Novel approaches to incorporating pharmacoeconomic studies into phase III clinical trials for Alzheimer's disease

Abstract

The societal and individual costs of Alzheimer's disease are significant, worldwide. As the world ages, these costs are increasing rapidly, while health systems face finite budgets. As a result, many regulators and payers will require or at least consider phase III cost-effectiveness data (in addition to safety and efficacy data) for drug approval and reimbursement, increasing the risks and costs of drug development. Incorporating pharmacoeconomic studies in phase III clinical trials for Alzheimer's disease presents a number of challenges. We propose several specific suggestions to improve the design of pharmacoeconomic studies in phase III clinical trials. We propose that acute episodes of care are key outcome measures for pharmacoeconomic studies. To improve the possibility of detecting a pharmacoeconomic impact in phase III, we suggest several strategies including; study designs for enrichment of pharmacoeconomic outcomes that include co-morbidity of patients; reducing variability of care that can affect pharmacoeconomic outcomes through standardized care management; employing administrative claims data to better capture meaningful pharmacoeconomic data; and extending clinical trials in open label follow-up periods in which pharmacoeconomic data are captured electronically by administrative claims. Specific aspects of power analysis for pharmacoeconomic studies are presented. The particular pharmacoeconomic challenges caused by the use of biomarkers in clinical trials, the increasing use of multinational studies, and the pharmacoeconomic challenges presented by biologicals in development for Alzheimer's disease are discussed. In summary, since we are entering an era in which pharmacoeconomic studies will be essential in drug development for supporting regulatory approval, payor reimbursement and integration of new therapies into clinical care, we must consider the design and incorporation of pharmacoeconomic studies in phase III clinical trials more seriously and more creatively.

Introduction

The societal and individual costs of Alzheimer's disease are already substantial, estimated at over $150B US dollars per year in the United States alone, and over $350B worldwide, following only heart disease and cancer (1, 2). These costs for Alzheimer's disease will increase significantly as populations rapidly age throughout the world, and is occurring as health systems throughout the world are struggling with finite health care budgets. Drug development requires 12 to 15 years and an investment of over $1 billion dollars (3). Many regulators and payers are beginning to require data demonstrating the cost-effectiveness of candidate drugs, in addition to safety and efficacy data, to consider drug approval and reimbursement. By requiring pharmacoeconomic data, society has introduced a final “black box” into the already risky, costly and complex process of drug development, with the added need for pharmacoeconomic and comparative effectiveness data. In some countries, safe and effective disease modifying drugs for Alzheimer's disease that do not demonstrate economic benefit may not be approved, or may not be reimbursed at prices that justify the costs of development. Agencies such as the UK's National Institute for Health and Clinical Excellence (NICE) (4) and comparative effectiveness initiatives in the United States (5, 6) and other countries (7) present a last hurdle that potentially decreases the number of investors and companies willing to engage in pharmaceutical research and development. Even when safety, efficacy and some pharmacoeconomic value have been demonstrated, societal willingness to pay can prevent adequate reimbursement that justifies the costs of drug development. In this worldwide context, pharmacoeconomic studies have become increasingly important for drug approval, pricing and reimbursement. In the U.S., expansions of access to health insurance under the just enacted health care reform legislation will almost certainly intensify the emphasis on more relevant and robust evidence, including evidence from pharmacoeconomic studies (8).

Here, we will address these issues directly, suggesting some specific ways to improve the design of pharmacoeconomic studies in phase III clinical trials for Alzheimer's disease.

Why focus on phase III for pharmacoeconomic studies?

There have been numerous attempts to demonstrate the pharmacoeconomic value of anti-dementia therapeutics through the use of several types of pharmacoeconomic studies (9, 10, 11). Modeling studies can be useful but are often seen as “abstract” to decision makers, and rarely are efforts made to validate the model outcomes with long term follow-up (12). Retrospective claims data are “real world,” often have the advantage of large sample sizes, and can be “hypothesis generating,” but cannot prove causality (13).

