But they didn’t.

Federal law requires government resources to be accessible to patients with disabilities. But patients can’t easily enlarge the text. “They all learned to get strong reading glasses and magnifying glasses,” said Sheffield, who retired in early October.

The difficulties are everyday reminders of a dire reality for patients in the VA system. More than a million patients are blind or have low vision. They rely on software to access prescriptions or send messages to their doctors. But often the technology fails them. Either the screens don’t allow users to zoom in on the text, or screen-reader software that translates text to speech isn’t compatible.

“None of the systems are accessible” to these patients, said , executive director of the .

Patients often struggle even to log into websites or enter basic information needed to check in for hospital visits, Overton said: “We find our community stops trying, checks out, and disengages. They become dependent on other individuals; they give up independence.”

Now, the developing VA medical record system, already bloated by outsize costs, has been until June 2023. So far, the project has threatened to exacerbate those issues.

While users in general have been affected by numerous incidents of downtime, delayed care, and missing information, barriers to access are particularly acute for blind and low-vision users — whether patients or workers within the health system. At least one Oregon-based employee has been offered aid — a helper assigned to read and click buttons — to navigate the system.

Over 1,000 Section 508 complaints are in a backlog to be assessed, or assigned to Oracle Cerner to fix, Veterans Affairs spokesperson . That is part of federal law guaranteeing people with disabilities access to government technology.

Hayes said the problems described by these complaints don’t prevent employees and patients with disabilities from using the system. The complaints — 469 of which have been assigned to Oracle Cerner to fix, he said — mean that users’ disabilities make it more difficult, to the point of requiring mitigation.

The project is under new management with big promises. North Kansas City, Missouri-based developer Cerner, which originally landed the VA contract, was recently taken over by database technology giant Oracle, which plans to overhaul the software, company executive Mike Sicilia said during a September Senate hearing. “We intend to rewrite” the system, he said. “We have found nothing that can’t be addressed in relatively short order.”

But that will happen under continued scrutiny. Rep. Mark Takano (D-Calif.), chair of the House Veterans Affairs Committee, said his panel would continue to oversee the department’s compliance with accessibility standards. “Whether they work for VA or receive health care and benefits, the needs of veterans must be addressed by companies that want to work with the VA,” he said.

Takano, along with fellow Democrats Sens. Bob Casey of Pennsylvania and Jon Tester of Montana, sent a letter Oct. 7 to VA Secretary Denis McDonough noting the significant gaps in the agency’s systems, and urging VA to engage with all disabled veterans, not merely those who are blind.

VA was alerted early and often that Cerner’s software posed problems for blind- and low-vision users, interviews and a review of records show. As early as 2015, when the Department of Defense and VA were exploring purchasing new systems, the National Federation of the Blind submitted letters to both departments, and Cerner, expressing concerns that the product would be unusable for clinicians and patients.

Alerts also came from inside VA. “We pointed out to Cerner that their system was really dependent on vision and that it was a major problem. The icons are really, really small,” , a VA anesthesiologist who participated in one of the agency’s user groups to provide input for the eventual design of the system.

The Cerner system, he told the agency and KHN, is user-unfriendly. On the clinician side, it requires multiple high-resolution monitors to display a patient’s entire record, and VA facilities don’t always enjoy that wealth of equipment. “It would be very hard for visually impaired people, or normal people wearing bifocals, to use,” he concluded.

Before the software was rolled out, the system also failed a test with an employee working with a team at Oregon’s devoted to helping blind patients develop skills and independence, said Carolyn Schwab, president of the American Federation of Government Employees Local 1042.

In the testing, the system didn’t work with adaptive equipment, like text-to-speech software, she said. Despite receiving these complaints about the system, VA and Cerner “implemented it anyway.” Recently, when a regional AFGE president asked VA why they used the software — despite the federal mandates — he received no response, Schwab said.

Some within the company also thought there would be struggles. Two former Cerner employees said the standard medical record system was getting long in the tooth when VA signed an agreement to purchase and customize the product.

Because it was built on old code, the software was difficult to patch when problems were discovered, the employees said. What’s more, according to the employees, Cerner took a doggedly incremental approach to fixing errors. If someone complained about a malfunctioning button on a page filled with other potholes, the company would fix just that button — not the whole page, the employees said.

VA spokesperson Hayes denied the claims, saying the developer and department try to address problems holistically. Cerner did not respond to multiple requests for comment.

Accessibility errors are as present in private sector medical record systems as public. Cerner patched up a bug with the Safari web browser’s rendering of its patient portal when the Massachusetts Institute of Technology’s student clinic threatened legal action, the former employees said. (“MIT Medical does not, as a general practice, discuss individual vendor contracts or services,” said spokesperson David Tytell.)

Legal threats — with and routinely facing lawsuits — are the most obvious symptom of a lack of accessibility within the U.S. health care system.

Deep inaccessibility plagues the burgeoning telehealth sector. A from the American Federation for the Blind found that 57% of respondents struggled to use providers’ proprietary telehealth platforms. Some resorted to FaceTime. Many said they were unable to log in or couldn’t read information transmitted through chat sidebars.

Existing federal regulations could, in theory, be used to enforce higher standards of accessibility in health technology. The Department of Health and Human Services Office for Civil Rights issued guidance during the pandemic on making telehealth technologies easier to use for patients with disabilities. And other agencies could start leaning on hospitals, because they are recipients of government dollars or federal vendors, to make sure their offerings work for such patients.

That might not happen. These regulations could prove toothless, advocates warn. While there are several laws on the books, the advocates argue that enforcement and tougher regulations have not been forthcoming. “The concern from stakeholders is: Are you going to slow-walk this again?” said , director of government relations at Powers Law, a Washington, D.C., law firm.

Building in accessibility has historically benefited all users. Voice assistance technology was originally developed to help blind- and low-vision users before winning widespread popularity with gadgets like Siri and Alexa.

Disability advocates believe vendors often push technology ahead without properly considering the impact on the people who will rely on it. “In the rush to be the first one, they put accessibility on the back burner,” said , a disability rights attorney with Brown, Goldstein & Levy, a civil rights law firm.

Ñî¹óåú´«Ã½Ò•îl Health News is a national newsroom that produces in-depth journalism about health issues and is one of the core operating programs at KFF—an independent source of health policy research, polling, and journalism. Learn more about .

USE OUR CONTENT

This story can be republished for free (details).

• Read Tahir’s “Embedded Bias: How Medical Records Sow Discrimination”

Ñî¹óåú´«Ã½Ò•îl Health News is a national newsroom that produces in-depth journalism about health issues and is one of the core operating programs at KFF—an independent source of health policy research, polling, and journalism. Learn more about .

USE OUR CONTENT

This story can be republished for free (details).

Since 2014, health systems around the country have partnered with Apple to tap into the mountains of data the company’s devices generate from patients. But most are still experimenting with these tools. While some doctors appreciate seeing records of home-monitored blood pressure, exercise and the like between visits, for others the data is more of a burden than an asset.

Over 100 types of data are available in Apple’s health app through iPhone, Apple Watch and third-party apps. In June, Apple said patients whose doctors work with one of the participating in the new feature will be able to send them tracked data like , .

Some see great promise in building “pipes” between a patient’s phone and the health records viewed by their clinicians. Apple is “democratizing the flow of health data” between doctors and patients, said Anil Sethi, a former Apple health director and current CEO of , a startup that manages health data for cancer patients.

But Apple’s announcement was shrouded in ambiguity and short on particulars. The company would not provide a complete list of the data patients can share with doctors and declined to comment for this article. Previous Apple moves to get more data into the hands of doctors have been announced with great fanfare, but questions remain as to how many health care providers are using the data and to what effect, and whether success stories are the norm or outliers. To date, rigorous studies showing clear health benefits from monitoring these types of data .

Although Apple has built pipes enabling patients to share growing amounts of data with medical professionals, it’s unclear how much data flows through them.

