Uncontrolled spread, p.29
Uncontrolled Spread, page 29
On September 20, 2017, Hurricane Maria made landfall in Puerto Rico and overwhelmed the island. Maria was a category 4 storm when it hit Puerto Rico and traversed the entire long axis of the island, bringing wind gusts of up to 113 miles per hour. Puerto Rico’s power grid was completely destroyed by the storm, leaving millions without electricity.
I was commissioner of the FDA when this tragedy struck, and the death and devastation inflicted on the island’s residents was our urgent concern. The FDA helped in the effort to restore essential services and public health. However, the impact on the island’s manufacturing sector also had broad implications that put at risk the public health of our entire nation. Puerto Rico was home to a large concentration of advanced drug and medical device manufacturing facilities, and these sites were shut down as a result of the destruction. The disruption risked the supply of critical medical products for all of the United States.
The clustering of these facilities in Puerto Rico was not by chance. It was the result of a policy enacted in 1976 that gave drug and device makers special tax incentives that made it economically favorable for them to locate their manufacturing on the island and then ship their products to the US mainland. It was part of a coordinated effort to move life-science manufacturing to the island as a way to support its economy. As a result of these efforts, the manufacture of medical products became a key part of Puerto Rico’s economy, and the island became responsible for the production of about 10 percent of all the prescription drugs consumed in the US.34
The FDA was working closely with the Puerto Rican government and manufacturers to assist in the hurricane recovery efforts. On Friday evening, September 22, I got word that a key dam had become vulnerable and might burst. The resulting floodwaters could have taken out a number of warehouses that had stocks of important cancer medicines and other therapeutics. If the drugs were lost, we feared it could trigger shortages of some lifesaving medicines.
I worked with the FDA’s chief of staff, Lauren Silvis, to secure permission to move the drugs out of the warehouse by air, but the airport was partially closed, and obtaining landing rights was difficult. We didn’t have a lot of time. I phoned Andrew Bremberg, who at that time was head of the president’s Domestic Policy Council, at the White House, to ask whom I should call for help. He connected me with Chris Krebs, an acting undersecretary at the Department of Homeland Security. Krebs would go on to head the Cybersecurity and Infrastructure Security Agency, a stand-alone federal agency that is an operational component under DHS oversight, and he would gain international attention when his efforts to affirm the integrity of the 2020 US presidential election put him in public conflict with President Trump and the president fired Krebs from his job. When I reached Krebs late on the night of September 22, he was at his brother-in-law’s wedding reception. I explained the problem. “Let me take my kids home and call you,” he replied. He called me thirty minutes later and spent the rest of the evening helping me secure the permissions we needed to move the products off the island. Thankfully, the dam ended up holding.
The crisis that night underscored some of our deeper vulnerabilities. We had located the manufacturing for some of our most important medicines to a small number of geographically concentrated sites. In some cases, we didn’t have an adequate plan B if our existing supply chain were disrupted. This wasn’t an issue particular to Puerto Rico. Across the drug industry, critical products are often made in just one or a few plants. A disruption in a single manufacturing site can cause a cascading series of shortages. After 9/11, we paid some attention to these risks. There was a focus on situations where a single manufacturing site was responsible for the production of most, if not all, of a critical medical product. Back then, the concern was that a terrorist attack on one of these sites could trigger a public health disaster. So, in situations where such risks existed, we started to quietly approach drug makers and encouraged them to build additional facilities. A deliberate attack on any one of these sites could cause drug shortages that could lead to mass casualties.35 One area of our focus at the FDA was the supply chain for insulin. It turned out that much of the country’s supply came from a single manufacturing site. That’s not the case anymore, but back then, we were worried about the risk that a deliberate attack on that site could disrupt the majority of the nation’s insulin supply. So the FDA worked closely with the drug manufacturer to help support the company in its efforts to build an additional manufacturing facility.
But that work to build redundant manufacturing sites for key medicines was a limited affair, and ultimately, attention faded and the production of a lot of critical medicines became geographically consolidated again. A week after Hurricane Maria, I traveled to Puerto Rico and visited with our FDA staff on the island. Many were without power and supplies, but they were working to keep the agency’s office operating because they were dedicated to the agency’s mission and knew how critical it was to help make sure FDA regulated products on the island, like food, blood, and medicines were safe and available to support the recovery efforts. Puerto Rico was the location of a large FDA field office where about one hundred of our staff worked. The agency’s presence reflected Puerto Rico’s importance to America’s drug supply chain. Before leaving for that trip, working with Lauren Silvis, the two of us had called the chief executives of each of the companies that maintained a substantial manufacturing facility on the island. We wanted to hear from the CEOs firsthand how much of an impact the storm had on their plants, the consequence to their operations, the public health implications of their outages, and what steps the FDA could take to help them resume critical functions. Our highest priority was to help the people of Puerto Rico. However, for the facilities that manufactured critical drugs, where shortages could have devastating nationwide consequences, we were also working directly with the Federal Emergency Management Agency to help the plants get back online. In some cases, we were helping plants obtain the fuel, generators, and diesel storage tanks they needed. Chris Krebs became a huge help to the FDA and to the people of Puerto Rico in achieving this mission.
