AI Adoption Requires Organizational Alignment and Trust, Say Healthcare Leaders

⁢Updated June 22, 2025
⁤ ⁣

Artificial⁣ intelligence is poised to revolutionize healthcare, but the primary ⁤obstacle⁤ isn’t the technology itself. According to⁢ panelists at ⁣ViVe25, the real challenge lies in ensuring that healthcare organizations and their clinical staff are⁤ prepared to embrace AI.

During a session titled “I’ve Got 99 ⁤Problems,but⁢ Tech Ain’t One,” healthcare⁢ leaders explored ⁢the practical⁢ challenges of AI ‍adoption,including aligning organizational⁣ goals,addressing workforce concerns,and finding the right balance between automation and human oversight.

“Technology is actually the easiest part,” said Dr.⁤ Nishit Patel, ⁣VP and CMIO at Tampa General Hospital. “The real challenge is ‍getting physicians,nurses,and frontline‍ staff comfortable with using AI ‍in their daily work.”

Nishit Patel,MD,VP and CMIO,Tampa General Hospital

Panelists included Dr. Angel Mena, Chief⁣ Medical Officer at symplr; Dr. Shoma desai, Executive⁤ Director of Digital Innovation at Cedars-Sinai Medical Centre; and Dr.⁢ Rebecca Miksad, Chief Medical Officer at Color⁣ Health.

Before AI can realize its full potential, healthcare organizations must agree on priorities and ensure ‍that frontline staff are adequately trained ⁤to use it effectively.

Mena cited a national survey revealing a significant lack of alignment among technology and operational leaders. “more than 50% of the leaders we surveyed were not on ⁤the ⁣same page about AI priorities,” he said. “Before implementing ⁢any AI solution,⁤ we need to be aligned⁣ on what problem we are solving.”

Desai echoed this sentiment, noting that‍ AI projects often fail due⁤ to organizational structures⁤ that don’t support their deployment. “if there’s‍ no clear strategy on where AI fits into clinical and operational workflows, it creates friction,” she said.‍ “We need to integrate AI into existing processes in a way that makes sense for clinicians, rather than⁣ forcing⁤ them⁢ to change how they work.”

The Human Factor: Overcoming ⁢Workforce Resistance

Even when an organization ⁤is ⁢ready for AI, individual resistance can hinder adoption. Patel shared an example from Tampa General, where an AI tool designed to assist with denials and appeals processing initially boosted productivity by 20%. However,some employees hesitated to use the tool,fearing job displacement.

“One ⁤individual actually had a slight decrease in productivity because they were afraid AI would replace their job,” Patel said. “even though the organization embraced AI, we had to address these concerns at the individual level ⁣to ensure full adoption.”

Mena emphasized ‍the importance of building trust⁤ during AI implementation.”People fear AI because ⁣they worry ⁣about job security,” he said. “We must reassure ⁢them⁣ that AI is here to assist, not replace.”

This fear extends to⁣ patients, who might potentially‍ be uneasy about AI’s role in clinical decision-making. Patel noted that while ⁣many in healthcare technology are optimistic about AI,⁣ the ⁢general public remains skeptical.

“over 60% of patients are terrified of AI being used in healthcare,” he said. “They worry that an algorithm will decide their fate without human oversight. Especially post-COVID,trust⁤ in the healthcare system has been eroded,and we have to rebuild that trust carefully.”

Desai stressed the importance of piloting AI solutions with key stakeholders before system-wide implementation. ⁣”We started small, tested repeatedly, and ensured that leadership and frontline users were involved at every step,” she said. ⁢This iterative approach allowed Cedars-Sinai ⁢to refine AI tools and build ⁤confidence in ‍their effectiveness.

One area where Cedars-Sinai has seen success ⁢is AI-driven clinical documentation. “We’re leveraging AI for ⁣intake, ambient listening, and clinical summarization,” Desai said. “by streamlining documentation, we ⁣give physicians more time to focus on patient ⁤care.”

Patel compared AI adoption to the introduction of bedside ultrasound. “At first, there was skepticism, but ⁣now you⁢ can’t imagine an emergency department without it,” he said. “AI will ⁣follow ⁣the same trajectory-it will start as a tool some are hesitant to use, but over time, it will become indispensable.”

AI’s Potential: Clinical and Operational Wins

Beyond⁢ documentation, AI is demonstrating its value in both clinical and ⁤operational settings. ⁣Patel highlighted ‍Tampa General’s AI-driven sepsis program, which has substantially reduced mortality rates.

“Sepsis is one⁤ of the leading causes of death in hospitals, with a typical mortality rate of 15-18%,” he said. ⁢”Through AI‍ and process improvements, we’ve been able ⁣to bring that down to⁢ under 7%. That’s saving real lives-over 400⁤ patients have gone home to their families because of these advancements.”

On the operational ⁢side,AI is addressing inefficiencies in prior authorization and scheduling. Patel shared how Tampa General partnered with a⁤ technology vendor to⁢ streamline authorization for procedures.

“Previously, ⁣we had staff⁤ manually reviewing notes and resubmitting authorizations, a process that could take up ⁣to 48 hours,” he said.”Now, AI automates much of that process, reducing delays and administrative burden.”

Miksad highlighted how AI is helping to standardize cancer care.”AI can scan patient charts and flag missing data, reducing the time it takes to‍ gather details for treatment decisions,” she said.”It’s not about replacing oncologists-it’s about ⁢making sure they have all the information they ⁤need, faster.”

As AI adoption grows, healthcare organizations must establish ⁢robust governance frameworks to ensure patient safety and data integrity.

“AI is only as⁢ good as the ⁣data it learns from,” Patel said. “We have to ensure our data is accurate,⁣ unbiased, and accessible, or we ⁣risk amplifying‍ existing disparities.”

Miksad ⁤emphasized the importance of traceability in AI-driven decisions. “Physicians and patients need to understand how AI reaches its conclusions,” she said. “If a recommendation is based on a⁤ lab value, ‍we need‍ to⁢ be able to trace that value back to its source, just like the FDA requires for drug approvals.”

The panelists agreed that human oversight remains essential, ‍at least ⁢for now. “Even if AI were perfect tomorrow, society⁢ isn’t ready to accept fully automated decision-making ‍in healthcare,” Patel said. “Patients need to know that a human is involved⁣ in their care.”

What’s next

Miksad said ‍the key to successful AI adoption⁣ lies in striking the right ⁢balance between automation and human expertise.”AI can‍ handle the routine 80% of cases, but we need‍ to ensure that the remaining 20%-the complex, high-risk cases-receive the⁢ human attention they deserve,” she said.

Patel reinforced the need ⁢for healthcare leaders to be proactive ‍in AI⁤ adoption.”There’s risk in deploying AI, but there’s also risk in doing nothing,” he said. “If we get this right, we can make care safer, more efficient, and more cost-effective. But if we ‍get it wrong,we ⁢risk eroding patient trust even further.”

mena added a ⁣note on the importance of collaboration. “The future of AI in healthcare isn’t about ⁣any one company or‍ health system-it’s ⁣about partnerships,” he said. “By working together, we can build AI solutions⁤ that ⁣truly enhance patient care and support the clinicians ‍who provide it.”