Introduction

Infectious microbial organisms are becoming increasingly resistant to the existing arsenal of antibiotic drugs. Antibiotics are indispensable in treating serious infections like tuberculosis, meningitis and pneumonia, preventing surgical site infections and managing immunocompromised individuals.1, 2 It is estimated that antimicrobial resistance (AMR) is directly responsible for 23 000 deaths annually in the United States and more than 25 000 in the European Union.3, 4 A conservative estimate of the economic cost of bacterial resistance is $55 billion dollars annually in the United States alone.5 Despite the necessity for new antibiotics, the development pipeline is constrained, especially for those that tackle lethal multidrug-resistant Gram-negative bacteria.6 Pharmaceutical and biotechnology firms are averse to investing in new classes of antibiotics because the market is risky and relatively unprofitable. There are promising new methods of antibiotic discovery and some novel antibiotics in development that may not translate into marketable products if firms do not perceive profit potential.7, 8

The antibiotics market has a number of characteristics that makes it financially unattractive to developers. First, antibiotics are less profitable than other drug categories because national conservation programs limit sales, antimicrobials become progressively ineffective due to AMR, there is an established generics market with many substitutes, reimbursement systems encourage the use of the cheapest drug and antibiotics are often prescribed for a brief duration.9, 10 Second, the regulatory requirements for market approval in the United States and European Union have been uncertain and prone to change, creating additional development risk.11 Third, many pharmaceutical companies have reallocated scientific talent and capacity to more profitable opportunities, thereby diminishing what antibiotic expertise and economies of scale they originally possessed.6, 12 Therefore, firms need to be further incentivized to invest in the discovery and development processes necessary to create marketable novel antibacterial drugs.

Investment in antibiotics can be incentivized through two broad strategies known as push and pull mechanisms.12, 13 Push mechanisms reduce a firm’s cost of researching and developing new drugs by distributing the expenditures across multiple parties. Examples of push incentives include increasing access to research, providing research grants, offering tax incentives and establishing public–private partnerships for sharing research and development (R&D) outlays. In contrast, pull mechanisms reward successful development of a drug by increasing or ensuring future revenue. This may be in the form of outcome-based rewards such as monetary prizes, advanced market commitments and patent buyouts, or as lego-regulatory policies that accelerate the market approval process, extend market exclusivity rights and increase reimbursement prices. In addition, a combination of complimentary push and pull incentives can be used in a hybrid approach. Proposed hybrid approaches include the Antibiotic Conservation Effectiveness Program and the Options Market for Antibiotics.14, 15

The European Observatory on Health Systems and Policies conducted the last major review and assessment of these push and pull incentives in 2010.12 Since then, numerous initiatives, programs and collaborations have been implemented with the goal of developing innovative business models for antibiotics. The World Health Organization (WHO) is developing its Global Action Plan under stewardship of the Scientific and Technical Advisory Group on Antimicrobial Resistance.16 The European Commission’s Innovative Medicines Initiative (IMI) is Europe’s largest public–private incentive program that supports the rapid development of safe and effective medicines for patients.17 The IMI has established the New Drugs for Bad Bugs (ND4BB) initiative that aims to remove the barriers associated with antibiotic drug discovery and development through collaboration.18 Notably, DRIVE-AB is a subsidiary program within the ND4BB initiative specifically focused on developing new economic models for antibiotic development.19 The European Commission and European Investment Bank are also launching a new risk sharing and direct loans program to finance development of new antibiotics (L Matthiessen, 2015, personal communication). In the United Kingdom, Prime Minister David Cameron commissioned the O’Neill Review on Antimicrobial Resistance and in the United States President Barack Obama released the National Action Plan for Combating Antibiotic-Resistant Bacteria.20, 21 Think tanks such as Chatham House, Brookings Institute, Pew Charitable Trusts and Wellcome Trust are involved in the global effort through active working groups and conferences that bring together policy and business leaders from around the world.22

The purpose of this article is to systematically review published incentive strategies to promote antibiotic research and critically assess the advantages and disadvantages of each. A framework is also proposed to assist policy makers in selecting appropriate incentives. This approach focuses on correcting the key market failures that perpetuate minimal investment in the field, while addressing antibiotic stewardship and patient access concerns, and accounting for implementation constraints.

