One way in which governments, health technology assessment (HTA) bodies, and healthcare decision makers can seek to encourage the development of truly new medicines (and new drug classes) is to recognize and reward innovation. In addition to the market advantages granted by providing better outcomes than existing therapies, innovation is commonly rewarded by the acceptance of a premium price for a new product during reimbursement and price-negotiation processes; ideally these processes should stimulate ongoing innovation while obtaining good value for money. A recent report in the United Kingdom has suggested that a focus solely on price minimization, rather than on product quality and entire life-cycle cost optimization, could reduce the incentives for innovation and potentially the attractiveness of the country as a setting for researching and developing pharmaceutical treatments (1). In addition, focusing only on price control may shift manufacturers’ incentives toward the development of high-cost drugs with large additional benefits, at the expense of incremental innovation in highly competitive areas, including common diseases.
To recognize important innovation in medicines, decision-making bodies must use explicit or implicit definitions of what characteristics constitute rewardable innovation. Ideally, agreement on such a definition across countries would provide a consistent incentive to manufacturers to conduct research into new methods of treating diseases, and simplify drug development, reducing costs and prices. Several frameworks for assessing the value of new medicines exist, for example, the American Society of Clinical Oncology (ASCO) and the European Society for Medical Oncology (ESMO) have produced value frameworks based on clinical benefit, toxicity, and impact on health-related quality of life (Reference Schnipper and Bastian2–Reference Cherny, Sullivan and Dafni4). However, these frameworks do not specifically address innovativeness, and there appears to be little consensus on which types of medicines are in fact innovative.
Innovation in health care does not fully follow the same pattern as other industries. In many sectors, innovation is typically associated over the long term with a reduction in costs as well as an improvement in the end product. This is not routinely the case in health care: a new product is often substantially different from existing therapies, and the improvements in patient outcomes that result from the use of innovative new medicines tend to be accompanied by increased expenditure by the healthcare system (Reference Cox, Kamal and Jankiewicz5;Reference Penington and Stubbings6). It is, therefore, important to consider what type or level of innovation justifies an elevated price (Reference Aronson, Ferner and Hughes7). All healthcare systems have finite resources, and there is a risk that inappropriate rewarding of innovation (for example, paying a premium for a drug that does not improve outcomes simply because it is new) could prevent funds being better spent elsewhere, and, therefore, lead to an overall decrease in patient outcomes at both an individual and a population level (Reference Claxton, Martin and Soares8). In addition, recent developments in the personalization of treatment for many conditions may mean that in the future reimbursement may need to focus on outcomes (at an individual or population level) rather than the acquisition costs of drugs; this may require a new approach to the HTA process.
The recent growth in development of new pharmaceutical products that are deemed innovative and efficacious, but are expensive, is accompanied by (and partly the cause of) an increasing focus in Europe, and increasingly in the United States, on the cost and affordability of health care (Reference McCabe, Bergmann and Bosanquet9). However, discussions about new medicines do not typically involve consideration of cost as a key component of innovation, and there is a risk that, in many countries, innovation in pharmaceuticals may not meet the needs of the wider healthcare system (and by extension, of society as a whole) (Reference Buxton and Chambers10). The aim of this systematic review is to investigate how innovation is defined with respect to new medicines, and to assess the extent to which published definitions of innovation incorporate the impact of new medicines on healthcare costs.
METHODS
Search Strategy
A series of systematic literature searches was conducted on May 25, 2016. Full terms used in all searches are listed in Supplementary Tables 1–5.
MEDLINE, Embase, and Embase Alert databases were searched using ProQuest Dialog (Ann Arbor, MI). The search strategy included multiple free-text terms covering the definition of innovation, combined with terms referring to health care, drugs, medicines, or pharmaceuticals. Additional searches combined healthcare terms with Medical Subject Heading (MeSH) and Emtree thesaurus index terms. Search results were limited to studies published from January 1, 2010, and filtered to exclude studies published only as conference abstracts. No language restriction was applied. The EconLit database was searched using the American Economics Association interface (https://www.aeaweb.org/econlit/). EconLit search terms combined “innovation”/”innovative” with terms including definition, health, drug, and medicine. EconLit search results were restricted to journal articles published in English from January 1, 2010.
