Addressing health inequities proved one overarching theme at the Healthcare Information and Management Systems Society trade show last week, as health IT experts shared ideas for addressing algorithmic bias and using data to tackle challenges like food and housing insecurity. Other themes included cybersecurity, telehealth’s role in the healthcare ecosystem and data-sharing in the public health sector.
HIMSS22 took place last week in Orlando, Florida, just seven months after last year’s trade show in August, which was delayed from March 2021 due to the COVID-19 pandemic.
Here are five highlights Modern Healthcare saw at HIMSS22.
1. Addressing algorithmic bias. While many healthcare and technology experts touted the benefits of artificial intelligence and other analytics tools, questions were raised on how to ensure algorithms deployed into clinical care don’t carry unintended biases or unexpected consequences for patients.
In a session on ethical and safety implications of AI, Jessica Newman, director of the AI Security Initiative at the Center For Long-Term Cybersecurity at UC Berkeley, cited a 2019 study that found a widely used algorithm for population health management—a predictive model that doesn’t use AI—underestimated the health needs of the sickest Black patients.
Bias and fairness are a “pervasive challenge for AI systems, since they typically learn from imperfect datasets that include human bias and historical bias,” she said.
During a session on health equity, U.S. Department of Health and Human Services Secretary Xavier Becerra said he directed the department’s Office of the National Coordinator for Health Information Technology to take a “deep look” at algorithmic bias. During a separate ONC town hall, ONC chief Micky Tripathi said his team is looking at algorithmic bias with a focus on the role that electronic health records software plays.
“EHR data is increasingly the data that’s used to train and develop algorithms,” he said. “It’s the vehicle that injects the results of algorithms … into administrative and clinical decision-making.
Download Modern Healthcare’s app to stay informed when industry news breaks.
2. Cybersecurity is about people—not just technology. Hospitals should think about preparing for cyberattacks the same way they think about other disaster preparedness plans, by creating a comprehensive business continuity plan that outlines what to do if the organization is hit by a hacker, said Julie Chua, branch chief of the risk management program at HHS.
That should include a plan for continuing to deliver clinical care and communicating with the public, and establishing and practicing it before an actual cybersecurity incident. It’s important for everyone in an organization to know who’s in charge of each part of an incident response plan, she added.
“Cybersecurity risk is not solely an IT issue,” Chua said during a panel discussion on cybersecurity.
As hospitals face a growing number of ransomware attacks, which can disrupt and delay patient care, it’s critical to train clinicians not only on cybersecurity best practices, but also on how to continue care for patients during system downtime, said Dr. Christian Dameff, a clinical informaticist and security researcher at UC San Diego, during a session on cybersecurity and patient safety.
Everything from EHRs to imaging technology can go offline during a cyberattack, he said.
“It’s about building disaster recovery and resilience among your clinical workforce,” Dameff said. “You have to train them how to use tools around them when the technology’s gone.
3. Telehealth’s potential to revamp the healthcare ecosystem. Telehealth use soared in the early days of the COVID-19 pandemic, as the Centers for Medicare and Medicaid Services significantly relaxed restrictions on telehealth. Since then, telehealth visits have plateaued at a lower level than 2020, but are still notably higher than pre-pandemic.
Telehealth is here to stay, according to healthcare executives at HIMSS22.
But the technology shouldn’t be used as a substitute for in-person care, said Jodie Lesh, chief transformation officer at Kaiser Permanente, during a panel discussion on healthcare transformation. “It will create a different ecosystem,” Lesh said. “But it’s hard to see that right now, because we’re in these incremental steps.”
John Glaser, executive in residence at Harvard Medical School and another participant on the panel, likened it to how automobiles not only substituted for other slower modes of transportation, but eventually paved the way for suburbs to exist, since people were able to live farther away from where they worked.
“What’s the second-order effect here?” Glaser said of telehealth, offering that more virtual care could lead to widespread hospital-at-home programs.
4. Tying social determinants into care plans. Healthcare organizations discussed how they’re using data to inform programs that tackle social determinants of health—including a University of Virginia Health System program to dispatch community health workers to areas with high rates of chronic diseases and a Highmark Health program to screen patients for housing, food and other challenges.
There has been an increasing number of contracts between healthcare organizations and community-based organizations in recent years, consistent with the move toward value-based care, said Kelly Cronin, deputy administrator of the Center for Innovation and Partnership in HHS’ Administration for Community Living, during a session on social care.
Those contracts can involve sharing IT systems so hospitals can send electronic referrals directly to a community-based or social service organization.
For hospitals, the first step of such programs is to collect data on which patients might need support from a community-based organization, followed by figuring out a way to refer those patients to the appropriate resources. Sending those referrals requires organizations to use interoperable software and shared data standards.
Groups like the Gravity Project are working to develop and sets standards for social determinants data.
ONC has also been adding social determinants into its data standards, including the U.S. Core Data for Interoperability and interoperability standards advisory.
Cronin said the industry needs to figure out financing for healthcare programs that refer patients to community-based organizations, since many community-based organizations don’t have the money to invest into IT tools and might not even have the workforce to take on additional referrals. ACL in 2020 launched a challenge to encourage community leaders to partner with healthcare organizations to develop successful and scalable approaches, which the agency is still working on.
5. Interoperability for public health. ONC’s Tripathi and Dr. Daniel Jernigan, deputy director for public health science and surveillance at Centers for Disease Control and Prevention, discussed an effort to improve IT infrastructure for public health agencies, dubbed “North Star” architecture.
ONC and CDC officials have previously said lack of interoperability between and among healthcare and public health entities hindered COVID-19 response.
As part of the agencies’ effort to modernize public health data systems, ONC is assisting CDC with developing a cloud-hosted environment that will make it easier to exchange information among state, territorial, local and tribal public health departments and the CDC, including shared tools and applications that are made available to various agencies.
“We think of those as building blocks,” Tripathi said of applications for core infrastructure like patient matching and data normalization, which could be used across agencies.
The program will also tie in existing data-exchange work from ONC, like the Trusted Exchange Framework and Common Agreement.