What we did was meaningful. I was drawn to start this publication because it presented an opportunity to break out of the typical crisis-driven flow and ebb of journalism about the field. Harvard wasn’t a publisher, but it was in the business of sharing knowledge, and I thought we could do for public health what Harvard Business Review does for business. I believed there was no public health without the public, and while it took some feints and half-steps to figure out what that meant for our journalism, we eventually settled on assessing every story idea with a simple question: “What would this story change?” Implicit was a corollary question: “And for whom?”

In the meantime, we relaunched the magazine as a digital publication, built out a social media presence, and launched a weekly newsletter. We co-sponsored a well-attended structural racism symposium and special issue, a series on public health data, a Public Health in Action collaboration with The Studio at the Harvard T.H. Chan School, and an event on artificial intelligence with Johns Hopkins Bloomberg’s Global Health Now. We had momentum—visitors to the site almost tripled in last year’s fourth quarter versus the prior year. Almost 15,000 people signed up for Harvard Public Health Weekly, close to 90 percent of them not connected to Harvard.

Readers ate up pieces on processed foods, the health effects of alcohol, and mental health. You also read our beautifully written and photographed narratives like the 10^th anniversary of the Flint water crisis or our look at Christian Happi’s bold aims for African science, and public health’s role in the recent Puerto Rican elections.

Our goal was to publish stories that would help improve health outcomes. That’s hard to measure in three-and-a-half years. But over 40 percent of you opened the newsletter in a typical week. In the last year, readers shared our articles more than 2 million times on social media. We’ve had at least 25 stories republished on other sites and 40 mentions in newsletters. Our stories have been cited in other publications and used in classrooms.

Public health outcomes change slowly, so it’s harder to measure real-world impact. I would love to hear from you about trying an idea you read about in HPH, or even if you just shared the idea with a colleague. Did you use an article from HPH in a class or a meeting? It would be great to hear from you at our inbox, magazine@hsph.harvard.edu. It will be live for a few more months. So will the site, and I encourage you to download articles you found useful.

The pandemic sparked a surge of public health journalism. These are the sites and newsletters I follow closely or scan regularly, and recommend to you:

• Axios Vitals
• Devex CheckUp
• The Examination
• Global Health NOW
• The Harvard Chan Studio
• KFF Health News
• MindSite News
• Public Health Watch
• The Healthiest Goldfish
• Your Local Epidemiologist
• ProPublica
• STAT Morning Rounds
• TradeOffs (podcast)

Also, the new Healthbeat is off to a promising start, focusing for now on Atlanta and New York City.

I have had a long and varied journalism career, much of it spent chronicling the vast impact of high technology. I have never done more meaningful and important work than what we were doing at Harvard Public Health. I am so thankful to the school, colleagues past and present and our fabulous advisory board, everyone who gave me informal counsel, and all the readers who reached out. I rue that we won’t be able to continue. But a wonderful thing about public health is its focus on the public. It is political with a small ‘p,’ rooted in communities.

What’s most important is that you in the public health community (and in the public) stay engaged in doing the good work you do. Keep telling your stories!

Onward,
Michael F. Fitzgerald

The post A farewell to HPH readers appeared first on Harvard Public Health Magazine.

Here are a few striking examples of advocacy in this new genre of graphic art.

Expand Image”>

Candy

In the late 1800s, urbanization increased people’s dependence on manufactured foods. Manufacturers often used cheaper, potentially harmful ingredients to cut costs.

Candy makers frequently added toxic mineral additives to their products to create vibrant colors. This 1858 satirical illustration depicts Death stirring candy tainted with arsenic and plaster of Paris.

The problem was widespread. In 1884, New York City’s Board of Health destroyed over 72,000 pounds of adulterated confectionery. Despite growing awareness of these risks exposed by chemists, regulatory efforts were initially weak because businesses opposed them. The lack of food safety regulations continued until the early 20th century, when the progressive movement and works like Upton Sinclair’s The Jungle finally prompted significant reforms. In 1906, Congress passed the Pure Food and Drug Act, protecting consumers from dangerous food additives and establishing more rigorous food safety standards.

Expand Image”>

Revolvers

The epidemic of gun violence in the U.S. has deep historical roots: Even in the 1880s, the widespread availability of guns to ordinary people was recognized as a growing threat. This danger emerged from several technological innovations in the mid-19th century.

The rotating cylinder that could hold six bullets was revolutionary when the Colt revolver was mass-produced in the mid-1800s. Before this innovation, firearms were far more cumbersome. They featured long barrels and required manually loading gunpowder and shot for each use.

The new revolvers changed everything. They had shorter barrels, rotating cylinders with multiple chambers, and used pre-manufactured bullets that combined gunpowder and projectile in a single cartridge.

These smaller weapons could be easily carried or concealed and quickly fired multiple times without reloading—democratizing gun ownership but creating new public safety risks.

The wild-haired figure in the Emporium for Cranks resembles Charles Guiteau, who shot President James Garfield with a revolver in 1881. Garfield died 79 days later from complications of the wounds. Newspapers dubbed Guiteau, who believed he was owed a consulship by Garfield’s administration, “a crank.”

Expand Image”>

Fire

Two innovations in mid-19th century cities promised safer living: brick exterior walls replacing wooden construction, and Elisha Otis’s “safety elevator,” first installed in 1857, which prevented cars from falling. These advances lead to a boom in high-rise construction starting in the mid 1880s.

However, developers ignored crucial risks. While brick walls wouldn’t burn, wooden floors and joists remained flammable. This 1884 cartoon depicts another tragic oversight: Fire departments didn’t yet have strong enough water pressure to reach the upper floors of high-rise buildings.

Photo credits: Bert Hansen Collection of Medicine and Public Health in Graphic Art, Ms. Coll. 67 of the Medical Historical Library, Cushing/Whitney Medical Library of Yale University

The post Editorial cartoonists were early U.S. public health advocates appeared first on Harvard Public Health Magazine.

Kennedy’s history of spreading disinformation about vaccines threatens to undermine decades of scientific progress and public trust. As the nation starts to grapple with the implications of this seismic shift, the role of accurate, evidence-based communication that actually reaches people has become more urgent than ever.

