Neil Hawkins

Supporting Decisions in Clinical Development

Health technology assessment (HTA) is a field of analysis that evaluates the medical, social, ethical, and economic implications of development, diffusion, and use of health technology. The aim of HTA is to aid decision-making by providing information about the likely outcomes associated with different courses of action, and it’s increasingly used to aid decisions regarding resource allocation with respect to new technologies. HTA comprises three core activities: 

the identification and review of evidence, 

the synthesis of evidence from multiple sources, 

and the development of decision-analytic models that seek to explicitly predict the outcomes of different courses of action in terms of endpoints that decision-makers can value and trade-off (such as resource use and measures of how patients feel, function and survive).

In addition to aiding decisions regarding resource allocation, HTA can aid decision-making during the clinical development process. The value of any clinical development process is a function of the likelihood that it will affect change (e.g. the adoption of new technology) and the perceived value of that chance. HTA activities can an aid decisions regarding investment, RCT design, and the need to conduct ancillary, non-experimental studies to confirm the value of new technologies. HTA aids these decisions by improving estimates of the value of candidate research programs. 

However, failure rates are high during clinical development, and there may be a reluctance to invest in additional activities outside the core RCT program given the risk that an individual product may not make it to market and generate revenue. This can lead to HTA activities only being commissioned late in the clinical development process. However, this means that the results of HTA activities are not available when decisions need to be made during the clinical development process.

The concept of Net Present Value (NPV) is useful when evaluating potential HTA activities during different stages of the clinical development process. The NPV of a project is the sum of the present values of the incoming and outgoing cash flows associated with a project. If we consider early HTA activities, the outgoing cash flows are the costs of the proposed activities. There will be little discounting of these as they are incurred now and in the near future. The incoming cash flows are the difference in expected profits from the new technology if we conduct and do not conduct the proposed HTA activities. These will be heavily discounted as they both occur in the future, are subject to time discounting and are inherently uncertain. The earlier in the clinical development process HTA activities are undertaken, the greater the discount. 

This suggests that HTA activities should be left as late as possible. There is however one important constraint; the HTA activities must be completed before the decisions they are intended to inform. If we wish to use the results of HTA activities to inform trial design decisions, they must be completed early in the clinical development process. If the HTA is being used to inform decisions regarding the need for ancillary studies to support reimbursement, HTA activities must be completed in sufficient time to allow such studies to be commissioned and completed. 

Overall, early HTA activities need to be fit for purpose, and no more. If they are 'over-engineered', they are likely to be judged not cost-effective and their potential value will be lost.

-Neil Hawkins | Vice President, Health Economics, ICON Commercialization & Outcomes

Articles

Health technology assessment is a field of analysis that evaluates the medical, social, ethical, and economic implications of development, diffusion, and use of health technology.

Health Technology Assessment in Clinical Development: When to Bridge Research and Decision-Making


  The costs of drug development are increasing, driven in large part by increasing failure rates in all phases of clinical development. According to one recent report,

 the number of new drugs approved per billion US dollars spent has halved roughly every nine years since 1950.

A number of explanations have been put forward to account for this reduction in the research productivity. These include the continual need to produce technologies that are better than those we already have, and increased caution on the part of regulators. Whatever the cause, there is a pressing need to ensure that decisions are made effectively to optimize success rates and ensure that those products that do enter the market realize their commercial potential.

Many factors, of course, determine a drug’s chances of success. Marketing approval is dependent upon the characteristics of the treatment and on decisions made by manufacturers and regulators. The usage of a product and the revenues it generates are then determined by physician and patient preferences and payer reimbursement policies.

But before these external decisions come into play, manufacturers themselves must make many decisions over the course of the clinical development and the commercialization processes. Which pre-clinical development leads to pursue? Which products to move into Phase I, II and III clinical development? And for which indications and patient populations? How many subjects should be included in each trial and how long should they last? Which trial comparators and endpoints should be included?

Following approval there are more decisions around product pricing, marketing, launch strategy, and post-marketing research.

Such decisions often have long-term consequences. For example, decisions regarding trial design will affect not only the regulatory approval process, but also what information is available to payers, physicians, and patients for assessing the product’s value. In turn, this affects the product’s long-term revenue prospects.

Ultimately, a pharmaceutical, biotech, or device company’s success depends on knowing which products to develop and which to abandon. The financial consequences of pursuing the development of a product that ultimately fails in late stage trials or that underperforms once on the market can be devastating to a company. Conversely, it is also expensive to overlook potentially profitable products. To borrow diagnostic terms, decisions should ideally be both sensitive and specific so that unsuccessful products can be abandoned early without sacrificing potential winners.


