Physician assistants feel ready for patients, less ready for paperwork and AI rules

Medical practices hiring physician assistants may need to rethink what “ready on day one” really means and where practice managers should concentrate onboarding time.

A Wolters Kluwer survey of 203 practicing PAs found strong confidence in patient-facing work but weaker readiness for the administrative and technology demands that dominate many clinics. Nearly all respondents (96%) said they felt prepared for patient interactions when they started practicing, but only 80% felt ready for documentation tasks and just 29% said they were “very prepared” for documentation.

For practice managers, the message is less about clinical training and more about building reliable systems for documentation, payer friction and safe technology adoption.

Documentation is the recurring “workflow friction”

The report links onboarding stress to the reality that documentation and administrative work can pull clinicians away from patient care, even when the tasks are unavoidable.

In an interview about the survey, Kelly Villella, a Wolters Kluwer segment leader and director of product management, described that friction in practical terms: “They have to write things down … [and] type things up. It has to go into an EHR. They want to focus on the patient,” she said, adding that new clinicians can also struggle because “every place that she has rotated has a different system.”

The survey’s bottom line for clinic operators: even well-prepared graduates may still need structured, repeatable support to become efficient inside a specific practice’s EHR, templates, workflows and documentation standards.

AI is moving faster than policy

The survey also highlights how quickly artificial intelligence is entering day-to-day clinical work — and how far behind many workplaces are in setting guardrails.

More than half of respondents (56%) reported daily AI use, with most using it for documentation and EHR streamlining (61%) and nearly half using AI-enabled ambient scribe tools for patient note-taking (48%). Yet only 32% said they have clear employer guidelines for safe AI use, and just 18% described those rules as strict.

Villella urged practice leaders not to assume new hires share the same baseline understanding of what’s appropriate. “Don’t assume, because it’s so nascent and evolving, that every clinician coming in has the same understanding,” she said.

She also warned against what she called “shadow AI,” staff using open, unapproved tools in ways that could create compliance and privacy risks. “There are all kinds of open sources proliferating, and you might think … ‘let me run my notes through this.’ That is not acceptable,” Villella said, calling for clearer onboarding language about “acceptable AI” versus tools that might be fine at home but not in a clinical setting.

For managers, the operational takeaway is straightforward: if the practice is adopting AI, or even if staff are using it informally, leaders need a written policy, training and vendor vetting.

That recommendation also aligns with broader federal and standards activity. The National Institute of Standards and Technology’s AI Risk Management Framework encourages organizations to actively manage risks tied to how AI is designed, used and overseen. And the Office of the National Coordinator for Health IT’s “HTI-1” rule includes requirements aimed at transparency for predictive algorithms and AI used in certified health IT the systems used broadly across hospitals and physician offices.

Insurance headaches remain a top pain point

Practice managers may not be surprised that payer friction is still a central theme.

In the Wolters Kluwer survey, 70% of PAs said the profession has changed in the past three years, and the most frequently cited shift (42%) was “navigating insurance issues,” followed by AI adoption (40%).

That finding fits into a larger picture of administrative workload across medical practices. The American Medical Association’s 2024 prior authorization survey found that 94% of physicians said prior authorization delays access to necessary care, and 24% reported that prior authorization led to a serious adverse event for a patient, including hospitalization or other severe harm.

CAQH, which tracks administrative transaction efficiency, has also pointed to an ongoing gap between available electronic standards and real-world adoption including prior authorization workflows.

Federal policy is moving in the same direction. CMS finalized an interoperability and prior authorization rule designed to modernize data exchange and reduce burden, including API-based requirements and timelines that affect payers and provider workflows.

For practice managers, those developments matter because they shape hiring, staffing and training decisions: even if the practice’s clinicians are strong clinically, administrative complexity can still be the bottleneck that determines productivity, patient access and burnout.

A fast-growing workforce with onboarding implications

The stakes are growing because the PA workforce is expanding quickly.

The U.S. Bureau of Labor Statistics projects employment of physician assistants will grow 20% from 2024 to 2034, with about 12,000 openings per year on average. Wolters Kluwer’s report also cites rapid workforce growth and expects the profession to keep expanding in response to access needs and clinician shortages.

That means more practices, especially primary care and multispecialty groups, will be onboarding new PAs regularly, not occasionally.

What practice managers can do next week

Asked what a practice manager could do differently “next week,” Villella didn’t point first to buying a new tool. She pointed to process.

Her advice: “Sit down and document your current policy and revisit that … make a plan to evaluate with your vendors … [and] explicitly write some policy for the different clinicians on your team about what they can and cannot use today.”

In practical terms, that can translate into a short, manager-owned checklist:

• Standardize documentation onboarding: templates, macros, EHR “how we do it here,” and timeboxed coaching focused on speed and accuracy.
• Create an AI acceptable-use policy: define approved tools, prohibited uses (especially anything involving PHI in unapproved systems), and required “human-in-the-loop” checks.
• Vet vendors for workflow integration: prioritize AI features built into existing, trusted systems over ad hoc tool use, and bake training into rollout plans.
• Train for payer reality: set expectations on prior authorization processes, escalation paths, documentation requirements that reduce denials, and who owns which steps.

The Wolters Kluwer survey suggests PAs want that support from employers. Its conclusion calls for clearer workplace guardrails for safe digital tool use and more on-the-job AI training tailored to clinical needs.

For practice managers, the opportunity is to turn what often feels like “extra training” into a repeatable system, the kind that helps new clinicians spend less time wrestling with workflows and more time practicing medicine.