What Is a Valid Mapping Algorithm in Cost-Utility Analyses? A Response From a Missing Data Perspective

ObjectivesAlthough numerous mapping algorithms from a non–preference-based measure to a target health utility measure have been developed and applied in cost-utility analyses (CUAs), conditions for a mapping algorithm to work well in a CUA are still unclear. In this research, we formulate the mapping problem as a missing data problem and clarify these conditions.MethodsWe defined a valid mapping algorithm based on the purpose of mapping (ie, not for prediction but for CUA), and derived a sufficient set of conditions for a valid mapping algorithm. We also conducted a simulation study to investigate properties of a mapping algorithm under situations where the conditions are satisfied and violated.ResultsThe derived sufficient conditions indicate that the complete overlap of the source measure with the target health utility measure is important and that a covariate that is omitted from a mapping algorithm but has an effect on the target health utility measure not captured by the source measure may invalidate a mapping algorithm. The conditions cannot be verified from data in a CUA but can be supported using external data. A simulation study showed that when at least 1 of the 3 conditions was violated, a mapping algorithm provided biased health utility estimates in a CUA, and that prediction accuracy did not necessarily reflect performance of a mapping algorithm in a CUA.ConclusionThe derived conditions provide a fundamental basis for better practices in developing and selecting a mapping algorithm.     

Quality of Life Related to Oral versus Subcutaneous Iron Chelation: A Time Trade-off Study

ObjectiveTo investigate the utility associated with subcutaneous infusion (deferoxamine) compared with once-daily oral administration (deferasirox) of iron chelation therapy.MethodsInterviews using the time trade-off technique were used to estimate preferences (utility) for health states by finding the point at which respondents were indifferent between a longer but lower quality of life (QoL) and a shorter time in full health. Participants (n = 110) were community-based, 51% women, median age 35 years, from four regions in Sydney, Australia. Respondents rated three health states involving equal outcomes for people with thalassemia but with different treatment modalities for iron chelation; an “anchor state” describing a patient receiving iron chelation without administration mode specified, anchor state plus iron chelation via subcutaneous infusion, and anchor state plus iron chelation through once-daily oral medication.ResultsOn an interval scale between 0 (death) and 1 (full health), median (interquartile range) utility of 0.80 (0.65–0.95) for the anchor state, 0.66 (0.45–0.87) for subcutaneous infusion, and 0.93 (0.80–0.97) for once-daily oral administration was obtained. The mean (median) difference of 0.23 (0.27) between the two treatments was statistically significant (Wilcoxon-signed rank test, P < 0.001). Subcutaneous infusion was associated with a mean (median) utility 0.13 (0.14) lower than the anchor state (P < 0.001), and once-daily oral treatment had a utility 0.10 (0.13) higher (P < 0.001).ConclusionCommunity respondents associate oral administration of an iron chelator such as deferasirox with enhanced QoL compared with subcutaneous treatment. Assuming equal safety and efficacy, QoL gains from once-daily oral treatment compared with subcutaneous infusion are significant.     

Qaly league tables: Handle with care

This paper examines some of the difficulties in using QALY league tables in priority setting. Such tables sometimes are seen as being ‘the’ way to prioritise in health care and in particular, at present, with respect to priority setting among purchasers in the UK NHS. However the paper highlights the fact that the base on which such tables is built is small—relatively few studies in the English language using CUA have been conducted anywhere. Further, four issues which require handling with care are set out: (i) the relevant measure of cost in QALY league tables has to be restricted to health service resource use; (ii) the relevant measure of benefit in QALY league tables is clearly restricted to QALYs, thereby the utility of health gains and indeed the maximisation of the utility of health gains; (iii) in incorporating the results of CUA studies into QALY league tables there is a need for greater clarification on what the margin constitutes; and (iv) those who might use CUA results in QALY league tables need to ascertain whether the original context of the study will allow the results to be transferred to the local context of the decision maker. The paper suggests that there is a need to be quite clear what goal QALY league tables serve. The authors argue that the only legitimate (and clearly important) goal of QALY league tables is the maximisation of the utility of health gains within a health service budget. The thinking underlying QALY league tables, adjusted to take account of the caveats in this paper, is the key to rational priority setting at a local level. It is this thinking that is to be emphasised rather than the use of ‘imported’ league tables or the use of results from CUA studies conducted elsewhere in the country or indeed in other countries. If results from elsewhere are to be used, the study context has to be examined carefully to establish the extent of its relevance to the local circumstances facing the purchasing authority.     

