HTA AND EFFICIENCY

The introduction of a “fourth hurdle” for pharmaceuticals has expanded the requirements for a marketing license from quality, safety, and efficacy to also include evidence on the economic impact for national and regional health services in terms of cost-effectiveness and budget impact. This “fourth hurdle” and the HTA processes should enable society to make informed decisions regarding health-care technologies so that society's resources are used in the most cost-effective way to maximize the level of health produced given the available resources. Of crucial importance in this process is resource allocation in terms of which technologies are prioritized for assessment and the subsequent evaluation and implementation of the HTA information provided. In the implementation process, decision-makers need appropriate information to enable them to make decisions to improve efficiency within the health-care system.

The four European HTA systems presented in the reviews clearly differ in the prioritization processes for the identification of technologies to assess and evaluate. For England and Wales, HTA prioritization is clearly described with precise prioritization criteria for the different HTA producers (18). The French review (15) outlines the process and the key prioritization criteria for ANAES. The entire Dutch review (2), in fact, focuses on a discussion of the lack of a formal prioritization process. It is even stated that “there is no central direction on which technologies are targeted for HTA.” The Swedish review (3) comments that the SBU prioritization process became more explicit and well structured after 1999, but the article gives no clear description of the prioritization criteria.

The scope of the HTA system and the allocation of HTA investments is of huge importance for the potential impact of the HTA system on overall efficiency. Why is it, for example, that the major proportion of HTA investments is made on pharmaceuticals when pharmaceuticals account for approximately 10–15 percent of total health-care costs? Towse and Pritchard (19) note that, from its establishment in April 1999 to January 2002, two thirds of NICE appraisals have been on pharmaceuticals. One might ask what explicit considerations have been made on the marginal return of investment on HTA (both from the societal and the health-care perspective) that arise from focusing to this extent on pharmaceuticals compared with nonpharmaceutical items. Are the relatively large proportion of pharmaceutical appraisals due to the fact that the impact and return of these investments are so much greater than appraisal investments in other sectors of the health-care system, or is it simply because it is less politically sensitive and “easier” to perform HTAs on pharmaceuticals? The issue of a potentially too narrow scope of the HTA system in England and Wales is discussed by Stevens and Milne (18), where they point out that the HTA system has largely excluded assessments of both service delivery issues and public health interventions. They conclude that “A wider scope [of the HTA system], while desirable would of course require increased resources for the HTA system as a whole.” This might be true, but, before this, the entire HTA system may need to be reassessed and reprioritized with HTA investment allocations redistributed accordingly so that the effects of any assessments and policy interventions are maximized.

Additionally, there are NHS concerns about whether the “right” technologies are referred to NICE and also concerns regarding the opportunity costs of positive NICE recommendations (19). Given global budget constraints, NICE recommendations tend to preempt growth money that could be used for more cost-effective purposes. NICE should perhaps take a broader perspective and assess and appraise established technologies that may not be cost-effective and whose discontinuance, therefore, could release resources for other more cost-effective treatments (19). Sculpher et al. (17) also note that it will be necessary to expand the appraisal process, particularly to identify widely used technologies that are not cost-effective, to release resources for new interventions. On the same topic, Cookson et al. (4) criticize the reluctance of NICE to rank the appraised technologies in any form of hierarchy—not even a grading “high,” “medium,” or “low” cost-effectiveness. The consequence of this may be that health authorities are forced to fund marginally cost-effective technologies approved by NICE and, thereby, divert funding away from more cost-effective existing technologies. Gafni and Birch (10) also touch on this issue in their discussion and criticism of the NICE methodological guidelines. They argue that the lack of an explicit requirement to incorporate all opportunity costs in the NICE submission guidelines may lead to (i) uncontrolled increases in NHS expenditure without any increase in total health improvements, (ii) increased inequities in the availability of services, and (iii) concerns about the sustainability of public funding for new technologies.

HTA AS A COST-DRIVER

In many European countries, HTA is required as part of the price and reimbursement submission after medical approval to market a drug. In France, the requirements are formal (through the transparency committee and the economic committee), whilst in The Netherlands, they are voluntary until 2005. In Sweden, they are recommended but the newly established pharmaceutical benefits board responsible for drug reimbursement decisions will make the requirements more formal. In England and Wales, there is no formal requirement for HTA information, but the introduction of NICE has caused the need to supply HTA information for compounds that are considered in NICE appraisals. A positive recommendation by NICE will be of major importance for the company promoting the recommended drug, given the new regulation that the NHS has to provide the budget for the recommended drug within three months of the issue of the recommendation (18).

In addition to the country-specific national requirements, the industry has to deal with and provide information to multiple regional, local, and hospital-specific HTA agencies. These multilayer requirements add to the complexity and costs of providing the required information. From an industry perspective, HTA requires resources and the country-specific heterogeneity and multilayer requirements will only increase the costs of producing that information. Ultimately, increased costs to the industry may lead to higher prices of pharmaceuticals for patients, governments, and society as a whole.

As a response to the increased and diverse requirements for HTA information, the pharmaceutical industry has increased investments in health economics/pharmacoeconomics. DiMasi et al. (7) report a clear trend for both an increased and an earlier involvement of health economics/pharmacoeconomics throughout the product life cycle in terms of:

• an earlier strategic integration in phase I–II regarding Go/No-Go decisions for continuing product development;
• an earlier development of integrated price and reimbursement strategies; and
• an increased involvement in the later phase III and postapproval/postlaunch studies.