Study Selection
Search hits from the three databases were combined, and duplicates were removed using EndNote software (Thomson Reuters, New York, NY). Remaining duplicate articles and conference abstracts were manually removed and titles and abstracts screened for eligibility. Articles were included if they described a definition of innovation with respect to new medicines, or referred to a relevant definition published elsewhere. Articles presenting definitions of innovation in medical devices, surgical techniques, or service delivery were considered to be outside the scope of this review, and were excluded. Full-text versions of articles that passed title/abstract screening were retrieved for further review, and studies not meeting the inclusion criteria were excluded (Supplementary Table 6).
Where studies referred to a relevant definition of innovation published elsewhere, the cited references were screened for inclusion in the review; no date restriction was applied to references identified through citation searching.
Data Extraction and Analysis
Definitions of innovation were extracted from all included references. In addition, terms used to describe components of innovation in each of the included definitions were identified. Because a large number of terms were identified, many of which described similar concepts (for example, “therapeutic benefit” and “therapeutic value”), similar definition terms were clustered together into ten groups; therefore, the resultant “dimensions of innovation” are derived from the identified terms, rather than being defined a priori. The mapping of definition terms onto the ten dimensions is shown in Supplementary Table 7. Definitions described in multiple publications were counted more than once in this analysis, but to avoid double-counting the same reference, the results of a previous systematic review of innovation (Reference Kesselheim, Wang and Avorn11) were excluded.
Supplementary Searches
Manual searches of relevant Web sites, including European HTA bodies, the European Medicines Agency, and key professional societies (for example, ASCO and ESMO), were conducted to identify stated policies or methods for assessing innovation. For HTA body Web sites, individual product assessments were not searched. Similarly, for professional societies, conference proceedings were excluded. A full list of Web sites searched and the search terms used is presented in the Supplementary Table 8. Because the Web site searches were not fully systematic, definitions identified from these sources were not included in the analysis of the dimensions of innovation.
RESULTS
Search Results
In total, 2,844 articles were retrieved in the database searches. After removing duplicates from the records, the titles and abstracts of 2,419 unique articles were screened. In total, twenty-eight articles were identified as being potentially relevant to the review objectives, and full-text versions were obtained. Of these, thirteen articles were excluded (Supplementary Table 6). Citation searching identified a further forty-two potentially relevant articles, of which twenty-four were excluded at the title/abstract screening stage, and three were excluded following full-text review. The flow of studies through the screening process is shown in Figure 1.
Study selection flow diagram.
Published Definitions of Innovation
In total, thirty-six published articles describing definitions of innovation were included in the review (Reference Aronson, Ferner and Hughes7;Reference Kesselheim, Wang and Avorn11–Reference Autret-Leca45) (Table 1). Of these, one study was a previous systematic review of how innovation is defined in drug development; the majority of articles identified defined drug innovation in terms of the number of yearly approvals of new drugs or patents (Reference Kesselheim, Wang and Avorn11). Because all of the studies in this previous review are included in the present analysis, this study was excluded from the analysis of the dimensions of innovation in published definitions to avoid double-counting. The remaining thirty-five studies described forty-four definitions of innovation, with thirty-five different definitions identified in total.
Definitions of Innovation Identified in the Published Literature
CNS, central nervous system; DRG, diagnosis-related group; EMA; European Medicines Agency; EPAR; European public assessment report; EVITA, EValuation of pharmaceutical Innovations with regard to Therapeutic Advantage; FDA, Food and Drugs Administration; mAb, monoclonal antibody; n/a, not applicable; NCE, new chemical entity; NICE, National Institute of Health and Care Excellence; NME, new molecular entity; NNT, number needed to treat; QALY, quality-adjusted life-year; UK, United Kingdom; USA, United States of America; VBP, value-based pricing.