In this new landscape, social media creators have emerged as the frontline of public health communication. Often trusted more than traditional institutions, these creators wield significant influence over how health information is disseminated and understood by not only the masses but also the hardest-to-reach populations. Yet many creators have told me they lack the tools and training to verify and translate health information and instead rely on quick internet searches, which can inadvertently spread inaccurate content. This lack of accessible science creates a void that both unintentional misinformation and deliberate disinformation readily fill. Equipping creators to combat mis- and disinformation is no longer optional. It’s essential.

Health misinformation disproportionately affects youth, people of color, and low-income communities, who often rely on social media for accessible health information. A recent study from the Centers for Disease Control and Prevention revealed that misinformation has significantly decreased vaccination rates in some communities, contributing to the resurgence of preventable diseases like measles and whooping cough.

Additionally, social media users are more likely to encounter and believe health misinformation. We know that a majority of people who use social media for health advice reported in a 2023 poll hearing and believing at least one false claim about COVID-19 or vaccines, compared to only four in ten who don’t rely on social media for health advice. These vulnerable groups often lack the resources to verify the information they see online. Even well-intentioned creators—those not spreading deliberate disinformation—struggle to simplify complex, jargon-heavy science for their audiences. How can we empower them to share accurate, impactful health messages?

While tools like commercial AI can summarize content and fact-checking services can identify false claims, these often fall short when it comes to offering creators audience-specific, evidence-based material that’s ready for sharing. Inspired by the need for a more effective solution, people have been developing strategies to simplify complex research. AI tools that summarize and organize academic papers have cropped up in the past year.

An organization I founded, Science to People, is working on a tool called VeriSci that uses AI to transform peer-reviewed health studies into usable content, with a language model specifically fine-tuned to best practices in science communication. This fall, YouTube announced it is working on something similar for creators in its partner programs. With social media companies’ commitment to fact-checking and removing misinformation and disinformation now waning, the need for independent, publicly accessible tools that offer scientific information in digestible language is clearer now more than ever before.

The demand for accessible and trustworthy health information is clear. For example, a recent experiment demonstrated the power of providing accurate messaging on mental health on TikTok, where videos tagged #mentalhealth have drawn more than 44 billion views. The researchers offered influencers digital toolkits that contained evidence-based mental health content in everyday language across several topics. They found that the creators who received the toolkits were significantly more likely to include mental health content supported by research in their videos. They also uncovered significant impacts. In the treatment groups, TikTok videos featuring the provided content attracted more than half a million additional views after the intervention. A follow-up study of the comments on these videos showed that these viewers had improved mental health literacy.

With support, creators could expand this impact across many health topics, reaching millions with accurate, culturally relevant information. Imagine a mental health advocate sharing evidence-based strategies to manage anxiety, or a sexual health educator presenting reliable birth control options tailored to their audience. Providing creators with science-driven information has the potential to improve health literacy and make a measurable difference in underserved communities.

As we enter this new era, it’s time to recognize that the frontline of public health has shifted to social media, where creators are leading the charge in sharing health information. Supporting these creators with innovative, research-backed resources is essential for combating misinformation and protecting public well-being.

These digital communicators have become our new public health allies. Empowering them with the right tools can make a significant difference in reaching diverse and often underserved audiences.

Top image: елена калиничева / Adobe Stock

The post Social media is the new public health frontline. Let’s treat it that way. appeared first on Harvard Public Health Magazine.

Preparing cancer patients for difficult decisions is an oncologist’s job. They don’t always remember to do it, however. At the University of Pennsylvania Health System, doctors are nudged to talk about a patient’s treatment and end-of-life preferences by an artificially intelligent algorithm that predicts the chances of death.

But it’s far from being a set-it-and-forget-it tool. A routine tech checkup revealed the algorithm decayed during the COVID-19 pandemic, getting 7 percentage points worse at predicting who would die, according to a 2022 study.

There were likely real-life impacts. Ravi Parikh, an Emory University oncologist who was the study’s lead author, told KFF Health News the tool failed hundreds of times to prompt doctors to initiate that important discussion—possibly heading off unnecessary chemotherapy—with patients who needed it.

He believes several algorithms designed to enhance medical care weakened during the pandemic, not just the one at Penn Medicine. “Many institutions are not routinely monitoring the performance” of their products, Parikh said.

Algorithm glitches are one facet of a dilemma that computer scientists and doctors have long acknowledged but that is starting to puzzle hospital executives and researchers: Artificial intelligence systems require consistent monitoring and staffing to put in place and to keep them working well.

In essence: You need people, and more machines, to make sure the new tools don’t mess up.

“Everybody thinks that AI will help us with our access and capacity and improve care and so on,” said Nigam Shah, chief data scientist at Stanford Health Care. “All of that is nice and good, but if it increases the cost of care by 20 percent, is that viable?”

Government officials worry hospitals lack the resources to put these technologies through their paces. “I have looked far and wide,” FDA Commissioner Robert Califf said at a recent agency panel on AI. “I do not believe there’s a single health system, in the United States, that’s capable of validating an AI algorithm that’s put into place in a clinical care system.”

AI is already widespread in health care. Algorithms are used to predict patients’ risk of death or deterioration, to suggest diagnoses or triage patients, to record and summarize visits to save doctors work, and to approve insurance claims.

If tech evangelists are right, the technology will become ubiquitous—and profitable. The investment firm Bessemer Venture Partners has identified some 20 health-focused AI startups on track to make $10 million in revenue each in a year. The FDA has approved nearly a thousand artificially intelligent products.

Evaluating whether these products work is challenging. Evaluating whether they continue to work—or have developed the software equivalent of a blown gasket or leaky engine—is even trickier.

Take a recent study at Yale Medicine evaluating six “early warning systems,” which alert clinicians when patients are likely to deteriorate rapidly. A supercomputer ran the data for several days, said Dana Edelson, a doctor at the University of Chicago and co-founder of a company that provided one algorithm for the study. The process was fruitful, showing huge differences in performance among the six products.

It’s not easy for hospitals and providers to select the best algorithms for their needs. The average doctor doesn’t have a supercomputer sitting around, and there is no Consumer Reports for AI.

“We have no standards,” said Jesse Ehrenfeld, immediate past president of the American Medical Association. “There is nothing I can point you to today that is a standard around how you evaluate, monitor, look at the performance of a model of an algorithm, AI-enabled or not, when it’s deployed.”

