The US declared monkeypox a public health emergency this month, but the decision may have come too late. Though states are now required to report cases, and commercial labs have an approved test, a testing bottleneck persists, and cases — which passed 10,000 confirmed cases this week — are likely still being underreported. Any effective public health response to an infectious disease is dependent on having accurate data. If the virus spreads to other populations, such as college dorms — where cases have already been reported — the testing bottleneck could ultimately make it impossible to contain the spread. Reliable demographic information is key to making the right choices for allocating limited tests and vaccines.

All of this feels like an uncanny echo of the early mishandling of Covid-19. Limited access to testing, a hobbled federal infrastructure to track cases, and the general lack of communication among different agencies and states complicated the federal government’s ability to make evidence-based public health decisions. Reporting lags on rising cases meant that lockdowns began too late to save tens of thousands of lives. Similarly, certain communities uniquely at risk, like Black and Hispanic people who lacked access to health care, were suffering higher rates of severe illness and death from Covid before policymakers had any way of knowing where to direct public health outreach.

But the roots of this deadly problem long predate monkeypox outbreaks or the Covid-19 pandemic. The US has always had a fragmented health care system, with widely disparate experiences for patients based on state, insurance company, or hospital chain. Without systems to reliably record and share population-level data between decision-makers, health care workers can’t focus on helping the patients who need it most. The consequences are worse for marginalized people — such as Indigenous people, people with disabilities, or youth at risk for teen pregnancy — who were already facing inadequate care before the pandemic.

It doesn’t have to be this way. The US has an opportunity to learn from the tough lessons of the last few years and build on work to improve transparency and data sharing. With monkeypox already a global public health emergency, it’s vital for the data to be available, promptly and accurately, to coordinate an effective public health response. This is how we can get there.

Why does data matter?

Evidence-based medicine — the practice of using observation, studies, and randomized controlled trials to test which treatments work — has transformed the medical field over the last century. But for that to work, as Covid showed, you need to have data to inform medical decisions.

The US has mandatory reporting systems for some contagious diseases, along with public health concerns like lead poisoning. This usually means that hospitals, clinics, and laboratories are required to report the location, severity of the illness, and treatment provided for any confirmed case. They also must document demographic information, such as race and ethnicity.

But that reporting is hobbled by the fact that there is no single agency responsible for the US health care system. Data is collected by federal agencies such as the Department of Health and Human Services — which houses the CDC, the Food and Drug Administration, and the Indian Health Service — as well as the Federal Emergency Management Agency (FEMA), which focuses on supplies and infrastructure for disaster preparedness. But communication among these agencies, the state health departments that report to them, and the hospitals and organizations where data is collected is often challenging, thanks to a fractured system made up of hundreds of different organizations.

Data comes in from over 900 health systems, or chains of hospitals under shared management; the largest include about 200 hospitals. But that’s just a fraction of the over 6,000 hospitals across the country. So when, for example, positive test results for Covid-19 or monkeypox, or cases of workplace exposure to pesticides, have to be reported to the state, public health boards in every state must coordinate with hundreds of different organizations and aggregate their data before they can share it with federal agencies. Except during an officially declared public health emergency — which, for monkeypox, is only a week old — the CDC has limited legal power to mandate reporting.

Data also isn’t collected the same way everywhere. There is a large number of different electronic health record systems currently in use in the US. They allow medical professionals to document a patient’s diagnosis and treatment, and in theory, share them more efficiently than in the days of paper-based records. But the software systems aren’t designed to be compatible with each other, so they cannot easily exchange data.

Even for a popular software platform like Epic, which covers about a third of hospital systems in the US, categories like a patient’s diagnosis — or even something as simple as their height or weight — are often customized for a particular hospital or chain. This makes for a more efficient workflow for the medical professionals on the ground, but it means that every hospital or chain is collecting slightly different information and organizing it differently. In order to piece the information together into a national picture that policymakers can actually use, each individual dataset has to be mapped onto a standardized format, a massive administrative burden that adds to delays.

For example, when I worked as a nurse in Canada, different hospitals in the same city used different recordkeeping software. Rather than digitally transferring data, other hospitals would fax a paper copy of their records, which had to be entered manually, leading to delay and data entry mistakes — and this was assuming that we knew the patient had been hospitalized there before. Getting the records of a patient’s medical history from primary care providers or clinics was even more challenging. It wasn’t uncommon for a single patient to end up with two or three duplicate charts, sometimes due to minor spelling errors in their name.

With hundreds of different organizations involved, it’s no wonder the US faces greater challenges in maintaining a complete and accurate national-level database than a country like the UK, with a centralized single-payer health care system. The sheer size and varied demographics of the US population add further challenges.

“The United States is incredibly diverse in many ways,” is how epidemiologist Katelyn Jetelina puts it. “You know, race, ethnicity, age, health status, state-level policies, rural, urban. There are so many [of what we call] confounders in epidemiology, so many important factors that will influence health and disease. What we see in New York City isn’t necessarily going to be generalizable or translatable to, for example, rural Texas.”