Several definitions of innovation were presented in more than one reference. In particular, four references described an algorithm based on the availability of existing treatments and therapeutic effect of a new therapy, with technological and pharmacological innovation included in the case of products for diseases responsive to previously available interventions (Reference Andria, Auriemma and Attanasio12–Reference Motola, De Ponti and Poluzzi15). This algorithm is used by the Agenzia Italiana del Farmaco (AIFA) in Italy. In brief, drugs for diseases without a recognized standard treatment are classed as important innovations if they have at least a partial benefit on clinical endpoints, but not if they provide only a minor or temporary benefit. In the case of diseases where subsets of patients have a limited response to existing treatments, a new therapy would need to provide a major benefit to be classed as an important innovation. For conditions responsive to existing treatment, new products could achieve a moderate innovation rating if they provide a major therapeutic improvement, but those with similar efficacy and safety to existing therapies are likely to be classed only as technological or pharmacological innovations (Reference Motola, De Ponti and Rossi14;46). The descriptions of this algorithm in two studies also included disease severity as a factor (Reference Andria, Auriemma and Attanasio12;Reference Motola, De Ponti and Poluzzi15).
In addition, five references described the approach to innovation taken by the National Institute for Health and Care Excellence (NICE). NICE considers a product to be innovative if it “adds demonstrable and distinctive benefits of a substantial nature which may not have been adequately captured in the reference case QALY (quality-adjusted life-year) measure” (Reference Bryan, Lee and Mitton16–Reference Rawlins, Barnett and Stevens20), and if the product offers a “step-change in terms of outcomes for patients” (Reference Bryan, Lee and Mitton16–Reference Ferner, Hughes and Aronson18). To be a “step-change,” a product must have a clinical benefit, particularly in the treatment of a condition with no existing effective treatment, or for which there is no consistently satisfactory treatment (Reference Ferner, Hughes and Aronson18).
A further three references described innovation in terms of a combination of the technology level and comparative effectiveness (Reference Heible21–Reference Sternitzke23). For example, one study considered technology level to be the technical level of innovation in relation to existing drugs for the same diseases; for new molecular entities, the technology level would be high. Comparative effectiveness was defined as the therapeutic value which the drug offers compared with competing products in the same therapeutic class (Reference Heible21); other authors have used a product's Food and Drugs Administration (FDA) review type (standard or priority) as a proxy for therapeutic potential (Reference Sorescu, Chandy and Prabhu22;Reference Sternitzke23). In this framework, substantial technical advances coupled with high therapeutic potential constitutes radical innovation. High levels of technical advance or therapeutic potential alone are classed as technological and market breakthroughs, respectively (Reference Sorescu, Chandy and Prabhu22).
The remaining twenty-seven definitions (Table 1) included a wide range of approaches to innovation. Some approaches are straightforward: one study defined a pharmaceutical innovation as any development that is intended to produce a therapeutic advance (Reference Wardell and DiRaddo41), and another considered innovation to be superiority over an active comparator in clinical trials (Reference Barbui, Cipriani and Lintas26). Other definitions included multiple aspects. For example, one proposed definition of innovation was “a medicinal product that provides, through a step-change, something novel, with the potential or proven ability to yield, for individuals and/or their society, a treatment not previously available or a clinically significant improvement in treatment, with large health gains and a favorable benefit to harm balance, at an acceptable cost” (Reference Aronson, Ferner and Hughes7). Similarly, an international workshop on drug innovation defined an innovation as a new active substance or biological entity that can improve the quality of patient management and outcomes, but also suggested that new indications, technological and manufacturing processes, formulations, and delivery systems could constitute innovations when applied to existing drugs (30).
In addition to the approach used by AIFA, two further studies described algorithms for assessing drug innovation (Reference Caprino and Russo27;Reference Puntmann, Schmacke and Melander37). Both algorithms were more complex than the AIFA method, but included similar elements, being broadly based on efficacy, safety, and unmet need (disease severity or absence of available treatments). One algorithm also incorporated the type of endpoint (surrogate vs clinical) used to demonstrate improvements in efficacy (Reference Puntmann, Schmacke and Melander37), while the other included the design of key clinical studies (e.g., use of an active comparator) and an assessment of adherence to a new product (Reference Caprino and Russo27).
