Impact pathways: improving medicine access through cross-national stockpiling

Abstract

Purpose

Pressure to address drug shortages is increasing, and calls for cross-national stockpiling solutions are on the rise. This paper argues that cross-national stockpiling presents an important opportunity for operations and supply chain management (OSCM) research to help improve medicine access in the short- and long-term.

Design/methodology/approach

To substantiate our claims for future research, we conducted an empirical analysis of shortage co-occurrence on twelve EU member-states, reviewed practice reports and academic literature, and held extensive discussions with pharmaceutical supply chain stakeholders.

Findings

We show that the potential for cross-national stockpiling is unmistakably there and that, although OSCM research has touched upon important facets of cross-national stockpiling, many open questions still require novel research.

Originality/value

We present opportunities for impactful research along various dimensions, including understanding barriers and stakeholder behavior, quantifying cost-effectiveness and potential knock-on effects, analyzing the positive and negative implications of product standardization and determining novel equitable mechanisms for allocation and financing cross-national stockpiles.

Citation

De Vries, H., Lemmens, S., Breugem, T. and Olivier, M. (2024), "Impact pathways: improving medicine access through cross-national stockpiling", International Journal of Operations & Production Management, Vol. 44 No. 13, pp. 364-372. https://doi.org/10.1108/IJOPM-07-2024-0550

1. Introduction

Medicine shortages are a growing problem in Europe. Twenty member states jointly had more than 6,000 shortages in 2019 alone (European Commission, 2021). This causes distress among patients, interruption of treatment, sub-optimal therapies, inferior efficacy, and in some cases even death. For example, in 2024, several countries faced shortages of Salbutanol inhalers – a life-saving product for patients with asthma, especially for children under 6 (European Medicines Agency, 2024). At least twelve European Countries have implemented national medicine stockpiles to mitigate these shortages and several others are considering implementation (Vogler and Fischer, 2020; Ministerie van Volksgezondheid, Welzijn en Sport, 2021). Stockpiles often imply stock-keeping obligations for suppliers (i.e. market authorization holders), wholesalers, and pharmacists. For example, France requires market authorization holders (henceforth called “suppliers”) of 422 medicines to keep four months of stock, German suppliers of generic medicines must keep six months of stock, and Dutch wholesalers and suppliers must keep four and six weeks of stock, respectively (Belgodère et al., 2023; Bundesministerium für Gesundheit, 2023).

Although countries consider stockpiling effective (70%) or very effective (30%) (European Commission, 2021), it also faces four key critiques. First, it creates an “arms race for scarce medicines”, inevitably leading to more shortages in some countries (van Oorschot et al., 2022; European Commission, 2021). Second, mitigating risks at the national level is deemed ineffective in Europe, as it is a single internal market (Torjesen, 2015). Third, stockpiling does not address the economic root causes of the problem (European Commission, 2021; Ministerie van Volksgezondheid, Welzijn en Sport, 2021; FDA, 2019), which may threaten its sustainability. Last, stockpiles at a national level may be very inefficient and expensive: “Manufacturers will have to produce greater volumes of products that may end up having to be destroyed as reserve stock goes unused and expires, and the excess stock will need to be appropriately warehoused” (European Commission, 2021). Pooling stockpiles may be much more efficient and enable countries to tap into them when needed. This could be in the form of (semi-) finished products for medicines, as suggested by the European Commission (European Commission, 2021). These could be kept in a physical stockpile owned by a cross-national entity, similar to the rescEU stockpile of medical emergency items (DG HERA, 2022). Alternatively, they could be kept in a virtual stockpile, which would be controlled by EU suppliers and wholesalers. Transshipments across countries could then be facilitated by a digital platform that visualizes inventories (DG HERA, 2022) and a mechanism through which countries submit requests for medicines to other countries (European Medicines Agency, 2023).

This impact pathways paper argues that Operations and Supply Chain Management (OSCM) could help answer many of the questions surrounding cross-national stockpiling. We first provide evidence for the potential of cross-national stockpiling using real-world shortage data from twelve European countries. We then present three pathways for future research to unlock this potential. These stem from over six years of research on drug shortages and interactions with pharmaceutical supply chain actors (ministries of health, insurers, wholesalers, and suppliers) as part of an international research project [1], as well as stakeholder reports and academic literature. We validated and extended these pathways at a three-round world café workshop with 38 academic and industry experts from various countries [2]. The workshop covered the case for cross-national stockpiles, implementation barriers, design choices, and the potential role of the EU. Answers were first provided in writing and discussed afterward. See the online Appendix for details on the workshop and how the data were analyzed. We note that while the problem of drug shortages directly inspired the pathways, this research could also benefit debates on cross-national stockpiles for other essential commodities, such as humanitarian relief items, cash, oil/gas, food, and water (cf. Rodríguez-Pereira et al., 2021; Algemene Rekenkamer, 2023).