With regard to clinical trials, there are considerable differences between pharmacoeconomic studies in phase III and in phase IV (14, 15). Phase III clinical trial data from efficacy studies are considered the “gold standard“ to demonstrate therapeutic efficacy, with high internal validity because there is randomization and blinding. However, from the perspective of pharmacoeconomics, phase III studies have more limited generalizability with regard to economic outcomes in the “real world” because of stringent patient selection criteria, rigorous attention to drug compliance, and the use of specialist clinical centers. Cost-effectiveness clinical trials that have the potential advantage of being “real world,” require large sample sizes, but have less internal validity than phase III studies. For any indication in any disease state, there have been few successful, well designed phase IV cost-effectiveness studies (16, 17, 18, 19). Further, since cost-effectiveness studies are most often performed in phase IV, they cannot be used for drug approval and for initial justification of pricing.

When the strengths and weaknesses of each of these approaches are considered, the most desirable means of generating pharmacoeconomic data for drug approval and reimbursement would be to demonstrate pharmacoeconomic outcomes in phase III. Considerable previous attention has been given to the methodology required to incorporate cost-effectiveness outcomes in phase III studies, but these discussions have generally not considered the unique aspects of incorporating cost-effectiveness as an outcome in phase III clinical trials for Alzheimer's disease (20, 21, 22).

Challenges unique to defining value in AD disease modifying treatment

Compared to some other diseases, Alzheimer's presents some relatively unique challenges from an economic evaluation perspective.

One problem is “ageism,” or bias against the elderly (23, 24, 25, 26, 27). The economic value of the life of an elderly person is often viewed as limited by regulators and payers, since even healthy elderly persons generally have a relatively short life expectancy and often limited or no productivity as traditionally defined (as “work”).

A further bias, which we have termed “dementia bias,” may result from lack of knowledge and recognition of the economic value of cognitive vitality in older persons. In addition, drugs for dementia do not have obvious links to medical costs. For example, in patients with Alzheimer's disease who have cognitive impairment, non-compliance with an anti-dementia therapy does not have an obvious link to medical costs the same way that not taking a diuretic may affect congestive heart failure and lead directly to a hospitalization.

An additional bias may be termed “delayed consequence bias.” This refers to the misperception that Alzheimer's disease is associated only with long-term economic consequences such as nursing home placement which may not occur until years after diagnosis and treatment. While these consequences are clearly important thresholds, they under-estimate the ongoing substantial costs associated with the disease prior to these events.

Reimbursement for disease-modifying therapies, which target disease progression, may be even more challenging. The practical value of delaying the progression of dementia may be viewed as limited, particularly if the cost of therapy is high. Slowing the rate of progression may not be viewed as necessarily beneficial, since the state of being demented itself is a seriously impaired and generally undesirable one (28). Patients may show no symptomatic improvement and cannot internally validate the decrease in rate of disease progression. Clearly, while the ultimate goal is to prevent dementia, at this time, therapies are being developed in patients with progressive dementia as a means of demonstrating efficacy in patients with the clinical disease.

Measuring quality of life in the demented is fraught with difficulties due to the patient's cognitive impairment. Quality-adjusted life years (QALY's) are used as effectiveness measures in cost-utility studies for cost-effectiveness. In patients with dementia, however, self-reporting for quality of life presents unique challenges. Patients may be unable to provide assessments. Moreover, quality of life evaluations may not be the same for patients and their caregivers (29, 30). In most countries and health systems, informal care is a major source of excess costs for Alzheimer's disease which may increase with disease progression, substantially affecting the quality of life for caregivers (2). Previous pharmacoeconomic studies have incorporated caregiver quality of life and costs. However, though there is considerable variance across health systems and countries, most payers (including the United States) do not pay for informal caregiving, and as a result, the relevance of informal caregiving costs to pharmacoeconomic evaluations for payers not at risk for these costs may be minimal. It has also been difficult to convince most payers that resultant caregiver formal medical costs are directly related to caregiving and should be considered an important pharmacoeconomic outcome for AD disease modifying therapeutics. Finally, while lost caregiver productivity may have societal interest, many caregivers (particularly elderly spouses) are not in the workforce.