Overall, twenty-five distinct definitions referred to clinical effectiveness, which was described using a variety of terms (Reference Aronson, Ferner and Hughes7;Reference Motola, De Ponti and Poluzzi15–Reference Green17;Reference Linley and Hughes19;Reference Heible21;Reference Adami, Ciampalini and Dell'Aera24–Reference Joppi, Bertele and Garattini31;Reference Lexchin33–Reference Salter, Zhou and Datta39;Reference Wardell and DiRaddo41). Terms relating to the availability of existing treatments (eleven definitions) (Reference Aronson, Ferner and Hughes7;Reference Mol, Arnardottir and Motola13;Reference Motola, De Ponti and Poluzzi15;Reference Bryan, Lee and Mitton16;Reference Caprino and Russo27;Reference Joppi, Bertele and Garattini31;Reference Lexchin33;Reference Soleimani and Zenios40;Reference Gonçalves, Maraninchi and Marino44) and to disease severity or unmet need (seven definitions) (Reference Motola, De Ponti and Poluzzi15–Reference Green17;Reference Gonzalez and Hofer28;Reference Morgan, Lopert and Greyson34;Reference O'Connor, McDonald and Lam35;Reference Puntmann, Schmacke and Melander37) were also commonly used. Novelty was mentioned in eight definitions (Reference Aronson, Ferner and Hughes7;Reference Ferner, Hughes and Aronson18;Reference Heible21;Reference Caprino and Russo27;30;Reference Kwong and Norton32;Reference Petrini36;Reference Nelson, Cohen, Greenberg and Kent42), while a further three referred to newness as a characteristic of innovative drugs (Reference Green17;Reference Gridchyna, Aulois-Griot, Maurain and Bégaud29;Reference Petrini36). Trial endpoints or the use of an active comparator were important factors in four definitions (Reference Adami, Ciampalini and Dell'Aera24;Reference Barbui, Cipriani and Lintas26;Reference Caprino and Russo27;Reference Puntmann, Schmacke and Melander37). Notably, only two definitions included cost as an element of innovation. The definition proposed by Aronson et al. in 2012 requires innovations to provide therapeutic advances “at an acceptable cost” (Reference Aronson, Ferner and Hughes7), while that described by Gonçalves et al. in 2016 suggests that “medical gains must be large with a favorable risk–benefit profile and ideally, acceptable costs” (Reference Gonçalves, Maraninchi and Marino44).
Additional Definitions of Innovation
Manual Web site searches identified definitions of innovation from NICE and AIFA that matched those in the published literature, as well as additional definitions from the Haute Autorité de Santé (HAS; France), the Tandvårds- och läkemedelsförmånsverket (TLV; Sweden), the Scottish Medicines Consortium (SMC), the Zorginstituut Nederland (ZINL), and the National Health Service in England (NHS England) (Table 2) (46–53). In France, HAS defines innovative products as those for which the manufacturers claim a moderate to major improvement of the clinical benefit compared with that provided by existing treatments (i.e., Amélioration du Service Médical Rendu [ASMR] of level I, II, or III) (Reference O'Connor, McDonald and Lam35). By contrast, in the Netherlands innovative medicines are considered to be those which are promising, but for which insufficient data are currently available to be able to grant positive advice (52). The TLV, the SMC, and NHS England all included newness or novelty in their definitions, but required this to be associated with added value (50), benefits for patients (51), or improvements in the quality of health and care (53), respectively. In Germany, innovation is not part of the legal framework for the assessment of new drugs, which focusses on the additional benefit provided.
Additional Definitions of Innovation Identified in Web Site Searches
ASMR, Amélioration du Service Médical Rendu; QALY, quality-adjusted life-year.
