The German special path in the assessment of cost-benefit ratios by IQWiG

Almost two years ago, the legislature decided on the basis for the cost-benefit assessment of medicines in Sections 35b, 31 Paragraph 2a and 139 Paragraphs 1 and 2 SGB V. Since April 1, 2007, the Institute for Quality and Efficiency in Health Care (IQWiG) can be commissioned by the Federal Joint Committee (G-BA) to carry out a benefit or cost-benefit assessment of medicines. This applies “to every medicine with patent-protected active ingredients that can be prescribed for the first time, as well as to other medicines that are important”.

>> Against this background, IQWiG designed a first draft (version 1.0) of a “methodology for the assessment of relationships between benefits and costs in the German statutory health insurance system” and made it available to the public for discussion on January 24, 2008. Although 46 associations, companies, institutions and experts were consistently critical of the methods presented, on October 14, 2008 IQWiG presented a second version (version 1.1) that was only revised in nuances. The only real innovations were essentially found in the so-called technical appendices on cost assessment, modeling and dealing with uncertainties, which accurately contain textbook knowledge and therefore partially contradict the main paper.
The revised version is still subject to sharp national and international criticism. The core of the proposed cost-benefit assessment methodology continues to be the so-called efficiency frontier. The IQWiG refers to the long history of application and the great importance that this concept has acquired in economics. In order to counteract emerging criticism in advance, IQWiG likes to point out the Nobel Prize for Markowitz's concept from 1952. In the last issue of the “Monitor Health Services Research” the head of IQWiG and two employees once again presented their reasons for designing the methods defined by IQWiG. From the perspective of the Association of Research-Based Drug Manufacturers, however, some points are not quite as simple as those presented by IQWiG.
The role of IQWiG
The Institute for Quality and Efficiency in Health Care (IQWiG), which is commissioned by the G-BA to evaluate the cost-benefit ratio, is not completely free to decide which methods it should use to carry out the evaluations. Instead, IQWiG has to carry out this “on the basis of the international standard(s) recognized in the respective specialist circles… of health economics”. This reference in particular in Sections 35b and 139a Paragraph 4 SGB V is a clear indication that the broad and partly controversial discussion on the IQWiG methods used to date has also reached and irritated the legislature. With the reference to compatibility with the standards internationally recognized in specialist circles, a criterion is introduced which at least provides a guideline for the methods to be used. At the same time, it may open up the opportunity for greater international harmonization of the methods used, as is the case, for example. B. is aimed at through various initiatives of the European Union.
At this point, IQWiG could have used the current state of science and used the methods for incremental cost-effectiveness analysis documented in numerous publications as the basis for their own methods. These are accepted in national (Schulenburg et al. 2007a) and international professional circles (Hjelmgren et al. 2001). In particular, the global professional association for health economic evaluations, ISPOR, represents the current status of cost-benefit assessment methodology with its collection of methods (ISPOR 2009). As an example, please refer to the comprehensive overview compiled by the German Institute for Medical Documentation and Information (DIMDI) (Zentner et al. 2005).
On the contrary, IQWiG has decided not to use these long-established methods that are recognized in specialist circles, but to commission an international advisory board to develop its own methodology. What is interesting is that this advisory board was made up of proven experts in benefit measurement, modeling and health technology assessment, but only with one specialist economist, who also resigned from the advisory board after a short time.
In the last “Monitor Health Services Research” the reasons for a different German methodological model were once again explained by IQWiG: (1) standards that are based on value judgments, (2) lack of spending limits set at national level in the German statutory health insurance system, (3) lack of prioritization and lack of weighing up different alternatives.
The IQWiG obviously accepts that there are international health economics standards, but criticizes that these are based on value judgments. But what did the legislature have in mind when it came to the phrase “recognized in specialist circles”? And doesn’t that apply equally to evidence-based medicine? Haven't the current method developments in EbM - such as GRADE - recognized how much EbM is based on value judgments, which therefore need to be made transparent? (Atkins 2004)
Equally surprising is the argument that the international standards cannot apply to Germany, since only here there is no nationally set limit for health spending and the laws of economics are therefore invalid, since decisions have to be made not under limited resources, but under infinite ones Resources. But isn't there a de facto requirement for contribution stability in the statutory health insurance system and haven't we introduced the health fund since January 1, 2009? The contribution rate is determined exclusively by a regulation of the Federal Ministry of Health via legal ordinance (Section 241 Paragraph 2 SGB V). In addition, the Bundestag and Bundesrat decide on tax subsidies. What should result from this other than a nationally specified spending limit?
