Benjamin M. Craig

Cost-effectiveness of gastric bypass for severe obesity

PURPOSE To estimate the cost-effectiveness of gastric bypass in the treatment of severe obesity. SUBJECTS AND METHODS We performed a cost-effectiveness analysis of gastric bypass versus no treatment from the payer perspective. We discounted quality-adjusted life-years (QALYs), life-years, and cost during the patients lifetime. Our target group comprised women and men aged 35 to 55 years with a body mass index between 40 and 50 kg/m(2), and who did not have cardiovascular disease and in whom conservative bariatric therapies had been unsuccessful. RESULTS The base case cost-effectiveness ratios ranged from 5000 dollars to 16,100 dollars per QALY for women and from 10,000 dollars to 35,600 dollars per QALY for men, depending on age and initial body mass index. In a few subgroups of older, less obese men, variation in parameters such as loss of excess weight, obesity-related quality of life, complication rates, and perioperative mortality affected the cost-effectiveness ratios. Parameter variation did not result in meaningful changes in the remaining patients. CONCLUSION Gastric bypass is a cost-effective alternative to no treatment, providing substantial lifetime benefits in patients who are severely obese.

Deriving a Preference-Based Measure for Cancer Using the EORTC QLQ-C30

OBJECTIVE The European Organization for Research and Treatment of Cancer Core Quality of Life Questionnaire (EORTC QLQ-C30) is one of the most commonly used measures in cancer care but in its current form cannot be used in economic evaluation because it does not incorporate preferences. We address this gap by estimating a preference-based measure for cancer from the EORTC QLQ-C30. METHODS Factor analysis, Rasch analysis, and other psychometric analyses were undertaken on a clinical trial dataset of 655 patients with multiple myeloma to derive a health state classification system amenable to valuation. Second a valuation study was conducted of 350 members of the UK general population using time trade-off. Mean and individual-level multivariate regression models were fitted to derive preference weights for the classification system. RESULTS The health state classification system has eight dimensions (physical functioning, role functioning, social functioning, emotional functioning, pain, fatigue and sleep disturbance, nausea, constipation, and diarrhea) with four or five levels each. Regression models have few inconsistencies (0 to 2) in estimated preference weights and small mean absolute error ranges (0.046 to 0.054). CONCLUSIONS It is feasible to derive a preference-based measure from the EORTC QLQ-C30 for use in economic evaluation. Future research will extend this to other countries and replicate across other patient groups.

Argentine Valuation of the EQ-5D Health States

OBJECTIVE To develop a set of health state values based on EuroQol EQ-5D instrument for the Argentine general population. METHODS Consecutive subjects attending six primary care centers in Argentina were selected based on quota sampling and were interviewed using the EuroQol Group protocol for measurement and valuation of health studies. Initially, the respondents were randomly assigned a unique card set; however, to improve efficiency, the subjects were later randomly assigned to one of three fixed sets of EQ-5D states. Using the visual analog scale (VAS) and time-trade off (TTO) responses for these states, we estimated a valuation model using ordinary least squares regression clustered by respondent. Predicted values for EQ-5D health states are compared with published values for the United States. RESULTS Six hundred eleven subjects were interviewed by 14 trained interviewers, rendering 6887 TTO and 6892 VAS responses. The model had an R(2) of 0.897 and 0.928 for TTO and VAS, respectively. The mean absolute difference between observed and predicted values was 0.039 for TTO and 0.020 for VAS, each showing a Lins concordance coefficient more than 0.98. Argentine and US TTO-predicted values were highly correlated (Pearsons rho = 0.963), although the average absolute difference was clinically meaningful (0.06), rejecting the US values for nearly two-thirds of the states (62.8%). The Argentine population placed lower values on mild states and higher values on severe states. CONCLUSION This study provides an Argentine value set that could be used locally or regionally, with meaningful and significant differences with that of the United States. Health policy in Latin America must incorporate local values for sovereignty and validity.

