Samwise Healthcare IT Newsletter
Sunday, July 19, 2026
AMA Convenes New SNOMED CT to CPT Mapping Initiative
The American Medical Association has launched an initiative to develop SNOMED CT to CPT terminology mappings, aimed at streamlining electronic prior authorization ahead of the January 1, 2027, FHIR-based compliance deadline. The project will enable real-time electronic prior auth by connecting clinical terminology with administrative coding, supporting the CMS Health Tech Ecosystem. AMA is calling on payers, EHR developers, clearinghouses, prior authorization vendors, and standards organizations to participate in collaborative testing and implementation. AMA President Dr. Willie Underwood III said the effort aims to support a “seamless, standards-based ecosystem for physicians, hospitals, payers, and patients.”
Sources: Healthcare IT News Share ↗ ✉︎ Email 💬 Text
CHAI Launches PULSE Initiative to Bring GenAI to Public Health Agencies
The Coalition for Health AI announced PULSE—its Public Health Use Case and Learning Scaling Engine—a national initiative to help state, tribal, local, and territorial public health agencies responsibly adopt generative AI. The program will engage up to 2,000 public health practitioners across ten jurisdictions, piloting five key use cases including biosurveillance, SDoH mapping, and multilingual communications. OpenAI and Anthropic donated 10 enterprise licenses with up to 2,000 seats, while Accenture will manage onboarding. Pilots are expected to launch this fall, with playbooks for broader adoption to follow in 2027. According to NACCHO, nearly 40% of local health departments currently do not use AI.
Sources: Healthcare IT News Share ↗ ✉︎ Email 💬 Text
AI With Human Support Cuts No-Shows 24% at Graybill Medical Group
Graybill Medical Group achieved a 50% reduction in front-office costs and a 24% decrease in patient no-shows by layering AI into existing workflows rather than replacing them. The physician-led community practice partnered with Third Way Health, which combines AI agents and human expertise, to address communication center bottlenecks. Rather than buying software first, Graybill mapped operational problems before introducing any technology. Staff retained responsibility for conversations requiring empathy and judgment, while AI handled repetitive scheduling and routing tasks. COO Jamie Reddick said the experience confirmed that “technology should support an operational strategy rather than become the strategy itself.”
Sources: Healthcare IT News Share ↗ ✉︎ Email 💬 Text
White House Launches Gold Eagle AI Initiative to Defend Critical Infrastructure
The Trump Administration unveiled Gold Eagle, a government-wide initiative using frontier AI to detect and patch cybersecurity vulnerabilities across critical infrastructure “at a speed and scale never seen before.” Established via a June 2 executive order on promoting advanced AI innovation and security, Gold Eagle coordinates vulnerability scanning and security verification across software and networks, including healthcare systems. The effort involves the Department of Treasury, DHS, and CISA alongside private-sector partners. National Cyber Director Sean Cairncross said the initiative will reduce duplicative scanning and deliver prioritized, actionable remediation information to cyber defenders across all critical infrastructure sectors.
Sources: Healthcare IT News Share ↗ ✉︎ Email 💬 Text
'Headless AI' Coordination Layer Could Reshape Healthcare Operations
Luma Health co-founder Aditya Bansod argues the next major step for healthcare AI is “headless AI”—a coordination layer that works across applications, gathering information from multiple systems and completing work that today requires employees to navigate disconnected platforms. Rather than adding AI inside existing systems, the approach would connect EHRs, payer portals, and scheduling tools, enabling real-time prior authorization lookups and referral management without interface-hopping. Bansod stressed governance is essential: AI agents should carry individual identities and role-based access controls, with full audit trails at both application and platform levels. His vision is AI absorbing administrative coordination so healthcare workers focus on patient care.
Sources: Healthcare IT News Share ↗ ✉︎ Email 💬 Text
The Hidden Disparity Built Into Healthcare Interoperability
Healthcare interoperability has an equity problem: patients with sensitive records—behavioral health, substance use, or immigration status—must currently choose between sharing everything or opting out entirely. Dr. Hannah Galvin, chief health information officer at Cambridge Health Alliance, co-founded the nonprofit Shift to address this all-or-nothing gap. Shift is developing standards-based data segmentation that lets patients share portions of their record by category, with consent managed through portals like MyChart. The technical challenge is significant and funding remains the biggest barrier, Galvin says. Until granular consent tools exist, interoperability inadvertently creates care quality disparities for vulnerable populations who opt out to protect sensitive data.
Sources: Becker’s Hospital Review Share ↗ ✉︎ Email 💬 Text
Health AI Regulation Gaps Span Ambient Scribes and Prior Authorization
Most health AI tools in use today fall outside the oversight framework patients and executives might expect, according to a Bipartisan Policy Center report. The most common applications—ambient scribes and payer prior authorization tools—are not FDA-regulated medical devices. Oversight is fragmented across FDA, CMS, the HHS Office for Civil Rights, FTC, and ONC, with state rules layered on top. This complexity creates a chilling effect on clinical AI for disease detection. Developers have an incentive to avoid FDA classification, and the current administration has proposed rolling back ONC’s AI transparency requirement for EHR-embedded tools. Bipartisan consensus on a federal framework exists in principle, but not in practice.
Sources: Becker’s Hospital Review Share ↗ ✉︎ Email 💬 Text
The Misstep Health Systems Without Dedicated AI Teams Can't Afford
Smaller health systems without dedicated AI teams are successfully deploying AI—but governance discipline matters more than budget size. Reid Health deployed 318 AI tools generating roughly $7.2 million in annual value without a chief AI officer, by routing every tool through a NIST AI Risk Management Framework evaluation with clinical and operational voices, not just IT. Willis Knighton Health reports zero failed deployments by deploying only solutions proven elsewhere and defining target value before investigating any tool. Midland Health CIO Shanidy Shook put it plainly: “We don’t have the luxury of chasing shiny objects. A misstep…creates clinician distrust.” Disciplined governance structures are the common thread, not staffing size.
Sources: Becker’s Hospital Review Share ↗ ✉︎ Email 💬 Text
Curated by JD · samwise.agency