For the earlier research and development phases, the inclusion of health economic and patient-reported quality of life end points require changes in clinical trials. The main implications and cost-drivers of the expanded scope of clinical trials are that evaluations of cost and effectiveness end points generally require:

• longer-term follow-up periods for the measurements to be meaningful from an economic and effectiveness perspective; and
• larger sample sizes for statistical inferences to be performed with required levels of statistical significance when resource use/cost end points are included in trials. It is well known that resource use variables generally tend to have a larger variance than efficacy/effectiveness due to the skewness in the distribution of the resource use variables, where a small proportion of patients can account for the major part of the observed resource use/cost in the study.

For the later phases in a product life-cycle the main cost-driver is the increased demand from government authorities and agencies for the industry to provide information from postlaunch real-world naturalistic studies on the costs, effectiveness, and cost-effectiveness of the drugs (as opposed to the highly controlled setting under which clinical trials are conducted in terms of selected patient populations and controlled protocol-driven treatment). In England and Wales, for example, NICE is focused on the cost-effectiveness of technologies in their everyday use. It is clear from the “Guidance Document for Manufacturers and Sponsors Making a Submission for a Technology Appraisal,” NICE is clearly indicating that the ideal source of effectiveness data should come from “prospective, randomized, controlled trials with a naturalistic design” (section 2.1.3), which suggests that companies, in effect, are expected to collect information, in parallel with or as part of the phase III registration trials, that give NICE a reasonable picture of how the product is likely to perform in routine clinical care (19). Sculpher et al. (17) conclude that companies are likely to invest more in pragmatic clinical trials with a broader population base to collect the data on costs and health outcomes that are relevant to NICE's decision making.

IMPLEMENTATION BARRIERS

The pharmaceutical industry is a major supplier of HTA, which is to a large extent produced by scientifically qualified in-house health economics departments. The industry-produced evaluations will provide an important piece of evidence for various health technology appraisal decisions.

As discussed by Schubert (16) there is a lack of internationally agreed HTA guidelines. There are, indeed, some common grounds among the different country-specific HTA guidelines on the general classification of analysis approach into cost-effectiveness, cost-utility, and cost-benefit categories, but the detailed guidelines then differ regarding the required perspective (societal or health-care system) and the required approach to conduct the analysis. From an industry perspective, HTA requires resources and this heterogeneity will only increase the costs of producing HTA information. The heterogeneity across different countries regarding the HTA requirements and criteria complicates the industry's work regarding planning and preparation of the required HTA information. It also means that trade-offs and prioritizations have to be made within the pharmaceutical industry because not every aspect of every HTA requirement in every country can be accounted for. Inevitably some issues have to be left for later research. To avoid wasting scarce resources, there is a need for increased transparency regarding the requirements. Additionally, there is a need for more collaboration and partnership between the industry, decision-makers, and HTA agencies and authorities to clarify the expectations and possibilities, as well as to reach an agreement and common understanding on the limitations regarding what HTA information can be delivered at different stages of the product's life cycle.

HTA systems are linked to the requirement for information to control the price and reimbursement of pharmaceuticals across Europe. If the HTA system is used as an excuse to contain costs and limit access to new technologies instead of guiding prioritization and an effective allocation of limited resources, the resulting outcome cannot be expected to produce anything but a suboptimal resource allocation. Relatively costly interventions may be relatively efficient in terms of the amount of health produced per unit of resources used. A major limitation in achieving an efficient resource allocation is the nontransparent decision-making process and the existing budget “silos” between budgets for drugs, devices, and in- and outpatient care settings. The full potential and value of HTA can only be realized if silos in health-care delivery and financing are replaced by integrated delivery and financing systems. There are several examples where the introduction of new and seemingly more costly technologies is efficient in terms of the net benefits they produce. In an information leaflet from the Association of the British Pharmaceutical Industry (ABPI) 2001 on “The cost of medicines—good value for patients,” specific UK estimates are presented on improvements in treatments since the 1950s. Medicines, for conditions such as asthma, epilepsy, high blood pressure, bronchitis, ulcers, and mental illness have both saved lives and brought savings in terms of a reduced need for hospital treatment (1). Based on a cost-benefit analysis approach, Cutler and McCellan (6) conclude from a U.S. perspective that technology change and medical spending as a whole is worth the increased costs of care. Four examples of health technologies/indications (heart attacks, low birth-weight infants, treatment of depression, and cataracts) are discussed, where the estimated benefit of technological change is much greater than the costs. Only for one considered technology/indication (breast cancer) are the costs and benefits of similar magnitude.

In line with this, Kleinke (13) observes that the fact that pharmacy costs are rising faster than medical cost inflation in the U.S. health-care system represents a major structural shift from provision of traditional medical services to the consumption of new breakthrough medical products. He points out that many drugs pay “dividends” to the health-care system that far exceed their costs and that calls by public and private payers for price and utilization controls would be ineffective and counterproductive.

A final major implementation hurdle for a successful use of HTA results is the lack of understanding of economic information and economic evaluations by health-care decision makers and clinicians and the following suboptimal use of supplied economic information found in several studies (5;8;9;11). Based on these findings, we are tempted to agree with Hutton and Brown (12), who wrote “Exactly how decisions are made on service planning and resource allocation within the NHS without substantive economics input is a question that seems eminent for future research” and “In the absence of a clear understanding of how decisions are actually made by health authorities, economists might be more questioning of the validity of criticism of their work.”