When I called each of the CEOs, the status reports varied. Some of their facilities had been badly flooded or damaged by the high winds. All of the plants had their electricity knocked out and were running on backup generators powered by diesel fuel—if they were operating at all. Most had only partially restored their functions. In many cases, the manufacturing sites were running enough of their key infrastructure like air-conditioning systems in order to protect their facilities, but they wouldn’t be able to resume substantial production anytime soon. Each of the sites I surveyed had been greatly impacted. There was one exception: Amgen.
The biotech company operated a large manufacturing facility in Puerto Rico, where it produced the drug Neupogen. The medicine is principally used to help the body make white blood cells, chiefly for patients who are receiving cancer chemotherapy. It works by stimulating the production of new white blood cells in the bone marrow. Neupogen is also vital to national security because it can be used to save the lives of people who have been exposed to radiation by helping the body reconstitute bone marrow.36 It’s for its clinical application in a radiological or nuclear emergency or incident that Neupogen has become a critical part of our national defense, a therapeutic hedge against an act of terrorism.37 If a dirty bomb or other radiological weapon were detonated in a populated city, there would be a surge in demand for Neupogen as a tool to help rescue people harmed by the radiation. For this reason, the US government has added Neupogen to the Strategic National Stockpile, to ensure there will always be an uninterrupted supply of this critical medicine.
When I got Amgen’s CEO, Bob Bradway, on the phone, he described the status of his facility. Located in a remote part of the island, it was operating virtually uninterrupted. The manufacturing site was so hardened that it could probably have churned out Neupogen almost indefinitely, even if it were kept off the electric grid for many months. Staffing the facility was his urgent concern, along with providing for the safety and well-being of his workers and their families. But his plant’s physical structure was fully secure. He had ample generators and enough diesel fuel stored in massive tanks to operate his manufacturing site for long stretches of time.
After a day spent on the phone with CEOs who were piecing together shoestring strategies to keep their plants limping along at partial capacity, Amgen stood out. I asked Bradway why his company had invested enormous resources in building a facility capable of withstanding a major disaster. He said that Amgen had made a guarantee to the federal government that there would never be a disruption in the supply of the drug. To help fulfill that promise, it had created a domestic network of hardened manufacturing sites and distributed these facilities around the nation. Amgen also maintained some excess inventory at its plants so there would always be a “stockpile” in the event of a crisis. The drug was a biologic, so it couldn’t be stored indefinitely. It wouldn’t have made sense for the US government to buy a whole bunch of it and put it in a warehouse. Instead, reserves were kept at the manufacturing sites. When Neupogen came off the production line it didn’t get immediately shipped out. Some of it went on a shelf at the plant, held in a careful storage unit, while product that had been manufactured weeks earlier was shipped out. This was a vendor-managed inventory, maintained on behalf of the federal government. Through this approach, a small but constant stockpile of Neupogen was being continuously refreshed.38
Amgen’s efforts provide an instructive lesson for how we can harden ourselves in preparation for a pandemic and secure the capacity to expand diagnostic testing in a crisis. COVID showed that testing is critical to identifying and controlling the epidemic spread of a new pathogen. The approach that Amgen took to guarantee a continuous supply of Neupogen could also apply to securing an adequate supply of testing: build the residual capacity we need into the system.
The backbone of our testing preparedness is the network of about one hundred public health labs spread across the nation. But these are mostly small labs that do specialized surveillance work. At full tilt, their available capacity to screen for SARS-CoV-2 at the height of the crisis peaked at about ten thousand tests a day—enough to do initial investigative work of a slowly spreading pathogen, and to offer a critical tripwire to a new pathogen, but not enough for a fast-moving epidemic.39 Even if we invested money to expand the capacity of this network, these labs couldn’t maintain enough idle volume to be able to provide sufficient surge capacity in a crisis. However, they offer a critical starting point for building a system that can guarantee better preparedness for the future.
The seeds of a more robust system had already been planted years earlier. Leveraging these public health labs, the US invested in a lab response network following 9/11, but it was largely a bespoke effort to evaluate bioterrorist threats. It wasn’t meant for the large-scale screening needed to keep up with a galloping epidemic. Writing in The Lancet in 2011, Dr. Ali S. Khan, the director of the CDC’s Office of Public Health Preparedness and Response, observed that “Large-scale and unpredictable natural, accidental, or intentionally caused disease outbreaks and environmental disasters need many of the same routine surveillance, laboratory, risk communication, and other core public health systems. The flexibility of the Laboratory Response Network shown during the anthrax attacks, for example, has also played a key part in validating BioWatch results, and for responses ranging from severe acute respiratory syndrome (SARS), monkeypox, West Nile virus, and H5N1, to investigations for ricin and saxitoxin poisonings and numerous exposures to mercury.”40 To support the national testing needs that would be required if an epidemic ever got under way, this system needed to be expanded on.