Perhaps the most common AI product in doctors’ offices is called ambient documentation, a tech-enabled assistant that listens to and summarizes patient visits. Last year, investors at Rock Health tracked $353 million flowing into these documentation companies. But, Ehrenfeld said, “There is no standard right now for comparing the output of these tools.”

And that’s a problem, when even small errors can be devastating. A team at Stanford University tried using large language models—the technology underlying popular AI tools like ChatGPT—to summarize patients’ medical history. They compared the results with what a physician would write.

“Even in the best case, the models had a 35 percent error rate,” said Stanford’s Shah. In medicine, “when you’re writing a summary and you forget one word, like ‘fever’—I mean, that’s a problem, right?”

Sometimes the reasons algorithms fail are fairly logical. For example, changes to underlying data can erode their effectiveness, like when hospitals switch lab providers.

Sometimes, however, the pitfalls yawn open for no apparent reason.

Sandy Aronson, a tech executive at Mass General Brigham’s personalized medicine program in Boston, said that when his team tested one application meant to help genetic counselors locate relevant literature about DNA variants, the product suffered “nondeterminism”—that is, when asked the same question multiple times in a short period, it gave different results.

Aronson is excited about the potential for large language models to summarize knowledge for overburdened genetic counselors, but “the technology needs to improve.”

If metrics and standards are sparse and errors can crop up for strange reasons, what are institutions to do? Invest lots of resources. At Stanford, Shah said, it took eight to 10 months and 115 man-hours just to audit two models for fairness and reliability.

Experts interviewed by KFF Health News floated the idea of artificial intelligence monitoring artificial intelligence, with some (human) data whiz monitoring both. All acknowledged that would require organizations to spend even more money—a tough ask given the realities of hospital budgets and the limited supply of AI tech specialists.

“It’s great to have a vision where we’re melting icebergs in order to have a model monitoring their model,” Shah said. “But is that really what I wanted? How many more people are we going to need?”

Image: Viorika / iStock

The post Health care AI, intended to save money, turns out to require a lot of expensive humans appeared first on Harvard Public Health Magazine.

The post Elevator Pitch: Project S.A.R.A.H. appeared first on Harvard Public Health Magazine.

Camey Christenson, who worked with Father Joe’s Villages at the time, remembers conversations with the leaders building the new health information exchange (HIE). “They said, ‘Simmer down! We have enough to focus on with actually sharing health data,’” she recalls.

So her group joined with others and built what they called a “community information exchange” (CIE) and won a $1 million grant from San Diego’s Alliance Health Care Foundation to roll it out under 211 San Diego, the organization that runs a hotline providing information on health, disaster, and community services.

The hope, Christenson says, was to start “blowing the roofs off the silos of information.” Providers could give better care—and save everyone money—if they could see the rest of the “ecosystem” in which their clients lived. Providers could also see whether clients were following through on referrals, which organizations had capacity for more clients, and what interventions worked.

San Diego launched the first version of the CIE in 2014; today, it maintains records for around 380,000 people, or more than 10 percent of the county’s population. The CIE helps coordinate care across 138 partners, ranging from all the area’s major health insurers to LGBTQ organizations to the YMCA. The system can also alert providers to crisis moments, such as a client entering the shelter system or winding up in the emergency department.

Over time, the CIE has become more integrated with the HIE that originally excluded it. Today, health care providers can access the CIE’s records through a special tab in the HIE dashboard. (In 2022, San Diego 211 took over the local HIE, which now shares the same board and CEO as the CIE.)

The community information exchange is already improving outcomes. One 2017 study run by the University of California, San Diego found that enrollment in the CIE significantly reduced emergency medical transports among seniors, saving an estimated $1.3 million in medical costs across San Diego County over a three-year period.

The San Diego community exchange is a model of what a CIE can be, and it has sparked a spread of the concept. It coordinates and provides tools and mentorship for a national network of more than 30 other places building their own exchanges. San Diego 211 actually trademarked the term “Community Information Exchange” to head off vendors using the term in a misleading way, Christenson said; other similar efforts across the country use names like “community referral networks.”

One effort it’s assisting is in Chicago, where an exchange is being built by the Illinois Public Health Institute (IPHI), with support from the Chicago Department of Public Health and the Cook County Bureau of Economic Development. As in San Diego, Chicago’s exchange is initially focusing on people experiencing housing insecurity, but IPHI is building the system so it can include other populations. It’s also making it HIPAA compliant so that it’s prepared from the outset to handle medical data as it grows. Because the Chicago metropolitan area is not currently served by an HIE, the CIE will be the first data tool capable of capturing information about individuals who use the health care system the most. Officials expect to deploy the exchange next year.

Health care institutions are already key stakeholders. They want the CIE to succeed in large part because it has the potential to drive down costs. In particular, care organizations anticipate savings by diverting high-use individuals—many of whom are unhoused—from their emergency departments, says Waldo Mikels-Carrasco, who oversees the CIE program at IPHI and previously helped run a health information exchange in neighboring Indiana.

The 2010 Affordable Care Act pushed health care providers and insurers to shift their care approach to improving population health—a recognition, at least in part, that keeping patients healthy means they use less care, which is less expensive than solely focusing on patients after they fall ill. Many hospitals responded by employing case managers and community health managers to refer patients to services. But there was no way to track who was being sent where, nor to record what interventions worked. Many small social service organizations were overwhelmed, and hospitals quickly found they didn’t have the tools to make a population health approach work.

“All these nonprofits out there, little community-based organizations that are totally understaffed. They’re totally not set up to get a ton more referrals, and certainly not to provide data back,” says Mikels-Carrasco. A CIE, on the other hand, can show case managers which organizations have capacity for referrals and how well various interventions work.

Changes the Biden administration made to waivers under Medicaid’s Section 1115, which allows states to pilot new care strategies, also promise to bring more funding to such interventions if better records are kept on them. The Biden administration encouraged waivers that address health-related social needs, allowing states to reimburse providers for services like housing or food assistance that could keep Medicaid recipients healthier.

“If we could get information back, we could show that [a referral] helped the person, and then … CMS will reimburse us,” Mikels-Carrasco says.