Last but not least, the IQWiG points to the lack of economic prioritization or lack of weighing up different alternatives. In fact, it is debatable whether we in Germany have a performance assessment in the healthcare system that is based on central priorities. However, why this should have an impact on the decision for or against the acceptance of international health economics standards remains unclear.

The concept of the efficient frontier
Now the IQWiG – as it itself confirmed – has set aside the requirements of the legislature and interpreted the law in a home-style manner. The result is the concept of the efficiency frontier, which the chairman of the international IQWiG advisory board, a doctor from the USA, recommended to IQWiG. The concept is based on the ideas of modern portfolio theory according to Henry Markowitz, which until recently played an important role in the compilation and evaluation of securities portfolios or loan portfolios, the problematic interpretation of which is currently shaking the economy worldwide.
To determine the efficiency limit of IQWiG, all existing treatment alternatives are entered graphically as points in a two-dimensional coordinate system. The benefit (ordinate) is determined specific to the indication and the costs (abscissa) are determined from the perspective of the insured community of the statutory health insurance (GKV). The linear connection of the individual alternatives graphically results in the efficiency limit for an existing treatment field within a defined indication. What is now important is how IQWiG wants to derive recommendations for new treatment options from this: New interventions must be at least as efficient as (1) the oldest and most ineffective alternative in relation to doing nothing (2) a hypothetical average alternative of all existing therapies or (3) the last alternative in relation to the second to last. Efficiency is always required.
The update of the existing market situation to derive principles of action for determining a maximum reimbursement amount apparently follows the regulations on the fixed amount. Here, a price regulation is derived based on the market weight. However, this is wrong from the start for markets with patent-protected innovations. The efficiency frontier postulates that in an indication in which there is generic competition, even the maximum possible benefit should only cost a few cents and vice versa.
Regardless of this fundamental problem, the concept of the efficient frontier has further methodological inadequacies. So all points on the efficiency frontier have lower utility compared to the last point. Even under the premise that a separate benefit assessment beforehand would not be methodologically questionable, it still follows that a new intervention according to this dogma shows the highest absolute benefit of all alternatives used for comparison. However, since the recommended course of action proposed by IQWiG only provides for an update of the market situation on the basis of the last to the penultimate intervention, a comparison with old and ancient alternatives, which have an even lower absolute benefit, is pointless. In fact, method paper 1.1. only a straight line is drawn or extended from the “penultimate” to the “last” product. Why extending the gradient now seems necessary, even though a concave form of the efficiency frontier seems highly likely in both theory and practice (principle of diminishing marginal utility), remains unanswered in the methods paper. It should be noted in passing that the slope of this line in the efficiency frontier model represents an explicit threshold, i.e. how much benefit (life) to society is worth how much cost.
The efficiency frontier instrument presented by IQWiG is problematic in practical implementation and susceptible to strategic action. With old therapies, the problem will arise that no sufficient evidence-based benefit or even no benefit at all can be proven for these interventions (which may have been on the market for a very long time), since such proof was not necessary upon approval or market entry. It is also questionable whether, even with existing evidence, this is comparable to newer therapeutic approaches.
In addition to the problem of comparability, there is a fundamental strategy vulnerability when extending the efficiency frontier. A change in the benefit-to-cost ratio for the penultimate intervention changes the slope of the efficient frontier and thus changes the division of the decision space for more costly interventions. This means that the manufacturer of the penultimate (or third most recent) alternative has a decisive role in making the market entry barriers higher or lower depending on their own considerations. This is clear in Figure 1 below. Such a procedure or such a distribution of roles raises at least some antitrust law questions that are in no way addressed or acknowledged in the methods paper (Fig. 1.)
The QALY concept and the IQWiG
In addition to the basic concept of the efficiency frontier, the IQWiG's now four methodological papers on cost-benefit assessment (method paper and three technical appendices) contain further individual specifications, which largely either contradict economic theory, do not follow the legal requirements or provide an alternative opinion to the main paper.