Modeling Ranking, Time Trade-Off, and Visual Analog Scale Values for EQ-5D Health States A Review and Comparison of Methods

Background:There is rising interest in eliciting health state valuations using rankings. Due to their relative simplicity, ordinal measurement methods may offer an attractive practical alternative to cardinal methods, such as time trade-off (TTO) and visual analog scale (VAS). In this article, we explore alternative models for estimating cardinal health state values from rank responses in a unique multicountry database. We highlight an estimation challenge pertaining to health states just below perfect health (the “nonoptimal gap”) and propose an analytic solution to ameliorate this problem. Methods:Using a standardized protocol developed by the EuroQol Group, rank, VAS, and TTO responses were collected for 43 health states in 8 countries: Slovenia, Argentina, Denmark, Japan, Netherlands, Spain, United Kingdom, and United States, yielding a sample of 179,431 state responses from 11,483 subjects. States were described using the EQ-5D system, which allows for 3 different possible levels on 5 different dimensions of health. We estimated conditional logit and probit regression models for rank responses. The regressions included 17 health state attribute variables reflecting specific levels on each dimension and counts of different levels across dimensions. This flexible specification accommodates previously published valuation models, such as models applied in the United Kingdom and United States. In addition to fitting standard conditional logit and probit models, which assume equal variance across health states (homoscedasticity), we examined a heteroscedastic probit model that assumes no variance for the 2 points anchoring the scale (“optimal health” and “dead”) and relaxes the equal-variance assumption for all other states. Rank-based predictions for the 243 unique states defined by the EQ-5D system were compared with predictions from conventional linear models fitted to TTO and VAS responses. Results:By construction, the TTO and VAS models assume no variance around the anchoring states of optimal health and dead. Mimicking this assumption in the probit rank models helps dissolve the nonoptimal gap. For all other states, variances in TTO and VAS were negatively associated with mean values, which contradict the assumption of homoscedasticity. Estimated health state values from the heteroscedastic probit model for the ranking data were highly correlated with predictions from both TTO and VAS models for the 243 EQ-5D states. Between VAS and rank-based estimates, Lins &rgr;, a measure of agreement, was over 0.98 with a mean absolute difference of 0.028. Corresponding measures of agreement between rank and TTO estimates were 0.96 and 0.12, which is similar to the agreement between VAS and TTO. Conclusions:Rank-based valuation techniques, which offer advantages of flexibility, generalizability, and ease of administration, may be attractive substitutes for TTO and VAS in the measurement of societal values for health outcomes.

US Valuation of Health Outcomes Measured Using the PROMIS-29

OBJECTIVES Health valuation studies enhance economic evaluations of treatments by estimating the value of health-related quality of life (HRQOL). The Patient-Reported Outcomes Measurement Information System (PROMIS) includes a 29-item short-form HRQOL measure, the PROMIS-29. METHODS To value PROMIS-29 responses on a quality-adjusted life-year scale, we conducted a national survey (N = 7557) using quota sampling based on the US 2010 Census. Based on 541 paired comparisons with over 350 responses each, pair-specific probabilities were incorporated into a weighted least-squared estimator. RESULTS All losses in HRQOL influenced choice; however, respondents valued losses in physical function, anxiety, depression, sleep, and pain more than those in fatigue and social functioning. CONCLUSIONS This article introduces a novel approach to valuing HRQOL for economic evaluations using paired comparisons and provides a tool to translate PROMIS-29 responses into quality-adjusted life-years.

The episodic random utility model unifies time trade-off and discrete choice approaches in health state valuation

BackgroundTo present an episodic random utility model that unifies time trade-off and discrete choice approaches in health state valuation.MethodsFirst, we introduce two alternative random utility models (RUMs) for health preferences: the episodic RUM and the more common instant RUM. For the interpretation of time trade-off (TTO) responses, we show that the episodic model implies a coefficient estimator, and the instant model implies a mean slope estimator. Secondly, we demonstrate these estimators and the differences between the estimates for 42 health states using TTO responses from the seminal Measurement and Valuation in Health (MVH) study conducted in the United Kingdom. Mean slopes are estimates with and without Dolans transformation of worse-than-death (WTD) responses. Finally, we demonstrate an exploded probit estimator, an extension of the coefficient estimator for discrete choice data that accommodates both TTO and rank responses.ResultsBy construction, mean slopes are less than or equal to coefficients, because slopes are fractions and, therefore, magnify downward errors in WTD responses. The Dolan transformation of WTD responses causes mean slopes to increase in similarity to coefficient estimates, yet they are not equivalent (i.e., absolute mean difference = 0.179). Unlike mean slopes, coefficient estimates demonstrate strong concordance with rank-based predictions (Lins rho = 0.91). Combining TTO and rank responses under the exploded probit model improves the identification of health state values, decreasing the average width of confidence intervals from 0.057 to 0.041 compared to TTO only results.ConclusionThe episodic RUM expands upon the theoretical framework underlying health state valuation and contributes to health econometrics by motivating the selection of coefficient and exploded probit estimators for the analysis of TTO and rank responses. In future MVH surveys, sample size requirements may be reduced through the incorporation of multiple responses under a single estimator.