Discrepancies between chained and classic utilities induced by anchoring with occasional adjustments.

BACKGROUND: Classic utility assessment uses death and perfect health as end points. Chained utility assessment uses other health states as endpoints. It has been previously noted that these 2 assessment procedures lead to different utilities. PURPOSE: The author attempts to explain these discrepancies between chained and classic assessments. METHOD: Previous data are plotted in a uniform way to facilitate comparison. Using time trade-off and paired-comparison data, the author estimates the extent to which respondents adjust their responses when end points are varied. Data were obtained in various samples: in healthy volunteers from the general public, in students, and in women at high risk for breast cancer seeking genetic counseling. RESULTS: The author obtained 741 valid data records from a total of 106 participants. The data replicate the pattern found previously. When compared to classic utilities, (1) chained utilities are smaller (larger) when the best (worst) endpoint varies and (2) the discrepancies become smaller for utilities near 0 and 1. The data reveal that there is a distinct failure to adjust responses when the end points are varied, as if the responses anchor on some master health scale. The latter finding explains the robust pattern of discrepancies. CONCLUSION: Decision analyses that use a mix of classic and chained utilities are not on firm ground. One should be wary of normative interpretations of new value assessment procedures. Alternative interpretations of the findings are discussed.     

Assessing Differences in Utility Scores: A Comparison of Four Widely Used Preference-Based Instruments

OBJECTIVES: To characterize the differences in utility scores (dUTY) among four commonly used preference-based Health-Related Quality of Life instruments, to evaluate the potential impact of these differences on cost-utility analyses (CUA), and to determine if sociodemographic/clinical factors influenced the magnitude of these differences. METHODS: Consenting adult Chinese, Malay and Indian subjects in Singapore were interviewed using Singapore English, Chinese, Malay or Tamil versions of the EQ-5D, Health Utilities Index Mark 2 (HUI2) and Mark 3 (HUI3), and SF-6D. Agreement between instruments was assessed using Bland-Altman (BA) plots. Changes in incremental cost-utility ratio (ICUR) from dUTY were investigated using eight hypothetical decision trees. The influence of sociodemographic/clinical factors on dUTY between instrument pairs was studied using multiple linear regression (MLR) models for English-speaking subjects (circumventing structural zero issues). RESULTS: In 667 subjects (median age 48 years, 59% female), median utility scores ranged from 0.80 (95% confidence interval [CI] 0.80, 0.85) for the EQ-5D to 0.89 (95% CI 0.88, 0.89) for the SF-6D. BA plots: Mean differences (95% CI) exceeded the clinically important difference (CID) of 0.04 for four of six pairwise comparisons, with the exception of the HUI2/EQ-5D (0.03, CI: 0.02, 0.04) and SF-6D/HUI2 (0.02, CI: 0.006, 0.02). Decision trees: The ICER ranged from $94,661/QALY (quality-adjusted life-year; 6.3% difference from base case) to 100,693 dollars/QALY (0.3% difference from base case). MLR: Chronic medical conditions, marital status, and Family Functioning Measures scores significantly (P-value < 0.05) influenced dUTY for several instrument pairs. CONCLUSION: Although CIDs in utility measurements were present for different preference-based instruments, the impact of these differences on CUA appeared relatively minor. Chronic medical conditions, marital status, and family functioning influenced the magnitude of these differences.     