In 2014, Apple released HealthKit, a tool enabling health systems to pull in patients’ health data, with their permission, from their iPhones. At the time, then-Mayo Clinic CEO John Noseworthy this would “revolutionize how the health industry interacts with people.” But a Mayo spokesperson told KHN the system’s use of HealthKit is now limited.

Cedars-Sinai Medical Center in 2015 that, through HealthKit, more than 87,000 patients had been able to share their data, an arrangement Cook on a quarterly earnings call. A Cedars-Sinai spokesperson declined to comment on what became of this project.

Even Apple’s attempts to use its own employees’ app data to improve their medical care have yet to pan out. The that an Apple initiative testing a new primary care service for doctors to monitor Apple employees’ health through their devices had stalled. The company said many of the story’s assertions were inaccurate.

There have been a few reports of success. Ochsner Health in Louisiana reported that patients in a hypertension management program that provided health coaching while monitoring blood pressure data from mobile phones were more likely than a control group , . The health system now also has remote monitoring programs for diabetes, chronic obstructive pulmonary disease and expectant mothers, an Ochsner spokesperson said.

And Epic, the nation’s largest health records company, said that over 100 of its large health system clients are using Apple HealthKit to capture data from home monitoring devices like blood pressure cuffs.

But patient-generated data has not been widely adopted in health care, said Dr. Benjamin Rosner, an associate professor of medicine at the University of California-San Francisco. He and others point out major hurdles.

One, Rosner said, is that evidence is that monitoring such data improves health.

Another is that doctors generally aren’t reimbursed by health insurers for reviewing data that patients collect at home.

“In America, we generally pay doctors and health systems to see patients in front of them and do things to them when they show up,” said Matthew Holt, a health technology startup adviser.

In instances in which doctors can be reimbursed for remotely monitoring patients, like those with certain chronic conditions, the payment is usually low, Rosner said.

And many doctors already feel inundated with patient health information and electronic health record tasks.

“Primary care doctors are overwhelmed by their inboxes,” said Dr. Rebekah Gardner, an associate professor of medicine at Brown University. “Before people start buying Apple Watches and sending all their sleep hours, let’s show that this improves health.”

She said she wants to see more rigorous, independently funded studies showing that monitoring data from wearable devices makes people healthier or improves their care.

Liability concerns weigh on some doctors’ minds. Dr. Oguchi Nkwocha, a community health center physician-executive in Salinas, California, worries he could be held liable if he missed something in “a diary of data,” but said he might be more open to data that was analyzed and presented with predictive insights.

Apple isn’t the only tech company that has struggled to make health app data-sharing mainstream. Both and enabled patients to share their data in their personal health record products over a decade ago but shut down these businesses because of limited user adoption, Holt noted.

Optimists believe that, eventually, research will show that more forms of data monitoring lead to better health and that technology could help make the data more digestible for doctors. Then, Apple might succeed in making its apps part of medicine — assuming the payment system changes in a way that gives providers more incentives to identify problems early and intervene before people get critically ill, Holt said.

“This is exciting for the future of chronic care management,” Dr. David Cho, a UCLA Health cardiologist, said of the new feature. With data at his fingertips on risk factors like exercise, diet and blood pressure, he believes he could help his patients manage chronic conditions more easily. That data, combined with virtual visits, could mean fewer office visits.

Apple’s announcement that it can integrate patient-generated data into the electronic medical record could be critical for doctors who want to see their patients’ self-collected information but don’t have time to hunt for it, said Dr. Seth Berkowitz, who leads a remote monitoring app pilot program at Beth Israel Deaconess Medical Center in Boston.

Some patients welcome a feature that would make it easy to share data with their doctors. Jen Horonjeff, a New York City-based autoimmune disorder patient and health care startup CEO, recently discovered by using an Apple Watch tracker that her heart rate, which doctors had described as irregular, registered as normal.

“I would absolutely send this to my physicians,” Horonjeff said, noting that her data would give doctors an accurate baseline of her heart rate if she were hospitalized.

But Gary Wolf of Berkeley, California, co-founder of the Quantified Self, a movement of people who track their health and other personal data, said that finding a doctor trained to make decisions with “fine-resolution data” is impossible.

Without more evidence that getting health app data to doctors is clinically beneficial, it will be hard to assess whether Apple is succeeding, said Neil Sehgal, assistant professor of health policy at the University of Maryland.

“Right now, we don’t know if there are consequences if you don’t put your Apple Watch data into your electronic medical record,” he said.

If evidence ultimately shows a benefit to sharing this information with doctors, he said, “that benefit will be concentrated among people who can buy the $1,000 phone and $400 watch.”

This story was produced by , which publishes , an editorially independent service of the .

Ñî¹óåú´«Ã½Ò•îl Health News is a national newsroom that produces in-depth journalism about health issues and is one of the core operating programs at KFF—an independent source of health policy research, polling, and journalism. Learn more about .

USE OUR CONTENT

This story can be republished for free (details).

The data being collected is so scattered that there’s little insight into which health care workers, or first responders, have been among the people getting the initial vaccines, as intended — or how many doses instead have gone to people who should be much further down the list.

The gaps — which experts say reflect decades of underfunding of public health programs — could mean that well-connected people and health personnel who have no contact with patients are getting vaccines before front-line workers, who are at much higher risk for illness. Federal and state officials prioritized health workers plus residents and staffs of nursing homes for the first wave of shots.

Although officials leading President Joe Biden’s covid response have pledged to tackle racial inequities as they seek to control the pandemic, lapses in reporting race or ethnicity could hinder efforts to identify and track whether minorities hit especially hard by the pandemic are getting shots at a high-enough rate to achieve hoped-for levels of herd immunity. So far, limited data in multiple states shows Black residents are getting vaccinated at lower rates than whites.

“Every state knows where they’ve sent vaccine, and every provider has to report inventory. But as far as who is being vaccinated, that one is a little more tricky,” said Claire Hannan, executive director of the Association of Immunization Managers.

Data that eventually makes its way to the Centers for Disease Control and Prevention and other federal systems is “only going to be as good as whatever you can get out of the vaccine registries” that vary by state, said Dr. Marcus Plescia, chief medical officer for the Association of State and Territorial Health Officials. “They’re all different and, going into this, they were all at different stages of how robust they were.”

There are 64 immunization registries in the United States that gather information for states, territories and a handful of large cities — and they aren’t connected. Meanwhile, real-time data in the U.S. public health system is virtually nonexistent, Plescia said.

Reporters at KHN examined the data being gathered versus what the CDC says is supposed to be collected for every person vaccinated, which includes: name, address, sex, date of birth, race and ethnicity, the date and location where they were vaccinated, and the shot they received (currently only two products are available, from Pfizer-BioNTech and Moderna). Not on its list: occupation, even though initial vaccine distribution largely hinges on place of work, prioritizing health care personnel, long-term care facilities and then other essential workers such as teachers, grocery store workers and firefighters.

Dr. Katherine Poehling, a pediatrician at the Wake Forest School of Medicine who’s on the CDC advisory committee that issued vaccine priority recommendations, declined to comment on whether occupation should have been a required element for reporting to the CDC.

“I think you can always wish for more data, but really what we’re going for is vaccinating everybody that wants to be vaccinated,” she said. “The fact that there was something available on day one was really remarkable,” she said, referring to a database that could track vaccine shipments and allocations by state.

Still, gaps are evident, including holes in CDC rules for reporting race and ethnicity. Race and ethnicity information are missing from at least hundreds of thousands of vaccine doses that have already been administered and reported to state public health authorities.

Texas’ vaccine data on Wednesday showed that race or ethnicity was unknown for more than 700,000 people. Virginia’s dashboard shows that data was missing for nearly 300,000 vaccinations, or 52% of vaccine doses, as of Tuesday. The same was true for tens of thousands of vaccinations in Colorado and Maryland.

In Minnesota, state law prohibits the sharing of data on race and ethnicity.

“It is important how many shots are administered, but it is critical that we get good race and ethnicity information about who is receiving it so we can identify disparities and other problems,” said Janet Hamilton, executive director of the Council of State and Territorial Epidemiologists.