In addition, for geriatric patients with limited life expectancy, QALYs or the related metric, disability-adjusted life years (DALYS), may not be the sole or even appropriate outcome measure (23). Recent work has investigated the use of functional outcome measures for determining the value of therapy and for pharmacoeconomic outcomes (31, 32, 33).

Willingness to pay metrics offer some advantages, but can be a particularly difficult issue in geriatrics and in dementia therapeutics because of elicitation challenges. In addition, there may be an “ageism” bias against the aged. In the developing world, for example, where the majority of new cases of Alzheimer's disease will occur in the 21st century, the majority of the population is under 25 years old, creating conflicting needs between generations. Such intergenerational conflict may only be resolved by effective patient group advocacy, a risky and unfortunate final endgame to the process of drug development.

Acute Episodes of Care as a Key Pharmacoeconomic Outcome in Alzheimer's disease

To address these key societal health economic issues, a common, significant, pharmacoeconomic outcome measure is needed in phase III AD clinical trials that can be employed across health systems and countries, and will be easily understood and appreciated. We propose that hospitalization, and more broadly, acute episodes of care that include emergency room visits, hospitalization, post-acute care in skilled nursing facilities and post-acute home care services related to the acute event (such as home skilled nursing and home health aides, as well as rehabilitation) may be a relevant metric. There is extensive evidence that both annual hospitalization rates, as well as post-acute skilled nursing rates, are increased in community dwelling patients with Alzheimer's disease and related dementias (34, 35, 36, 37, 38, 39, 40, 41). In addition, not only are hospitalization rates increased, but hospital lengths of stay (and therefore annual numbers of hospital days) are also increased significantly (see figure 1). A summary metric for an acute episode of care could include total number of “acute episode of care” days that would include days spent in hospital and in post-acute settings (both post-acute inpatient and outpatient services). This metric could have particular value in multi-national clinical trials where health systems vary significantly with regard to the nature and utilization of acute and post-acute care.

Figure 1.

Annual Acute Care Costs, Admissions and Hospital Days from “Real World” Claims: AD Versus Controls

Furthermore, there is some evidence that event rates for hospitalization and for annual hospital days are lower in clinical trials (as may be expected from this selected population, with potentially fewer comorbidities and better clinical care) (42, 43). Nevertheless, some studies suggest that even the currently available symptomatic agents can reduce hospitalization rates and annual hospital days (figure 2) (44).

Figure 2.

Annual Hospitalization Rates and Hospital Days in Clinical Trials

Stage of disease is another issue to be considered in hospitalization rates for Alzheimer's disease. The causes of hospitalization vary by disease stage. In mild to moderate AD, patients may have some degree of functional impairment affecting basic activities of daily (ADL), and instrumental activities of daily living (45, 46). For community-dwelling dementia patients not on Alzheimer's disease therapy, annual healthcare costs increased by $714 for each additional instrumental activities of daily living impairment and $1,541 for each additional ADL impairment (47), demonstrating a direct relationship between function in dementia patients and their healthcare costs. Mild to moderate patients are prone to accidents, medication mismanagement and poor control of comorbidities, preventable hospitalizations, behavior disorders, and begin to need caregiver assistance and supervision. Late stage Alzheimer's disease patients frequently have multiple complications of dementia, such as infections (48), leading to hospitalization and institutionalization. At end stage, hospital costs may actually decline (49). Pharmacy costs increase as patients require cognitive enhancers, psychotropics, and antidepressants. Depending on the stage of disease being targeted in the clinical trial (mild cognitive impairment, mild to moderate, moderate to severe), these differences in causes of acute episodes of care should be considered in a particular pharmacoeconomic study design, and in the design of any care management strategy that is included in the study (see below).

