Blab AI
KLAS Report Shows AI Adoption Improves Clinician EHR Experience but Highlights Training Gap
A fresh KLAS Arch Collaborative 2026 report, titled AI Insights from the Arch Collaborative 2026: An Early Look at How Epic Customers Are Leveraging AI to Improve the Clinician Experience, delivers the first large‑scale evidence that artificial intelligence (AI) tools embedded in electronic health records (EHRs) can lift clinician satisfaction and perceived operational efficiency.
The study canvassed more than 121,000 clinicians across 92 U.S. health systems that use Epic, the country’s dominant EHR vendor. Twelve of those organizations were flagged as “progressive Epic‑backed” and supplied detailed usage data. The resulting picture is clear: clinicians who employ at least one AI tool rate the EHR’s ability to support operational efficiency 7 percentage points higher than those who do not. In terms of overall experience, the average Net EHR Experience Score (NEES) for AI users sits at 72.2, compared with 64.9 for non‑users.
Yet the data also reveal a hard‑wired ceiling. Satisfaction climbs steadily as clinicians adopt up to four distinct AI tools, but the perceived value plateaus when a fifth or sixth tool enters the mix. The authors call this a “saturation limit,” describing the effect as “workspace noise” that arises when too many uncoordinated AI workflows vie for a clinician’s attention.
Training proves to be the missing linchpin. Less than one‑quarter of clinicians who have adopted AI tools report that they received adequate formal training on how to validate and incorporate AI‑generated content into their daily workflows. The report notes that adequate training correlates with a 20‑point jump in overall system satisfaction.
Role‑specific usage patterns also emerge. Physicians and advanced practice providers (APPs) lean on AI primarily for ambient visit‑note drafting (54 % and 52 % respectively) and patient‑history summarization (51 % and 52 %). Nurses, by contrast, rely on AI mainly for shift and patient‑stay summaries (30 %) and for managing high‑volume patient‑message workflows. Allied health professionals focus on semantic access to the patient story, using AI‑generated patient‑history summaries (48 %) and advanced EHR search tools (23 %).
When the report dissects efficiency gains by task, the biggest gains stem from focused operational workflows. Order creation sees a 9 % efficiency spread, while summarizing shifts or patient stays yields a 6 % spread. EHR data discovery tools add another 5 %.
The implications are clear: health systems must move beyond a blanket strategy of installing AI tools. Instead, they should craft role‑specific training pathways and feedback loops that align AI capabilities with each clinician group’s unique workflow pain points. The authors argue that without such targeted enablement, the promise of AI to curb clinician burnout and documentation burden remains unrealized.
Epic’s own communications indicate that the company is expanding AI integration across its platform, aiming to embed AI directly into clinical, administrative, and patient‑facing workflows. The KLAS report supplies a data‑driven lens on how those efforts are playing out in practice.
In sum, the KLAS Arch Collaborative 2026 report confirms that AI tools can elevate clinician experience and operational efficiency, but it also underscores a significant training gap and a saturation point beyond which additional tools add little value. Health systems that wish to realize sustained benefits will need to focus on role‑specific deployment and comprehensive training.
The study canvassed more than 121,000 clinicians across 92 U.S. health systems that use Epic, the country’s dominant EHR vendor. Twelve of those organizations were flagged as “progressive Epic‑backed” and supplied detailed usage data. The resulting picture is clear: clinicians who employ at least one AI tool rate the EHR’s ability to support operational efficiency 7 percentage points higher than those who do not. In terms of overall experience, the average Net EHR Experience Score (NEES) for AI users sits at 72.2, compared with 64.9 for non‑users.
Yet the data also reveal a hard‑wired ceiling. Satisfaction climbs steadily as clinicians adopt up to four distinct AI tools, but the perceived value plateaus when a fifth or sixth tool enters the mix. The authors call this a “saturation limit,” describing the effect as “workspace noise” that arises when too many uncoordinated AI workflows vie for a clinician’s attention.
Training proves to be the missing linchpin. Less than one‑quarter of clinicians who have adopted AI tools report that they received adequate formal training on how to validate and incorporate AI‑generated content into their daily workflows. The report notes that adequate training correlates with a 20‑point jump in overall system satisfaction.
Role‑specific usage patterns also emerge. Physicians and advanced practice providers (APPs) lean on AI primarily for ambient visit‑note drafting (54 % and 52 % respectively) and patient‑history summarization (51 % and 52 %). Nurses, by contrast, rely on AI mainly for shift and patient‑stay summaries (30 %) and for managing high‑volume patient‑message workflows. Allied health professionals focus on semantic access to the patient story, using AI‑generated patient‑history summaries (48 %) and advanced EHR search tools (23 %).
When the report dissects efficiency gains by task, the biggest gains stem from focused operational workflows. Order creation sees a 9 % efficiency spread, while summarizing shifts or patient stays yields a 6 % spread. EHR data discovery tools add another 5 %.
The implications are clear: health systems must move beyond a blanket strategy of installing AI tools. Instead, they should craft role‑specific training pathways and feedback loops that align AI capabilities with each clinician group’s unique workflow pain points. The authors argue that without such targeted enablement, the promise of AI to curb clinician burnout and documentation burden remains unrealized.
Epic’s own communications indicate that the company is expanding AI integration across its platform, aiming to embed AI directly into clinical, administrative, and patient‑facing workflows. The KLAS report supplies a data‑driven lens on how those efforts are playing out in practice.
In sum, the KLAS Arch Collaborative 2026 report confirms that AI tools can elevate clinician experience and operational efficiency, but it also underscores a significant training gap and a saturation point beyond which additional tools add little value. Health systems that wish to realize sustained benefits will need to focus on role‑specific deployment and comprehensive training.