  A range of consultancy and research services support the clinical development and commercialization process. These complementary services help manufacturers decide which candidates to pursue and how to design the development process. They also provide additional information that supplements the results of the trial programs and helps clinicians, patients, and payers determine the value of new treatments.

These complementary services broadly fit in two groups: research activities that collect new information or synthesize existing information; and consultancy activities that help decision makers interpret the available data. The common thread running through all of them is that they aim to improve decision-making by increasing the quantity and variety of data, making more efficient use of available data, and providing tools that support the decision making process—all to better characterize the potential value of new treatments.

These services include reviews of how existing products are priced and used and surveys of stakeholders such as patients, clinicians, and payers. Together they provide valuable information on what information is likely to be required by decision-makers and on the prospects for a new product.

It is important to note that clinical trials themselves are unlikely to provide all the information necessary to determine the relative value of a product

 (Sculpher et al. 2006). For instance, they are unlikely to include all relevant comparators or to measure all the clinical endpoints that represent how a patient feels, functions, or survives over a sufficient time horizon to reflect the potential benefits and drawbacks of treatment. Nor do they always include a sufficient number of subjects to measure all effects with adequate precision.

In some cases, clinical trials may not be the best strategy to answer a particular question, and other research methodologies may be required. These include epidemiological studies of treatment patterns and disease burden, research to develop and validate patient-reported outcome (PRO) instruments, and studies to measure resource utilization. Importantly, these complementary studies facilitate extrapolations from the clinical trial data. For example, they may allow trial endpoints (such as measures of response) to be extrapolated to clinical endpoints (such as survival and measures of quality of life).

These extrapolations are essential when assessing the value of a new product compared to existing ones. They also provide information that helps refine estimates of the potential future value of products in development. In addition to collecting new information, it is also important to identify and synthesize data from existing sources. This synthesis may involve describing available data, performing formal qualitative analyses, conducting a statistical meta-analysis, and developing formal analytic or cost-effectiveness models. Careful collation, review, distillation, and presentation of the existing evidence can help avoid some of the common cognitive biases associated with decision-making. These include anchoring on early evidence, making decisions based on easily recalled evidence (availability bias), overestimating success, and underestimating the impact of uncertainties.

More efficient use of available information can be made via structured analytics and modeling, such as:

Developing cost-effectiveness models to quantify the comparative value of new treatments

Using multi-criteria decision analyses that allow the explicit examination of trade-offs between different treatment options

Simulating the outcomes, pay-offs and uncertainties associated with different development options through models of the drug development process

Grading evidence to understand the strengths and deficiencies of the existing evidence base


  It is vitally important to take a strategic view when making decisions throughout the clinical development process. If critical information has not been collected within the trial program, it will typically not be possible to collect it post-approval in time to support launch.

There is, however, a paradox: due to the attrition of candidate products as they progress through the clinical development process, it is more “expensive” to fund additional research early on in the clinical development process. If, for example, 90 percent of drugs entering Phase II fail to reach market, an expenditure of US$10 in Phase II is the equivalent of an expenditure of US$100 at launch.

However, these considerations need to be balanced against the potential benefits of investing early in the process to increase efficiency and magnify the pay-off later on. As a consequence, research and consultancy activities conducted during the clinical development process need to be carefully designed in order to maximize their cost-effectiveness. For example, reviews of evidence should not be conducted to the exacting standards required by the various Health Technology Assessment (HTA) agencies, but rather they should focus on identifying and summarizing the evidence available from the most important sources. In general, these can be readily identified from existing reviews by consulting with subject area experts, supplemented by simple searches. Cost-effectiveness models may be deterministic and can often focus on comparisons within trials or on very simple indirect comparisons. In all cases, research should be commissioned and studies designed with clearly defined questions in mind. The decisions to be supported by the research need to be clearly identified. There also needs to be an understanding of the overall process and context in which these decisions are embedded. Essentially there needs to be an understanding of how research and consultancy activities are to be applied in the overall development and commercialization process.

Gather information from diverse sources in order to avoid availability bias (making decisions based only on easily recalled evidence);

Provide research that is accessible and not overly complicated;

Focus on the salient issues rather than on all issues;

As the development and commercialization process proceeds, ongoing research and consultancy activities help to refine estimates of a product’s future revenues and anticipated returns on investment.

-Neil Hawkins l Vice President, Health Economics, ICON Commercialization & Outcomes

References
  1. Scannell, J.W. et al., 2012. PERSPECTIVES. Nature Reviews Drug Discovery,
  11(3), pp.191–200.
  2. Sculpher, M.J. et al., 2006. Whither trial-based economic evaluation for health
  care decision making? Health Economics, 15(7), pp. 677–687.

The costs of drug development are increasing, driven in large part by increasing failure rates in all phases of clinical development.