Josh Kitfield

From a patient perspective, some of the most impactful innovations are simple. For example, since joining KP as a member, one of things that "wowed me" was the ability to self-schedule appointments. No need to call a telephone number and wait on hold just to schedule an appointment. While this is tremendously convenient, the primary risk of self-scheduling is missing acute conditions that require an immediate response. While healthcare orgs will include a disclaimer to call 911; it puts the onus on patients. I wanted to highlight this innovative program from the SOUTHERN CALIFORNIA PERMANENTE MEDICAL GROUP who developed and deployed (at a scale) a technology-enabled service that allows patients to seek the right care at the right time. The online workflow is as follows: 1) A patient who wants to schedule an appointment is directed to enter their reason for visit 2) this is interpreted by the "Clinical Alert System" to detect high-acuity conditions (e.g., chest pain or abdominal pain) that also integrates pt clinical data to provide a customized assessment. 3) This prompts specific clarifying questions and when appropriate, a recommendation is made to contact a human agent for clinical escalation, rather than schedule an appointment. 4) For those with a low-acuity concern; the "Care Navigation System" guides the pt to the appropriate next step - scheduling an appt, conducting an asynchronous visit (e.g. "e-visit) or triage to another non-visit modality (e.g., referral to pharmacy system for med refill). This ensures that high-acuity pts are flagged and quickly escalated (rather than scheduling an appt in the future); and other pts are routed to the right care method to resolve their complaint. See a comprehensive write-up published in NPJ Digital Medicine. https://lnkd.in/gk7KUujq Kudos to the team that developed, tested, deployed and continue to manage this program. Demonstrates, yet again, KP's ability to deploy AI systems responsibly and at scale. Dinh Nguyen, MD, MSHI, Khang Nguyen, Sinjin L.

103

2 Comments

I am delighted to announce the first episode of my new podcast Healthcare AI Pioneers. I interview guest Nassib Chamoun and Brenda Campbell about Houston Methodist innovative approach to integrating AI at the point of care within Epic in partnership with the Healthcare Data Analytics, Inc.. The focus is on identifying EHR insights in patients who are at high risk of readmission and/or have advanced illness, and by surfacing the data, allow for more tailored care management and improved transitions in care.  It’s a great example of AI working at scale and producing tangible results including markedly lower readmissions and costs of care. This is why I started this podcast in the first place. So that stakeholders in the community of people trying to make AI work at scale can learn from the success of others. This is a fantastic story to share. I hope you enjoy it. If you like it, please consider subscribing and/or reposting. Some of the key takeaways: ·     To gain frontline adoption, it is important to have clear and trusted AI risk scores ·     Implementing adaptable tools that use real-time data using AI is possible ·     Predictive analytics must fit into real workflows to make them effective https://lnkd.in/e2zbSJvx US Acute Care Solutions Robert Wachter Noah Pines

20

2 Comments

Prior Authorization AI implementations from Availity, Cohere, Optum, and others report impressive automation numbers. For example, Availity: 80% touchless processing and Cohere: 90%. These numbers are focused on how often the agent reached the payer and submitted a decision. I started wondering: what about knowing how the decision was reached? The HL7 Da Vinci Project has created implementation guides that define the workflow of an integratable, interoperable prior authorization process that can be used in both clinical and pharma applications. I used their guidance to architect an agentic application for prior authorization. In a human process, you can ask an employee how a decision was reached. It's a bit different when you are talking to an AI Agent. When I dug into it, the question became surprisingly hard to answer: *Which version of which coverage criteria was the agent following on the date of that denial?* Not "we believe it was following policy X." The actual version. Logged. Verifiable. Da Vinci defines the workflow — not the implementation. And when it comes to AI-generated decisions in PA, that implementation gap has real consequences. Payer coverage criteria arrive as PDFs. Vendors maintain proprietary copies, manually updated. There's no push notification when a payer changes its criteria. No version log tied to each decision. That gap has a name: CHAI-PA-TRANS-003, Context Version Auditability. It's a named compliance requirement from the Coalition for Health AI, developed by 100+ experts across UnitedHealth, CVS Health, Blue Cross Blue Shield, Mayo Clinic, and Stanford. And it's not the only pressure point: - CMS-0057-F: Denial reasons must cite specific policy provisions. Public reporting of PA metrics begins March 31, 2026. - WISeR: Federal AI PA pilot across Medicare in six states, under direct monitoring through 2031. - State legislation: Texas, Arizona, and Maryland now require documented human oversight for AI adverse determinations. I wrote up what I found — the gap in the current architecture, why the three dominant approaches to criteria management all fail the audit question, and what a governed context layer looks like in practice. https://lnkd.in/eZNunnxB

1

📽️ RTM in 2026: From Technology Add-On to Core Care Strategy RTM isn’t just evolving - care delivery is. In our latest webinar with Vikram Sethuraman, Bryan Lang PT, DPT, MHA, OCS and Mary Rose Strickland - we brought together clinical, operational, and product leaders to unpack what Remote Therapeutic Monitoring looks like in 2026 and why CMS updates signal something much bigger than code expansion. The key takeaway? ➡️ RTM is not a billing strategy. ➡️ It’s a care delivery strategy. As clinicians increasingly manage care longitudinally - not just during visits, but between them - RTM becomes a foundational tool for: •Improving adherence between visits •Increasing plan-of-care completion •Strengthening patient engagement •Driving outcomes with real-time insight •Supporting sustainable, scalable clinical models With CMS separating device supply from treatment management and reinforcing time-based expectations, it’s clear: RTM is meant to support continuous clinical engagement, not function as a tech add-on. During the discussion, our panel explored: •What changed in 2026 and why •The three primary RTM implementation models (full-service, self-service, hybrid) •Common operational friction points and how to mitigate them •Real-world rollout lessons from clinic operators •Financial sustainability benchmarks •How RTM supports PTs stepping further into first-contact and population health roles The clinics succeeding with RTM aren’t layering on software. They’re redesigning workflows, aligning teams, and embedding RTM into their care model. As healthcare shifts toward outcomes-driven, longitudinal care, RTM is quickly becoming a core pillar of modern practice. 🎥 Watch the full webinar here: https://lnkd.in/ejC5vksJ #RemoteTherapeuticMonitoring #MSKCare #RehabTherapy #RTM #CMS #HealthcarePolicy #DigitalCare #PatientEngagement Prompt Health