The CDC declined to say how many of the vaccine records it had received were missing the information. In response to questions, CDC spokesperson Kristen Nordlund said the agency plans to publish race, ethnicity and other demographic data next week.

The Department of Health and Human Services did not respond to multiple requests for comment.

Dr. Marcella Nunez-Smith, chair of the Biden administration’s covid-19 health equity task force, on Wednesday conceded that the racial and ethnicity data is “incomplete” but said it wasn’t the only way to gauge progress of the vaccine rollout on the ground.

“We can think about things like neighborhoods and communities as metrics and ways to track as well,” she said. “We’re building our equity dashboard right now, and we’ll rely on government sources as well as sources of data external to government.”

The ongoing struggle for complete data shows how little has changed for the CDC since the virus appeared in the U.S. one year ago and its early efforts to collect data identifying covid-infected people were widely panned.

So far, the CDC has publicly stated how many vaccines have been distributed nationwide and how many doses administered. includes a breakdown of how many shots have been given by state and in long-term care facilities. Walgreens and CVS together have given more than 2.5 million doses in nursing homes and other long-term care facilities, though neither company has released data on race or whether the shots were given to patients or staffers.

State and federal health officials know where vaccines go as officials must track inventory by facility. Several states have released breakdowns of doses administered by the type of institution, providing a window into how many shots are being used in hospitals, nursing homes, pharmacies, primary care practices, public health departments and tribal health sites. And when signing up for an appointment, individuals may be asked to provide their occupation to attest they qualify for a shot under a state’s rules at a given time.

and require providers to submit data on the occupations of vaccine recipients, in a break with CDC practice. But several states contacted by KHN said they do not collect that information, such as Idaho, Michigan, Minnesota, Texas and Virginia.

Electronic health records manufacturers that provide software to hospitals and other facilities said they are scrambling to modify the software to accommodate data reporting requirements that vary by state.

Occupation is one example. Another: Texas law requires the state to collect information on all medications given “in response to a declared disaster or public health emergency,” said State Health Services spokesperson Chris Van Deusen.

Leigh Burchell, vice president of policy and government affairs at the EHR firm Allscripts, said these variations are “obstacles none of us has tackled before,” though she thinks that, overall, “successes outweigh failures” as companies have had to adjust quickly during the pandemic.

EHR systems can connect to state registries, which ultimately send vaccine tracking data to the CDC. A lack of “a coordinated, national public health infrastructure” continues to be a problem that “forces everyone to work less efficiently than would be optimal,” Burchell said.

Health IT consultant Reed Gelzer said the situation reflects the 30-year-plus failure of the public health system to modernize data collection. He said officials need look no further than chronic problems tracking childhood immunizations, handled in some states at the county level, and in others at the state level, often poorly. A national system to track immunizations has never existed, which he argues should have been discussed before the vaccine rollout.

“As far as I know, even in the earliest days of the pandemic, nobody did stress-testing of the information system,” Gelzer said.

Cerner, a major electronic health records company, says that some hospitals are using an existing workplace health system to track employees who have been vaccinated while others create a patient record for vaccinated employees as well as for patients. The systems can capture demographic details, but the data fields to do that have to be turned on and it’s unclear whether its client hospitals have done so.

The CDC and other federal agencies rely on a complicated web of systems to get data about who’s been vaccinated. State and local vaccine registries, known as immunization information systems, are the most comprehensive source of records and the “source of truth,” Hannan said.

Those registries have long-standing connections to providers’ electronic health records, said Rebecca Coyle, executive director of the American Immunization Registry Association. But they aren’t meant to capture certain information, such as a patient’s medical history and occupation.

Those state and local registries transmit data to an HHS-owned clearinghouse, where personal details are redacted.

The clearinghouse gets data from other sources, too. These include a new CDC vaccination clinic mobile app called VAMS, as well as pharmacies, prisons and federal agencies like the Department of Veterans Affairs and the Indian Health Service.

A limited slice of the data then moves to another CDC repository known as the “Data Lake,” where it can be analyzed and reported to the CDC and , a separate software platform developed by federal contractor Palantir for former President Donald Trump’s Operation Warp Speed effort. The Data Lake also receives information on shipment and vaccine orders from the CDC’s VTrckS system.

On top of that dizzying array of tools, many states use another, third-party software system, PrepMod, to manage vaccine inventory, appointments and reporting.

When asked whether not having data on occupations could hinder tracking whether priority groups have received their shots, Nordlund of the CDC said it’s unnecessary to vaccinate all individuals in one phase before initiating the next.

“This means ideally hitting a sweet spot that maximizes getting vaccine into arms while also being mindful of the priority groups,” Nordlund said, “especially because these are people who are higher risk for complications from covid-19 or are more likely to be exposed to the virus because of their jobs.”

Lawmakers recently attempted to address the nation’s antiquated public health data infrastructure, partly by appropriating $500 million under the CARES Act to the CDC. In an August letter to Rep. Lucy McBath (D-Ga.), former CDC director Dr. Robert Redfield said the agency would use the funds to update how state and health departments report data to federal officials, improve the CDC’s own data infrastructure, and develop new standards for public health reporting.

Additionally, tucked into the massive year-end spending bill Congress passed in late December was a requirement that HHS expand and improve public health data systems used by the CDC and award grants to state and local health departments to upgrade their infrastructure.

The Biden administration has made promises to strengthen the federal government’s approach to data collection on vaccination efforts.

KHN data reporter Hannah Recht and KHN correspondent Lauren Weber contributed to this report.

Ñî¹óåú´«Ã½Ò•îl Health News is a national newsroom that produces in-depth journalism about health issues and is one of the core operating programs at KFF—an independent source of health policy research, polling, and journalism. Learn more about .

USE OUR CONTENT

This story can be republished for free (details).

Yet pooling data from the digital records systems in thousands of hospitals has proved a technical nightmare thus far. That’s largely because software built by rival technology firms often cannot retrieve and share information to help doctors judge which coronavirus treatments are helping patients recover.

“I’m stunned at EHR vendors’ inability to consistently pull data from their systems,” said Dale Sanders, chief technology officer of , a medical data analytics company. “It’s absolutely hampering our ability to understand and react to COVID.”

Over the past decade, federal officials have spent some $36 billion switching from paper to electronic health records, or EHRs, expecting, among other things, to harness volumes of medical data to reveal which treatments work best.

EHRs document every step doctors or other health care workers take in treating a COVID patient, from medicines prescribed to signs of progress or setbacks. Data collected from large numbers of patients could quickly yield answers about which treatments are succeeding.

But the pandemic is bringing into stark relief just how far the nation is from achieving the promised benefits, critics say.

Dr. Richard Cook, a research scientist and health care safety specialist, traces the data problems to missteps dating to the rollout of EHR, which began in earnest in 2009 and has been controversial ever since because commercial players produced ― and hospitals bought — systems that have proved more suited to billing than public health. “This was a boondoggle from the get-go, and the promoters knew it at the time,” Cook said.

Although some health systems are beginning to draw on EHR data to spot coronavirus trends and beneficial treatments, most health organizations around the country cannot readily do so.

“If we had a national database, we’d get a readout quickly about responses to [COVID-19] treatments,” said Dr. Eric Topol, director of the Scripps Research Translational Institute.

Medical researchers favor studies that test the efficacy of a drug in a formal clinical trial, and trials are underway for a variety of possible COVID-fighting medicines, including hydroxychloroquine. The results could take months or more, however, and doctors treating critically ill patients have few options in the meantime.

Topol said “real-world” evidence drawn from computerized records of COVID patients, while not as reliable as a clinical trial, is “still very useful” to help guide medical decisions.

Medical data has been hard to tease out because much of it resides in electronic “silos,” which government officials have not required technology companies to open up and eliminate.

“We’ll see piecemeal readouts of small numbers from individual health systems,” Topol said, but “don’t have the important data that we need.”

