4

Potential Opportunities for Restructuring Strategic National Stockpile Scope, Governance, and Decision Making

POTENTIAL OPPORTUNITY TO REEXAMINE THE MISSION AND SCOPE OF SNS

Tom Inglesby, chief executive officer and director, University of Pittsburgh Medical Center (UPMC) Center for Health Security, maintained that the Strategic National Stockpile (SNS) is improving, commending its financial accountability mechanisms, inventory control, security systems, low administrative costs, business practices, number of products developed through the Biomedical Advanced Research and Development Authority (BARDA), its decision and review processes, and its regulatory interactions. Key challenges to be addressed going forward, according to Inglesby, concern the SNS mission and scope. Even if it is limited to chemical, biological, radiological, and nuclear (CBRN) threats alone, it is still almost beyond sustainability in terms of current cost environment and the development of the stockpile to meet requirements; this is compounded with all hazards, emerging infectious diseases (EIDs), and drug shortages: “Ultimately, even though we talk about that scope and mission like it is a science or a programmatic decision, I think it is absolutely a political decision and in fact a national security governance decision.”

Inglesby advised that the SNS should focus on events that the country cannot deal with without the SNS: CBRN; influenza; EIDs; and events that have the potential to lead to massive loss of life, crisis, and instability. He commented that whatever decision is made should be political, with assets and structure that are aligned. Carlin commented that the SNS mission and scope must be limited largely because of funding. She also advised that its

remit should be CBRN, flu, and EIDs. However, because EIDs are more common, she suggested that perhaps they should be moved to another non-stockpile mechanism altogether because sustainability makes SNS capacity for EID stockpiling likely to be infeasible. Carlin cautioned about a near-term funding issue as products transition from Project BioShield to the SNS, noting that Congress has been informed but will likely still be surprised with the cost of the SNS moving forward. However, because the standing committee mandate came from Congress, it provides a platform to speak to Congress directly and make realistic budget requests.

As to whether the SNS should have a role in responding to day-to-day (non-CBRN) events and routine supply-chain shortages, Burel explained that although the original intent was for the SNS to be able to address CBRN threats, that scope has continued to expand to cover all types of hazardous events, natural disasters, pandemic influenza, and EIDs. Although day-to-day drug shortages do not currently fall under the auspices of the SNS, in the role as a specialty distributor of pharmaceuticals, the SNS can respond to a need on a one-off basis for life-saving purposes within the health care system if the SNS holds a product that is available nowhere else (see Box 4-1).

Sosin commented that a national resource such as the SNS, which has tremendous potential in terms of skilled personnel, materiel, and relationships in the medical and pharmaceutical distribution network, should be available in crisis situations: “Those who appropriate funds and expect stewardship and good response, and those who are responsible for those responses, appreciate the ability to utilize those resources. I think there is a natural expectation that an insurance product like a stockpile will not just sit on the shelf and be forgotten.”

Jacobson raised the issue of who should be responsible for the management of the SNS in a crisis situation to maximize a response (see Box 4-2) that requires coordinating many different components of items, people, places, and logistics. Jason Frederick, vice president of operations, FedEx Custom Critical, and Mitch Mattingly, president, Metro Logistics, Inc., both replied that they receive instructions directly from the Centers for Disease Control and Prevention (CDC).¹ Burel noted that once the decision is made to deploy, SNS does everything necessary to deploy it all the way to the state level with support from their third-party logistics providers. Then the states and localities take responsibility for dispensing to the public. Performance is monitored by the DSNS, which intervenes if things go wrong.

___________________

¹ Frederick commented that CDC is their number one priority customer when there is an event, but that they also hear from other government agencies trying to circumvent the system to lock down capacity.