Comparison of health state utility values derived using time trade-off, rank and discrete choice data anchored on the full health-dead scale

Recent years have seen increasing interest in the use of ordinal methods to elicit health state utility values as an alternative to conventional methods such as standard gamble and time trade-off (TTO). However, in order to use these ordinal methods to produce health state values for use in cost-effectiveness analysis using cost per quality adjusted life year (QALY) analysis, these values must be anchored on the full health-dead scale. The paper reports on two feasibility studies that use two approaches to anchor health state utility values derived from discrete choice data on the full health-dead scale: normalising using (1) the TTO value of the worst state and (2) the coefficient on the 'dead' dummy variable. Health state utility values obtained using rank and discrete choice data are compared to more commonly used TTO utility values for two condition-specific preference-based measures; asthma and overactive bladder. Ordinal methods were found to offer a promising alternative to conventional cardinal methods of standard gamble and TTO. There remains a large and important research agenda to address.     

Is there a case for using visual analogue scale valuations in cost‐utility analysis?

This paper critically reviews theoretical and empirical propositions regarding visual analogue scale (VAS) valuations of health states and their use in cost-utility analysis (CUA). A widely repeated assertion in the economic evaluation literature is the inferiority, on theoretical grounds, of VAS valuations. Five common criticisms are: VAS lacks a theoretical foundation; VAS values are not 'choice based'; VAS values are not consistent with utility-under-uncertainty requirements; context and range effects observed in VAS valuation data mean that they cannot even be considered to represent measurable value functions; and when completing a VAS, people are not trying to express values.We address each of these points: the VAS does have a theoretical basis, being entirely consistent with the non-welfarist foundations of QALYs and CUA; the 'choiceless' nature of the VAS is incorrectly judged by stated preference criteria relevant to monetary rather than health state valuations, and VAS valuations do in any case involve an element of choice; because valuations are intended for use in social decision-making, it may be advantageous that VAS values are elicited under conditions of certainty; although there are measurement problems with the VAS, means such as better design and transformations of data can deal with these; and with any method of eliciting values, it is unrealistic to expect people consciously to think in terms of social science constructs such as utilities.Moreover, there are problems, both theoretical and empirical, with alternative methods. Selection of the appropriate valuation method should be based on empirical performance, and in this the VAS has important advantages. We conclude that there are strong grounds for disputing the consensus view against the VAS and challenge those who hold it to deploy more convincing arguments and evidence in favour of alternative methods. However, we identify areas where further research is required to establish and consolidate the potential of the VAS as a valuation method.     

The minimal clinically important difference determined using item response theory models: an attempt to solve the issue of the association with baseline score

ObjectivesDetermining the minimal clinically important difference (MCID) of questionnaires on an interval scale, the trait level (TL) scale, using item response theory (IRT) models could overcome its association with baseline severity. The aim of this study was to compare the sensitivity (Se), specificity (Sp), and predictive values (PVs) of the MCID determined on the score scale (MCID-Sc) or the TL scale (MCID-TL).Study Design and SettingThe MCID-Sc and MCID-TL of the MOS-SF36 general health subscale were determined for deterioration and improvement on a cohort of 1,170 patients using an anchor-based method and a partial credit model. The Se, Sp, and PV were calculated using the global rating of change (the anchor) as the gold standard test.ResultsThe MCID-Sc magnitude was smaller for improvement (1.58 points) than for deterioration (−7.91 points). The Se, Sp, and PV were similar for MCID-Sc and MCID-TL in both cases. However, if the MCID was defined on the score scale as a function of a range of baseline scores, its Se, Sp, and PV were consistently higher.ConclusionThis study reinforces the recommendations concerning the use of an MCID-Sc defined as a function of a range of baseline scores.     

Estimating CHU-9D Utility Scores from the WAItE: A Mapping Algorithm for Economic Evaluation

Background

The Weight-Specific Adolescent Instrument for Economic Evaluation (WAItE) is a new condition-specific patient reported outcome measure that incorporates the views of adolescents in assessing the impact of above healthy weight status on key aspects of their lives. Presently it is not possible to use the WAItE to calculate quality adjusted life years (QALYs) for cost-utility analysis (CUA), given that utility scores are not available for health states described by the WAItE.