Sanders, whose firm is a member of the , a business-sponsored group promoting coronavirus data-sharing and analysis, said federal health officials lost precious time by failing to address this need as early as mid-January.

He said the federal Centers for Disease Control and Prevention, or CDC, should have devised a COVID data-collection plan using standardized terminology so hospitals with incompatible EHRs could compare notes on the fast-paced pandemic.

The CDC did not respond to written requests seeking comment. A spokesman for the Health and Human Services office that coordinates health information technology policy said: “This is a novel disease so the health care system did not know what data we needed to collect ― we are learning that the system needs to build out reporting information on multiple clinical features.”

Still, several of the top EHR manufacturers have joined the data-sharing coalition, which is pledging to at least partially fill the information void. The group has access to COVID data from about two dozen health systems and is expecting to add more.

“This is the first attempt at this that I’m aware of where inherently competitive EHR vendors have come together to work together with clinical researchers,” said Dr. Brian Anderson, chief digital health physician with the MITRE Corp., a nonprofit technology group that formed the coalition in late March.

Anderson said the coalition is “getting close” to being able to share some results from reports of treating people with convalescent plasma recovered from patients who have survived COVID-19. The group is also examining treatment data on the drug as it irons out some of the technical difficulties that complicated its analysis of hydroxychloroquine. Last week, the Food and Drug Administration that hydroxychloroquine could cause heart problems and should be used only in a hospital or clinical trial.

There are other signs the EHR industry is relaxing its grip on medical data in response to the emergency. Major EHR vendor Cerner Corp. has researchers access to some types of COVID-19 data, including “clinical complications and outcomes that could help drive important medical decisions.”

And some health systems have begun publishing data drawn from EHRs. One released this month, for instance, tracked the outcome of 5,700 coronavirus patients treated at 12 hospitals in a New York City health system and found that 88% of patients placed on ventilators had died. All the hospitals shared the same records vendor.

“In crisis, people seek data and authorities demand it,” said Cook, the health care safety specialist. But, he said, “it is not possible to build such a system on demand.”

Ross Koppel, a professor at the University of Pennsylvania and longtime EHR safety expert, said that the COVID-19 pandemic illustrates both “strengths and disappointments” of the digital systems.

While health systems using a single vendor have been able to pool data, Koppel said, the industry has battled regulators seeking to adopt common standards, a practice known as interoperability.

“That failure to mine these oceans of invaluable data reflects the power of the vendors to prevent government requirements for data standards and interoperability,” he said.

Limits in electronic data collection systems also are hindering COVID-19 public health and surveillance efforts.

Officials said they are sometimes required to manually fill out and fax some forms, wasting valuable time. Some information must be printed out from EHRs and reentered by public health authorities because it cannot be sent electronically.

Certain CDC , such as Person Under Investigation COVID case reports, can take up to 30 minutes to complete. Other forms exchanged between hospitals and laboratories often are missing critical information, leading to delays in contacting patients and identifying people they had close contact with. In some states, demographic information on race and ethnicity is missing 85% of the time, and patients’ addresses, half the time, according to Janet Hamilton, executive director of the Council of State and Territorial Epidemiologists.

“We’re using yesterday’s technology for the biggest public health emergency in our lifetimes,” Hamilton said. “COVID has demonstrated for people what we’ve known all along. You can’t leave public health at the end of the line.”

The government’s health IT chief says a new administrative to promote interoperability and bar EHR manufacturers from impeding the flow of information will take time to change behavior.

“If this were to have happened three or four years in the future when we have interoperability … we would be in a much better spot here. But unfortunately, that’s not quite the case, but we’re still keeping our work going,” Donald Rucker, national coordinator for health information technology, said during an April 15 virtual .

Ñî¹óåú´«Ã½Ò•îl Health News is a national newsroom that produces in-depth journalism about health issues and is one of the core operating programs at KFF—an independent source of health policy research, polling, and journalism. Learn more about .

USE OUR CONTENT

This story can be republished for free (details).

• A large number of members of Congress are 65 or older — as many as 48 of the senators and more than 150 House members ― and should be taking precautionary measures to avoid the coronavirus, including not attending large gatherings. But there are no provisions for handling votes and congressional business remotely.
• In his speech Wednesday night, President Donald Trump highlighted his hopes of getting Congress to pass a payroll tax holiday to help consumers affected by the virus or suffering from the illness it causes, COVID-19, and to help stimulate the economy. But that proposal seems to be running into bipartisan objections. One problem: It doesn’t help people whose pay is lost because of the loss of work, such as gig workers.
• Democrats have offered instead an ambitious package of initiatives to help cover expenses for workers who may be out of work because of the outbreak. But they have to hammer out many details to get Republican approval.
• Insurance companies have agreed to the government’s request that they cover testing at no cost to customers, but many people could still face out-of-pocket expenses for treatment.
• House Speaker Nancy Pelosi unveiled a plan this week to upgrade consumer provisions in the Affordable Care Act. She did not take sides in the party’s debate on switching to a “Medicare for All” single-payer plan versus adding a government-run public option to the ACA. But her efforts are designed to help House members campaign on health care fixes that are popular with voters.
• The federal government released a rule this week to make electronic health records more easily transferrable and allow consumers to have computer apps that hold their records. Although the concept of getting records sent between medical offices more easily is broadly popular, the effort could raise privacy issues, industry groups allege.

Ñî¹óåú´«Ã½Ò•îl Health News is a national newsroom that produces in-depth journalism about health issues and is one of the core operating programs at KFF—an independent source of health policy research, polling, and journalism. Learn more about .

USE OUR CONTENT

This story can be republished for free (details).

“Patients should have control of their records, period. Now that’s becoming a reality,” said Health and Human Services Secretary Alex Azar. “These rules are the start of a new chapter in how patients experience American health care.”

Officials said the rules likely will give patients a greater say in health care decisions and put an end to a long-standing practice in which some doctors and hospitals resist handing complete medical files over to patients upon demand. Many of the provisions are set to take effect in 2022.

“The days of patients being kept in the dark are over,” said Centers for Medicare & Medicaid Services Administrator Seema Verma. “In today’s digital age, our health system’s data-sharing capacity shouldn’t be mired in the Stone Age.”

Yet the new rules also have raised concerns about privacy as technology companies, such as Google, Microsoft, Apple and Amazon, open up new markets for providing medical records through mobile apps. Major EHR vendor Epic, for instance, has warned that freer flow of medical records could spur the unwanted sale of data or other unauthorized uses.

“Family members may be shocked to find that their most personal health data has been mined and sold by data brokers and is now known by others, Epic CEO Judy Faulkner wrote last June in opposing the rules.

Administration officials said they have taken privacy considerations into account and would require developers to attest to plans to protect the security and use of medical data.

Verma took a swipe at Epic in an interview with KHN and Fortune.

“We’re not afraid to take on special interests to do what’s right for patients. Some people disagree because they want to keep the data,” she said. “The reality is that patient data belongs to patients. It doesn’t belong to EHR companies.”

Verma said the nation’s health care system remains “hugely expensive and inefficient as repeat tests drive up costs and, perhaps most importantly, doctors are forced to provide care with an incomplete clinical picture, especially at a time when the health care systems could be under stress.”

“With the handling of the COVID virus, the urgent need for coordinated integrated care could not be clearer,” she said.

The Department of Health and Human Services on Monday issued two rules to give patients greater access to their electronic health records.

Among the key provisions:

• Patients must be able to access their medical records on a smartphone at no cost and can share those records as they choose.
• Health systems must be able to exchange information about patients’ past medical treatments or conditions.
• “Information blocking” practices (that is, anti-competitive behaviors) by health care providers, developers of electronic medical records and others are prohibited.
•  Electronic medical record certification requirements are updated so that health professionals can discuss safety and usability concerns without being bound by gag clauses in software sales contracts.
• Insurers are required to share health claim data with patients on Medicare and Medicaid through a mobile app.
• Insurers must advise patients of their network of health providers through an app.