If acute episodes of care are a key pharmacoeconomic outcome for Alzheimer's disease, what are the issues in designing phase III clinical trial for a disease modifying therapy for Alzheimer's disease to demonstrate cost-effectiveness?

Improving pharmacoeconomic outcomes in clinical trials that employ biomarkers

The drive to use biomarkers such as neuroimaging as outcomes measures in disease modifying clinical trials is greatly needed in order to demonstrate biological efficacy, particularly in proof-of-concept phase II studies, providing data on which phase II candidates to advance to phase III. Biomarkers enable smaller sample sizes and may reduce the duration of AD clinical trials, potentially reducing their costs (50, 51). Biomarkers can also be helpful in Phase III to support disease-modification claims and to differentiate the labeling of compounds from symptomatic agents (51). In some cases, biomarkers function to monitor adverse events such as cerebral edema in Phase III trials and would impact pharmacoeconomic formulations (52).

However, as sample sizes are reduced and the time of study shortened, the probability of observing a meaningful pharmacoeconomic outcome is reduced. This is because event rates such as hospitalization in clinical trials are low, and even lower than in community based samples (see figure 2) (42, 43). In addition, as study length is shortened, the lag time to pharmacoeconomic events becomes more problematic.

Pharmacoeconomic event capture rates can be increased by extending the duration of the trial, adding an open label or follow-up period to the trial, or enriching the trial. Since it may be expected that pharmacoeconomic outcomes may lag biomarker outcomes, adding an open label follow-up period may be particularly valuable in biomarker studies (see figure 3). While a phase III open label extension provides additional time to observe event rates, the absence of a placebo control group for this period limits the ability to draw firm conclusions from this part of the study.

Figure 3.

Hybrid Designs for Increasing the Power of Economic Studies in Phase III Efficacy Trials for AD

Challenges in economic outcomes in multi-national clinical trials

Quality of care, systems of care, availability of formal and informal care, and the cost of care, vary enormously across the spectrum of the developed and developing world, and even within regions of countries (21, 53). Attempting to conduct a multi-national clinical trial in which standard pharmacoeconomic outcomes are employed presents almost insurmountable challenges. Costs for services are highly variable, and system wide availability and use of services also varies greatly. These variable inputs mean the value of a given treatment is inconsistent across societies, and a single pricing model cannot apply internationally based on a single trial. While this often leads to multi-national phase III clinical trials without significant pharmacoeconomic outcomes, the requirement for a pharmacoeconomic trial in each and every country is expensive and time consuming. Country-by-country analyses of pharmacoeconomic measures where the sample sizes are large enough may also facilitate drawing informed conclusions. This strategy depends on planning the sample sizes for each country or region as part of the trial design.

To advance the conduct of phase III pharmacoeconomic multinational studies, we propose the development of “harmonized” standards that apply and translate well across health systems. For example, one harmonized standard might be “episodes of acute care” we describe above, since hospitalization is certainly a similar pharmacoeconomic event across regions and countries. Multinational efforts at harmonizing standards for economic studies in phase III clinical trials for Alzheimer's disease should be considered an important agenda for consideration by pharmaceuticals companies conducting such studies.

Using clinical practice guidelines and disease management to offset variable care management effects

Care management has been demonstrated to have a significant impact on outcomes in Alzheimer's disease (54, 55). Previous studies have demonstrated that the use of guidelines, telephonic nursing interventions, and other structured approaches can have effects on outcomes that are on the order of magnitude of currently available drugs (56, 57, 58). These interventions have had significant health economic impact, such as delay to nursing home placement (59).

The “sentinel effect” of simply participating in a clinical trial impacts outcomes because patients get advice and care from experts, raising the floor of care management compared to an average patient receiving care in the community setting. This sentinel effect not only affects clinical outcomes, but can be expected to impact pharmacoeconomic outcomes, such as the cost of care for comorbidities, and hospitalization rates. One approach to resolving this issue has been the use of “real world, cost-effectiveness” studies (15). However, as discussed previously, there have been few, if any, successful “real world” effectiveness (as opposed to efficacy) studies, study designs are still evolving, and these studies generally occur in phase IV, after regulatory approval and pricing have occurred (16, 17, 18, 19).

