Gérard de Pouvourville

The efficiency frontier approach to economic evaluation of health-care interventions

BACKGROUND IQWiG commissioned an international panel of experts to develop methods for the assessment of the relation of benefits to costs in the German statutory health-care system. PROPOSED METHODS The panel recommended that IQWiG inform German decision makers of the net costs and value of additional benefits of an intervention in the context of relevant other interventions in that indication. To facilitate guidance regarding maximum reimbursement, this information is presented in an efficiency plot with costs on the horizontal axis and value of benefits on the vertical. The efficiency frontier links the interventions that are not dominated and provides guidance. A technology that places on the frontier or to the left is reasonably efficient, while one falling to the right requires further justification for reimbursement at that price. This information does not automatically give the maximum reimbursement, as other considerations may be relevant. Given that the estimates are for a specific indication, they do not address priority setting across the health-care system. CONCLUSION This approach informs decision makers about efficiency of interventions, conforms to the mandate and is consistent with basic economic principles. Empirical testing of its feasibility and usefulness is required.

Can economic evaluations be made more transferable

Several commentators have identified the lack of generalisability and transferability of economic evaluation results. The aims of this study were: (a) to develop a checklist to assess the level of generalisability and transferability of economic evaluations; (b) to assess the generalisability and transferability of economic evaluations between the UK and France using the checklist; (c) to identify reasons for any lack of transferability and generalisability; (d) to assess how the transferability and generalisability of economic evaluations can be improved; and (e) to outline ways in which databases of economic evaluations and journals can assist in this area. The checklist was developed using previous work and the templates of the NHS EED and CODECS databases. A sub-checklist of essential items was then derived. Validation of the two checklists was undertaken with Health Economists participating in the EURONHEED project. Economic evaluations involving the UK and France were then located and assessed using the checklist. A summary score for each study was calculated based on the percentage of correctly reported (applicable) points, and the results in the empirical analysis compared to identify differences. The extended checklist includes 42 items, and the sub-checklist 16 items. Twenty-five economic evaluations met the inclusion criteria for the empirical analysis. In the extended checklist the mean score was 66.9±13.6%. The results for the sub-checklist were very similar. The analysis revealed that costing, assessments of generalisability by the author(s), assessment of data variability, discounting, study population, and the reporting of effectiveness are areas that need more attention. Differences in cost-effectiveness results are often accounted for by price or organisational differences. The developed checklists are useful in assessing the generalisability and transferability of economic evaluations. In order to improve the generalisability and transferability of economic evaluations authors need to be more explicit and detailed in describing and reporting their studies. If they are to provide added value to their users, international databases of economic evaluations should systematically assess the generalisability and transferability of studies. Further research is in progress on producing a weighted version of the checklist.

Valuing EQ-5D using Time Trade-Off in France

ObjectivesWhile a French language version of the EQ-5D exists, to date, there has been no French value set to accompany it. The objective of our study was then to derive the French TTO value set of the EQ-5D.MethodsA total of 452 respondents aged over 18 were recruited who were representative of the French population with regard to age, gender, and socio-professional group. The direct valuation of 24 health states was first obtained by Time Trade-Off (TTO), and the negative TTO values were bounded using the monotonic transformation. Several alternative model specifications were investigated to estimate the values for all 243 states in the EQ-5D descriptive system. Only the best fitting model is presented in this paper. The analysis was conducted at an individual level to make the maximum use of the available data, and we estimated mixed models with random intercept. Models were compared through the Akaike information criterion (AIC), the mean absolute error (MAE), and the Pearson correlation coefficient between the observed and the predicted values of each model.ResultsAfter exclusion, 443 respondents took part in the study. The best fitting model included the same variables as the N3-model used in UK.ConclusionThis study provides the French value set of the EQ-5D based on the stated preferences of the French general public facilitating cost-effectiveness analysis.