Research methodology

A systematic review of the literature was performed to identify specific policies, mechanisms, incentives and business models for stimulating R&D in antibiotics using guidelines from the Centre for Reviews and Dissemination.23 From this literature search, strategies were identified and classified using the push–pull framework and then their advantages and disadvantages were evaluated. Literature was initially sourced from peer-reviewed journals, augmented with gray literature, and then validated through expert opinion. Gray literature is literature not formally published such as conference proceedings, reports, legal documents and press releases.

Identification, screening and eligibility assessment of peer-reviewed and gray literature

The search protocol for peer-reviewed journals (Figure 1) was operationalized through MEDLINE via PubMed, Scopus, Econlit, Business Source Complete and CINAHL. Where possible, search results were filtered to include only literature that focused on humans, was published in the past 10 years, was in English and was either a journal article, review, systematic review, conference report or interview.

Figure 1
figure 1

Generalized search protocol.

Following compilation of initial search results, the literature was first screened using ineligibility criteria applied to titles and abstracts. Articles were deemed ineligible if they focused on clinical settings, scientific research, prescribing practices, antibiotic stewardship and any criteria that was refined in the initial search but was not applied to all databases. The second screening involved reading each article and assessing eligibility. Literature was deemed eligible and relevant to this review if it discussed one or more antibiotic R&D incentive methods.

Gray literature was screened based on eligibility as it was identified and added directly to the compilation of relevant literature. The literature search began by identifying several key review articles and searching their references for articles not already identified. Gray literature was further identified through a Google search for articles, PowerPoint presentations, advocacy statements and conference listings. Websites of key advocacy groups, think tanks and policy committees were identified and further searched for sponsored literature.

Analysis

A total of 46 unique incentive strategies were identified from the literature search. The strategies encompassed single incentives and policies as well as multifaceted business models combining multiple incentives, policies and conservation mechanisms.

Expert opinion

This set of 46 strategies was presented to experts in the field including academics, advocates, industry professionals and policy makers. A total of 26 experts were initially approached, 9 experts provided feedback and 1 new strategy was added to the consolidated list. Therefore, a total of 47 strategies were reviewed.

Postreview critical analysis

Following compilation of all the incentive strategies, the incentives were then critically analyzed using criteria identified from the literature as important to creating an effective incentive package. This critical analysis forms the basis for a framework for selecting an optimal incentive package.

Identification and evaluation of incentive strategies

Push strategies

Push mechanisms (Figure 2) seek to make drug development more attractive to firms by lowering their costs of generating a new drug. These incentives are useful because they reduce the barriers to entry that preclude participation by small- and medium-sized enterprises (SMEs).24 These smaller firms develop a majority of new drugs, yet frequently lack the capital to translate early preclinical research into clinical development.11, 24, 25, 26, 27 Anti-infectives, including antibiotics, have higher success rates than other drug categories in the final phases of development.28 Therefore, early push funding can help companies reach the R&D stages that are likely to produce marketable products. In addition, an early-stage R&D payment is more valuable than an equal pull incentive paid at a later date because of the time value of money. Spellberg et al.29 found that an early subsidy could be as much as 95% smaller than an equally effective future reward. Finally, these policy subsidies may be linked to discrete R&D stages or drug characteristics to ensure alignment of developer goals with public priorities.13

Figure 2
figure 2

Push strategies. IMI, Innovative Medicines Initiative; ND4BB, New Drugs for Bad Bugs; R&D, research and development.