Further complicating both clinical and pharmacoeconomic outcomes in global phase III clinical trials is the wide variability in the standards of care across sites and countries. Clinical trials to date have not addressed this issue, allowing for the variability of care to be essentially random. Perhaps, this is one reason why recent studies have found wide variability in the placebo groups rate of progression, greatly effecting efficacy outcomes, and no doubt affecting pharmacoeconomic outcomes (60).

Dealing directly with this issue of variability in care has advantages and disadvantages. Standardizing care management across all sites can reduce variability, but should also improve the level of care patients receive. Conceivably, this may reduce pharmacoeconomic event rates. However, perhaps by standardizing care across all clinical trial sites, the true efficacy and pharmacoeconomic effectiveness of a pharmaceutical in a phase III clinical trial setting can be more reliably and consistently measured, especially in Alzheimer's disease where care management can have a very large impact on outcomes. The application of disease management in clinical trial designs has been previously discussed (61).

Standardization of care through the use of guidelines, care management or disease management approaches further raises the issue of whether the study is relevant to the “real world.” There are two responses to this objection. First, clinical trials are almost never considered “real world” in any event, so there should be little issue with having a “better” clinical trial.

Second, and perhaps of greater significance, is that if a clinical trial incorporates a standardized care management or disease management protocol into the efficacy study, it is conceivable this could have considerable marketing and “real world” clinical advantages if the drug is approved. There is substantial evidence and precedent for clinical value of marketing drugs in combination with care management. Indeed, for almost all chronic diseases, including Alzheimer's disease, drug therapy must be combined with clinical care to obtain optimum outcomes.

Implementing a standardized care management or disease management strategy in a phase III clinical trial employing guidelines for care can be based on available clinical practice guidelines (62, 63, 64). The use of telephonic care for patients with Alzheimer's disease in clinical trials may be particularly useful and less costly to insure quality care in the interim between visits (65, 66). As mentioned above, stage of disease in the trial should be carefully considered in designing these efforts.

Quality of care can vary widely across regions, nations and cultures. Case management is a common approach to coordinating care and insuring that quality care is provided, and that caregivers are active and educated participants. Across regions and countries, case managers may have various background, skills, and training. Case management strategies could be simplified, specified, and integrated across trial sites in Phase III multinational Alzheimer's clinical trials. Clinical instruments such as the interRAI designed to standardize assessment of the complex needs of the geriatric patient (67, 68) as well as standardized guidelines for the care of patients with Alzheimer's disease (69, 70) can be employed to assess patients' disabilities and evaluate their needs. Case management can coordinate care for comorbidities with other professionals. The effectiveness of their actions can be monitored through training and oversight.

Methodologically, implementing standardized care as a new aspect of a clinical trial could lead to improvement in the placebo group and greater difficulty demonstrating a drug-placebo difference. Initiating the new standards during a placebo-run in period would minimize this risk. Participating in trials also tends to improve care and enhance placebo effects and this “trial effect” might be mitigated by a placebo run-in with standardized care.

Capturing better pharmacoeconomic data in a phase III clinical trial using claims data

The use of electronic claims data and the potential of electronic medical records to advance pharmacoeconomic studies in clinical trials has great potential (20). Health economic data are currently collected at clinical visits primarily by self report from caregivers. Such data can be unreliable, and does not necessarily capture services actually paid for and therefore relevant to a specific health system. It is possible to collect electronic claims data as a more reliable and consistent measure of direct medical costs paid by the system of care under study and accrued during the time of trial. This can be done by incorporating the request for claims data into the Institutional Review Board submission, and the informed consent. In addition, if an open-label period is employed, capturing electronic claims can reduce the costs and facilitate the gathering of health economic data. Thus, hybrid studies incorporating the collection of administrative claims into the phase III clinical trial presents an important opportunity for improving the comprehensive and reliable capture of health economic data in clinical trials, while potentially also reducing the cost of such data collection.