One example is IQWiG's handling of aggregated/generic benefit measures, a de facto standard in health economics. Although the IQWiG does not fundamentally rule out the use of cross-indication benefit measures, it is clear from the comments on the QALY concept in the methods paper and additional oral communications that cross-disease comparisons are in fact not desired. The rejection of the QALY concept is justified by the well-known limitations of the concept that have been widely researched for 40 years. However, the postulated approach of IQWiG leads to the problem within a single indication that an independent efficiency limit would have to be set up for each benefit parameter. It is difficult to make consistent decisions or recommendations to the joint self-government in this way.
Instead, the methods paper rightly points out the possibility of taking society's willingness to pay into account. However, reference is made elsewhere to the lack of a legal basis (preamble, pp. v-vi). Against this background, it remains more than questionable whether the decision-makers of joint self-administration will be able to make consistent and therefore fair decisions across indications using the efficiency frontier methodology. In fact, it is to be feared that indication-specific quasi-threshold values will instead arise that stand in the way of rational decision-making.
Since the inclusion of willingness to pay is therefore not possible, the use of sum scores that have yet to be developed is logical from an IQWiG perspective, as otherwise no clear recommendation about the overall benefit of an intervention is possible. In the health economics literature there is the well-documented case study of the cost-effectiveness assessment of chronic kidney disease. In 1968, Klarman, O'Francis and Rosenthal presented one of the first works on the creation of a sum score for the purpose of incorporating it into a cost-benefit assessment: the so-called QALY concept. Torrance later pointed out the diverse possible uses, which ultimately led to the formation of the EuroQol Group in 1987.
The QALY concept has been scientifically researched and constantly developed for over 40 years. Its strengths and limitations are well known. Anyone who starts developing independent, non-indication-wide sum scores without necessity is turning the wheel back 40 years. The result will be a concept very similar to QALY; a quasi-QALY depending on the indication, the methodological limitations of which can possibly be re-examined in the next 40 years.
Further benefit parameters
Within the IQWiG benefit assessment, only benefit parameters based on the best evidence are regularly accepted as reasons for inclusion and exclusion. Since the cost-benefit assessment should be carried out against the background of realistic resource consumption and effectiveness in everyday medical practice, benefit parameters that were collected on the basis of a lower but appropriate level of evidence would also have to be included here. However, if only the study situation according to the benefit assessment is relevant for the selection of the benefit parameters, this would lead to significant distortions. Important economic benefit parameters would not be recorded here because they are rarely the primary clinical endpoint of clinical trials (e.g. time of transition from outpatient to inpatient, time of deterioration in severity, etc.). Although it was noted in the methods paper that the economic perspective must also have an impact on the benefit assessment, there is nothing to be found about this in the general Methods 3.0.
The perspective of cost calculation
The methods paper is inconsistent with regard to the perspective to be taken: While the social perspective is clearly rejected in the main document, in the technical appendix to the cost calculation it is given a comparable status to the insured person's perspective. This contradiction is neither acknowledged nor resolved in the main document. The fact that indirect costs in the form of productivity losses due to incapacity to work should also be ignored in this way does not fit with the perspective adopted by the insured community. Ultimately, employers co-finance health insurance contributions on an equal basis and therefore also have a legitimate interest in taking days of incapacity for work into account. It is not just for these reasons that the authors of a workshop conducted by the BMG in 2007 came to the conclusion that areas of expenditure beyond the statutory health insurance perspective also need to be taken into account. From the VFA's perspective, this can only be ensured through a social cost perspective. On the one hand, this would take into account all areas of expenditure in the five existing social insurances (health, nursing care, accident, pension and unemployment insurance) and, on the other hand, it would also include the productivity losses relevant to employers due to absence from work due to illness.
Another critical point within the IQWiG concept is the period of time to be taken into account in the cost-benefit assessment. If only the costs for the entire duration of an illness are taken into account for each intervention, but the benefit is limited to the evidence period determined in the benefit assessment, Massive distortions in the construction of the efficiency frontier are to be expected. For example, if a drug leads to significantly improved survival rates in severe sepsis in a study over 28 days, then on the one hand only up to 28 days of survival would now be included in the assessment and not the rest of the life saved. On the other hand, the lifetime costs of survivors would be reported as a negative balance of survival. Modeling should internalize this disadvantage using appropriate methods. Modeling was introduced into medical and health economic decision-making precisely for the purpose of scientifically based extrapolation of existing evidence. The scientific standard here is that modeling of both costs and benefits is carried out over the same period of time.