U.K. utility weights for the EORTC QLU‐C10D

The EORTC QLU‐C10D is a new multi‐attribute utility instrument derived from the widely used cancer‐specific quality of life questionnaire, EORTC QLQ‐C30. It contains 10 dimensions (physical functioning, role functioning, social functioning, emotional functioning, pain, fatigue, sleep, appetite, nausea, bowel problems), each with four levels. The aim of this study was to provide U.K. general population utility weights for the QLU‐C10D. A U.K. online panel was quota‐sampled to align the sample to the general population proportions of sex and age (≥18 years). The online valuation survey included a discrete choice experiment (DCE). Each participant was asked to complete 16 choice‐pairs, each comprising two QLU‐C10D health states plus duration. DCE data were analysed using conditional logistic regression to generate utility weights. Data from 2,187 respondents who completed at least one choice set were included in the DCE analysis. The final U.K. QLU‐C10D utility weights comprised decrements for each level of each health dimension. For nine of the 10 dimensions (all except appetite), the expected monotonic pattern was observed across levels: Utility decreased as severity increased. For the final model, consistent monotonicity was achieved by merging inconsistent adjacent levels for appetite. The largest utility decrements were associated with physical functioning and pain. The worst possible health state (the worst level of each dimension) is ?0.083, which is considered slightly worse than being dead. The U.K.‐specific utility weights will enable cost–utility analysis (CUA) for the economic evaluation of new oncology therapies and technologies in the United Kingdom, where CUA is commonly used to inform resource allocation.     

Are decisions using cost-utility analyses robust to choice of SF-36/SF-12 preference-based algorithm?

Background  

Cost utility analysis (CUA) using SF-36/SF-12 data has been facilitated by the development of several preference-based algorithms. The purpose of this study was to illustrate how decision-making could be affected by the choice of preference-based algorithms for the SF-36 and SF-12, and provide some guidance on selecting an appropriate algorithm.     

Comparison of the minimally important difference for two health state utility measures: EQ-5D and SF-6D

Background: The SF-6D and EQ-5D are both preference-based measures of health. Empirical work is required to determine what the smallest change is in utility scores that can be regarded as important and whether this change in utility value is constant across measures and conditions. Objectives: To use distribution and anchor-based methods to determine and compare the minimally important difference (MID) for the SF-6D and EQ-5D for various datasets. Methods: The SF-6D is scored on a 0.29–1.00 scale and the EQ-5D on a −0.59–1.00 scale, with a score of 1.00 on both, indicating ‘full health’. Patients were followed for a period of time, then asked, using question 2 of the SF-36 as our anchor, if their general health is much better (5), somewhat better (4), stayed the same (3), somewhat worse (2) or much worse (1) compared to the last time they were assessed. We considered patients whose global rating score was 4 or 2 as having experienced some change equivalent to the MID. This paper describes and compares the MID and standardised response mean (SRM) for the SF-6D and EQ-5D from eight longitudinal studies in 11 patient groups that used both instruments. Results: From the 11 reviewed studies, the MID for the SF-6D ranged from 0.011 to 0.097, mean 0.041. The corresponding SRMs ranged from 0.12 to 0.87, mean 0.39 and were mainly in the ‘small to moderate’ range using Cohen’s criteria, supporting the MID results. The mean MID for the EQ-5D was 0.074 (range −0.011–0.140) and the SRMs ranged from −0.05 to 0.43, mean 0.24. The mean MID for the EQ-5D was almost double that of the mean MID for the SF-6D. Conclusions: There is evidence that the MID for these two utility measures are not equal and differ in absolute values. The EQ-5D scale has approximately twice the range of the SF-6D scale. Therefore, the estimates of the MID for each scale appear to be proportionally equivalent in the context of the range of utility scores for each scale. Further empirical work is required to see whether or not this holds true for other utility measures, patient groups and populations.     