Enrichment designs for pharmacoeconomic studies in clinical trials

Most elderly (>90%) with AD have at least one medical comorbidity that significantly impacts costs (eg, diabetes, congestive heart failure (CHF)), and 60% have more than 3 comorbidities (13, 71, 72). These studies indicate that cognitive impairment can drive costs in AD patients with medical co-morbidities. Thus, economic studies in AD should include patients with multiple comorbidities to reflect the “real world” geriatric patient population with dementia, but this is often not the case, since patients with multiple co-morbidities may be excluded from trials intentionally, or through selection bias. For analysis of such studies, it may be useful to employ comorbidity indexes (such as the Charlson index) to control for comorbidities as an important source of excess costs in Alzheimer's patients (69).

Clinical trials designed to investigate the impact of Alzheimer's disease modifying therapies on medical comorbidity costs (particularly those with high pharmacoeconomic event rates such as hospitalization) could be done. For example, congestive heart failure is the most common cause of hospital admission in the United States. CHF patients with dementia have higher costs than those without dementia (73, 74, 75). We and others have suggested currently available cognitive enhancers may improve medical management for co-morbid diseases such as diabetes and CHF (76, 77). The impact of disease modifying therapies on improving medical management for comorbid diseases may be a relevant intermediate economic outcome in clinical trials.

Event rates of hospitalization in patients with CHF and dementia are higher than in patients with CHF alone, and much higher than in patients with dementia alone (see figure 4). These data suggest that the opportunity for reducing hospitalization rates in patients with CHF and dementia is far greater than in patients with dementia alone, particularly in the mild to moderate stages of the disease. Perhaps, clinical trials with a focus on the pharmacoeconomic value of anti-dementia drugs could be done in such enriched populations with co-morbid illness. In other disease states, there are examples of enrichment models for clinical trials that can be adapted to clinical trials of disease modifying therapies for Alzheimer's disease (78, 79).

Figure 4.

Enrichment Strategies: Increased Hospital Utilization in Patients with AD and CHF

Because economic events rates and costs for patients with mild cognitive impairment (MCI) may be even lower than those with dementia, the cost-effectiveness of drugs for prevention trials focused on MCI will be especially challenging. For these studies, effects on comorbidity management and medication compliance, including the use of enrichment designs to demonstrate pharmacoeconomic outcomes could be especially valuable (80).

Power analyses for economic outcomes in Alzheimer's disease modifying clinical trials

The incorporation of pharmacoeconomic outcomes into phase III clinical trials should be considered as seriously as clinical outcomes. Until now, funders and investigators generally have been unwilling to commit the additional resources to a clinical trial to adequately power pharmacoeconomic outcomes if the sample size required for answering the economic question is larger than that required for answering the clinical question. Economics have not been considered important enough for the additional expense and effort. With the increasing emphasis on pharmacoeconomics and comparative effectiveness worldwide, this bias must change.

Power analyses for pharmacoeconomic outcomes should be hypothesis driven. For example, for current disease modifying trials in Alzheimer's disease, power analyses for clinical outcomes may be modeled around a 30% slowing in the rate of disease progression during an 18 month period of study (51, 81). However, is it also realistic to expect a 30% change in health economic outcomes (like hospitalization) during this period? Or, is the change in health economic outcomes expected to be less (eg. a 10% decline in hospitalization rates)? Will the change in health economic outcomes lag the time of the trial for detecting disease modification, arguing for the open label or extended observation period? Will considerably larger sample sizes be needed to obtain a result, since the sample size required to answer economic questions is generally larger than the sample size required for clinical outcomes? The basic principles of sample size estimation for cost-effectiveness analysis are well developed (20). As with sample size estimation for clinical outcomes, there is no certainty about the parameter values that should be used in the power analysis.