Donald Rucker, who coordinates health information technology policy for HHS, said the new rule “will allow patients the ability to manage their health care the same way they manage their finances or the travel or other parts of their life on their smartphone.”

While Epic, the maker of the most-used electronic health records software, led a campaign to derail the rules, its chief competitor, Cerner Corp., argued the rules were long overdue.

“Consumers should have the right to access the health care information their providers have about them and dictate where they want it to go. Although existing laws allow patients to access their data, it doesn’t work,” Cerner CEO Brent Shafer said in a statement.

The rules also attempt to prevent EHR vendors from silencing critics of their software products. The government wants to encourage doctors and other users of EHR technology to share their experiences about software problems by prohibiting so-called gag clauses in sales contracts. That could free users to criticize EHR systems, including more open discussion of flaws, software glitches and other breakdowns.

“Botched Operation,” an investigation published by Kaiser Health News and last year, found that the federal government has spent more than $36 billion on the EHR initiative. Thousands of reports of deaths, injuries and near misses linked to digital systems have piled up in databases over the past decade — while many patients have reported difficulties getting copies of their complete electronic files, the investigation found.

Consumers have long sought to be more in the loop on health care decisions in a user-friendly form. Many specifics about how that will happen, including how patients would make sense of complex pricing policies for purchasing health care and insurance and assessing quality, remain unclear, however.

To cut down on exorbitant “surprise” medical bills, Verma said, the CMS’ new rule would require insurers to let patients know which medical providers are in their networks. One study found that such bills — often not covered by insurance — have struck more than half of American adults.

For well over a decade, federal officials have struggled to set up a digital records network capable of sharing medical data and patient records. In 2004, President George W. Bush said he hoped to have a digital record for most Americans within five years. In early 2009, the Obama administration funneled some $36 billion in economic stimulus money to help doctors and hospitals buy the software needed to replace paper medical files.

Despite the slow progress, federal officials remain optimistic that digital records will save the nation billions of dollars while reducing medical errors, unnecessary medical testing and other waste — and encouraging more Americans to take a bigger role in managing their health care by comparing prices.

Trump administration officials on Monday sought to blame the Obama administration for creating what they called a “tower of Babylon,” in which doctors and health systems couldn’t seamlessly share patient information or “talk to one another.”

“It’s led to a tremendous amount of frustration on the part of medical professionals and patients as physicians, interacting with patients, oftentimes spend more time looking at computer screens than they do into the eyes of the people they’re trying to heal,” said White House official Joe Grogan.

Ñî¹óåú´«Ã½Ò•îl Health News is a national newsroom that produces in-depth journalism about health issues and is one of the core operating programs at KFF—an independent source of health policy research, polling, and journalism. Learn more about .

USE OUR CONTENT

This story can be republished for free (details).

Wouldn’t life be easier if you could view your full medical history with a few taps on your smartphone?

The consolidation of medical records may be on its way, as technology companies prod the health care industry to embrace an internet-based common standard for storing and sharing patient information. It’s known as FHIR and pronounced “fire” — a catchier way of saying Fast Healthcare Interoperability Resources.

Industry analysts say the rapidly growing demand for freer exchange of health care information is creating an electronic health record market estimated to reach . With numbers like that bandied about, it should come as no surprise that Silicon Valley tech giants Apple and Google are lining up for a slice of the pie — as are other technology behemoths, including Amazon and Microsoft. Those corporations, and many smaller companies and startups, offer FHIR-based apps and services to consumers and health industry professionals.

The idea behind FHIR is to share specific pieces of information, such as symptoms, procedures or diagnoses, without passing along entire documents. Each discrete chunk of data has a unique identifier, which makes it possible for patients, doctors and researchers to get the information they need on any device or browser, regardless of where the data is stored.

Proponents of the new standard say it should ensure that health care information can be exchanged seamlessly among providers across the industry — from a urine-testing laboratory in Los Angeles, for example, to a pediatrician in Redding, California.

When tech companies can agree to use an industry standard, the adoption of new technology accelerates: Think Bluetooth or USB. It’s a confusing tangle when that doesn’t happen: Think laptop power cords.

Adoption of the medical record-sharing standard may begin to accelerate with the rollout of this year requiring health care providers that receive payments from the U.S. government — Medicare and Medicaid — to use FHIR-compatible apps for patient data.

FHIR has met considerable resistance, however. Until recently, there were few business incentives and limited advocacy by health care providers to create the necessary demand for FHIR’s adoption, said Micky Tripathi, chairperson of the advisory council of , a not-for-profit developer of electronic health information standards that created FHIR.

“Change will come incrementally over time, not overnight,” Tripathi said. “The older standards, though inferior, have the advantage that they are in use today and the cost of ripping and replacing them is not worth the added benefit.”

Some hospitals and medical clinics put the brakes on the very concept of data sharing by engaging in “information blocking” in an effort to retain patients. In 2016, Congress enacted financial penalties to stop them from doing so, but their resistance has not been eliminated.

“There have been roadblocks that prevent innovations and data from being widely shared where it could benefit patients,” said Julia Adler-Milstein, director of the Center for Clinical Informatics and Improvement Research at the University of California-San Francisco.

Beyond its potential to revolutionize medical records requests, FHIR may also provide the first reliable gateway for patient-generated health information from millions of smartwatches, fitness trackers and blood pressure monitors to merge with clinical data in doctors’ offices, people in the industry say.

The need for a standard arose from of electronic health records at medical sites ranging from hospitals and doctors’ offices to urgent care clinics and nursing homes.

Digitizing health records was intended to clean up the chaos of paper-based medical histories, allowing information to be shared more readily. But many practitioners still , and among those who have made the digital switch, medical records often lie isolated in electronic silos.

In 2014, proposed leveraging how the internet works to break open the silos. FHIR evolved quickly, creating what its product director, Grahame Grieve, of international health information exchange.

At UCSF, FHIR is being used in a study to track the weight of newborns hour-by-hour. UCSF’s Healthy Start program integrates weight data with other information about each newborn to alert doctors when one of them may be struggling.

Established medical technology companies and a host of startups are salivating over FHIR, because eventually it could give them paths to lucrative uses of data, including for personalized medicine, population health and medical genetics as well as in emerging technologies such as machine learning and artificial intelligence.

“We want to have FHIR in our analytics and machine-learning tools,” Aashima Gupta, Google Cloud’s director of Global Healthcare Solutions, said at a conference in Orlando, Florida, in February.

Tech giants are already striving to benefit from the growing use of FHIR:

• Microsoft recently released , its data-on-demand offering, to attract health care clients to its cloud services.
• Google joined with the American Medical Association in an effort to improve coordination among health care systems with FHIR and develop methods of collecting and managing patient-generated health data.
• Apple’s Health Records app uses FHIR to let consumers download data from their health care providers.
• Amazon Comprehend Medical works through Amazon Web Services to offer guidance for health care data specialists using FHIR.

Six big tech companies — Microsoft, Google, Amazon, IBM, Oracle and Salesforce — have also joined to support FHIR and broader sharing of health care data through a government-endorsed project called , which is intended to make it easier for patients to view and download their health records.

Consumer advocates and cybersecurity experts warn that personal health information shared on the web could be compromised. They want to make sure the risk is minimized before any widespread rollout of FHIR products. Patients do not have a say in how their health providers store medical information, but patients can request their records be sent in the format they prefer, including paper.

Facilitating access to all that data for both patients and providers without first determining how to keep it secure may open a Pandora’s box that can never be shut, warned David Finn, executive vice president of strategic innovation for , a Mission Viejo, California, and Austin, Texas-based cybersecurity consulting firm.

“We have to change the way we think about data. It is our most valuable asset. But we have not adjusted our thinking about data to how the bad guys think of it,” Finn said. “Until we think about what you could do maliciously with that information, I’m afraid we will not catch up with them.”

This story was produced by , which publishes , an editorially independent service of the .

Ñî¹óåú´«Ã½Ò•îl Health News is a national newsroom that produces in-depth journalism about health issues and is one of the core operating programs at KFF—an independent source of health policy research, polling, and journalism. Learn more about .