Dimensions of Innovation in Published Definitions
The results of reducing the terms used to describe innovation from all forty-four definitions identified in the literature (including duplicates) to ten dimensions of innovation are shown in Figure 2. Therapeutic benefit was the most commonly occurring dimension, found in forty definitions. No other dimension was seen in more than half of the definitions, with novelty and the availability of existing treatments the second and third most common dimensions. Unmet need and safety were each factors in ten definitions. Administration (which includes factors such as convenience and adherence) occurred less frequently than newness.
Dimensions of innovation, number of occurrences in identified definitions.
DISCUSSION
In total, the systematic search found forty-four definitions of innovation, with a further seven identified through manual Web site searching. Some definitions occurred more than once; in particular, the algorithm used by AIFA in Italy (Reference Andria, Auriemma and Attanasio12–Reference Motola, De Ponti and Poluzzi15) and the approach used by NICE in England (Reference Bryan, Lee and Mitton16–Reference Rawlins, Barnett and Stevens20) were both mentioned in several publications.
Analysis of the dimensions of innovation included in the definitions identified suggested that the therapeutic benefit offered by a new product is generally considered to be the most important factor in categorizing a new medicine as innovative. However, quantification of therapeutic benefit is consistently absent from definitions, and is left to subjective interpretation. Under several definitions, drugs for indications where no previous therapies exist and those with novel structures or mechanisms of action would also be considered innovative, although novelty of structure of mechanism alone is not rewarded by HTA bodies, payers, or clinicians. Notably, only two definitions included cost, and these required only that the cost of an innovation be “acceptable” (Reference Aronson, Ferner and Hughes7;Reference Gonçalves, Maraninchi and Marino44).
Our findings are similar to those of a 2013 systematic review of innovation in drug development, which found that the majority of articles identified defined innovation in terms of the number of new drugs (21/42; 50 percent) or patents (4/42; 10 percent); the three studies that included economic elements assessed innovation in terms of productivity, cost-effectiveness, and market share (Reference Kesselheim, Wang and Avorn11). The potential for innovations to reduce costs is rarely described in the literature. One study identified in this review was a 2009 systematic review of the cost-effectiveness of innovations (defined according to novelty), which found that of 2,128 published cost-effectiveness ratios, 1,533 (72 percent) described innovations that increased cost and improved health (Reference Nelson, Cohen, Greenberg and Kent42). Only nine comparisons (0.4 percent) described innovations that were decrementally cost-effective, that is, that saved more than $100,000 per QALY lost (Reference Nelson, Cohen, Greenberg and Kent42). Of interest, these nine interventions were not new drugs, but comprised changes in devices, procedures, physiotherapy, and behavioral interventions, as well as, in one case, watchful waiting in place of surgical intervention (Reference Nelson, Cohen, Greenberg and Kent42).
The definitions described most frequently in the literature were those used by two HTA bodies, AIFA and NICE (Reference Andria, Auriemma and Attanasio12–Reference Rawlins, Barnett and Stevens20;46;47), and in general the definitions used during HTA processes appear to follow a similar trend. The absence of costs from their analysis of innovation is somewhat surprising. We hypothesize that this reflects a two-step process, with price negotiations taking place after an assessment of clinical value (in some countries innovation, together with other factors, is incorporated through acceptance of a higher cost per QALY threshold), but it may contribute to the tendency for new innovations to be associated with increasing costs.
Several studies addressed the type of innovation displayed by a new product, in terms of both technical differences and therapeutic potential (Reference Heible21–Reference Sternitzke23). For example, an entirely new molecular entity with a high degree of additional clinical benefit might constitute a radical innovation, whereas a small chemical change providing only minor additional benefits would be described as an incremental innovation (Reference Heible21–Reference Sternitzke23).