To keep up with the needed testing volume, such capacity has to be already in place. We need to create a reserve capability that encompasses the full spectrum of testing platforms and consumables that we’ll need in a crisis. It can’t just be mothballed somewhere, either. It needs to be kept operational. This reserve capacity should be thought of as similar to the way that Amgen created those hardened facilities for the continuous production of Neupogen, making sure there was always a supply of the medication available, and that we had the ability to ramp up production in the event of a crisis. We should take the same approach when it comes to testing capacity, and create facilities that have excess volume that can be expanded in a crisis.
The best way to secure this capability is to leverage the nation’s existing labs. Under such a scheme, the government could contract with the large commercial labs and ask them to create some additional testing capacity in their existing facilities. This excess capacity would always be available in the event of a crisis. Other smaller labs could also provide some of this volume, but the three largest commercial labs (LabCorp, Quest, and BioReference Laboratories) with their enormous throughput and facilities, could become the core pillars of this new system.
Under this approach, the federal government would subsidize these labs to build excess capacity inside their existing sites. The labs would also be paid to keep a fresh stockpile of testing consumables (reagents, swabs, etc.) to support a surge in testing demand. Demand for lab services already fluctuates by season, and there’s always some standby capacity available to handle a natural surge in the demand for testing that occurs each winter. We can build on that concept to increase the size of that reserve. Here is how it could work:
Right now, the commercial labs might maintain about three thousand testing machines in their largest reference sites. They typically operate these facilities at 80 percent capacity for maximal productivity. Their present focus is to operate these facilities for maximum efficiency. That helps keep costs down. Instead, there needs to be an equal focus on operating these sites with greater resiliency. Some surge capacity needs to be built into the system as a hedge against a major crisis.
This could mean, for example, operating five thousand machines but running each of them at 50 percent capacity. Running more machines at a reduced volume isn’t maximizing their efficient use. You’re leaving a lot of expensive machinery operating below capacity. However, that excess volume is now available to be tapped in the event of a public health crisis. Addressing a surge in the demand for testing then becomes a function of just running the existing machines a little harder. In this way, the surge capacity is baked into the infrastructure and can be lit up in the event of a crisis. It’s kept hot. The federal government can pay the labs for the additional cost of maintaining this excess capacity, as part of our national security preparedness. Asking labs to run more machines at a reduced capacity is preferable to building a giant lab and mothballing it. This kind of sophisticated equipment needs to be changed out every five to seven years, so it must be operating to make sure it’s up to date, and it needs to be in the hands of a lab that’s able to properly staff these facilities and implement those regular upgrades.
It’s also critical that we distribute this infrastructure around the nation. Diagnostic labs need to be close to population centers to maintain their efficiency. Testing large volumes of patient samples is as much a challenge of logistics as it’s a question of having enough testing machines and consumables to process the samples. One of the challenges during the peak of the COVID crisis was just the physical processing of the samples. Getting the patient samples shipped around the country to a small number of reference labs created bottlenecks. National reference labs were strained to take in and process the many thousands of additional packages that were arriving every day by FedEx, filled with swabs. To reduce testing delays, it’s important to have the testing sites spread around the country, so that they’re closer to the points of care. State public health labs can fill a key role in a crisis, helping to direct this traffic in patient samples across different private and public labs, enabling states to make maximum use of existing supplies and testing capacity.
As for the hardware, different testing platforms are needed to test for different kinds of pathogens. For detecting new viruses, labs need to maintain PCR and sequencing machines. Labs also need equipment to perform antibody tests. The platforms used for such tests can be especially important for detecting infections found primarily in the blood, like the Zika infection.41 To have a proper infrastructure that prepares us for the full range of potential threats, we need to support a diverse base of complementary testing equipment.
We also need to think about the front end—the system for collecting patient samples. In the case of COVID, we couldn’t fully leverage the large footprint that commercial labs like Quest and LabCorp maintain because those facilities were designed primarily for collecting blood samples. COVID was a respiratory pathogen that required nasal swabs. So, we had to create new testing sites that were able to handle huge volumes of swabs. South Korea, following its experience with MERS, built and maintained a lot of the infrastructure they’d need in order to erect a massive screening program for a respiratory pathogen. That’s why it was able to establish those drive-through sites so quickly. Initially, at the outset of a crisis, the testing is likely to get done at these kinds of large testing centers. This was the experience with COVID. But over time, to give more people access to timely screening, the goal must be to move testing into a distributed model where it’s done at medical offices, at retail locations like pharmacies, and at home.
There were other obstacles to putting in place a better testing infrastructure that will need to be solved as part of a new pandemic blueprint. One was the way that the CDC claimed intellectual property rights over the tests it designed. The agency’s scientists are allowed to file patents on assays they develop for public health purposes, including in emergencies. The CDC can then seek royalties on the tests.42 However, this practice can complicate the CDC’s ability to quickly hand off tests to commercial manufacturers. That’s what happened with COVID. Commercial kit makers worried about how the CDC’s claims of intellectual property over the tests could impede the ability to market the tests, and get reimbursed. Negotiating the terms for licensure can eat up valuable time in a crisis. That’s precisely what occurred during COVID.