Margaret Brandeau, Coleman F. Fung Professor of Engineering, Stanford University, commented that the SNS is part of a much broader system, within which it has much responsibility but not much authority (PHEMCE places items in the stockpile with the expectation that the stockpile must respond quickly to all hazard-type crises within 48 hours, etc.). She suggested an actionable response: quantifying the mission expansion and connecting it to funding. That is, as part of the political process of determining the role of and the funding for the SNS, push back by laying out the core mission and explaining that the expansion of the core mission must lead to increased funding. Brandeau observed there are financial trade-offs that should be considered in the context of widening of the mission of the SNS, saying “If we do not have enough money to replace these expiring things then it will be at the cost of not covering this. If we do not prepare for this

particular hazardous scenario that we now have found, it will cost or save this much money.”

Skivington observed that the increased demands on the SNS as it has evolved have changed the value proposition of the stockpile. He suggested that there may be substantial value in a review of the SNS’s governance, given its administrative complexity, to analyze where components of the process of maintaining and operating a stockpile are working well, where things can be improved, and how the benefits provided by having an SNS can be optimized.

O’Toole advocated for keeping SNS in the national security space, because the latter has different salience, priorities, and budget lines that garner the attention of Congress and the Executive Branch in a way that public health can never do: “We are in a moment of great jeopardy. I do not remember a time when America has had as many potential adversaries, let alone an adversary as brutal as ISIL [Islamic State of Iraq and the Levant] seems to be. They have avowed that they are interested in bioweapons.” Thus, she argued that it is important for the standing committee to proceed carefully in making recommendations, in order to help the Department of Health and Human Services (HHS) restructure the SNS without seeming to have failed.

Umair Shah, executive director, Harris County Public Health & Environmental Services, Harris County, Texas, pointed to turnover of expertise and experience as a critical issue, for people involved with the SNS at the

state and local levels in particular. He suggested finding systemic strategies to allow new personnel to be smoothly incorporated and educated about institutional knowledge using strategies proven to have worked elsewhere.

Burhans suggested an integrated, unified budgeting process among all agencies participating in PHEMCE, the SNS itself, and those responsible for the last mile so that there are discrete appropriations for each activity that does not compete or conflict with each other, causing difficult decisions for directors about budgeting (e.g., lab equipment versus SNS drill). A discrete budgeting process would provide transparency as to which resources are being directed to which aspect of the countermeasures enterprise and allow adjustments to be made to resources as needed to respond to gaps.

Carlin remarked that the idea of unifying budgets around this issue is starting to become a common refrain that is echoed in the Blue Ribbon Panel Recommendation 4, which calls for a unified budget for biodefense (not only the SNS). She cautioned that unifying the budget for the SNS would require unification of all aspects of biodefense and preparedness.

Burhans suggested breaking down the budget not only at the state and local levels, but across the spectrum. BARDA provides outputs; those outputs are often inputs into the SNS. Therefore, he suggested, it would make sense to align those budgets and products so that the output of one part of the process becomes part of another, which enables predicting what those costs are. O’Toole clarified that this is different from the integrated biodefense budget concept proposed by the Blue Ribbon Commission, which is to look across the entire government.

POTENTIAL OPPORTUNITY FOR PROMOTING VISIBILITY OF THE PUBLIC HEALTH ENTERPRISE

It is evident that public health enterprise is important, held Parker, who suggested implementing a strategy to bring it to greater public prominence. He further suggested another standing committee on public health in the United States: “that committee could work really hard to tie public health into national security, which is needed desperately. . . . I would recommend strongly that we do something to bring attention to the entity called public health in America,” which Parker elaborated is suffering a lack of funding and attention that is perhaps a consequence of its own reliability. Starr agreed: “The first year that we got 100 on our technical assistance review, there was a great deal of rejoicing within the city. Then the next year, [same score,] a little less [rejoicing]. By the third year, no one cared anymore. It is true. You are a victim of your own success.” Paul Petersen, past SNS coordinator and current director of the Emergency Preparedness Program, Tennessee Department of Health, suggested the use of storytelling and videos (similar to examples at PHE.gov or PublicHealthEmergency.gov)

to demonstrate the impact of the SNS and show how the job gets done; it also allows government to let people know what is being done with funds to protect them.

Parker suggested that if public health structuring was given more priority, some SNS issues would fall into place. For instance, some models (e.g., Federal Aviation Administration, Selective Service) are run by counties or states, but can be federalized at the flip of a switch, which is needed in the event of an EID disaster or terrorist incident.