The Robert Wood Johnson Foundation has provided funding to many data-sharing initiatives, including the ones in San Diego and Chicago. Hilary Heishman, a deputy director at the foundation, says because these exchanges can be built by a community from the ground up, they address many of the trust issues that hamper other health information initiatives. That can make them a tool for underserved communities to gather data on their own needs much faster than a top-down system could detect them. “Public health needs sources of information that are more driven by the community that they’re trying to support and serve,” Heishman says. Grassroots organizations are using CIEs to find ways to “use data to tell the story of how our community is doing and where the opportunities are, where the challenges are.”

These data initiatives can be a key tool for finding “patterns in the noise,” she says. And that’s “what public health should be doing all the time.”

Source images: Adobe Stock

The post Community information exchanges quench health data droughts appeared first on Harvard Public Health Magazine.

Targeting infectious diseases like influenza or lead exposure would have been easier: State laws required reporting those illnesses to public health officials. But asthma is a chronic disease that health care providers weren’t required to report. How could caseworkers find the children they wanted to help?

Clifford S. Mitchell, director of the office that led the program, first looked to Medicaid claims data since the asthma program was intended for low-income kids. But that data was only reported every six months, and then it had to go to a vendor to be analyzed—which meant as much as a year may have passed between a child’s visit to the hospital and the first outreach by a caseworker. By then, contact information for the families often no longer worked. If their children hadn’t just had an asthmatic crisis, parents often saw no need for a caseworker to visit their home. “We [had] a conundrum, because we have this service,” Mitchell remembers, “but we can’t really efficiently get to the people who are most likely to receive the service favorably and ask for it.”

He found a solution in a system that wasn’t created for public health at all: a health information exchange, initially set up in the mid-2000s to organize and share patients’ electronic health records between their health care providers. Maryland’s health information exchange, known as CRISP, received hospitalization reports every night. It could identify Mitchell’s target patients in real time.

It took some effort to work out technical and privacy issues to integrate CRISP’s data into systems used by the state health department and county-level case workers, and the program slowed dramatically during the COVID-19 pandemic. But on September 8, 2022, information about children eligible for the asthma program began moving automatically from the CRISP database to the county caseworkers who could reach out to families. Caseworkers were suddenly enrolling about 25 percent of children the system flagged for the program—a success rate roughly ten times higher than when they were working with Medicaid claims records.

“I am one lucky guy,” Mitchell says. He has the real-time ability to track an illness and deploy an intervention that most public health workers only dream about. He is also working in a state that has taken steps to make it especially easy to adapt a health information exchange (HIE) built primarily for health care providers so that it serves a public health function.

Maryland is one of 43 states that rely on HIEs for at least one public health function, according to a 2022 study published by the Civitas Networks for Health, a national collaborative of organizations working on HIEs. Providers, payers, and public health officials are gradually recognizing the whole health system, public and private, can benefit if they share tools with one another.

For many years, there have been two entirely separate conversations about data modernization in health care—one among public health agencies used to working with disease registries based on reporting requirements imposed on health care providers, and a parallel conversation among doctors, hospitals, and health plans working to streamline patient care. There’s a growing recognition, especially after the pandemic, that both communities are wrestling with the same question, and there are many ways they could work together—to make practitioners’ lives easier and to improve the nation’s health.

Using CRISP for poor kids with asthma would have seemed far-fetched when it began taking shape in about a decade earlier, because it was originally created to shuffle the medical paperwork of seniors. The data-sharing initiative spawned from the frustrations of John C. Erickson, the founder and then-chairman of a chain of elder care facilities called Erickson Senior Living. He didn’t like that when nursing home residents went to the hospital, the hospital’s doctors “would know nothing about them,” recalls David Horrocks, then Erickson’s chief information officer, “and then they come back, and our docs would know nothing about what happened in the hospital.”

Most information-driven industries were then rapidly adopting technology, like email, that’s now commonplace; medicine was still in the Stone Age, with practitioners relying on paper records and fax machines. Erickson thought there had to be a better way. In 2006, he enlisted Maryland’s three largest hospital systems to launch pilot programs to digitally share seniors’ medical records, and he put up $5 million of his own money to get it off the ground. Horrocks, as Erickson’s tech chief, was tasked with building the platform. Over several years, CRISP was developed.

It coalesced as lawmakers were trying to drag the health care industry toward the latest innovations of the information age. A major shot in the arm came as part of the economic stimulus package signed by President Barack Obama in 2009, which included the Health Information Technology for Economic and Clinical Health Act (HI TECH for short). The legislation codified a position in the Department of Health and Human Services (HHS) to coordinate health technology efforts and appropriated $35 billion over ten years to encourage the adoption of electronic health records—and HIEs to share records between providers.

None of these efforts paid much attention to public health; policymakers were focused on making providers more efficient, bringing down the cost of care, and improving patient outcomes. The federal government did not specify what form the HIEs should take, so they evolved in many different ways. Some were for-profit companies, some were independent nonprofits, and some sat inside state or county agencies. Some covered entire states while others were organized at a regional level or only included a subset of providers. Some were more focused on networking hospitals; others were more focused on serving health insurance companies. Collaborations between different stakeholders were often uncomfortable, because they required competing providers to share data with one another, as well as greater transparency between providers and insurers. In some states, privacy laws had to be modified to facilitate information exchange. And in some parts of the country, these challenges were so insurmountable that they are still not served by an HIE.

Maryland was fertile ground for building an especially robust HIE. The state already had a unique payment model in which an independent commission appointed by the governor sets payment rates for hospitals. The administration of then-Governor Martin O’Malley helped get all the state’s hospitals on board with CRISP, in 2009, and provided funding to keep it going. Maryland made a major change to its payment system in 2014 that allowed hospitals to keep more revenue by keeping patients healthier and using fewer services. That in turn meant hospitals needed better data tools to track their patients, and CRISP was ready to fill this need.

CRISP gave every hospital and provider a tracking dashboard and even real-time alerts when a patient went to the ER or was admitted to the hospital. Over time, these tools expanded to gather data from medical and mental health case managers. “If there’s somebody who can get help so they’re not admitted every week to the hospital, going from ER to ER, finding out who that person is and helping them becomes a really good thing for the health care system,” says Joshua Sharfstein, a medical doctor and professor at Johns Hopkins Bloomberg School of Public Health, who was Maryland’s health secretary as this system was being created in 2014.