USE OUR CONTENT

This story can be republished for free (details).

Health products powered by , or AI, are streaming into our lives, from virtual to and .

boasted that its AI could Others say computer systems that will make radiologists obsolete.

“There’s nothing that I’ve seen in my 30-plus years studying medicine that could be as impactful and transformative” as AI, said Dr. Eric Topol, a cardiologist and executive vice president of Scripps Research in La Jolla, Calif. AI can help doctors interpret , and of the , and could potentially take over many mundane medical chores, freeing doctors to spend more time talking to patients, Topol said.

Even the Food and Drug Administration ― which has approved more than in the past five years ― says “the potential of digital health is

Yet many health industry experts fear AI-based products won’t be able . Many and fear that the tech industry, which lives by the mantra “,” is ― and that regulators aren’t doing enough to keep consumers safe.

Early experiments in AI provide a reason for caution, said Mildred Cho, a professor of pediatrics at Stanford’s Center for Biomedical Ethics.

Systems developed in one hospital often flop when deployed in a different facility, Cho said. Software used in the care of has been shown to discriminate against minorities. And AI systems sometimes learn to make predictions based on factors that have less to do with disease than the used, the or whether a patient was . In one case, AI software incorrectly concluded that people with pneumonia were less likely to die ― an error that could have led doctors to deprive asthma patients of the extra care they need.

“It’s only a matter of time before something like this leads to a serious health problem,” said Dr. Steven Nissen, chairman of cardiology at the Cleveland Clinic.

Medical AI, which pulled in in venture capital funding in the third quarter alone, is “nearly at the peak of inflated expectations,” concluded a July report from the research company . “As the reality gets tested, there will likely be a rough slide into the trough of disillusionment.”

That reality check could come in the form of disappointing results when AI products are ushered into the real world. Even Topol, the author of “Deep Medicine: How Artificial Intelligence Can Make Healthcare Human Again,” acknowledges that many AI products are little more than hot air. “It’s a mixed bag,” he said.

Experts such as Dr. Bob Kocher, a partner at the venture capital firm Venrock, are blunter. “Most AI products have to support them,” Kocher said. Some risks won’t become apparent until an AI system has been used by large numbers of patients. “We’re going to keep discovering a whole bunch of risks and unintended consequences of using AI on medical data,” Kocher said.

None of the AI products sold in the U.S. have been tested in randomized clinical trials, the strongest source of medical evidence, Topol said. The first and only randomized trial of an AI system ― which found that colonoscopy with computer-aided diagnosis found more small polyps than standard colonoscopy ― was published online in .

Few tech startups publish their research in peer-reviewed journals, which allow other scientists to scrutinize their work, according to a in the European Journal of Clinical Investigation. Such “stealth research” ― described only in press releases or promotional events ― often overstates a company’s accomplishments.

And although software developers may boast about the accuracy of their AI devices, experts note that AI models are mostly tested on computers, not in or other medical facilities. Using unproven software “may make patients into unwitting guinea pigs,” said Dr. Ron Li, medical informatics director for AI clinical integration at Stanford Health Care.

AI systems that learn to recognize patterns in data are often described as because even their developers don’t know how they have reached their conclusions. Given that AI is so new ― and many of its risks unknown ― the field needs , said Pilar Ossorio, a professor of law and bioethics at the University of Wisconsin-Madison.

Yet the majority of AI devices don’t require FDA approval.

“None of the companies that I have invested in are covered by the FDA regulations,” Kocher said.

passed by Congress in 2016 ― and championed by the ― exempts from federal review, including certain fitness apps, electronic health records and tools that help doctors make medical decisions.

There’s been on whether the 320,000 medical apps now in use actually improve health, according to a report on AI published Dec. 17 by the .

“Almost none of the [AI] stuff marketed to patients really works,” said , professor of medical ethics and health policy in the Perelman School of Medicine at the University of Pennsylvania.

The FDA has long focused its attention on devices that pose the . And consumer advocates acknowledge that some devices ― such as ones that help people count their daily steps ― need less scrutiny than ones that diagnose or treat disease.

Some software developers don’t bother to apply for FDA clearance or authorization, even when legally required, according to

Industry analysts say that AI developers have little interest in conducting expensive and time-consuming trials. “It’s not the main concern of these firms to submit themselves to rigorous evaluation that would be published in a peer-reviewed journal,” said Joachim Roski, a principal at Booz Allen Hamilton, a technology consulting firm, and “That’s not how the U.S. economy works.”

But Oren Etzioni, chief executive officer at the Allen Institute for AI in Seattle, said AI developers have a financial incentive to make sure their medical products are safe.

“If failing fast means a whole bunch of people will die, I don’t think we want to fail fast,” Etzioni said. “Nobody is going to be happy, including investors, if people die or are severely hurt.”

Relaxing Standards At The FDA

The FDA has come under fire in recent years for allowing the sale of dangerous medical devices, which have been linked by the to over the past decade.

Many of these devices were cleared for use through a controversial process called the , which allows companies to market “moderate-risk” products with no clinical testing as long as they’re deemed similar to existing devices.

In 2011, a committee of the concluded the 510(k) process is so fundamentally flawed that the FDA should throw it out and start over.

Instead, the FDA is using the process to greenlight AI devices.

Of the 14 AI products authorized by the FDA in 2017 and 2018, 11 were cleared through the 510(k) process, according to a . None of these appear to have had new clinical testing, the study said. The FDA cleared an designed to help diagnose liver and lung cancer in 2018 based on its similarity to approved 20 years earlier. That software had itself been cleared because it was deemed “substantially equivalent” to .

AI products cleared by the FDA today are largely “locked,” so that their calculations and results will not change after they enter the market, said Bakul Patel, director for digital health at the FDA’s Center for Devices and Radiological Health. The FDA has not yet authorized “unlocked” AI devices, whose results could vary from month to month in ways that developers cannot predict.

To deal with the flood of AI products, the FDA is testing a radically different approach to digital device regulation, focusing on evaluating companies, not products.

The FDA’s pilot , launched in 2017, is designed to “reduce the time and cost of market entry for software developers,” imposing the “least burdensome” system possible. FDA officials say they want to keep pace with AI software developers, who much more frequently than makers of traditional devices, such as X-ray machines.

Scott Gottlieb said in 2017 while he was FDA commissioner that government regulators need to make sure its approach to innovative products “is efficient and that it fosters, not impedes, innovation.”

Under the plan, the FDA would pre-certify companies that “demonstrate a culture of quality and organizational excellence,” which would allow them to provide about devices.

Pre-certified companies could then release devices with a “streamlined” review ― or no FDA review at all. Once products are on the market, companies will be responsible for and reporting back to the FDA. have been selected for the pilot: Apple, FitBit, Samsung, Johnson & Johnson, Pear Therapeutics, Phosphorus, Roche, Tidepool and Verily Life Sciences.

High-risk products, such as , will still get a comprehensive FDA evaluation. “We definitely don’t want patients to be hurt,” said Patel, who noted that devices cleared through pre-certification can be recalled if needed. “There are a lot of guardrails still in place.”

But research shows that even have been recalled due to serious risks to patients, said Diana Zuckerman, president of the National Center for Health Research. “People could be harmed because something wasn’t required to be proven accurate or safe before it is widely used.”

Johnson & Johnson, for example, has recalled and .

In a , the American Medical Association and others have questioned the wisdom of allowing companies to monitor their own performance and product safety.

“The honor system is not a regulatory regime,” said Dr. Jesse Ehrenfeld, who chairs the physician group’s board of trustees.

In an October letter to the FDA, Sens. Elizabeth Warren (D-Mass.), Tina Smith (D-Minn.) and Patty Murray (D-Wash.) questioned the agency’s ability to ensure company safety reports are “accurate, timely and based on all available information.”

When Good Algorithms Go Bad

Some AI devices are more carefully tested than others.