Relationship between compliance and persistence with osteoporosis medications and fracture risk in primary health care in France: A retrospective case—control analysis

BACKGROUND Nonadherence to treatment is an important determinant of long-term outcomes in women with osteoporosis. OBJECTIVES This study was conducted to investigate the association between adherence and osteoporotic fracture risk and to identify optimal thresholds for good compliance and persistence. A secondary objective was to perform a preliminary evaluation of the cost consequences of adherence. METHOD This was a retrospective case-control analysis. Data were derived from the Thales prescription database, which contains information on >1.6 million patients in the primary health care setting in France. Cases were women aged >or=50 years who had an osteoporosis-related fracture in 2006. For each case, 5 matched controls were randomly selected. Both compliance and persistence aspects of treatment adherence were examined. Compliance was estimated based on the medication possession ratio (MPR). Persistence was calculated as the time from the initial filling of a prescription for osteoporosis medication until its discontinuation. RESULTS The mean (SD) MPR was lower in cases compared with controls (58.8% [34.7%] vs 72.1% [28.8%], respectively; P < 0.001). Cases were more likely than controls to discontinue osteoporosis treatment (50.0% vs 25.3%; P < 0.001), yielding a significantly lower proportion of patients who were still persistent at 1 year (34.1% vs 40.9%; P < 0.001). MPR was the best predictor of fracture risk, with an area under the receiver-operating-characteristic curve that was higher than that for persistence (0.59 vs 0.55). The optimal MPR threshold for predicting fracture risk was >or=68.0%. Compared with less-compliant women, women who achieved this threshold had a 51% reduction in fracture risk. The difference in annual drug expenditure between women achieving this threshold and those who did not was approximately euro300. The optimal threshold for persistence with therapy was at least 6 months. Attaining this threshold was associated with a 28% reduction in fracture risk compared with less-persistent women. CONCLUSIONS In this study, better treatment adherence was associated with a greater reduction in fracture risk. Compliance appeared to predict fracture risk better than did persistence.

Nutritional, lifestyle and environmental factors in ocular hypertension and primary open‐angle glaucoma: an exploratory case–control study

Purpose:  To evaluate known and potential risk factors, including nutritional, lifestyle and environmental factors, differentiating patients with high‐tension primary open‐angle glaucoma (POAG) from control subjects with ocular hypertension (OHT).

Risk-sharing agreements for innovative drugs

The concept of risk-sharing has often been discussed and sometimes as a new form of contractual agreement between payers and the pharmaceutical industry for setting the value of an innovation conditional to demonstration of its effectiveness and efficiency in real life. The most prominent application of this concept in recent years has been the agreement in the United Kingdom between NICE and the industry for the use of β-interferons and glatiramere for the treatment of multiple sclerosis. In this case the industry was asked to demonstrate its claims in terms of the impact of the treatment on quality of life and of its efficiency, measured by an incremental cost-effectiveness ratio in pounds sterling per qualityadjusted life-years. Apart from this obvious example, however, the concept has until now had only very few other applications, unless one includes under this term the price-volume agreements that are used in countries in which prices are regulated. The present contribution addresses four questions. First, we examine why this concept has emerged, problems it is supposed to answer, and by whom. Second, we consider the prerequisites to its implementation. Finally, we discuss the pros and cons of risksharing for both parties involved, the payer and the industry.

Guidelines for completing the EURONHEED transferability information checklists

The structure and methods to complete and derive a quality score from the European Network of Health Economic Evaluation Databases (EURONHEED) transferability information checklists for published economic evaluations were reported and discussed in a previous paper (Boulenger et al. in Eur J Health Econ 6, 334–346, 2005). Within the same paper, the use of the checklists was illustrated through their application to a sample of economic evaluations conducted in France and UK. The transferability information subchecklist, consisting of 16 items from the original 42-point checklist, and methods used to derive it, were validated among 16 health economists across Europe participating in the EURONHEED project. Recent correspondence with other researchers, however, indicates that the checklists are now being utilised in empirical work and the methodology of assessing transferability and generalisability. This supplementary paper provides full details of the guidelines that have been developed and recently updated by the authors, such that the overall and subchecklists can be more widely and consistently completed and utilised. We also briefly discuss associated issues such as weighting of items in the checklists and give further clarifications regarding what we consider the most appropriate applications of the checklists to be.