2. Why cross-national stockpiling?

Cross-national stockpiling is particularly attractive when medicine shortages occur in only one or a few countries simultaneously. Stocks can then be flexibly allocated to the few countries that need them. We analyzed co-occurrence using data from national medicine shortage catalogs of EU member states. To be included in our analysis, the shortage catalog should be publicly accessible, define shortages clearly, and reveals the Anatomical Therapeutical Chemical (ATC) code of the product in shortage (e.g. the name, the active pharmaceutical ingredients (APIs), or the ATC code itself). Based on these criteria, we selected the following twelve countries: Austria, Belgium, Denmark, France, Germany, Hungary, Ireland, Italy, Netherlands, Norway, Spain, and Sweden. We merged the different datasets based on the Anatomical Therapeutic Chemical (ATC) code to identify whether a shortage co-occurs in multiple countries. Since most of countries only report active shortages, we selected one specific date to compare shortages across countries: May 15th, 2021. We refer to the online Appendix for details on the data-cleaning process and a discussion on the comparability of the datasets. The datasets from the twelve countries yield 4,029 unique shortages corresponding to 1,033 unique ATC codes. Figure 1 summarizes the amount of overlap.

The key insight from Figure 1 is that the level of overlap is rather limited. Most of the 1,033 ATC codes experiencing a shortage were reported in a single country. When a shortage did impact multiple countries, it usually only affected a few (two, three, or four) countries. Almost all (97.4%) shortages impacted six or fewer countries. Only two (0.002%) shortages affected more than eight countries.

This provides a strong argument for cross-national stockpiling. While the level of overlap differs per type of medicine, the finding of limited overlap, and the potential of cross-national stockpiling, holds across all types of medicines (see the online Appendix).

3. Pathways for future research

This limited overlap provides a strong case for policy makers to explore opportunities for cross-national stockpiling. Our workshop and stakeholder discussions yielded broad consensus on the need for some form of cross-national stockpiling “at least for essential medicines” and “at least for countries that have experience with collaboration”. They also raised many questions regarding the scope, design, governance, and cost-effectiveness of cross-national stockpiles. We argue that many of these questions offer opportunities for impactful OSCM research, and that many fundamentally differ from well-studied similar problems in the literature. This section outlines these questions and opportunities and proposes three future research pathways. The first pathway is primarily practice-oriented, while the second and third identify new OSCM problems arising in the cross-national stockpiling context.

3.1 Pathway 1: enlarging the evidence base on the impact and feasibility of cross-national stockpiling

Three key questions must be answered to move stakeholders towards cross-national stockpiling: (1) What barriers exist, and (how) can they be addressed? (2) How would stakeholders be affected and how would their behavior evolve? (3) Are investments in cross-national stockpiling indeed cost-effective? While potential interventions to address drug shortages have been extensively discussed (Ergun et al., 2023), grey and academic literature that thoroughly examines feasibility and stakeholder behaviors and quantifies cost-effectiveness is nearly absent (de Vries et al., 2021). For example, workshop participants noted that stock pooling and transfers of several medicines is hindered when countries differ in “treatment regimens”, “strengths”, “national needs”, and “legislations”, but evidence on how this affects overall cost-effectiveness is lacking. One notable exception is a report on stockpiling of antibiotics in the EU (DG HERA, 2022). However, this solely focuses on antibiotics, and does not quantify cost-effectiveness nor the impact of various design choices (see also Pathway 2).

The lack of evidence increases the risk that implementation opportunities are overlooked or that suboptimal design choices are made, especially in light of the many potential knock-on effects (Ergun et al., 2023; van Oorschot et al., 2022). We advocate further research that (1) maps the key stakeholders, their behaviors, and their views on the operational, legal, and financial feasibility of various cross-national stockpiling archetypes (see Pathway 2) and (2) develops models that can help establish sound evidence on the cost-effectiveness of cross-national stockpiling. The latter would allow for a comparison between cross-national stockpiles and other interventions to address drug shortages, such as national stockpiles, flexible (local) production capacity, enforcing multi-sourcing, changing procurement practices, and adjusting price regulations (de Vries et al., 2021; European Commission, 2021). While the OSCM community has a long history of analyzing inventory pooling, we argue below that existing models do not incorporate the complex dynamics of cross-national stockpiling or the key knock-on effects on national stockpiles and parallel trade (van Oorschot et al., 2022). We see a significant role for research that develops game theoretic models or detailed simulation or system dynamic models that capture such dynamics and knock-on effects. We also advocate research that uses publicly available datasets to estimate key model parameters such as the “cost” of a drug shortage, holding costs, and the joint distribution of the probability and the duration of shortages in multiple countries (Ergun et al., 2023).