Surprisingly, although the availability of existing treatments was included in eleven definitions of innovation (Reference Aronson, Ferner and Hughes7;Reference Mol, Arnardottir and Motola13;Reference Motola, De Ponti and Poluzzi15;Reference Bryan, Lee and Mitton16;Reference Caprino and Russo27;Reference Joppi, Bertele and Garattini31;Reference Lexchin33;Reference Soleimani and Zenios40;Reference Gonçalves, Maraninchi and Marino44), there was no specific mention of drugs for orphan diseases in any of the definitions identified. It is unclear whether the innovativeness of orphan drugs is fully captured by the definitions currently in use; in particular, these may not take into account the possibility that the funding of orphan drugs may lead to the future development of treatments for broader patient populations.
From an economic perspective, all current definitions of innovative medicines are incomplete. Inclusion of drug costs may allow the construction of a partial economic definition, but the nature of approval and reimbursement processes means that innovation in medicine is inherently different from innovation in other industries, which relies substantially on diffusion (Reference Berwick54).
In industries with functioning markets, a product's characteristics and price may lead it to become a drastic innovation (making current products obsolete) or a nondrastic innovation (improving on current products, or reducing costs); the definition of a product as innovative is confirmed after launch, and is typically dependent on uptake. The reward for developing an innovative product is, therefore, its uptake by a large number of consumers. By contrast, of the studies identified in the literature, only a single review article referred to innovation being determined by the success (societal or commercial) of a new drug (Reference Cadranel, Créquit and Vieira43). The innovativeness of a new medicine is typically decided by HTA bodies or regulators alongside assessment of therapeutic benefit, that is, before the medicine is in widespread use, and without necessarily addressing the economic implications. There is often no formal attempt to quantify therapeutic benefit, and despite some use of multi-criteria decision analysis methods, decisions are often dependent on rough comparisons between therapeutic areas and/or patient characteristics.
Utility/QALY-based approaches are used in some countries, but have been shown to fail in certain areas, leading to the implementation of various adjustments (end-of-life criteria, the UK Cancer Drugs Fund, specific conditions for ultra-orphan diseases, etc.) (Reference Schlander, Garattini and Kolominsky-Rabas55). When the innovative nature of a product leads to acceptance of a high price, affordability issues may lead to restrictions on use; consequently, some new drugs (e.g., Sovaldi® for hepatitis C) may meet an HTA definition of innovation, and may be a commercial success, but not be considered to be a drastic innovation in an economic sense due to limited uptake. In other words, because there is not a perfect market for pharmaceuticals (market failure), the price is not determined by the equilibrium between the willingness to pay consumers and the profit expectations of manufacturers, leading to underutilization of a drastic innovation.
Another potential consequence of the way the innovativeness of new medicines is assessed is that it is possible for a new product to be classified as not innovative, because the active ingredient has been used in other indications for many years; nevertheless, the demonstration of efficacy in a new indication may be a major advance scientifically, and may provide substantial therapeutic benefits for patients. This situation, and the resulting challenges to drug prices (as, for example, occurred following the authorization of dimethyl fumarate for relapsing–remitting multiple sclerosis) (Reference Toumi and Jadot56) may both delay patients’ access to effective therapies and reduce the incentive for manufacturers to investigate potential drug repurposing opportunities. The assumption in some definitions that a product must be novel to be innovative may, therefore, lead to a failure to fully recognize clinical benefits. Notably, the German approach of considering only the additional benefit provided by a new product avoids this situation.
Conversely, there is no consistent relationship between research and development activities, which by their nature involve a form of innovation, and whether the new medicines that result are considered innovative. Although novelty is included in several of the definitions of innovation identified in the systematic search, it is not typically rewarded per se by HTA bodies, while payers and clinicians are likely to prioritize improvements in patient outcomes over novelty. Consequently, there is a distinction between the concept of innovation during research and that of a new medicine being innovative from an HTA perspective.