There is a possibility that push incentives will fund projects that fail.13 In addition, developers have asymmetrically more information than funders regarding a particular project’s development. Thus, there is an opportunity for developers to misrepresent their probability of success and project goals in order to attain financing.12 Finally, research subsidies may damage operational efficiency by reducing financial pressure to economize or funder guidance may overly constrain the innovative capability of a developer.12, 13 The advantages and disadvantages of each push strategy, along with all other strategies, can be seen in Table 1.

Table 1 Evaluation of incentive strategies: advantages and disadvantages

Pull strategies

Outcome-based pull strategies

Outcome-based pull incentives (Figure 3) raise project valuation by increasing future revenue through promised monetary rewards. In contrast to push mechanisms, outcome-based pull incentives only compensate successful development. Given that all R&D risk is borne by the developer, firms are motivated to maximize efficiency and adhere to efficacy requirements set by the funder.13, 30

Figure 3
figure 3

Outcome-based pull strategies.

However, financial risk and uncertainty are substantial deterrents for many potential market participants. This is particularly relevant to SMEs that may lack the resources to move from early-stage research to late-stage clinical trials. It is also difficult to determine an appropriate magnitude for the prize that must sufficiently motivate developers while remaining cost effective. In addition, it is a challenge to define the optimal set of drug characteristics linked to the reward so that they are neither perversely specific nor too general.12 Finally, an effective outcome-based pull system relies on a government that is willing to stand by long-term guarantees.

Lego-regulatory pull strategies

Lego-regulatory pull incentives (Figure 4) are government policies that indirectly facilitate higher market returns for firms that launch a new antibiotic. Similar to outcome-based mechanisms, lego-regulatory strategies reward only successful research and thereby maximize R&D efficiency and motivation. In addition, by basing the incentive on market factors such as price and market exclusivity, lego-regulatory mechanisms circumvent the issue of determining an appropriate reward size.12

Figure 4
figure 4

Lego-regulatory pull strategies. AMR, antimicrobial resistance; IP, intellectual property; US, United States.

However, like outcome-based mechanisms, the financial risk of R&D is borne by the developer, thus excluding those firms that do not have substantial capital to market a promising antibiotic. Furthermore, many lego-regulatory mechanisms involve market exclusivity extensions that may dampen competition and innovation. When patents are extended, generic manufacturers are prevented from entering the market earlier and originators are less inclined to develop successive antibiotics that could cannibalize their exclusive market.12

Hybrid strategies

Each push, outcome-based pull and lego-regulatory pull mechanism has distinct advantages and disadvantages, but none provide a comprehensive solution to address the market failures outlined above. There is an increasing consensus that a single approach is not an adequate solution.12, 31 Therefore, a combination of the above incentives or a hybrid strategy (Figure 5) that balances the varying attributes of the mechanisms may be needed.

Figure 5
figure 5

Hybrid strategies. AMC, Advanced Market Commitment; EU, European Union; LPAD, Limited Population Antibacterial Drug; NTAP, new technology add-on payment; PDP, product-development partnership; P4P, pay-for-performance; US, United States.

Mechanisms to fund incentives

Some proposed strategies focus on how to fund the incentives discussed above (Figure 6). These mechanisms are not incentives themselves, but could be used to augment an incentive package and relieve some of the financial burden inherent to incentivizing R&D of antibiotics.

Figure 6
figure 6

Mechanisms to fund incentives. R&D, research and development.

Framework for selection of incentive strategies

Market criteria to create an attractive and supportive environment for investment

A plethora of potential incentive strategies exist, each with their own merits, drawbacks and level of complexity. Therefore, a framework would be useful to assist policy makers in selecting a comprehensive and effective incentive package. Given the serious market failures outlined earlier, the key goal of any antibiotics incentive package is the creation of an attractive and supportive environment for investment. To achieve this, the following market criteria must be met:

  1. 1

    Improve the overall net present value (NPV) for new antibiotic projects.

  2. 2

    Enable greater participation of SMEs.