17

3 Comments

Healthcare Isn’t Missing More Data — It’s Missing Salient. Every healthcare organization is overflowing with data, yet leaders still ask: Why don’t our dashboards match? Why are numbers inconsistent? Why does it take weeks to get answers? Because this was never a “data problem.” It’s a structure and performance problem that most tools can’t solve. And that’s exactly where Salient is different. While others deliver dashboards, cubes, or “better reporting,” Salient delivers the outcome: a full performance optimization solution that connects technology + business acumen + AI into one engine. With Salient, organizations gain: a memory-based architecture for real-time exploration a unified model and semantic layer that keeps definitions consistent business logic tied to real human roles and real workflows AI that accelerates insight instead of adding noise The result? matching dashboards reliable metrics faster decisions earlier visibility into shifts measurable ROI across people, process, and technology Anyone can offer structured data. Only Salient delivers the clarity, alignment, and operational intelligence that healthcare has been chasing for years. If you're ready to replace fragmentation with true performance optimization, let’s talk.

8

Healthcare Isn’t Short on Data; It’s Short on Context Every system spits out dashboards, scorecards, and metrics. But more numbers don’t make decisions smarter—clarity does. Clinicians don’t need another report. They need signals that tell them what action to take. Healthcare CIOs must shift the narrative from “more data” to “less noise, more meaning.” The organizations that win in 2026 won’t be the ones drowning in BI tools—they’ll be the ones who: ⦁ connect data to meaningful outcomes ⦁ slice information into decision-ready insights ⦁ streamline alerts so clinicians act, not react The question isn’t Can we measure it? The question is What does it actually tell us to do? Stop feeding dashboards. Start feeding decisions. #HealthcareLeadership #HealthcareIT #DataWithPurpose #ClinicianExperience #HealthcareAnalytics #DecisionScience

13

2 Comments

Progress tests leadership.   Every industry—including healthcare—evolves over time. New technologies emerge. Customer expectations shift. Business models adapt. The question isn’t whether change will come, but how we respond to it.   In a recent article by McKinsey & Company, Arvind Krishna, Chairman and CEO of IBM shared a perspective that resonated with me: waiting for certainty is often how organizations get disrupted.   Deferring hard decisions and holding onto misaligned work or layering new technology onto outdated processes rarely creates momentum. Real progress comes from rethinking how work gets done and aligning strategy, talent, and culture around where you’re going.   AI is a powerful lever. But only if we rethink how work gets done, especially across back-office and administrative functions. Transformation ultimately comes down to alignment, so strategy, talent, and culture moving together with intention.   Leadership isn’t proven by waiting for clarity. It’s proven by the courage to move. Enjoy the full article here - https://lnkd.in/efXcQJm8

11

As we kick off 2026, healthcare leaders and AI innovators need to pay close attention to the evolving regulatory landscape. New state laws are reshaping how AI is integrated into patient care, emphasizing transparency and informed consent to build trust and protect patients. Take Texas's Responsible Artificial Intelligence Governance Act (TRAIGA), effective January 1, 2026: It mandates that healthcare providers deliver written notifications to patients when AI is involved in services or treatments, before or on the day of service, with exceptions for emergencies. This is a game-changer for ensuring patients understand AI's role in their care. California is leading with similar measures. AB 489, also effective January 1, 2026, prohibits AI systems from implying they hold healthcare licenses and requires clear disclosures when AI interacts with patients, preventing any confusion about who's providing care. Additionally, SB 243 targets companion chatbots (often used in therapy or support), requiring upfront notifications that users are engaging with AI and safeguards against harmful content, especially for minors. Other states are following suit with proposals gaining momentum. In Ohio, HB 525 (introduced late 2025) aims to restrict AI in therapy, banning independent decisions or direct client interactions without human oversight, while prioritizing transparency. Pennsylvania is exploring comparable bills focused on informed consent and restrictions in sensitive areas like mental health. And beyond these, states like Illinois (with its 2025 law on AI in psychotherapy) and Florida (pre-filing for 2026) are pushing for human review, consent requirements, and limits on AI in high-stakes scenarios. As someone passionate about AI-driven patient design, I see these regulations as opportunities to innovate responsibly, creating tools that empower patients while meeting compliance head-on. What's your take? How is your organization adapting to these changes? #AIinHealthcare #HealthcareGovernance #PatientTransparency #HealthTech #AIEthics

5

1 Comment

Many health systems talk about improving quality. Fewer operationalize it. In our webinar with the population health and quality leadership team at Houston Methodist, we will hold a candid discussion about what it actually takes to move quality reporting out of a compliance silo and into everyday operations across a large ACO. What stands out the most: 1. Quality improvement is not a reporting exercise. It is an organizational discipline. 2. Tracking and reporting must be embedded in operations, not reviewed at year end. 3. Translating CMS complexity into clinical clarity is leadership work. 4. Physician engagement requires focus and prioritization. Join us for what is sure to be an enlightening discussion. #CMS #HoustonMethodist #MSSP #HealthcareQuality

2