Samwise Healthcare IT Newsletter — Thursday, July 10, 2026

Samwise Healthcare IT Newsletter

Thursday, July 10, 2026

Healthcare IT  ·  Cybersecurity  ·  Policy  ·  AI Analytics  ·  Interoperability
All your morning news, carefully curated and summarized daily
EHR/EMRWORKFORCE

Epic Distributes Presidential Duties Among Four Leaders as Sumit Rana Prepares August Exit

Epic Systems announced Tuesday that four senior leaders — Seth Howard, Mark Lipsky, Erv Walter, and Garrett Adams — will absorb the responsibilities of outgoing president Sumit Rana, who leaves August 14. Rana, a 30-year Epic veteran widely regarded as a probable successor to founder and CEO Judy Faulkner, disclosed his departure in a July 3 email, citing the death of his father last November and a wish to support his mother in India. Faulkner affirmed the company’s depth in a July 7 statement: “We also have an outstanding group of seasoned people throughout Epic whose leadership is critical to the company’s success.”

Sources: Healthcare IT News   ✉︎ Email 💬 Text

AI/ANALYTICS

NewYork-Presbyterian Deploys OpenEvidence Clinical AI Across 10 Hospitals and 350 Sites

NewYork-Presbyterian announced July 9 it will deploy OpenEvidence’s clinical AI search tool across all 10 of its hospitals and nearly 350 care sites in New York City and Westchester County. The enterprise-wide rollout also extends access to academic partners Columbia University Vagelos College of Physicians and Surgeons and Weill Cornell Medicine. OpenEvidence provides a conversational AI tool that answers clinical questions using peer-reviewed research and evidence-based guidelines. The deployment covers 45,000 employees and affiliated physicians. It represents one of the broader system-wide rollouts of evidence-based clinical AI at a major U.S. health system, which operates across New York City’s five boroughs and Westchester County.

Sources: Becker's Hospital Review   ✉︎ Email 💬 Text

AI/ANALYTICS

Healthcare CIOs Warned That Confident AI Outputs Pose Growing Clinical Risk Beyond Deployment

As generative AI becomes embedded in clinical workflows, healthcare CIOs face a new governance challenge: AI systems that produce plausible but inaccurate outputs with apparent authority. Lior Eshel, CEO of medical imaging company TestDynamics, warns that a model cannot flag its own hallucination, making independent verification essential. Eshel recommends hospitals compile a full inventory of deployed AI tools, secure contractual rights to independent performance testing, and validate systems against local patient populations rather than vendor-reported benchmarks. “Shared accountability is the only workable model,” he told Healthcare IT News, “and it carries a precondition: shared visibility.” He envisions AI governance becoming a continuous operational discipline rather than a one-time procurement exercise.

Sources: Healthcare IT News   ✉︎ Email 💬 Text

AI/ANALYTICS

Healthcare Leaders Divided on Autonomous AI in Medicine, but Agree Accountability Must Come First

Three healthcare executives offer divergent timelines for autonomous AI in clinical practice but agree that accountability — not model performance — will define the outcome. Dr. Niki Panich of Penguin AI notes that narrow autonomous AI already exists: FDA-authorized diabetic retinopathy screening systems have operated without physician involvement since 2018, though he says a general AI clinician is “a decade-plus” away. Dr. Jay Anders of Medicomp Systems argues clinicians must remain responsible: “We love to personify AI as a human replacement; however, humans are the ones who must remain responsible.” Bob Taylor of Altera Digital Health calls for AI to function as a “decision-support engine” grounded in physician-approved clinical standards.

Sources: Healthcare IT News   ✉︎ Email 💬 Text

EHR/EMR

KLAS: Epic Expands Global EHR Share While Oracle Suffers Major Losses Including Failed Sweden Implementation

The global electronic health record market fell to a near five-year low in 2025, with 189 decisions affecting 447 hospitals and more than 81,000 beds, according to KLAS Research’s Global Hospital EHR Market Share 2026 report released Tuesday. Epic grew its international presence through its EHR-as-a-service program, while Oracle Health recorded 34 hospital and nine organizational losses — more than half tied to a failed implementation in Sweden’s Västra Götaland Region, which ended its Oracle contract after seven years. About one-third of Oracle customers KLAS surveyed said the vendor is no longer part of their long-term plans. IQVIA had its strongest year yet in Europe and the Middle East.

Sources: Healthcare IT News   ✉︎ Email 💬 Text

WORKFORCEAI/ANALYTICS

Nurse AI Use Nearly Triples in One Year but Most Remain Skeptical of Output Accuracy

Registered nurses’ AI use in clinical workflows nearly tripled in one year — from 15% to 44% — per Incredible Health’s 2026 Annual State of Nursing report, which surveyed 2,240 U.S. nurses. Those using AI regularly are less likely to fear job displacement (37%), but 83% say AI output is “rarely or only sometimes accurate enough to act on without checking it first.” Only 8% of nurses report their employer has set a clear AI strategy, and one in five say AI tools “just appear” without explanation from leadership. Among nurses using AI for clinical tasks, 28% use Epic AI, 8% use ambient documentation tools, and 4% use OpenEvidence.

Sources: Healthcare IT News   ✉︎ Email 💬 Text

AI/ANALYTICS

SE Health CDO: Treating Data Like Cash Is Prerequisite to Trustworthy Healthcare AI

Robert Slepin, chief digital officer at SE Health, one of Canada’s largest home health providers, argues healthcare leaders manage financial assets with precision but treat data with far less rigor. “From my experience, healthcare providers do not govern and manage data assets as well as they oversee and administer financial resources such as cash,” Slepin told Healthcare IT News. SE Health has established a Data Governance Council to clarify ownership and resolve contested definitions before major system go-lives. Slepin, also a CIO advisor in Epic’s emeritus program, says poorly governed data raises the likelihood that AI systems will produce biased or confidently false results.

Sources: Healthcare IT News   ✉︎ Email 💬 Text

AI/ANALYTICSTELEHEALTH

Baptist Health Partners With Vega Health to Deploy AI Patient Screening for Hospital-at-Home Program

Jacksonville, Fla.-based Baptist Health has partnered with Vega Health to deploy an AI screening tool for its Hospital at Home program, which launched in February at Baptist Medical Center Jacksonville. Vega Health’s large language model scans inpatient charts against Baptist Health’s eligibility criteria to identify candidates for home-based hospital care, then routes them to clinical staff, per a July 9 release. The platform runs inside Baptist Health’s secure cloud, keeping patient data in-house. The partnership plans two additional use cases: surfacing medication history during emergency-department-to-inpatient transitions and using predictive analytics to flag health concerns earlier.

Sources: Becker's Hospital Review   ✉︎ Email 💬 Text