Until the US started using commercial labs to ramp up testing capacity for monkeypox in late June, samples could only be processed at state public health labs, with a cumbersome process. Hot spots like New York were overstretched, while other states’ labs sat idle. The delays and poor coordination between clinics and city health departments meant that contact tracing happened too late to contain the spread. If the spread had been caught earlier, patients would have been more likely to minimize their risk and seek out testing and treatment if they were exposed, and there would have been more advance warning on ordering a vaccine supply.

Undertesting doesn’t just affect the case numbers reported, but hurts patients’ access to treatment. Tecovirimat, or TPOXX, an antiviral drug that is most effective for treating monkeypox if started early, can’t be prescribed until a test comes back positive, and since it’s not officially approved by the FDA for monkeypox treatment, doctors need to jump through bureaucratic hoops to prescribe it. This leaves many patients suffering from untreated painful lesions for days or weeks.

As Jetelina pointed out in a Substack post, monkeypox doesn’t need to go the same way that Covid did; it’s a known disease, with a vaccine already developed, and spreads via close contact rather than being airborne. But the slow initial response, disorganized due to lack of information, means that the window of opportunity to contain monkeypox is closing.

Where do we go from here?

However difficult the growing pains, there has been real progress made on data collection since the first US Covid-19 cases in early 2020.

The National Covid Cohort Collaborative, a project run by the National Institutes of Health which gathers clinical data on Covid-19, was stood up during the pandemic. Joni Rutter, the collaborative’s acting director, describes the challenges they faced when combining hundreds of disparate data sources around the pandemic: “Even when you’re talking about height, one site will send us data in inches. One site will send it in centimeters.”

For more complicated questions, the process is even more fraught. Long Covid, for example, is linked to more than 200 distinct symptoms of varying severity, but screening tools generally include only some of these, their definitions vary between different hospitals and clinics, and doctors often won’t document every symptom a patient experiences. As a result, estimates on the risk of long Covid vary from as much as one in two Covid-19 cases to one in 20. It’s also particularly important for the Collaborative’s dataset to accurately reflect the diversity of the US population, a challenge their team has worked hard on. “It really helps us to get access to rural communities and more minority communities,” Rutter says.

The NIH’s efforts to build the Collaborative database in the right way were a major step forward, one that should be more widely adopted. More than 2,000 scientists are using the group’s centralized database system to ask critical questions about Covid, like rates of reinfection, characteristics of long Covid, and differences in outcomes between urban and rural patients. Meanwhile, the National Patient Safety Board, an advocacy group calling for a health care equivalent of the National Transportation Safety Board, hopes to improve tracking of medical errors and use machine learning to find underlying causes.

Other organizations are working on cleaning up the data at its source. The Advanced Research Projects Agency for Health, officially authorized in March 2022, is another NIH program based on the Defense Department’s famous research center DARPA, with the goal of promoting innovation and new technology in health care. Its initial work may include revamping electronic health records and letting hospitals migrate their data over to new and improved systems. In its 2022 National Covid-19 Preparedness Plan, the White House committed to improving data infrastructure by scaling up electronic case reporting systems to cover all states, in order to better track case counts and hospitalizations and link these to vaccination rates.

That isn’t enough, though. According to Karen Feinstein, spokesperson for the National Patient Safety Board, the entire approach to health care needs to change. One example to emulate could be the aviation industry; thanks to decades of recommendations from their safety board, which has scrupulously tracked airline data since 1967, accident and fatality rates in air travel have fallen drastically.

“We have all kinds of technology to keep our pilots and passengers safe on airlines and our astronauts safe as they go to and from the space station,” she said. “We know that the answer is to build a better airplane or to build a better spaceship, and to have the pilots and astronauts do the things for which they are trained and prepared. The problem we have in health care is that we haven’t yet built a better airplane.”

And building a “better airplane” for health care will involve reforming the current decentralized and fragmented recordkeeping. As Rutter sees it, “electronic health records need to evolve, and that’s going to be one of those things that I think we as a community, as consumers, need to help ensure does happen.” In the meantime, the National Covid Cohort Collaborative will continue with its current strategy of cleaning and combining the existing records, and is about to launch a section on monkeypox within its open-access database.

Jetelina believes that the federal agencies involved in public health responses need to be granted stronger legal authority to mandate standard reporting from states and hospital systems so they can come closer to the kind of constant surveillance the UK managed early on with Covid-19 and with monkeypox. She thinks the key is to “take out a lot of this red tape and bureaucratic paperwork, at least during a public health emergency, [and] respond much, much quicker.”

With monkeypox, the US can lean on the systems and infrastructure built during the Covid-19 pandemic, but some programs, like those that reimburse providers for treating uninsured patients or provide free Covid-19 tests, vaccines, and antiviral drugs to community health centers, were already scaled down after funding was decreased. In order to pull together a national response, the US needs straightforward, transparent data reporting that can be compared and combined on a national level.

The final difficulty will be in keeping this momentum going. The declaration of a new public health emergency for monkeypox will help keep federal funding flowing toward projects like the OpenData portal, but the need for better health care infrastructure won’t end when the emergency does. In a chronically underfunded public health system, short-term efforts may not be enough.

As Feinstein puts it, “the challenge we always have is something new that distracts the efforts toward reform, because we’ve gotten close to this in the past.” But with the lessons learned during the pandemic and new threats potentially on the horizon, she believes that “now is the time.”