  3. 3

    Encourage participation by large pharmaceutical companies.

  4. 4

    Facilitate cooperation and synergy across the antibiotic market.

Improve the overall NPV for new antibiotic projects

Net present value is the sum of all costs and revenues of a given project adjusted for the time value of money and risk of failure. It is a general measure of the profitability of a project. Sharma and Towse32 estimated the current risk-adjusted NPV for developing an antibiotic to be −$50 million. In contrast, the risk-adjusted NPV for a neurological drug is +$720 million, and for a musculoskeletal drug this figure soars to +$1.15 billion. As long as the NPV for antibiotic projects remains negative or relatively low, any company looking to maximize profits will not spend significant resources on this class of drugs. Financial incentives that increase revenues, decrease costs or lower the risk of R&D make investment more appealing to all market players. Sharma and Towse32 suggest that a reasonable target for NPV should be $200 million to make investment in antibiotics competitive with other therapeutic classes.

Enable greater participation of SMEs

Small biotech corporations and spinoffs from university research labs hold promising, novel ideas and actually make up a majority of pharmaceutical R&D market share. Munos24 found that between the early 1980s to early 2000s, the proportion of all new drugs attributable to SMEs had increased from 23% to 70%. However, SMEs have much smaller capital reserves than large pharmaceutical companies, hindering the transition from initial research to expensive late-stage trials required for market approval. This is particularly detrimental to antibiotic development because the probability of clinical trial success for an antibiotic is positively weighted toward the later phases.33 Thus, incentives that provide milestone payments, early seed money or reduce their cost of initial R&D allow smaller firms to move a promising antibiotic into the more favorable late-stage clinical trials.

Encourage participation by large pharmaceutical companies

Large market capitalization (large cap) pharmaceutical companies do not have the same capital restrictions faced by most SMEs. If a project is determined to be significantly profitable, then large pharmaceutical firms can often secure the needed funding. However, they are more concerned with the antibiotic market’s uncertainty with regard to size, risk, volatility and regulation. Large cap companies need annual revenues of $800 million for a drug to remain profitable.34 In contrast, SMEs often only need to generate revenues of $100 to $200 million per year.34 For this reason, large companies seek greater revenue certainty and regulatory transparency. These come from credible market commitments as well as large financial rewards for successful antibiotic development.

Facilitate cooperation and synergy across the antibiotic market

There is an opportunity to encourage cooperation and synergy among key industry, academic and government players in the antibiotic market. This involves sharing information, resources and expertise among stakeholders to create additional value in the market. The ideal incentive would reward collaboration by encouraging firms to cooperate to meet public health goals, share important human resources, streamline the supply chain and improve regulatory transparency. Not only do these incentives indirectly reduce the cost of antibiotic R&D, but they also help to align public and private priorities.

A market-based framework for incentive selection

The primary goal of an incentive package is to create an attractive and supportive market for investment in antibiotics. As discussed above, this is accomplished by improving the NPV of antibiotic R&D projects, lowering barriers for SME participation, encouraging large cap companies to invest in the market and facilitating cooperation and synergy among all stakeholders. The following framework (Table 2) is an incentive classification system based on these four market criteria.

Table 2 Market-based framework for selection of incentive strategies

It then follows that an incentive package that aims to create a supportive and attractive market for investment in antibiotics could be created through:

  1. 1

    A single type 1 incentive.

  2. 2

    A combination of type 2 and type 3 incentives.

  3. 3

    A combination of type 3, type 4 and type 5 incentives.

Type 6 incentives could be used, but tend to be weaker market incentives and may be less effective at generating investment and market interest. As seen in Table 3, each incentive has been classified into one of the six types, depending on its ability to meet the market criteria.

Table 3 Classification of incentive strategies using the market-based framework

Factoring in public health objectives: stewardship and access

Beyond creating a viable market for antibiotics, an incentive package should reinforce broader public health objectives pertaining to AMR. There are two key public health objectives that are interrelated with the economic aspects of an antibiotics market:

  1. 1

    Promoting antibiotic stewardship.