An for diabetic eye disease was studied in 900 patients at 10 primary care offices before being approved in 2018. The manufacturer, IDx Technologies, worked with the FDA for eight years to get the product right, said Dr. Michael Abramoff, the company’s founder and executive chairman.

The test, sold as IDx-DR, screens patients for diabetic retinopathy, a leading cause of blindness, and refers high-risk patients to eye specialists, who make a definitive diagnosis.

IDx-DR is the first “autonomous” AI product ― one that can make a screening decision without a doctor. The company is now installing it in primary care clinics and grocery stores, where it can be operated by employees with a high school diploma. Abramoff’s company has taken the unusual step of buying liability insurance to cover any patient injuries.

Yet some AI-based innovations intended to improve care have had the opposite effect.

A Canadian company, for example, developed AI to predict a person’s risk of Alzheimer’s based on their speech. Predictions were more accurate for some patients than others. “Difficulty finding the right word may be due to rather than to cognitive impairment,” said co-author Frank Rudzicz, an associate professor of computer science at the University of Toronto.

Doctors at hoped AI could help them use chest X-rays to predict which patients were at high risk of pneumonia. Although the system made accurate predictions from X-rays shot at Mount Sinai, the technology flopped when tested on images taken at other hospitals. Eventually, researchers realized the computer had merely learned to tell the difference between that hospital’s ― taken at a patient’s bedside ― with those taken in the radiology department. Doctors tend to use portable chest X-rays for patients too sick to leave their room, so it’s not surprising that these patients had a greater risk of lung infection.

DeepMind, a company owned by Google, has created an AI-based mobile app that can predict which hospitalized patients will develop up to 48 hours in advance. A blog post on the described the system, used at a London hospital, as a “game changer.” But the AI system also for every correct result, according to a . That may explain why patients’ kidney function , said associate professor of radiology at the Hospital of the University of Pennsylvania. Any benefit from early detection of serious kidney problems may have been diluted by a high rate of “overdiagnosis,” in which the AI system flagged borderline kidney issues that didn’t need treatment, Jha said. Google had no comment in response to Jha’s conclusions.

False positives can harm patients by prompting doctors to order unnecessary tests or withhold recommended treatments, Jha said. For example, a doctor worried about a patient’s kidneys might stop prescribing ibuprofen ― a generally safe pain reliever that poses a small risk to kidney function ― in favor of an opioid, which carries a serious risk of addiction.

As these studies show, software with impressive results in a computer lab can founder when tested in real time, Stanford’s Cho said. That’s because diseases are more complex ― and the health care system far more dysfunctional ― than many computer scientists anticipate.

Many AI developers cull because they hold huge amounts of detailed data, Cho said. But those developers often aren’t aware that they’re building atop a deeply broken system. Electronic health records were developed for billing, not patient care, and are filled with mistakes or missing data.

A KHN investigation published in March found sometimes life-threatening errors in patients’ medication lists, lab tests and allergies.

In view of the risks involved, doctors need to step in to protect their patients’ interests, said Dr. Vikas Saini, a cardiologist and president of the nonprofit Lown Institute, which advocates for wider access to health care.

“While it is the job of entrepreneurs to think big and take risks,” Saini said, “it is the job of doctors to protect their patients.”

Ñî¹óåú´«Ã½Ò•îl Health News is a national newsroom that produces in-depth journalism about health issues and is one of the core operating programs at KFF—an independent source of health policy research, polling, and journalism. Learn more about .

USE OUR CONTENT

This story can be republished for free (details).

The doctors at Midwest (City) Regional Medical Center in Oklahoma worried that the software failed to track some drug prescriptions or dosages properly, posing a “huge safety concern,” Lewis said. Lewis cited the alleged safety hazards in a whistleblower lawsuit that he and another former employee of Community Health Systems (CHS) filed against the Tennessee-based hospital chain in 2018.

The suit alleges that the company, which had $14 billion in annual in 2018, obtained millions of dollars in federal subsidies fraudulently by covering up dangerous flaws in these systems at the Oklahoma hospital and more than 120 others it owned or operated at the time.

The whistleblowers also allege that Medhost, the Tennessee firm that developed the software, concealed defects during government-mandated reviews that were supposed to ensure safety.

Both CHS and Medhost have the allegations and moved to dismiss the suit. The motions are pending. Last month, Department of Justice lawyers wrote in court filings that they were still investigating the matter and had not yet decided whether to take over the case.

The lawsuit is one of dozens filed by whistleblowers, doctors and hospitals alleging that some electronic health records (EHR) software used in hospitals and medical offices has hidden flaws that may pose a danger to patients — and that a substantial chunk of the $38 billion in federal subsidies went to companies that deceived the government about the quality of their products, an ongoing Fortune-KHN investigation shows. The subsidies were designed to persuade hospitals and doctors’ offices to install software that would track the medical history of every patient and share the information seamlessly with other health care providers.

But the software makers allegedly gamed the system, repeatedly. Three major EHR vendors have made multimillion-dollar settlement deals — totaling $357 million — over Justice Department investigations which include allegations that they rigged or otherwise gamed the government’s certification test. At least two other companies are under investigation.

Beyond those cases, federal officials have paid hundreds of millions of dollars in subsidies to doctors and hospitals that could not show they were even qualified to receive them, according to federal officials. Nearly 28% of doctors and 5% of hospitals who attested to meeting government standards later failed audits. Federal officials told Fortune and KHN that they have clawed back $941 million in improper subsidies.

“We’re entering an entirely new area of health care fraud,” John O’Brien, senior counsel with the Department of Health and Human Services Office of Inspector General, said in a July 2017 announcing a with eClinicalWorks, one of the nation’s leading sellers of EHRs for physicians.

The concern is not just over wasteful spending of tax dollars. EHRs monitor the medicines people take and their vital signs, so software glitches that prevent doctors from accessing files quickly, that mix up patients or send vital test results to the wrong file can contribute to serious injuries, or even deaths.

In March, Fortune and KHN revealed that thousands of injuries, deaths or near misses tied to software defects, user errors and other problems have piled up in various government-sponsored and private repositories.

“Ultimately, it’s about patients getting the right care,” Andrew Vanlandingham, the HHS inspector general’s senior counselor for health information technology, said in an interview. He said that investigators are “gearing up” for more scrutiny of the important industry, including closer monitoring to make sure that records software is safe.

Leaping Into The Digital Era

In 2009, Congress committed billions of dollars in economic stimulus funds to bring the era of paper medical records to a close. Officials hoped to cut down on medical errors caused by illegible paper records and draw on the power of massive troves of medical data to drive down the cost of health care and help develop improved treatments for disease.

The hastily devised plan offered Medicare doctors and other medical professionals up to $44,000 and $64,000 in subsidies if they bought the software and accepted patients on Medicaid, the federal health care program for low-income people.

The money was intended to help them pay vendors to install EHRs in their offices. Hospitals, which required more sophisticated and costlier software, could receive millions in subsidies, based on the number of inpatients treated. To give them a nudge, officials warned doctors and hospitals that failure to wire up would trigger gradual cuts in their Medicare payments. EHR vendors had to meet certification standards set by the HHS Office of the National Coordinator for Health Information Technology, or ONC.

Providers had to attest that their EHR software could perform a variety of functions, which the government described as making “meaningful use” of the technology.

Certification was essentially an open-book test in which the government gave vendors the questions in advance — for instance, the names of 16 or so drugs the system would have to prescribe electronically to pass. The Justice Department has alleged that some vendors simply doctored their software to pass the test — for example, programming the required codes for just the specified 16 drugs they would be tested on, rather than all medicines — as officials had expected.

Frank Poggio who recently retired from a 45-year career in health technology, saw the cases of fraud coming, he said, because the tests “were superficial, and if you wanted to game it, you could game it.”

Poggio said there were many weaknesses in the system that allowed a vendor to show a “prototype” as opposed to live software.