For public health workers, CRISP can map hospitalizations for conditions like asthma to identify hot spots with much more precision than previous tools—and with actionable information. CRISP also runs the prescription drug monitoring platform for the state Department of Health. Its role is so central to Maryland’s health system that it is recognized as what’s known as a Health Data Utility, bedrock infrastructure that provides health information to many players in the state’s health care and public health ecosystem.

HIEs like CRISP, if widely deployed, could radically improve the practice of public health in the US, says Craig Behm, CRISP’s CEO. For instance, heat maps are a mainstay of public health. But they identify very broad geographic hotspots, which Behm says often doesn’t tell public health workers very much they don’t already know. “The [bigger] problem is, ‘What are the compounding factors and health related social needs and different outcomes? And where do people get care? Those details are available!… We need to go deeper now and we can.”

That level of information is available with CRISP. But “public health has been so undervalued and underfunded for so long,” Behm says, “Public health [officials are] having trouble even understanding what modern infrastructure could do for them.”

The COVID pandemic made painfully clear the high cost of underinvestment in public health data infrastructure. Federal budget allocations also show what a low priority such resources have been for policymakers. In 2009, for example, Congress appropriated $35 billion to advance data tools for providers and insurers. But it didn’t earmark money for CDC data modernization until 2019 and has only appropriated a total of $1.35 billion for the effort over the last four years.

The COVID pandemic made clear just how useful CRISP could be for public health, Behm says. Very early in the pandemic, CRISP was processing massive amounts of data every night, tracking case reports, hospitalizations, and deaths. Lab workers were sending incomplete demographic data to local Departments of Health, but CRISP could match the health department registries with its own data set, tracking in real time essential data like the race and ethnicity of infected Marylanders. When public health workers needed phone numbers to do contact tracing, Behm recalls, “I was like, ‘Well, great! I have all the phone numbers in the world!”

In the scramble to respond to COVID, other states’ public health workers also turned to HIEs as partners—often in ways they never had before. HIEs were glad to help, but those relationships became strained as the crisis passed, says Jolie Ritzo, Civitas’s vice president for strategy. “The challenges feel bigger than they were,” Ritzo says.

HHS is now steering the next evolution in the health information ecosystem, laying out standards and certifications for national information exchanges that will allow records to follow a patient from one HIE to another. At the same time, the CDC is leading its own Data Modernization Initiative and has signed contracts for hundreds of millions of dollars with Palantir to build a data system of its own.

A major open question is how these two efforts will interact. Many HIE leaders hope that existing data exchange infrastructure can link the information networks many providers already use regularly with upgraded data systems for public health officials. Erica Galvez, CEO of Manifest Medex, California’s largest HIE, and a former official in the health technology office at HHS, says she believes many public health departments will look to existing HIEs as an “on ramp” to the nationwide network.

Claudia Williams, who was Manifest Medex’s founding director, says HIEs have already done a lot of work to manage the complicated relationships between stakeholders at the local, regional, and state levels. It would be in everybody’s interest, she thinks, for the CDC to try to engage with the HIE infrastructure that exists rather than trying to develop an entirely separate information system.

“The better place to start is to strengthen and incentivize what goes on at the state level,” says Williams, who also served as senior advisor of health innovation and technology in the Obama administration and is now the chief social impact officer of UC Berkeley’s School of Public Health.

But fewer than half the states have HIEs ready to take on full public health functions at this point, says Horrocks, who left CRISP in 2022 to build a state-wide HIE in New York. (Horrocks also coauthored the nationwide study of HIEs published by Civitas in 2022.) Still, he believes they’re the best structure for public health to rely on in the long term.

“HIEs do something that’s very unique: We hold fully identified data in trust, and that lets us put datasets together,” Horrocks said. “I don’t know who else can do that. We’re not going to trust the … the state government, and you’re certainly not going to trust CDC to hold everybody’s fully identified data. And, I don’t think you’re going to trust Google to do it, either, or Palantir. I think that’s a role that these HIEs can uniquely fill.”

Image: berCheck / Adobe Stock

The post Weaving data into the fabric of public health appeared first on Harvard Public Health Magazine.

Public health experts extolled the promise of artificial intelligence to solve longstanding public health problems in a panel discussion on AI but also raised concerns about its potential for exacerbating inequity.

An example of the technology’s potential was how AI was used by the Chicago Department of Public Health to make outbreak predictions for diseases, such as measles. The technology could be applied widely to forecast and prevent food-borne illnesses, said Micky Tripathi, acting chief AI officer at the Department of Health and Human Services. The United States has vast discrepancies in regulatory approaches at different levels of government, as well as in the size and sophistication of local public health staffs. “How do we figure out how these technologies can be democratized?” he asked. Minimizing such gaps is a primary concern for HHS as it prepares a strategic plan for AI.

Tripathi made his remarks as part of the panel “Making AI a Lifesaver,” held on October 8 at the Johns Hopkins University Bloomberg Center in Washington, D.C. The panel was cosponsored by Harvard Public Health, Global Health NOW, and Hopkins Bloomberg Public Health.

Another panelist, John Auerbach, senior vice president at the global consulting firm ICF, noted that AI could help small public health departments by streamlining tasks like filling out forms or using AI to decide which restaurants to inspect. But he asked, “how do you compensate for the fact that there’s not going to be sophisticated data capacity in a lot of locations?” He said using AI equitably might require a “slow” and “simple” approach oriented more toward everyday tasks than visionary applications.

The panelists delved into AI’s potential to shake up health care to improve both efficiency and outcomes of care. Possible uses range from vaccine and drug development to medical diagnostics and disease screening to providing personalized health messaging to patients. Right now, though, AI is primarily popping up in assisting diagnostics in radiology and in routine administrative applications. While there are myriad examples of AI pilots, things that scale are far less evident.

Disparities in health care resources hamper the equitable use of AI. For one, developing AI applications is costly. A single AI model can cost upwards of $1 million, beyond the reach of under-resourced health departments and hospital systems. One panelist said a dean at Stanford University told him the school had spent between $3 and $5 million on a single AI implementation. “Nobody can scale that for implementation, right?” said Jesse Ehrenfeld, a radiologist and immediate past president of the American Medical Association. Another panelist, Elizabeth Stuart, a biostatistician at the Johns Hopkins Bloomberg School of Public Health, noted that AI continues to draw on limited data sets, a problem in both research on and application of the technology. “We need to be really conscious of who is not in the data that we are using to develop these models, and then the implications of that for use in various settings,” Stuart said.