Specific pharmacoeconomic issues of biologicals

Biologicals such as immunotherapies are generally more expensive than oral agents. Nevertheless, the cost-effectiveness of biologics has been investigated and demonstrated (82, 83). The disease modifying effect and safety profile of biologics in Alzheimer's disease is not yet known. Biologicals, such as a passive immunization with a monoclonal antibody, may be especially expensive if employed ultimately for long-term preventive therapy (84). While an “active” vaccine might resolve some of the pharmacoeconomic concerns of a “passive” vaccine to prevent Alzheimer's disease, obviously the efficacy and safety of vaccines for prevention will require extensive testing.

Modeling studies could be helpful to create hypotheses and consider the differences in costs and comparative effectiveness of an AD disease modifying biological versus an oral disease-modifying therapy, or versus current standard of care with symptomatic agents. Since the cost-effectiveness equation always includes pricing of the drug, demonstrating cost-effectiveness for an expensive biological may be particularly difficult. We may expect that if current biologicals in clinical trials prove efficacious and reasonably safe and are among the first to market as disease modifying agents for AD, patient advocacy and societal willingness to pay, rather than clear answers from a pharmacoeconomic outcome in a phase III trial, may be expected to drive regulatory approvals and payer reimbursement.

Conclusions

Evidence from clinical trials consistently demonstrates that Alzheimer's therapies currently on the market can improve cognition, function, behavior and other clinical outcomes in some patients. From an individual perspective, these clinical outcomes have inherent value. Disease modifying drugs currently in clinical development for Alzheimer's disease are using similar clinical outcomes, as well as rate of change measures and biomarkers to demonstrate efficacy.

Economically, from a population perspective, the equation is far less clear. Accelerating changes in regulatory and payer requirements for economic data emphasize that we need to take pharmacoeconomics far more seriously in clinical development. Here we have suggested several novel ways to incorporate pharmacoeconomic studies into disease modifying clinical trials for Alzheimer's disease.

For Alzheimer's disease, like other medical illness such as CHF or diabetes, we believe that a key pharmacoeconomic outcome for regulators and payers is hospitalization. We have further suggested that bundling hospitalization as part of an “episode of acute care” might have advantages in demonstrating pharmacoeconomic effects of disease modifying therapies, particularly in multinational trials, and that reductions in costs for “episodes of acute care” of disease modifying Alzheimer's drugs can be achieved with appropriate study design.

We have also offered some novel approaches to clinical trial design to improve the ability to incorporate pharmacoeconomic studies into phase III clinical trials. We have proposed that the focus should be on better designs for phase III clinical trials in the pre-approval time frame, so that these data can be employed for regulatory approval, pricing, and reimbursement.

The individual willingness to pay for improved cognition and function is great, but societal willingness to pay may be less enthusiastic for a variety of reasons, not only including perceptions of efficacy or safety concerns, but also because of ageism, dementia bias and other factors. If enough developed societies around the world demonstrate reluctance to pay for Alzheimer's disease treatments, innovation may be stifled. The pharmaceutical sector – also with limited budgets and resources – will be forced to make equally hard decisions about which drugs it will advance in development, based not only on the unmet market medical need, the potential clinical efficacy and safety of drugs development, but also on whether their efforts will result in a product that will be reimbursed by payers or will languish on pharmacy shelves for lack of willingness to pay.

Academia and industry could spend many tens of billions of dollars developing reasonably worthwhile disease-modifying drugs over many decades, only to find that the reimbursement of these drugs depends, at the end, on patient advocacy and societal politics to make such drugs available. The risks of this scenario are too high. Therefore, substantial serious attention must be paid to designing better pharmacoeconomic studies into phase III clinical trials, so that the value of anti-dementia disease modifying therapy can be clearly communicated to all concerned.

Acknowledgements: The authors thank Diana Shineman, PhD, Michael Grundman MD and Henry Glick, PhD for helpful discussions.

Disclosures: The authors report no disclosures that are relevant to this manuscript.