3.2 Pathway 2: designing cross-national stockpiling policies that ensure short- and long-term access

Implementing a cross-national stockpile induces three fundamental design choices:

Stockpiling archetype. The first is whether a medicine should be kept in (1) a physical stockpile operated by a cross-national entity or company, (2) a virtual stockpile – a system that visualizes and monitors stocks across countries and facilitates stock transfers, (3) a combination of these, or (4) not at all. Our workshop and stakeholder discussions suggest that consensus on this choice is lacking, and that the optimal archetype differs per medicine, in line with DG HERA (2022). This is not solely a matter of cost-effectiveness maximization – choosing the option that yields the largest reduction in shortages per monetary unit – but also of equity and market attractiveness. For example, the stockpile archetype can strongly affect how equitably the different countries are served (more on this in Pathway 3). Similarly, it can affect supplier incentives to keep or register a medicine on the market (DG HERA, 2022). The lack of such incentives is an important cause of shortages (De Weerdt et al., 2015). Although some research compares virtual and physical stockpiles (see Liu et al., 2016, and references therein), this stream commonly takes a cost and/or service level perspective only and considers a focal company that controls multiple warehouses. This differs from our setting, where different “warehouses” (i.e. the countries where the stocks are located) have a high level of autonomy. Future research could develop stylized and detailed simulation models that (1) account for the unique objectives and stakeholder landscape, (2) capture key trade-offs, (3) yield generic insights on what drives the performance of various stockpiling archetypes, and (4) derive specific insights for case studies (e.g. stockpiling in the EU).

Cross-national scope and quantities. The second key choice is which medicines to include and what quantity to stock in the cross-national stockpile. This problem bears some resemblance with dual sourcing inventory management, where inventory can be replenished from a fast but expensive source (cf. national stockpiles) and from a slower but cheaper source (cf. cross-national stockpiles). However, the corresponding literature considers optimization of sourcing decisions given that a second source exist for a given product instead of the more strategic decisions of whether it should exist and what quantity this source should stock (see Svoboda et al., 2021, for a literature review). These decisions are especially complex for virtual stockpiles. In this case, the cross-national entity does not directly set stock levels but can influence stock quantities through policies that specify minimum stock obligations for suppliers and/or wholesalers. This already occurs in many European countries.

Keeping such policies simple is critical, as this limits misinterpretations, reduces the complexity of implementation, and increases the feasibility of monitoring. Minimum stock levels are, therefore, commonly expressed in terms of the number of weeks or months of demand, with no or very little differentiation of this number across medicines (Ministerie van Volksgezondheid, Welzijn en Sport, 2021). Balancing the policy’s (cost)effectiveness and complexity presents a novel optimization problem that differs fundamentally from the optimization paradigm found in the inventory management literature (which optimizes inventory control parameters per product) (Ergun et al., 2023).

Another difference with commonly studied inventory control problems is that minimum stock policies can fundamentally alter the supplier landscape and affect availability in the long term. Cross-national stock requirements may reduce national stock obligations, making the corresponding markets more attractive and countering a trend whereby suppliers leave markets because of small margins, more stringent stock requirements, and failure-to-supply penalties (de Vries et al., 2021). However, if the cross-national stockpile makes stock transfers (i.e. parallel trade) easier, it can also make certain markets less attractive. Specifically, stock transfers from low-price to high-price countries (Pauwels et al., 2014) further erode suppliers’ margins and incentivize them to limit supplies to or even withdraw from low-price countries (Bart, 2008).

We, therefore, see a significant need for future research on policies for minimum stock obligations that balance cost-effectiveness and complexity, while incorporating key knock-on effects.

Product form and packaging. The form of the product being stockpiled presents a third key design choice. Three forms can be distinguished (cf. DG HERA, 2022): (1) APIs (Active Pharmaceutical Ingredients), (2) white label stock (i.e. products without labels and leaflets), (3) finished products. This postponement choice brings about trade-offs between flexibility, speed, and feasibility. For example, API stockpiling provides much flexibility in selecting the specific medicine to produce, but increases the time required to produce finished products and is difficult to monitor. A rich body of literature and models on postponement could be adapted to the case of national stockpiling to help support this choice (see Zinn, 2019, for a review). Such models could be parametrized by collecting data on flexibility, speed, and shortages for each product form.