There’s already a practical divide around AI emerging in public health departments. One survey of local health departments in the U.S. found that among those serving a population of over 500,000, 24 percent were already engaging in AI or had plans to do so, versus only five percent of smaller departments.

Avoiding an AI double standard is possible, the panelists said. One way to expand access is to develop AI platforms that are openly accessible and can seamlessly integrate with different health data sources and software across different care settings.

Several efforts are underway to bridge the AI gap. In January, the National Science Foundation unveiled the National Artificial Intelligence Research Resource pilot, a two-year program aimed at lowering the barriers for innovation in AI. The program connects successful applicants to infrastructure resources for developing new AI models.

Voluntary academic-led collaborations are also accelerating the adoption of AI in health care. Institutions such as the University of California health systems and Duke University are partnering with various health care providers to share AI research, validation practices, and standards for AI use. Tripathi said public-private partnerships in AI are essential, and because of the U.S.’s federalist nature, AI policy related to public health is certain to vary by state.

The panelists broadly agreed that there needs to be more transparency in how AI is being used. For starters, noted Ehrenfeld, better visibility into AI will help flag flaws that lead to inequity as well as make AI a more effective tool for public health workers. Stuart noted the clear need for training on AI’s ethical issues and applications presents a big opportunity for schools of public health and medical schools.

To counter AI’s transparency challenges, policymakers are working to improve the regulatory structures. Last October, the Biden administration signed an executive order to accelerate the ethical management of AI’s risks. It tasked HHS with drafting an AI action plan to oversee responsible AI implementation in healthcare.

Tripathi said strategies include a certification system for companies that sell electronic health records. To gain this imprimatur, vendors who build an AI application need to disclose the model’s training data set, maintenance strategies, and validation methods. The published information is “basically a nutrition label,” he said. If every vendor in the U.S. gained certification, it would cover 96 percent of hospitals and 78 percent of physician offices nationwide.

Tripathi noted that HHS plans to release its full strategy for AI in January.

One thing that appears unlikely is Congressional action to help standardize AI policy. “Congress doesn’t appear to be on the verge of having some national stance for any of the states,” Tripathi said. “The notion of states’ rights is something that, if anything, is becoming even more ingrained in this kind of policy.”

The post The path to more equitable AI appeared first on Harvard Public Health Magazine.

—Leah Rosenbaum

(Study in Hepatology, March 2024)

The post A blood test to predict fatty liver appeared first on Harvard Public Health Magazine.

ACEGID’s director, Christian Happi, wants his splendid HQ to deliver a bold message: Africa has transcended its colonial past, and African scientists are capable of leading the world. In fact, they already are.

Expand Image”>

This year, under Happi’s guidance, ACEGID is ramping up Sentinel, one of the first-ever public health programs to use cutting-edge CRISPR gene editing to surveil for diseases. It’s based on a diagnostic panel called CARMEN, short for Combinatorial Arrayed Reactions for Multiplexed Evaluation of Nucleic acids. Launched in 2020 by the Broad Institute and celebrated in Nature, CARMEN involves extracting viral RNA from patient blood plasma, making copies of it, then dropping these copies onto rubber chips, each one slightly larger than a smartphone and etched with thousands of “microwells.” CARMEN is powered by CRISPR’s most powerful protein, Cas13, which looks for specific viral sequences for a variety of diseases. Happi’s research teams use it to search for 16 blood-borne pathogens, among them Ebola, Lassa fever, West Nile virus, and Mpox.

ACEGID’s research teams collect patient samples at four rural clinics, then transport them to Ede. (It is also testing the technology in Sierra Leone.) The facility aims to ramp up to testing as many as 2,000 by the end of the year—a 12-fold increase, if it works. At that scale, experts believe, health authorities could start to discern where, say, Lassa is hot, or where an outbreak of Mpox is imminent, and whether public warnings and quarantines are needed. Currently, there’s no way to proactively monitor for diseases; local clinics simply alert authorities when they have diagnosed a patient.

Happi has a deep resume. A Harvard postdoc and a professor of molecular biology and genomics at Redeemer’s, in 2014 he diagnosed the first case of Ebola in Nigeria, and in 2020 directed the first sequencing of the SARS-CoV-2 genome by a lab in Africa.

At 56, he is a supremely confident man who still carries a small-town folksiness from growing up in rural Cameroon, where his father drove a bus. If you step into his office, you’re in for a geopolitical sermon. “For decades,” he says when I stop by on a sweltering afternoon in March, “researchers have been taking blood samples back to Europe, back to America. And it doesn’t work. We have to wait months for the results.”

Even as the Nigerian economy sputters (the inflation rate is currently 34 percent) and as the country’s leaders clamber for foreign investments, Happi has managed to bring in about $100 million in donations over the last five years—from the World Bank, the Elma Foundation and the Skoll Foundation, among others. These days, he juggles research with fundraising for ACEGID, traveling the globe making funding appeals. Whether he’s in London or Paris or Singapore, he wears a shirt in an African print to drive home his principal message: The world needs to shed colonial ideas about epidemiological research in Africa.

Happi is trying to make Africa scientifically self-sufficient. Since ACEGID’s founding in 2014, he’s trained molecular biologists from all 55 African countries except for Eritrea. As we speak, young scientists from Sao Tome and Lesotho are in the next room, learning about genomics. He sees each student as the seed of a revolution, and he sees himself as their stern taskmaster. “If you don’t go back to your country and share what you learned,” he told them earlier that day, “you’ve failed.”

Eventually, Happi tells me, Africa won’t just have a host of qualified scientists. It will, he says, have better laboratories. It will have more factories that make reagents and test tubes. “We shouldn’t have an inferiority complex,” he says. “Let us realize our dreams.”

I look up from my notebook and register what is by far the biggest work of art displayed in the whole room: a photograph that captures an ebullient Barack Obama shaking Happi’s hand.