Dr. Scott Monteith, a Michigan psychiatrist who served as an early certification juror, said he saw some limitations firsthand. He said one vendor took 30 minutes to produce a list of patients who had diabetes and also smoked, data he figured any computer program should be able to spit out in seconds. The vendor passed.

“That’s an example of how poorly thought-out the whole thing was,” said Monteith, who noted he was then, and still is, a big booster of EHRs.

Jeffery Daigrepont, a senior vice president at Coker Group, a firm that advises health providers on business decisions, said the government erred by handing out too much money in the early stages of the program, when many doctors and hospitals had not yet done much more than agree to participate.

“It was an upside-down pyramid,” he said. “You got the bulk of the money for doing the least amount of effort.”

Dr. John Halamka, a physician and Harvard Medical School professor who chaired the ONC standards committee, which wrote the certification rules, defended the process.

“The only problem [with certification] is that it presupposed that the product the vendor certified would be the same product they sold,” Halamka said. “It presupposes that people will go into the certification process and participate in good faith.”

That did not always happen in the rush to snatch up subsidy dollars, according to the whistleblowers’ suits. The Justice Department case against eClinicalWorks, which has 130,000 providers, accused the company of rigging tests to win certification, claims the company has denied. The company did not respond to numerous requests for comment.

The government accused Greenway Health, a Florida-based EHR developer with 75,000 providers, of doing the same thing. The DOJ’s complaint included a number of instant-message exchanges between Greenway employees in which they allegedly discuss their plan for gaming the certification process by “shortcutting some functionality” of the software. In February, Greenway Health for just over $57 million without admitting wrongdoing.

The whistleblower case filed by Lewis and former co-worker Joey Neiman accuses the CHS hospital chain of submitting more than $385 million in false claims for EHR stimulus payments between 2012 and 2014.

Visiting the Oklahoma hospital as part of a troubleshooting team in June 2015, Lewis heard that physicians worried flaws in the system could result in patients being sent home “with the wrong drugs, doses or instructions,” according to the suit.

Things got so bad that local doctors were threatening to admit patients elsewhere unless the hospital fixed the software problems, according to the suit.

In a statement, CHS said it had “complete confidence” in its records systems. “The allegations made in the lawsuit against our hospitals are completely without merit,” the company said. Medhost denied its software has flaws, noting in its : “Hundreds of facilities have successfully used our software over the years and continue to do so today.”

Few in the industry seemed surprised by such allegations. When news of the eClinicalWorks case broke, Farzad Mostashari, who led the ONC from 2011 to 2013, : “Let me be plain-spoken. eClinicalWorks is not the only EHR vendor who ‘flouted certification/misled’ customers. Other vendors better clean up.”

The Electronic Health Record Association, a trade group that represents more than 30 vendors, did not respond to a request for comment. However, vendors have argued that they faced a tangle of regulations that required them to meet constantly shifting standards that government officials often could not explain.

ONC officials declined to answer written questions. But in a statement, ONC said it takes steps to ensure that products “are safe for patients and usable by providers.”

System Glitches And Accusations Of ‘Gaming’ The System

While the ONC sets the standards, the federal Centers for Medicare & Medicaid Services (CMS) had the job of paying doctors and hospitals that attested to meeting the “meaningful use” criteria. As of September 2018, CMS had paid out $38.4 billion in these funds.

In 2012, CMS hired accounting firm Figliozzi and Co. of Garden City, N.Y., which audited almost 50,000 medical professionals. Nearly 28% failed, despite the fact that they had previously attested to meeting the standards. Hospitals did better, posting a 5% failure rate. CMS officials said they have recovered some $941 million in these improper payments. The losses to the Treasury are likely far higher because only 14% of the medical professionals and 40% of the hospitals receiving payments were audited.

Michael Arrigo, who has served as an expert witness in health IT-related fraud and medical malpractice cases, said that in some cases EHR vendors misled hospitals about the challenges of replacing paper records with computers.

Others rolled the dice, apparently hoping the program was so large and complicated that they were unlikely to be targeted for audit. “Sometimes [providers] got away with it until a whistleblower found out,” Arrigo said.

Reviewing state and federal court filings, Fortune and KHN found more than two dozen cases, many filed by hospitals against vendors, which depict chaotic EHR installations and safety concerns as they pursued meaningful-use dollars.

Parrish Medical Center, a 210-bed public hospital on Florida’s Space Coast, is one. In December 2010, the Titusville hospital contracted with McKesson’s Enterprise Information Solutions. One of America’s largest companies, McKesson said its product would be easy for doctors and nurses to learn and help them “deliver high-quality, safe patient care.”

But the deal collapsed, prompting a bitter court battle in which the hospital repeatedly assailed McKesson’s competence. For instance, the hospital alleged that bugs in the software caused it to create more than one record for the same patient, a flaw dubbed a “major safety issue.”

An expert hired by Parrish said he contacted eight other hospitals, including three in Florida, which had dumped McKesson due to what he called “poor or unsatisfactory customer service.”

The medical staff at one of those hospitals was “up in arms” because it took 63 mouse clicks to look up a patient’s lab results, according to the expert’s report.

Parrish later signed on with another EHR vendor and the suit has since been settled. Both Parrish and McKesson declined to comment for this story. McKesson sold its health IT business to Allscripts in October 2017. Earlier this year, Allscripts reported to the Securities and Exchange Commission that government attorneys have requested documents from the company as part of an investigation into McKesson’s certification.

In another lawsuit, Weirton Medical Center, a hospital in West Virginia, stated in a court filing that it submitted “inaccurate” meaningful-use data to the government ― though it blamed the vendor. The hospital alleged the system failed to identify a patient who was critically ill and in the hospital. The hospital declined to comment to KHN and Fortune about the case, which has been settled.

Hamstrung By Technology?

ONC officials said they keep no log of complaints they receive.

A published in JAMA this month found that 40% of the software that ONC singled out for post-marketing review had flaws that could lead to patient harm, including inaccurate drug codes, information displaying incorrectly and decimal points gone missing.

That’s “a concerning number, and we have to do something to address that,” said researcher Raj Ratwani, the director of the MedStar Health National Center for Human Factors in Healthcare and a co-author of the study. These systems were used in 786 hospitals and by 37,365 provider organizations, according to Ratwani, who said there’s no way to know how many defects have been fixed.

ONC has about 100 pieces of once-approved software products. But most were tiny market players that had few or no users and went out of business. PlatinumMD, which had just 48 “meaningful” users, is an example. In a 2014 whistleblower lawsuit, San Diego urologist Dr. Scott Brown alleged that PlatinumMD filed for $18,000 in subsidies on his behalf even though it had not yet fully installed his EHR. In February 2016, the defunct company’s owners settled the case without admitting liability by paying the government $180,000.

Another 132 government-certified products have been flagged for corrective action due to “non-conformities.” As for the technology that the government alleges was fraudulently certified, it’s still used in health care settings across the country.

While those vendors faced multimillion-dollar settlements and now must operate under the oversight of a government monitor, their technology was not taken off the market. Nor were they dumped by many customers who, for the most part, however dissatisfied, were stuck with it.

ONC seemed to acknowledge that decertifying a large vendor would cause a major disruption, noting in an October 2016 regulation: “Our first and foremost desire would be to work with developers to address any problems.”

In the regulations, ONC cited the costs medical providers would face should their EHR vendor shut down as ranging from $33,000 to as much as $650 million.

“It is very difficult to switch product,” said Steve Waldren, chief medical informatics officer for the American Academy of Family Physicians. “You couldn’t just go down the street and pick up another EHR, put it in and move your data over.”

He noted that beyond the considerable cost of the technology, providers would have to take time to learn a new system.

“ONC does seem to have a stance that removing some of these players from the market would be very disruptive,” said Brad Ulrich, a Tennessee health IT expert. “They are almost too big to fail.”

Ñî¹óåú´«Ã½Ò•îl Health News is a national newsroom that produces in-depth journalism about health issues and is one of the core operating programs at KFF—an independent source of health policy research, polling, and journalism. Learn more about .

USE OUR CONTENT

This story can be republished for free (details).