Do English and Chinese EQ-5D versions demonstrate measurement equivalence? an exploratory study

Background

Although multiple language versions of health-related quality of life instruments are often used interchangeably in clinical research, the measurement equivalence of these versions (especially using alphabet vs pictogram-based languages) has rarely been assessed. We therefore investigated the measurement equivalence of English and Chinese versions of the EQ-5D, a widely used utility-based outcome instrument.

Methods

In a cross-sectional study, either EQ-5D version was administered to consecutive outpatients with rheumatic diseases. Measurement equivalence of EQ-5D item responses and utility and visual analog scale (EQ-VAS) scores between these versions was assessed using multiple regression models (with and without adjusting for potential confounding variables), by comparing the 95% confidence interval (95%CI) of score differences between these versions with pre-defined equivalence margins. An equivalence margin defined a magnitude of score differences (10% and 5% of entire score ranges for item responses and utility/EQ-VAS scores, respectively) which was felt to be clinically unimportant.

Results

Sixty-six subjects completed the English and 48 subjects the Chinese EQ-5D. The 95%CI of the score differences between these versions overlapped with but did not fall completely within pre-defined equivalence margins for 4 EQ-5D items, utility and EQ-VAS scores. For example, the 95%CI of the adjusted score difference between these EQ-5D versions was -0.14 to +0.03 points for utility scores and -11.6 to +3.3 points for EQ-VAS scores (equivalence margins of -0.05 to +0.05 and -5.0 to +5.0 respectively).

Conclusion

These data provide promising evidence for the measurement equivalence of English and Chinese EQ-5D versions.

     

A Bayesian Approach for Incorporating the EQ-5D Visual Analog Scale When Estimating the Health-Related Quality of Life

ObjectivesThe EuroQoL 3-level version of EQ-5D and 5-level version of EQ-5D questionnaires are often used to quantify health states. They include ordinal responses across 5 health dimensions (EQ-5D index) and an EQ-visual analog scale (EQ-VAS) overall health rating. We investigated the value of incorporating the EQ-VAS to update health utility estimates using a Bayesian framework.MethodsWe created a joint bivariate normal EQ-VAS and EQ-5D index utility model and compared this to a univariate normal EQ-5D index utility model. We tested these models for 1026 Sri Lankan patients with chronic kidney disease and 94 Australian patients with wounds. We validated our approach by simulating EQ-VAS and EQ-5D index responses and applying our Bayesian model and then comparing the modeled estimates to our observed data.ResultsThe combined model showed a reduction in estimate uncertainty for all respondents. Compared with the EQ-5D index-only model, the mean utility for Sri Lankan respondents dropped from 0.556 (0.534-0.579) to 0.540 (0.521-0.559) in men and increased from 0.489 (0.461-0.518) to 0.528 (0.506-0.550) in women, with reduced credible interval width by 13% and 23%, respectively. The mean utility in Australian respondents moved from 0.715 (0.633-0.800) to 0.716 (0.652-0.782) in men, and 0.652 (0.581-0.723) to 0.652 (0.593-0.711) in women, with reduced credible interval width by 23% and 17%, respectively. The credible interval width for simulated data also narrowed, ranging from 8.3 to 8.5%.ConclusionsIncluding the EQ-VAS through Bayesian methods can add value by reducing requisite sample sizes and decision uncertainty using small amounts of additional data that is often collected but rarely used.     

Raising the quality of cost-utility analyses: lessons learnt and still to learn.

