Table of Contents
Hackensack Meridian Health’s path to becoming the first health network in the nation to earn The Joint Commission’s Responsible Use of AI in Healthcare Certification began years before the accrediting organization launched the program, culminating in the Hackensack Meridian Joint Commission AI certification announced July 29, about seven weeks after the voluntary program’s June 1 introduction.
Why the Hackensack Meridian Joint Commission AI Certification Wasn’t an Overnight Achievement
The Edison, N.J.-based health system had already established multidisciplinary governance structures to evaluate AI technologies, according to Joel Klein, MD, executive vice president and chief digital and information officer. “It’s not an overnight thing,” Klein told Becker’s.
What the Certification Actually Evaluates
The Joint Commission certification evaluates healthcare organizations across areas including AI governance and oversight, patient privacy, risk and bias assessment, ongoing monitoring, transparency and staff training, a broad scope covering both the technical and organizational dimensions of responsible AI deployment.
How the Hackensack Meridian Joint Commission AI Certification Survey Process Worked
For Hackensack Meridian, the review resembled other Joint Commission survey processes, Klein said. Surveyors held numerous sessions with the organization, conducted tracers with staff members and reviewed documentation outlining how the health system evaluates AI tools for factors such as safety and efficacy. They also examined contracts. “They wanted to look at everything,” Klein said, describing the process as comprehensive and credible.
How Governance Efforts Predated the Formal Certification Program
Hackensack Meridian’s AI governance efforts began years earlier as the technology started gaining traction in healthcare. Rather than leaving decisions to a small technology team, the health system assembled leaders from across the organization to evaluate proposed AI technologies and consider whether they could safely improve care, maintain appropriate clinician involvement and be defended to patients and their families. Klein estimated that dozens of people participate in AI governance and monitoring at Hackensack Meridian to varying degrees, including program managers who coordinate reviews and senior leaders who weigh in on individual technologies.
Why Administrative AI Tools Face the Same Scrutiny in This Governance Model
That governance also extends beyond AI tools that directly influence clinical care. While clinical AI may appear to warrant greater scrutiny than administrative applications, Klein said seemingly back-office technologies can also create risks for patients. An improperly deployed tool could, for example, result in incorrect supplies being ordered or interfere with a patient’s relationship with an insurer.
Why This Distinction Matters for Governance Scope
By extending the same rigorous evaluation to administrative and operational AI tools as to clinical ones, Hackensack Meridian’s approach reflects an understanding that patient harm can originate from seemingly low-stakes back-office systems just as readily as from tools directly involved in diagnosis or treatment.
Three Priorities Behind This Hackensack Meridian Joint Commission AI Certification Approach
For CIOs building their own AI governance programs, Klein pointed to three priorities. First, governance should be multidisciplinary rather than controlled by one or two decision-makers. At Hackensack Meridian, that includes input from compliance, legal, clinical, academic and nursing leaders, physicians and patient and family perspectives.
Building an Objective Evaluation Framework
Second, health systems should develop an objective framework for evaluating AI technologies, such as a scoring rubric, rather than relying solely on individual judgment. “Don’t just eyeball it. Don’t keep it subjective,” Klein said. A standardized approach can help organizations evaluate technologies more consistently as the number and types of AI tools continue to expand.
Setting Clear Boundaries on Vendor Data Access
Third, health systems need clear expectations for how technology vendors can use patient information. “You’ve got to remember that it’s not your data; it’s the patient’s data,” Klein said. He said health systems should carefully scrutinize vendor requests for long-term access to patient data, including requests to use information to improve or commercialize their products. Although some arrangements may be appropriate with sufficient safeguards, health systems should establish what uses they are willing to permit before signing contracts.
Why Discipline Becomes Harder Under Clinical Pressure
That discipline can become more difficult as clinicians and health system leaders push to adopt rapidly advancing technologies, Klein said. Healthcare professionals often see technologies that could potentially improve outcomes or reduce administrative burdens and understandably want access to them quickly. But speed should not override governance, he said. “If we just deploy something because of the perceived urgency, eventually you’re going to make a big mistake,” Klein said.
What This Hackensack Meridian Joint Commission AI Certification Means Going Forward
With Hackensack Meridian’s certification arriving just seven weeks after the Joint Commission’s program launched, other health systems pursuing this same certification should expect a comprehensive, multi-session survey process that examines governance structures, documentation, and vendor contracts as thoroughly as it examines the AI tools themselves. Given Klein’s emphasis that this achievement rested on years of prior governance-building rather than a rapid compliance push, health systems just beginning to formalize AI oversight may need a longer runway before pursuing certification themselves.
What to Watch Going Forward
As more health systems pursue The Joint Commission’s Responsible Use of AI in Healthcare Certification following Hackensack Meridian’s example, industry observers will likely watch which specific governance structures, multidisciplinary committees, objective scoring rubrics, or vendor data contract standards, prove most transferable to other organizations. Given Klein’s warning that deploying AI tools under perceived urgency risks a “big mistake,” this Hackensack Meridian Joint Commission AI certification case may reinforce a broader industry lesson that sustainable AI governance requires sustained institutional investment well before any formal certification process begins.
For more healthcare industry updates, insights and news, visit DistilINFO. Click here to subscribe to stay informed.