  2. 2

    Improving patient access to new antibiotics.

Promoting antibiotic stewardship

Research and development of antibiotics also needs to be sustainable in addition to being profitable. The traditional patent-based business model rewards developers through market exclusivity, providing the opportunity to command high prices. Once a patent expires, the market is flooded with generic drugs that compete based on sales volume in a race against impending resistance.10 This unsustainable business model reinforces the overmarketing and overconsumption of antibiotics that has contributed to high levels of AMR. Simply increasing developer return on investment does not address this problem directly. Numerous experts have proposed antibiotic business models that reinforce conservation efforts by completely severing a developer’s return on investment from sales volume and price (for example, Antibiotic Health Impact Fund (AHIF), Antibiotics as Public Goods and Rewarding Antibiotic Development and Responsible Stewardship Program (RADARS)). This concept is known as delinkage and is beneficial for three key reasons.10 First, it provides developers with a concrete return on investment that is extraneous to the market. Second, delinkage removes the motivation for developers to oversupply their antibiotic. Third, it facilitates access to new antibiotics for those who need them most. Other experts advocate the use of demand-side antibiotic usage fees to internalize the negative externalities accompanying antibiotic use (for example, Antibiotic Innovation and Conservation (AIC) fee and Antibiotic Innovation Funding Mechanism (AIFM)). This fee can then be used to finance other incentive mechanisms such as milestone payments or end prizes.

Improving patient access to new antibiotics

Patient access to new antibiotics plays an important role in controlling the spread of AMR and preserving existing antibiotics. However, under the current patent-based business model, developers are incentivized to distribute their new antibiotics based on ability to pay instead of need.10 This may not be a problem for countries with universal access to pharmaceuticals. However, for those countries without complete public pharmaceutical coverage, drug prices remain a significant hurdle to patient access. This issue can be complicated by conservation-related restrictions on antibiotic use as well as technical challenges with distribution. This may be overcome by transferring or licensing out a new antibiotic’s patent to the government along with the responsibility of distribution and equitable access (for example, patent buyout and payer license). Other proposals streamline the regulatory approval process to allow new antibiotics with significant therapeutic value to reach the market faster (for example, Limited Population Antibacterial Drug (LPAD) approval and special drug status).

Factoring in public health objectives

Selection of incentives using the above market framework must be done with consideration of public health goals. An incentive package that meets the four market criteria may not effectively support these public health goals. For instance, the type 1 incentive, special drug designation, does not align developer promotion and marketing goals with conservation priorities. In this case, an additional incentive or incentives are necessary to augment this package. Aspects of conservation could be encouraged through conditional grants and pay-for-performance (P4P) prizes alongside the special drug designation incentives that stimulate market investment. In some cases, incentives may directly contravene public health objectives. For example, market exclusivity extensions and value-based pricing directly incentivize firms to continue overmarketing antibiotics and distributing based on ability to pay. For this reason, these types of incentives may need to be altered or not included in the package. Market exclusivity extensions could be swapped out for conservation-based market exclusivity extensions and value-based pricing could require continual reassessment to reflect antibiotic effectiveness.

Factoring in implementation feasibility

Not only does any potential incentive package need to be comprehensive, it must also be feasible. Many of the proposals discussed have been developed on a theoretical level, but rarely tested or deployed. Although design of appropriate incentives is challenging, it pales in comparison with the political, regulatory, industry and financial hurdles that may be faced during implementation. A comprehensive strategy that is unwieldy, too complex and financially unreasonable provides no advantage. Therefore, more pragmatic design constraints must be considered. These will ultimately reflect a nation’s political priorities, operational realities and industry demands concerning:

  1. 1

    The size of the incentives.

  2. 2

    The timing of incentive delivery.

  3. 3

    Governance of the incentive package.

  4. 4

    International coordination.

