect approaches improve fit over the entire range of EQ-5D. Mean average error is 10% and 5% lower compared with the linear model, respectively. Root mean squared error is 3% and 2% lower. The mixture model demonstrates superior performance to the indirect method across almost the entire range of pain and HAQ. These lead to differences in cost-effectiveness of up to 20%. There are limited data from patients in the most severe HAQ health states. Modeling of EQ-5D from clinical measures is best performed directly using the bespoke mixture model. This substantially outperforms the indirect method in this example. Linear models are inappropriate, suffer from systematic bias, and generate values outside the feasible range.

QALY and Productivity Loss: Empirical Evidence for “Double Counting”

ObjectivesWhen quality-adjusted life-years are used for economic evaluation, the controversial issue of “double counting” of productivity loss emerges, particularly given the lack of empirical data. MethodsIn this study, we performed a Web-based, large-sample survey to address the issue of double counting. To determine the influence of income reduction on utility scores, we obtained utility scores of eight health states with three instruction types: a) no instruction, b) instructed to consider income reduction, and c) instructed not to consider income reduction (compensated). Respondents were randomly sampled from the online panel and asked to evaluate 1 of 24 patterns by both standard gamble and time trade-off methods. ResultsA total of 6551 respondents completed the questionnaire. First, despite the lack of instruction on income reduction, many respondents spontaneously assumed lost income. The proportion tended to be higher when considering more severe health states. Second, the degree of assumed income reduction was related to utility scores. For a 10% income reduction, respondents assumed a 0.02- to 0.04-decrease in utility score (both standard gamble and time trade-off methods). Third, utility scores did not change significantly when instruction was given not to consider income reduction (compensated) compared with when no instruction was given. ConclusionsAn assumed income reduction clearly influenced utility scores; however, compensation for lost income failed to sufficiently improve utility scores. In our view, the effect of income on utility scores does not only reflect wage loss. Our results suggest that the impact of double counting is negligible.