POTENTIAL OPPORTUNITY FOR IMPLEMENTING A SCIENTIFIC PERSPECTIVE FOR RISK ANALYSIS, INVENTORY, AND DECISION MAKING

Risk Analysis and Risk-Based Decision Making

Skivington pointed to the need to reexamine and restructure SNS governance and oversight, echoing Khan’s observation that here we are 17 years into the SNS and there has been no substantial change in this regard. He likened the current structure to a coordinated silo, but a silo nonetheless: he highlighted the need for a risk-based approach to ensure that every decision made is inclusive of the state and local officials, as well as being assessed post hoc for the purposes of improvement. Sosin countered that in his view, the government has in fact evolved. Rather, the issue is that risk is very difficult to quantify and contextualize, and the PHEMCE itself cannot make those determinations. The greatest challenge, he noted, is to identify the “risks we accept that we cannot prepare for,” that is, acknowledging and accepting that there is risk that will not be covered. Early targets set for CBRN events were too large and thus unsustainable, yet it has been difficult to walk those triggers back to a more sustainable level. This is a political decision, and “it isn’t . . . governance necessarily unless you include . . . governance within [the] political process.”

Jacobson construed the issue of risk-based management as a multicriteria problem to which the traditional use of cost–benefit analysis does not apply because averages are meaningless. “We are dealing with rare events, which have great cost, but may never occur.” He questioned what specifically is being done in the SNS context to evaluate quantitative and/or qualitative risk. Sosin responded that public health consequences are risk as defined in the intelligence community context, imperfect though that may be. Qualitative input from the intelligence community contributes to understanding and modeling the intentions and plans of groups intending to do harm and how those agents might be disseminated in populations.

Risk Assessment and Scientific Approaches for the SNS Inventory

Starr suggested reexamining the threat assessment to limit the scope of the stockpile even further if possible, with those decisions justified on the basis of sound science. To reduce pressure on public health in rural settings, requirements should be adjusted on the basis of scientifically sound threat profiles.

Parker noted that therapies can change over time, and we are at a scientific standpoint where we should review our therapeutics and regimens using new technologies and techniques (i.e., animal models) to see if we still have the same requirements. For example, one of the most expensive commodities in the stockpile is anthrax vaccine, which is there for both preventive and therapeutic reasons. Therefore, he questioned whether the recommended 60 days of antibiotics and anthrax vaccine is really warranted.² Applying innovation and science could change not only the cost structure of the SNS, but also the cold chain distribution, which is expensive.

Khan suggested that the SNS should also redefine its antibiotic holdings based on a portfolio that covers more microbes with fewer broad-spectrum drugs. He suggested that there are also things around margins that can help to deal with resource constraints; for instance, an effective solution to the problem of unsustainability could be for BARDA to assume a more powerful role, for example, by developing pathogen- or agent-independent therapeutics that target sepsis independent of bacterial or viral etiology. To deal with the problem of expiration dates, BARDA could also focus on new formulation strategies that make MCMs inert and thus infinitely stable until they are activated for use: “buy one and be done” to improve sustainability. Furthermore, developing a universal influenza vaccine would also reduce pressure on the SNS.

O’Toole reiterated that it is time to rethink and reexamine the sparsely substantiated assumptions about the critical premises of the SNS. There are revolutionary advances going on in bioscience and biotechnology, but they are stove piped in universities, and an entirely new strategy for research and development around bioscience is urgently needed, she said, not just for biodefense, but for the economy as well. According to the offensive weapons program data, we will not rescue anyone in the initial plume; it is better to discuss this now rather than after a bioterror attack.

Eva Lee, professor and director, Center for Operations Research in Medicine and Healthcare, H. Milton Stewart School of Industrial and Systems Engineering, Georgia Institute of Technology, emphasized the need for

___________________

² Based on one study of primates where one primate had spores in the mediastinum at 60 days.

innovation in vaccine design and MCM development, citing the difficulty in maintaining compliance with current regimens. Prophylactic MCMs are important, as are innovations in design and rapid testing to improve security, save lives, and protect the economy.