Expand Image”>

Adversity drives Happi’s ambition. When he was eight years old, he got malaria for the umpteenth time. His mother carried him to the local health clinic in Cameroon, and on the way back the pair stopped to rest under a tree. Happi’s siblings had also endured malaria multiple times, and some neighbor kids had died from it. He asked his mom, “Why isn’t there a cure for this disease?” When she said she didn’t know, Happi told her ‘When I grow up, I will find a cure.’”

Arguably, the prediction was a long shot. Happi was the sort of boy who, when sent to the store for milk, would stop at a soccer field, set the milk behind the goal, and play for a couple of hours. He lived amid strict discipline, though. At his Catholic school, he says, the teachers hit kids with sticks for making an error reciting multiplication tables. “And then they sent you home with a note,” he says, “and your parents whipped you again. The beatings worked,” he adds, chortling. “I always got 100 percent.”

During breaks from school, he worked as a laborer. At the local market, he transported bananas and plantains on his head; he carried blocks of concrete at construction sites; he picked cassavas and yams on local farms. In his spare time, he read voraciously, borrowing books from a wealthy friend. He was especially drawn to French comic books and American police stories.

He was less fond of being a postdoc in America. Happi arrived stateside with a PhD in molecular parasitology from the University of Ibadan, in Nigeria. But as he tells it, many of his colleagues regarded him condescendingly and dismissed Africa, saying of its people, “‘You’re not smart enough to make discoveries.’ For me,” he says, “those statements were catalytic.”

In 2011, Happi moved back to Nigeria, resolved to prove he could do world-class science on his home turf. His moment came in 2014, just weeks after he founded ACEGID along with Pardis Sabeti, a computational geneticist at the Broad Institute and a professor of immunology and infectious Diseases at the Harvard T.H. Chan School of Public Health.

One evening that July, while unwinding at home with his young family, he got a call saying a passenger who’d flown into Lagos that day had died with a fever. Could he sequence this man’s blood sample and see if it was Ebola?

Happi’s lab was 90 minutes away, over a rough, unlit road. It was a level 2 biosafety lab, meaning it lacked the high-end equipment required to handle highly lethal specimens such as Ebola, which should be handled in a level 4 context. But Happi thought of the conversation he’d had with his mother long ago, under the tree, about malaria. He remembered the kids who died in his village. He told his wife, Anise, “If I don’t make it back, take care of the children.”

He drove to the lab and worked until dawn to sequence the virus. Then he called Nigeria’s Minister of Health and said, “We have Ebola in the country.” Contract tracing kicked in. As the rest of West Africa endured 11,000 deaths from the virus, Nigeria, population 227 million, lost but eight people.

Expand Image”>

Expand Image”>

It’s another scorching day in March, and I’m watching ACEGID researchers wrestle a goat outside a mud house in a small, dusty community. We’re a three-hour drive from Ede, in Ondo State, a Connecticut-sized territory that has seen 16 deaths from Lassa fever so far this year, making it one of Nigeria’s hotspots for the disease, which is transmitted to humans largely through contact with rodent excretion. Goats contract and also help spread Lassa, which is expected to extend its deadly reach amid climate change.

Lassa fever killed Olesunkanmi Eniayewu, a 40-year-old cocoa farmer in Ondo, in 2019. His father, Oladipo, 75, tells me his son came home from harvesting and then displayed all the hallmark symptoms—a fever, dizziness, aching joints. After 12 days in the hospital being treated with the antiviral drug ribavarin, fluid welled up around his son’s heart, a common Lassa symptom. When doctors operated to remove the fluid, he died.

Oladipo Eniayewu is a frail and elegant bald man who moves about with a cane, his sight severely compromised by cataracts. When we meet, he’s looking dapper in a brightly printed buba and sokoto made from Ankara fabric, but his speech is quiet and reluctant. “He was my firstborn,” he says of his son, “and he was so generous. If I needed money, he’d give it to me. Wherever I wanted to go, he carried me in his car. I never thought he would die before me.”

Expand Image”>

Even though Lassa kills far more people than Ebola—about 10,000 die of Lassa each year—it remains sparsely studied. Scientists have sequenced it less than 10,000 times, according to the National Library of Medicine. (COVID, in contrast, has been sequenced over 15 million times.) As such, it’s a priority for ACEGID. In 2021, the center hired the director’s wife, Dr. Anise Happi, a former Harvard postdoc and University of Ibadan professor, to head up zoonotic research on Lassa. Her teams catch and dissect rats, both in Ondo and in Ebonyi, and the genetic material from Lassa-bearing animals figures in CARMEN tests: It’s the control against which blood samples are compared for Lassa infection.

Her teams are also taking blood samples from rodents—one study found that 50 percent of all rats they studied were Lassa-infected.

Anise Happi is currently working on a vaccine for Lassa, as are about 30 other research teams worldwide. She’ll launch the second phase of human testing in October. Still, as she sees it, Lassa will pervade Nigeria for a long time coming. “It’s not just in rats and humans,” she says. “Our research has shown that it’s also in lizards. It’s in goats. It’s in monkeys. If it just mutates a little bit more, it could become inhalatory. It’s not a disease you can easily control.”

In Ondo, many people tell me that pharmacists are insufficiently educated on Lassa; they often cause Lassa victims harm by giving them drugs used for malaria and typhoid fever. Meanwhile, Ondo’s myriad Pentecostal churches frequently champion prayer as the best treatment for disease. One morning, I meet 56-year-old Alaba Aladesanwa, and she tells me that in 2018 when she was delirious with Lassa, she went directly to her church and then lay there, sleeping and hallucinating, for five days as her fellow congregants prayed over her. “I thought I was going to die,” she says, “so I got a ride to the hospital on a motorcycle. Then my pastor called and asked me to come back to the church.” Aladesanwa refused him. She stayed in the hospital and recovered.

Her experience is not unusual, says Michael Olonite. “We Christians have a hard time accepting that science is the best course. I think Muslims are better at it,” he tells me while we’re sitting under a giant shady tree in the village of Ogbese. Olonite is a Catholic catechist who nearly died from Lassa in 2019. While in the hospital he met a fellow Lassa patient, a Muslim imam named Jimoh Tajudeen. The two men are now friends, and they work as Ogbese’s liaisons for ACEGID. When the center’s scientists need to approach, say, a local farmer, to test his goats for Lassa, they make introductions.