This systematic review has some limitations. First, the EconLit search was restricted to articles published in English, and some relevant economic literature published in other languages may have been missed. No language restriction was applied to the MEDLINE or Embase searches, and several studies published in French or Italian were reviewed; it is, therefore, likely that the relevant health economic literature is captured in the review. However, the economic dimension in pharmaceutical innovation may be under-represented in the literature, which focuses mainly on aspects related to research and development costs, and represents the perspective of manufacturers rather than HTA bodies. Second, some studies published before 2010 which were identified through citation searching were included, but the systematic search was limited to references from 2010 onward. It is likely that the older definitions included represent those which are still considered to be useful (and cited); older definitions not included in the review are, therefore, likely to be of lesser interest. Third, inclusion of references identified through citation searching in the analysis of the dimensions of innovation mean that this analysis should not be considered to be fully quantitative. However, this potential bias toward more popular definitions means that the results may be considered a reasonable indication of the weight currently given to particular aspects of innovation.
This review has focused on healthcare innovation with specific reference to new medicinal products. Future research is needed to investigate the factors that influence uptake of innovative medicines, and particularly whether the usage of some products is lower than expected due to greater consideration by payers of cost and affordability issues than innovativeness and cost effectiveness. The limitations of a focus by decision makers on drug acquisition costs are likely to become increasingly clear as the treatment of many diseases becomes personalized, with outcomes dependent on processes incorporating the use of diagnostic tests and the selection of the most suitable treatment for each patient, rather than on the efficacy of drugs alone; encouragement and rewarding of innovation in these areas is likely to require an approach different from that currently in use.
The innovativeness of devices, surgical interventions, and other procedures may be assessed differently (Reference Nelson, Cohen, Greenberg and Kent42), as may innovation in the development of drugs for orphan diseases, which was not described in the literature identified by our systematic search. Additional work will be needed to investigate innovation in other areas of health care, and to compare this with the way innovation is considered in the wider economic literature and in other industries. For example, diffusion of innovations may be more prevalent with regard to the use of particular procedures and devices than appears to be the case for new medicines.
RECOMMENDATIONS
In conclusion, overall, this systematic review of the published literature has found that the most commonly referred to aspect of innovation with respect to new medicines is therapeutic benefit; other key elements are unmet need, safety, and the availability of existing treatments. Novelty (of structure or mechanism of action) was also a common component of the definitions identified. However, we do not believe that novelty alone, in the absence of added therapeutic benefit, should constitute a rewardable innovation, as this may prevent funds being better spent elsewhere. For products that improve patient outcomes, the extent to which novelty should increase the acceptable price is not straightforward: by encouraging research and development, rewarding novelty may lead to long-term gains in terms of new medicines, at the expense of short-term increases in expenditure. The weight given to novelty may, therefore, be a political judgement based on societal values, and require a perspective broader than that of the individual product and condition under consideration. It is clear that, other than therapeutic benefit, there is little agreement on what characteristics of new medicines constitute rewardable innovation.
Finally, alignment across countries and among regulators, HTA bodies, and payers would help manufacturers define research policies that can drive innovation and lead to new methods of treating diseases, but may be challenging, as judgements about what aspects of innovation should be rewarded vary among stakeholders, and depend on political and societal factors.
SUPPLEMENTARY MATERIAL
Supplementary Table 1: https://doi.org/10.1017/S0266462318000259
Supplementary Table 2: https://doi.org/10.1017/S0266462318000259
Supplementary Table 3: https://doi.org/10.1017/S0266462318000259
Supplementary Table 4: https://doi.org/10.1017/S0266462318000259
Supplementary Table 5: https://doi.org/10.1017/S0266462318000259
Supplementary Table 6: https://doi.org/10.1017/S0266462318000259
Supplementary Table 7: https://doi.org/10.1017/S0266462318000259
Supplementary Table 8: https://doi.org/10.1017/S0266462318000259
CONFLICTS OF INTEREST
At the request of the AGORA Think Tank, support for the systematic review was provided by Paul Overton (Beacon Medical Communications Ltd) and Natalie Shalet (NAS Healthcare Solutions Ltd), whose organizations received project funding from Novartis Pharma AG. D.C., M.F., P.O., and N.S. report personal fees from Novartis Pharma AG related to the content of this study; P.O. and N.S. report project funding from Novartis Pharma AG outside the submitted work.