Informing Resource Decision Making for the SNS

Rocco Casagrande, managing director at Gryphon Scientific, presented the results of a Department of Homeland Security (DHS)–CDC partnered risk mitigation study for informing resource decision making for the SNS. Its purpose was to leverage information in the DHS Integrated CBRN Terrorism Risk Assessment (ITRA; see Box 4-3) to provide quantitative information that is useful when making decisions about procurements for the SNS. Essentially, the study aim was to provide a risk-based analysis for informing discussion on how and why the SNS is stocked, specifically, to determine what products, if included in the SNS, would reduce the risk associated with CBRN incidents and how much of these products would be needed to best reduce risk.

Casagrande explained that CBRN incidents can be small (hurting a handful of people) or they can be enormous, so the question at hand is how much of an MCM is needed to address the threat. Some MCMs are specific to a particular agent, and so investment decisions between MCMs should be prioritized: “How do you make an investment in a countermeasure that

only addresses agent X versus agent Y?” The effectiveness of some MCMs is dependent on the dose received by a victim, giving rise to decisions about which MCMs are likely to have real utility in a response. For example, if a victim gets an overwhelming dose of many chemical or radiological agents, the countermeasure will not change the medical outcome. Furthermore, the length of a prophylactic window is a function of dose for some agents (which itself depends on weapon payload and target). According to Casagrande, this requires determining the likelihood or risk that these victims are going to get an overwhelming dose of the agent versus a dose that is treatable.

Casagrande explained that the study methodology first employed the ITRA to generate several hundred biological and radiological attack scenarios that represent the risk of terrorism with these agents. The second phase also looked at chemical and nuclear attack scenarios, before looking at the CBRN space in total. Exposure information in these scenarios was used in models that can predict the ability of the public health response system to mitigate the consequences given the dispensing of MCMs in the SNS. Then other, equal-cost, SNS formularies were generated to determine how the dispensing of MCMs from these notional formularies mitigates consequences (risk is defined actuarially as probability × consequences). Then, to make recommendations, the performance of the current and alternative SNS formularies was compared against the suite of scenarios when considered together.

Casagrande noted that PHEMCE was heavily involved to determine what metrics are used to evaluate the performance of the SNS, as well as the predicted deployment and use of MCM from a federal level. From the local and state levels, several stakeholders were involved to help understand how MCMs are used in the last mile: what trigger decisions are, what their dispensing options are, etc.

According to Casagrande, the ITRA is a rich resource that contains millions of scenarios, each of which differs in terms of probability, consequences, agent, and target. To represent the risk space best, the investigators chose the riskiest scenarios that cover a variety of consequences, agents, and targets. The aim was to select scenarios where the combination of consequence and probability are actually going to drive risk (only the thousand out of a million or so scenarios). They selected a variety of scenarios that differ in terms of the probability and consequences (some were lower probability but higher consequence, and others that were higher probability and lower consequence) but also capture a variety of agents and many targets to represent the risk space (see Figure 4-1).

When sampling the risk assessments, the investigators ranked the ITRA simulations by their total risk and took samples through them: 100 scenarios from the median sample; 100 scenarios from each type of agent and CBRN;

100 samples from the low end, the P5, such that 95 percent have a greater total risk; 100 scenarios from the top end, the P95 such that 95 percent of the scenarios have a lower total risk. He construed the samples as bracketing the possibilities in terms of the threat, with the high end being considered the best possible day for the terrorists, which Casagrande explained as

On the low end, the P5, everything was going against them, so they got a lower total risk. The investigators made sure to capture that variety of uncertainty in what the terrorists might do or how luck or fortune might play in their favor or against them. Uncertainty in public health response was accounted for using three response capabilities: optimistic, average, and pessimistic (see Figure 4-2).