  5. 5

    Intellectual property rights.

There are obvious financial constraints on the size of the incentive, as well as differing philosophies on the role of direct government involvement. A related challenge concerns managing the selected incentive package. A new regulatory agency or governing body may be required to determine public health priorities, define the optimal number and depth of drug specifications linked to incentives, calculate socially fair rewards and monitor development progress. This is especially important as many of the recent proposals operate on a global scale and require coordination, input and agreement across borders (for example, AHIF, AIFM and WHO Global Consortium). This new organization could operate under a new agency or as part of an existing forum such as the G-20.

It is increasingly recognized that delinking sales volume from financial motivation to develop antibiotics is valuable to controlling AMR and facilitating equitable access to novel antibiotics.10 Several delinkage models such as patent buyouts, AHIF and AIFM are based on the concept of transferring intellectual property (IP) to the public domain that poses a major hurdle to implementation. From a public health perspective it makes sense to shift control of new antibiotic IP from the private to the public domain, but this change poses a risk to the industry. Many pharmaceutical companies want to keep patent rights because it provides additional assurance that costs can be recouped if incentives and policies are reneged or are inadequate. If transferring IP to the public domain is not feasible, these types of delinkage models become irrelevant. However, if this were the case, delinkage can still be created through incentives such as payer licenses, guaranteed revenue minimums or advanced market commitments (AMCs). Table 4 summarizes our assessment of each incentive strategy based on the market criteria, the public health objectives and implementation feasibility.

Table 4 Criteria-based assessment of incentive strategies

Example applications of the framework

Given the market failures endemic to antibiotic R&D, an incentive package should first correct market deficiencies. The framework outlined above is useful for this purpose. The package can then be augmented to address public health issues regarding antibiotic conservation and patient access. The following are three examples of the application of the above framework in devising an appropriate incentive strategy.

Scenario 1: A single type 1 incentive

The WHO’s Global Consortium for stimulating antibiotic R&D is a well-rounded, hybrid model with five parts: (1) support at the drug discovery stage through milestone prizes and an open source platform, (2) grants for academics, SMEs and big pharmaceutical firms to lower development barriers and risk, (3) patent buyout prizes for proven novel antibiotics, (4) public funding of clinical trials and (5) advance purchase commitments to preserve antibiotics.35, 36 The WHO model attempts to create a product-development partnership (PDP) across the entire pharmaceutical value chain, or what is referred to as a global consortium. The WHO Global Consortium explicitly addresses each of the six objectives. Early milestone payments enhance project NPV by reducing early costs that can have an even greater impact overall because of the time value of money. SME participation is explicitly encouraged with early-stage grants and an open source platform. Public funding of clinical trials appeals to large and small firms alike by reducing overall project costs and risk.37 Patent buyouts facilitate antibiotic stewardship by negating the need for excessive marketing or production. However, to be attractive, these end prizes would need to be sufficiently large, and calculating this in such a way to minimize waste while providing sufficient incentive may prove difficult. Patient access could be assured by partnering with worldwide generic producers who could keep costs low for patients.10 Still, the consortium itself, along with its financing for public trials and end prizes, would have to be publicly financed, shifting costs and risk to the public sector. Given the massive scope of this model, there are numerous implementation issues that pose serious challenges to overcome. These include attaining adequate public funding for grants, patent buyouts and clinical trials, coordinating a new global entity to manage the consortium and liaising with industry to reach an agreement on IP rights.10