Comparison of US panel vendors for online surveys.

Background Despite the increasing use of panel surveys, little is known about the differences in data quality across panels. Objective The aim of this study was to characterize panel survey companies and their respondents based on (1) the timeliness of response by panelists, (2) the reliability of the demographic information they self-report, and (3) the generalizability of the characteristics of panelists to the US general population. A secondary objective was to highlight several issues to consider when selecting a panel vendor. Methods We recruited a sample of US adults from 7 panel vendors using identical quotas and online surveys. All vendors met prespecified inclusion criteria. Panels were compared on the basis of how long the respondents took to complete the survey from time of initial invitation. To validate respondent identity, this study examined the proportion of consented respondents who failed to meet the technical criteria, failed to complete the screener questions, and provided discordant responses. Finally, characteristics of the respondents were compared to US census data and to the characteristics of other panels. Results Across the 7 panel vendors, 2% to 9% of panelists responded within 2 days of invitation; however, approximately 20% of the respondents failed the screener, largely because of the discordance between self-reported birth date and the birth date in panel entry data. Although geographic characteristics largely agreed with US Census estimates, each sample underrepresented adults who did not graduate from high school and/or had annual incomes less than US

Keep it simple: Ranking health states yields values similar to cardinal measurement approaches

15,000. Except for 1 vendor, panel vendor samples overlapped one another by approximately 20% (ie, 1 in 5 respondents participated through 2 or more panel vendors). Conclusions The results of this head-to-head comparison provide potential benchmarks in panel quality. The issues to consider when selecting panel vendors include responsiveness, failure to maintain sociodemographic diversity and validated data, and potential overlap between panels.

US Valuation of the SF-6D

OBJECTIVES To examine the relationship between ordinal and cardinal valuation of health states. STUDY DESIGN AND SETTING We analyzed rank, visual analog scale (VAS), and time trade-off (TTO) responses for 52 health states defined using the EQ-5D classification system developed by the EuroQol Group. We analyzed 179,431 responses from 11,483 subjects in eight countries: Slovenia, Argentina, Denmark, Japan, Netherlands, Spain, United Kingdom, and United States. We first compared responses across methods by frequency of ties and values below dead. Ordinal associations between methods were evaluated using Spearmans correlation and Kendalls tau. Next, we estimated numerical values from rank responses using country-specific conditional logit models. After anchoring predicted values on a common scale, we further investigated the cardinal relationships between rank, VAS, and TTO-based values using Pearsons rho and quadratic regression. RESULTS For each country, rank responses are less likely than TTO responses to be tied and to indicate that states are worse than dead. In all countries, rank responses show a strong linear correlation with both TTO (Pearsons rho=0.88-0.99) and VAS (rho=0.91-0.98) responses. However, rank-based values imply greater decrements in health for mild states than cardinal values. CONCLUSIONS Illiteracy and innumeracy can hinder implementation of complex preference elicitation techniques in diverse settings and populations. These results indicate that ranking exercises may provide an attractive alternative for health-state valuation.

Infertility evaluation and treatment among women in the United States

Background. The original SF-6D valuation study collected 3503 standard gambled responses from 611 UK respondents to predict quality-adjusted life year (QALY) values. Methods. Using 19,980 paired comparison responses from 666 US respondents and a stacked probit model, the 25 coefficients of the original SF-6D multiattribute utility (MAU) regression were estimated, such that each coefficient represents a QALY decrement. The US QALY predictions were compared with UK predictions using 8428 SF-6D states in the US Medicare Health Outcomes Survey (MHOS), 1998 to 2003. Results. Twenty-two of the 25 decrements in the SF-6D MAU regression are statistically significant. The remaining decrements are insignificant based on US and UK results. The US and UK QALY predictions for the MHOS SF-6D states are remarkably similar given differences in experimental design, format, and sampling (Lin’s coefficient of agreement, 0.941; absolute mean difference, 0.043). Limitations. The underlying theoretical framework for the study design and econometric analysis builds from the episodic random utility model and the concept of QALYs and inherits their limitations. Conclusions. This study enhances the potential for US comparative effectiveness research by translating SF-6D states into US QALYs as well as improves upon discrete choice experiment design and econometric methods for health valuation.