Casagrande explained that there are a range of ways that MCMs might actually be used, how quickly they might be dispensed, or what the trigger point for calling SNS would be. Treatment length, for example, can vary significantly according to how many doses of MCM a victim receives. As with the risk assessment sampling, the investigators took the most optimistic case and the most pessimistic case to bracket the uncertainty in the public health response. This approach provided an understanding of the upper and lower bounds of current, possible response capabilities, demonstrated which responses had the most impact, and the best possible reduction of risk with MCMs. Conclusions would only be robust if they held across all nine of those uncertainty combinations, the most and least sophisticated terrorists, and the high and low response capabilities. Furthermore, if some of the conclusions are only valid under certain circumstances they might possibly be leveraged.

To bound the analysis, the investigators tried to create an SNS opportunity space to determine what risk looks like (e.g., how many people die across how many different scenarios; see Figure 4-3), given the complete absence of the SNS. Nonexistence of SNS and MCM availability only in the local medical system could be used. The number of deaths from that scenario was then compared to the number with an SNS with unlimited resources. The area between these curves is the “opportunity space” that the SNS can occupy, with all of the alternative formularies between the two curves.

Referring to the investigators’ formula for risk, Lee questioned how it is possible to differentiate between events with high risk but low consequence and events with huge consequences in which the risk will be identical, because it is represented by a number. She commented that this is

critical in terms of looking at an investment portfolio: the key part is using that information to determine how you make investments downstream. Casagrande replied that they intentionally chose risk as the metric (as opposed to consequence or probability) to capture a spectrum of scenarios, some of which are less likely, but of much higher consequence than others that are lower consequence but much more likely. Investment decisions ideally would cover events that are extremely common but lower consequence as well as those that are really rare and extremely high consequence, but understanding the diversity of events (e.g., a thousand scenarios) will help to guide investment decisions and assess where the gaps are at present. Producing these scenarios on a risk basis enabled them to identify the types of scenarios that are not well prepared for and to analyze the similarities among those scenarios.

Jacobson contended that the investigators’ risk analysis has a fundamental flaw in that terrorists can choose the most optimistic setting for an attack, thus rendering the balance invalid. Casagrande explained that although terrorists would indeed like to have unlimited resources, in reality they do not have that choice. The optimistic/pessimistic sampling is to understand the best judgment of the intelligence community on their capabilities. There is also some built-in uncertainty: atmospheric conditions might not be ideal for issuing the attack; the chosen venue may not have many people in it. Many factors beyond terrorists’ control determine their final capabilities in terms of the weapons they are able to produce,

the target they are able to hit, and the consequences of the strike. There is extensive game theory on behalf of what the adversary is predicted to do; for example, they are not allowed to execute stupid attacks or ones that they know are going to fail (they are not allowed to try to attack a city by releasing something downwind as opposed to upwind, for instance).

Jaffe raised the topic of requirement setting; he commended the investigators’ work insofar as bounding the question and deciding how to determine consensus scenarios (which are the on-ramp of how to define the threat). But he cautioned against having this “wide-open aperture,” because we should be able to determine the best possible scenario for the terrorists and the worst possible for us, and vice versa. This information generates the exposure data for each of the different scenarios that are chosen for each of the different threats. From that point, modelers working in BARDA’s Analytical Decision Support area can assess the unmitigated to the mitigated benefit, forming the basis for the public health consequence assessments that are used for requirements setting—which, although not directly related to the formulary study, is part of the overall process. Requirement setting is necessary for determining quantities—how much will be used and what the product will look like in a stockpiling goal.

Casagrande clarified that the ITRA was intended to be a small piece of the overall decision-making process. Sosin explained that when looking across this spectrum of risk, it is possible to buy down 50 to 75 percent of the risk by aiming at 10 percent of the scale; he cautioned against stockpiling to very rare but catastrophic scenarios. He commented that a piece of this exercise was to better understand that spectrum of risk to make better decisions about what those targets might be. ITRA, by design, is about non-state terrorist risk and thus is relatively limited with respect to the spectrum of what the SNS might be. He highlighted as the most valuable part of the exercise the process of carefully considering each scenario and how it might play out in an actual event and response, as well as challenging assumptions and striving toward improvements in the concept of operations and best possible use of resources for MCMs.

This page intentionally left blank.