Tajudeen stops by to linger with us under the tree. He tells me how, when he contracted Lassa, “My whole body was aching. I was vomiting blood. The local pharmacist prescribed medicine for typhoid fever. This happens all the time.”

Eventually, the two clerics climb onto Olonite’s motorcycle to ride off to the local market, to talk to their neighbors about Lassa and its threats.

Expand Image”>

Expand Image”>

Happi arrives at ACEGID one March morning to find the building dark. He stalks out to the power station, where, amid 46 solar inverters (“the biggest array in Nigeria,” he claims), an ACEGID worker is cradling a burnt electrical cord in his hands, puzzling over what caused a minor power surge. “The system is still working,” he assures Happi. The tech explains they’ve shut it down as a precaution, until they figure out what torched the cord.

Happi is unimpressed. “If the system is working,” he asks the guy, “then why don’t we have power?” The testing equipment needs constant power, and in West Africa brownouts can happen hourly. That’s caused blood samples, which are optimally refrigerated at -80 degrees Celsius, to spoil. It’s also induced Happi to install solar panels at his remote clinics.

The director has a reputation for being a hard driver. “He’s, uh, very demanding,” says Philomena Eromon, his lab manager and former graduate student, laughing in a way at once affectionate and overwhelmed. Still, she seems to adore her boss. She is forever bringing him cold bottles of water. “Hydrate,” she says. “We need you.”

Later that day inside the lab, two scientists are staring at a $110,000 machine called a Fluidigm C1, whose screen bears the message “System Verification in Process.” The scientists are poised to insert a CARMEN panel carrying 96 samples into the machine, which will then look for genetic matches for eight separate pathogens, but the machine had to be shut down because of a power issue. It takes more than 10 minutes to come back online. Then the scientists find the lab has run out of some reagents needed for the tests. It will take a month for Broad to ship them and two months more for the institute to check a previous CARMEN test. The March test I witnessed finally wrapped up in May.

ACEGID has excellent technical equipment, but servicing it creates challenges not uncommon in Africa. Eromon, the lab manager, has a sample preparation machine that needs a new motor. She paid $30,000, up front, for the part late last year. Seven months later, the part had arrived but she was still waiting for a technician to come from South Africa to install it. A prominent African genomicist cautions against reading too much into ACEGID’s logistical challenges. “ACEGID is one of the best research groups I know in Africa,” says Tulio de Oliveira, the director of the South Africa-based Centre for Epidemic Response and Innovation. De Olivera and his team identified the Beta and Omicron variants of COVID.

Like Happi, de Oliveira is tired of questions about how science in Africa compares to the rest of the world. “Look, if you want to critique African science, you need to break things down,” he says. “We probably are not going to send a rocket into outer space, but when it comes to epidemics–no one knows more about them than we do. We’ve always had Ebola. We’ve always had Lassa fever. Add that with the latest technology and, yes, we can become leaders.”

Expand Image”>

Expand Image”>

Some of the challenges come from pushing something new. “There’s a lot of logistical issues,” says Elyse Stachler, the lead staff scientist on the CARMEN team at the Broad Institute in Cambridge, Massachusetts. “There’s also a lot of procurement issues getting the reagents needed to run samples. It’s not always easy to get them shipped to Nigeria. And so we’re now just finally starting to build up that capacity.”

CRISPR itself only surged into the headlines in 2020, when the discoverers of the CRISPR “scissors” (a Cas9 protein) won the Nobel Prize in Chemistry. Because scientists can use it to remove, insert, and also find genetic material, it’s frequently heralded as a potential “miracle fix” to life-threatening diseases. At the March 2023 International Summit on Human Genome Editing, in London, an American woman named Victoria Gray proclaimed that experimental CRISPR therapy had zapped her sickle cell disease and enabled her to “dream again without limitations.” Last December, the FDA approved two gene therapies for sickle cell and there’s now rising hope that gene therapy could also liberate patients afflicted with cystic fibrosis, hemophilia, and other illnesses.

But new technologies always face obstacles. The treatment that helped Gray is priced at $2.2 million, not counting the cost of hospital stays, and the price will likely be high for other diseases as well. CRISPR-based monitoring of diseases—what Happi is doing—will require high upfront investments but once in place could benefit multitudes at a low cost-per-person. Proponents see it as a powerful alternative to standard polymerase chain reaction (PCR) testing because it could enable researchers to test scores of samples in a single run. Last year, an article in the American Journal of Public Health noted that it could be especially helpful to “low-income minority communities, which often suffer disproportionate rates of infectious diseases,” in part because “these viruses are challenging and expensive to track.”

But CRISPR-based testing for COVID-19 is not yet commercially viable, despite optimistic predictions at the start of the pandemic. “It’s not clear to me how much sense it makes to use CRISPR. Why not use just the PCR?” asks Elena Ivanova Reipold, the technology innovation lead at FIND, a Swiss NGO working to expand access to reliable diagnostics. She says even though PCR is relatively inexpensive, low- and middle-income countries often struggle with logistical challenges when transporting blood samples. “What’s the point of having technology,” she asks, “if you can’t respond to an outbreak on the periphery of the country?”

Andrew Subica, the lead author on the American Journal of Public Health article, sees things differently. “Nigeria and Sierra Leone are where you need to be applying surveillance technologies the most,” says Subica, an associate professor of social medicine, population, and public health at University of California, Riverside. He adds that “unless someone figures out a way to make these technologies work in a given low-income country, they will never be applied there, and low-income areas will never get any kind of benefit from cutting-edge technologies.”

Subica sees Happi’s project as a golden opportunity. “It’s surprisingly rare,” he says, “that you have a new technology and you get to test it where all these issues need to be worked out.”

Disease science is a suspenseful process, its experiments often fraught with risk. Christian Happi has, quite possibly, more bravery than anyone else in genomics—not many scientists would have done what he did to sequence Ebola. But as he embarks on his work with CRISPR, he’s in the deep of the night all over again. And we’re just going to have to wait until dawn to see how it works out.

Adeleye Roseline Funke provided translation from Yoruba to English.

Top image: Christian Happi enters a lab at ACEGID in Ede, Nigeria in March.

The post The cutting edge of CRISPR is in Nigeria appeared first on Harvard Public Health Magazine.