Heightened awareness by health care funders of the need to find more efficient ways of using scarce health care resources has led to greater demand for evidence of cost-effectiveness. Implicit in this demand is that evidence is generated using clear reporting and accepted methods. The research reported here updates an earlier review of published cost-utility analyses (CUAs) to address whether previously identified gaps in reporting have diminished over time. Raising CUA standards requires systematic and regular reviews of published material to allow adequate monitoring and evaluation. There is also a need to 'appraise the appraisers' in the sense of reviewing peer-review processes. This is particularly so in those journals which are growing in importance as outlets for economic evaluation information. The findings from this study indicate continuing variation in the quality of reporting. At the lower end of this spectrum improvements could be made in the reporting of comparators, in the clarity of effectiveness evidence, in the assignment of utility weights to health states and in reporting of sensitivity analysis. CUAs published in peer-reviewed specialist medical journals were more likely to be lower in quality suggesting guidance on the appraisal of economic submissions needs to be extended to the editors of these particular journals. These findings could be used to help to target attempts to raise the quality of evidence-based CUA information.     

Health outcome measures used in cost-effectiveness studies: a review of original articles published between 1986 and 1996

Theoretically, the preferred type of health economic evaluation is the cost-benefit approach in which costs as well as benefits are measured in monetary units. This type of analysis is rarely found in practice, however, where cost-effectiveness analysis (CEA), cost-utility analysis (CUA) and other forms of economic evaluations are instead favored. The use of quality adjusted life-years (QALYs) or life-years gained, if applicable, is generally recommended in CUA/CEA because these measures will make possible broad comparisons with other studies as well as with norms regarding society's willingness-to-pay for health benefits. The purpose of this paper is to study the choice of health outcome measures and the extent to which results from CUA and CEA are discussed from such a willingness-to-pay perspective. Based on the analysis of a sample of 455 studies included in the Health Economic Evaluations Database (HEED), it is concluded that major differences exist in the choice of health outcome measures across disease categories. There is no evidence that QALYs or life-years gained have become more common over the years and CEAs using intermediary outcome measures are as common as those using life-years gained. Furthermore, studies using QALYs or life-years gained often lack a relevant discussion of society's willingness-to-pay per QALY or life-years gained.     

Anchor Points in Transitions to a New School Environment

This exposition provides a theoretical framework for designing and implementing prevention programs for students in school-related transitions (e.g., entering kindergarten, transition to high school). For this purpose, an anchor point is defined in an ecological and developmental perspective as an element of the person-in-environment system that facilitates transaction between the person and the environment. Depending upon the context, anchor points can lead to adaptive transaction or maladaptive transaction. Features and examples of anchor points are discussed and proposed are procedures to plant positive anchor points which lead to adaptive transaction and children's development in a new person-in-environment system. Finally, its strengths in prevention programs for school-related transitions are addressed.     

Economic Evaluation of New Models of Care: Does the Decision Change Between Cost-Utility Analysis and Multi-Criteria Decision Analysis?

ObjectivesTo experiment with new approaches of collaboration in healthcare delivery, local authorities implement new models of care. Regarding the local decision context of these models, multi-criteria decision analysis (MCDA) may be of added value to cost-utility analysis (CUA), because it covers a wider range of outcomes. This study compares the 2 methods using a side-by-side application.MethodsA new Dutch model of care, Primary Care Plus (PC+), was used as a case study to compare the results of CUA and MCDA. Data of patients referred to PC+ or care-as-usual were retrieved by questionnaires and administrative databases with a 3-month follow-up. Propensity score matching together with generalized linear regression models was used to reduce confounding. Univariate and probabilistic sensitivity analyses were performed to explore uncertainty in the results.ResultsAlthough both methods indicated PC+ as the dominant alternative, complementary differences were observed. MCDA provided additional evidence that PC+ improved access to care (standardized performance score of 0.742 vs 0.670) and that improvement in health-related quality of life was driven by the psychological well-being component (standardized performance score of 0.710 vs 0.704). Furthermore, MCDA estimated the budget required for PC+ to be affordable in addition to preferable (€521.42 per patient). Additionally, MCDA was less sensitive to the utility measures used.ConclusionsMCDA may facilitate an auditable and transparent evaluation of new models of care by providing additional information on a wider range of outcomes and incorporating affordability. However, more effort is needed to increase the usability of MCDA among local decision makers.