Transparency and participation of the (professional) public
as well as the manufacturer
In its self-portrayal, IQWiG has always enabled the highest level of transparency and participation. For this reason, the VFA presented a proposal for a transparent evaluation process in a report in 2007. A key element here was the scoping workshop, which is also known at NICE. The aim of scoping is to determine consensually the questions to be addressed, the choice of comparative therapies, the definition of patient-relevant objectives and outcomes as well as the necessary methodological framework. The demand for a scoping workshop has also been requested several times by scientific circles in Germany, but has not (so far) been implemented by IQWiG.
However, the methods paper contains neither statements about the actual processing process of the cost-benefit assessment nor about an active participation process. The process of benefit assessment and the possibility of consultation are only described in General Methods 3.0. A broad involvement of the interested public, especially scientific associations, patient representatives and manufacturers, is anchored in the law. In other countries, the realization has already matured to such an extent that parameters defined by consensus at the beginning of an assessment significantly shorten the discussion at the end.
Importance for health services research
The outcome of this method discussion will also have an impact on drug-related health services research. IQWiG's refusal to take note of all the evidence for assessments will in some cases lead to recommendations that cannot be supported by the decision-making bodies. Furthermore, this is not conducive to the acceptance of health services research studies. If the results are fundamentally of no relevance to the assessment, there is little point in carrying it out. There is a great need in Germany for data about everyday care, on the basis of which health economic analyzes can then be carried out. In this area in particular, the results of health services research cannot easily be transferred from other countries.
Conclusion
The slightly revised method paper 1.1 of the “Methodology for the assessment of relationships between benefits and costs in the German statutory health insurance system” presented by IQWiG on October 9, 2008 is rightly still subject to extensive national and international criticism. Although the IQWiG has slightly changed the decision zones to be derived from the efficiency frontier compared to version 1.0, this does not create more clarity about the recommendations to be expected, but on the contrary, it creates a greater degree of arbitrariness and uncertainty.
The two aspects of methodological arbitrariness and the resulting uncertainty regarding possible recommendations to decision-makers not only refer to the efficiency frontier, but also run through the entire text of the methods paper and through the technical appendices, which are now available for the first time. The IQWiG therefore fails to meet the requirement to provide clear recommendations for action in the form of clear methodological guidance. Only if the authors of pharmacoeconomic studies know which criteria individual studies must meet in order to carry out a cost-benefit assessment can these be anticipated. Since this is missed, methodological arbitrariness remains, which comes at the expense of transparent inclusion and exclusion criteria. Furthermore, the significance of the technical appendices remains unclear: Should they (a) explain the statements in the methods paper, (b) supplement them, (c) replace them or (d) serve as reference points? Due to the heterogeneity of the technical appendices, these questions remain unanswered and tend to cause confusion.
In March 2007,schulenburg et. al. presented a veritable, immediately implementable concept for introducing cost-benefit assessment in Germany. In autumn 2007, around 80 specialist economists and experts from service providers and service providers reached a consensus on fundamental principles for conducting health economic studies in Germany. Both were ignored. Two years later, we are still at a dead end in a method discussion with no prospect of a breakthrough. As early as 2006, Germany was the last country in Western Europe that did not use health economic analyzes to evaluate medical innovations. In the case of the other European countries, the differences in the assessments are more evident in the depth and only marginally in the methodology - Germany will largely differ here. What all countries have in common is that the assessment serves to inform a decision-maker, usually a ministry or a body of self-government, which then takes concrete action - incorporating its political value judgments. Intellectually inadequate and transparently manipulative evaluations not only make the decision-maker's work more difficult, but also weaken the credibility of the entire system. The VFA therefore interprets the said “international standards” in the law as a request to stop the process of reinventing the first wheel, which has been going on for several years, and to start better with a machine that has been running elsewhere for ten years and to intelligently adapt it to Germany. This was a successful concept when the DRG was introduced in Germany. <<