The diffusion of Health economics knowledge in Europe: The EURONHEED (European Network of Health Economics Evaluation Database) Project

This paper overviews the EURONHEED (EUROpean Network of Health Economics Evaluation Databases) project. Launched in 2003, this project is funded by the EU. Its aim is to create a network of national and international databases dedicated to health economic evaluation of health services and innovations. Seven centres (France, Germany, Italy, The Netherlands, Spain, Sweden and the UK) are involved covering 17 countries. The network is based on two existing databases, the French CODECS (COnnaissance et Decision en EConomie de la Santé) database, created in 2000 by the French Health Economists Association (Collège des Economistes de la Santé), and the UK NHS-EED (NHS Economic Electronic Database), run by the Centre for Reviews and Dissemination, University of York, York, England.The network will provide bibliographic records of published full health economic evaluation studies (cost-benefit, cost-utility and cost-effectiveness studies) as well as cost studies, methodological articles and review papers. Moreover, a structured abstract of full evaluation studies will be provided to users, allowing them access to a detailed description of each study and to a commentary stressing the implications and limits, for decision making, of the study. Access will be free of charge. The database features and its ease of access (via the internet: http://www.euronheed.org) should facilitate the diffusion of existing economic evidence on health services and the generalisation of common standards in the field at the European level, thereby improving the quality, generalisability and transferability of results across countries.

How the availability of recombinant human TSH has changed the management of patients who have thyroid cancer

Recombinant human TSH (rhTSH) is used in patients who have had surgery for thyroid cancer but are at low risk of recurrence. The rhTSH is used for the preparation of postoperative administration of 3.7 GBq (100 mCi) of radioiodine for thyroid-remnant ablation and for the determination of serum thyroglobulin levels during follow-up. In these two conditions, the efficiencies of levothyroxine withdrawal and rhTSH administration are similar; however, rhTSH can be administered during levothyroxine treatment, and its use avoids the hypothyroid period induced by levothyroxine withdrawal, reduces whole body exposure after radioiodine administration, avoids potential morbidity and maintains a better quality of life compared with hormone withdrawal.

A Markov model simulation of the impact of treatment persistence in postmenopausal osteoporosis.

Background . Osteoporosis is a common disorder of the skeleton that increases bone fragility and the risk of fracture. Bisphosphonates have become the reference treatment for postmenopausal osteoporosis because of their proven efficacy in reducing fracture rates. The effectiveness of bisphosphonates is, however, limited by poor treatment adherence and persistence with treatment. A model has been designed to simulate the impact of improving persistence rates on treatment effectiveness. Methods . The Markov model followed a cohort of patients over 10 years to estimate the total number of incident osteoporotic fractures by age for the overall population of women with diagnosed postmenopausal osteoporosis in France (mean age, 71.1 years ±9.6; range, 50-96 years). The impact of clinical efficacy, persistence, and residual treatment effects data on predicted fracture risk was also estimated in the model. Results . Predicted numbers of incident fractures appeared consistent with published data. Compared with no treatment, the relative risk of fracture over 10 years was 0.831 for weekly bisphosphonate treatment with an assumed persistence rate of 51% after 1 year (absolute risk reduction = 11.4%). This relative risk decreased to 0.731 (absolute risk reduction=18.1%) if hypothetical full-treatment persistence was achieved. In terms of public health, improving persistence with bisphosphonate treatment by only 20% could have the same impact as a 20.2% increase in clinical efficacy. The benefit associated with improved persistence declines as full persistence is approached. Conclusion . Improving persistence can increase treatment effectiveness. Giving greater priority to persistence interventions might have a greater impact on the health of osteoporotic women than advances in treatment efficacy.