Scenario 2: A combination of a type 2 and type 3 incentives

The Options Market for Antibiotics (OMA) model is a hybrid mechanism that allows government or nongovernmental organization purchasers to invest in a drug in early-stage development. In this model, funders may purchase the right to buy a specified number of antibiotics at a reduced price, if and when the antibiotic ever made it to market.15 In many ways, this could be considered a form of milestone payments, but with a future discount for options holders. The early payments, if large enough, could both improve the overall NPV and enable greater participation of SMEs. Larger pharmaceutical firms may be attracted by the risk-sharing element of the venture, in that funders may pay when antibiotics are in early clinical development. This also indirectly signals a potential commitment to purchase the product upon marketing approval. Lower prices or even marginal cost pricing at marketing approval will help to facilitate patient access. In addition, antibiotic stewardship can be promoted by combining the OMA with an AMC. Bulk purchasing commitments would shift control of sales volume to the sponsor and allow for appropriate distribution of the antibiotic. However, such a scheme would do little to directly facilitate cooperation among corporations, unless it was combined with modifications to anti-trust laws. If enacted in isolation, anti-trust waivers could hinder patient access to medicine by allowing collusion among producers to maintain artificially high prices. The goal of such reforms would be to promote cooperation and synergy across the antibiotic market.12 Although cooperation would be desirable in the early development phases, it would not be desirable in the marketing phase with regards to setting prices. Such reforms could be applied to the OMA model by allowing companies to share early-stage data, potentially increasing the transitional probabilities from one phase to the next in later development.

Scenario 3: A combination of type 3, type 4 and type 5 incentives

The Antibiotic Conservation Effectiveness (ACE) Program is a hybrid strategy that combines outcomes-based and lego-regulatory pull mechanisms with the objective of promoting antibiotic conservation. The Program has four key components: (1) P4P payments centered on public health and conservation goals, (2) conservation-based market exclusivity, (3) value-based reimbursement that ties drug pricing to the effectiveness of the drug and (4) anti-trust waivers that allow coordination of conservation activities between developers.14 Given the pull-centric nature of the ACE Program, this incentive package particularly targets large cap pharmaceutical companies. Therefore, it would be beneficial to augment this package with a SME-focused incentive such as direct funding. Antibiotic research addressing specific health priorities can be targeted through direct funding and can include expert technical and managerial help that may prove useful to SMEs with less experience. The ACE Program does not facilitate patient access nor promote cooperation and synergy between industry and the government. Thus, there is role for a LPAD approval system in this incentive package. Under the LPAD approval system, the safety and efficacy of an antibiotic targeting a newly resistant pathogen would be examined through smaller, faster and less costly clinical trials. LPAD-designated antibiotics would be limited to a narrow indication for which there is a particularly high patient need and therapeutic benefit. With this system the regulatory agency would provide significant guidance to the developer and continue monitoring the effectiveness of the drug beyond approval.38

In transitioning from single incentives to more complex, international business models, implementation becomes significantly more difficult. A feasible, yet comprehensive, incentive strategy likely will include a wide selection of smaller incentives that collectively address market and public health aspects as opposed to a revolutionary antibiotic business model. In our opinion, the ideal package would include several incentives that facilitate cooperation and synergy throughout the market, one or two R&D-linked push incentives and a large pull incentive rewarding successful development.

Conclusion

Antimicrobial resistance is a complex health policy problem. Multiple market failures make it financially unattractive for pharmaceutical and biotechnology companies to invest in antibiotic discovery and development. This problem is complicated by incentives to oversell antibiotics and distribute based on ability to pay instead of need. This systematic literature review has identified 47 incentives that could be used to encourage and accelerate R&D of novel antibiotics. These incentives have been classified using the push–pull framework and their individual advantages and disadvantages have been evaluated. However, given the large number of possible incentive schemes, a decision framework is needed to help select an effective package of incentives. An ideal solution will tackle the market deficiencies that have resulted in the stagnant market, address the public health priorities that reflect the growing need for a sustainable solution to AMR and operate within implementation constraints. Because of the complexity of the problem, we suggest first developing an incentive package that addresses the core market failures. This can be further enhanced to achieve public health objectives such as antibiotic conservation and patient access. The above framework acknowledges that there are multiple viable solutions to stimulating antibiotic discovery and development that will ultimately be determined by government priorities, industry demands and operational realities unique to a particular country.