A strongly related choice entails the harmonization of packaging by introducing multilingual (electronic) package leaflets (ePLs) (Skogman-Lindqvist et al., 2023). Harmonization can counter shortages by increasing market attractiveness and speeding up stock transfers between countries (DG HERA, 2022; Skogman-Lindqvist et al., 2023; Norwegian Medicines Agency, 2023). However, it can also increase parallel trade and induce corresponding knock-on effects (see above). Future modeling and implementation research are needed to understand the net impact of ePLs on shortages in light of such effects.

3.3 Pathway 3: designing equitable allocation and financing mechanisms that maximize social welfare and prevent opportunistic behavior

Independent of the design choices outlined in Pathways 1 and 2, two fundamental questions underlie any cross-national stockpile: (1) How should scarce medicines be allocated? or “Who does the sharing?”, (2) How should the financial costs of the stockpile be allocated? Workshop participants identified fair financing and allocation mechanisms as key for (successful) implementation. Accordingly, we see promising future research opportunities along three dimensions, with each dimension being a predecessor of the next:

Equity metrics. Medicine shortages have substantial negative societal impacts, making equity considerations highly relevant. For example, how should insufficient medicines be allocated equitably between two countries with different needs? Should “rich” countries contribute proportionally more to the stockpile than others? Such questions bear a resemblance to the allocation and financing mechanisms studied in the disaster preparedness literature (see, e.g. Rodríguez-Pereira et al., 2021). One key difference is that, in cross-national stockpiling, countries are likely to have sufficient means and therefore realistic outside options, such as maintaining an individual stockpile. Allocation mechanisms from collaborative game theory can accommodate outside options (Cruijssen et al., 2007; Guajardo and Rönnqvist, 2016). However, these mechanisms are often computationally intractable, and additional factors (such as economic means) are not easily incorporated (Rodríguez-Pereira et al., 2021). We therefore see an important opportunity in developing novel equity metrics that are practical (i.e. intuitive and easy to use) and account for the features relevant to cross-national stockpiling.

Incentive alignment. Countries should have incentives to report their needs truthfully, and resulting allocations should be equitable and in line with countries’ contributions. For example, how can we avoid that countries which purposefully pay little for medicines and are therefore more exposed to shortages, claim an unfairly large part of the stockpile? To design incentives, we see a need for research on mechanism design for cross-national stockpiling. Such mechanisms have been applied successfully in related contexts such as allocating surplus medicines (Zhang et al., 2020)) and humanitarian funding (Pedraza-Martinez et al., 2020). Besides the different operational context, we see a key challenge in accounting for the endogeneity of demand, i.e. the chosen mechanism will directly influence demand and hence the optimal size of the stockpile, as opposed to the commonly studied case where the amount of demand (or resources) is considered fixed.

Strategic trade-offs. To convince policy makers of the benefits of cross-national stockpiling, it is imperative to quantify the benefits. As discussed in Pathway 1, this includes the benefit of the increased flexibility it provides in mitigating shortages. Equally important is quantifying the tradeoff between cost (e.g. investment, coordination, and operational cost) and equity considerations. For example, are equitable allocation mechanisms substantially more costly than non-equitable ones? And can some allocation mechanisms improve equity at a low cost? We see opportunities for novel analytical methods extending existing work on the price of fairness (e.g. Bertsimas et al., 2011; Breugem and Van Wassenhove, 2022) and system dynamics and simulation-based methods using real-world data.

Future research would also benefit from empirical work on equity perceptions and preferences, a topic that has received limited attention in the OM literature (Breugem et al., 2023). To the best of our knowledge, it is an open question how equity preferences might differ across countries, especially when considering larger, and hence more diverse, collaboration initiatives.

4. Conclusion

Pressure to address drug shortages increases, and calls for cross-national stockpiling solutions are on the rise. In this impact pathways paper, we showed the potential for cross-national stockpiling is unmistakably there, and outlined impactful opportunities for OSCM research to help improve medicine access, including but not limited to understanding barriers and stakeholder behavior, quantifying cost-effectiveness and potential knock-on effects, analyzing the positive and negative implications of product standardization, and developing novel equitable mechanisms for allocation and financing cross-national stockpiles.

Figures

Figure 1

Shortage co-occurrence across EU countries

Notes

1. This involved dozens of meetings and dissemination events involving stakeholders from Belgium, Denmark, France, Germany, the Netherlands, Norway, Sweden, the UK, and the World Health Organization.

2. This took place at a research dissemination event of the MIA Project (https://www.bi.edu/research/centres-groups-and-other-initiatives/mia/the-mia-project/), held at BI Oslo on May 15th 2024.

Supplementary material

The supplementary material for this article can be found online.