Even mature EHR environments often have a clinician adoption problem, not just a technology problem. Many clinicians underuse available tools, shortcuts, templates, and workflow capabilities that could improve efficiency. Without ongoing training, coaching, and adoption support, physicians may develop inefficient workarounds that increase documentation time, administrative burden, and frustration.

1. Emphasize ongoing training. Offer regular EHR skills refreshers (e.g., monthly webinars, brief “tips and tricks” sessions, update briefings) and on-demand microlearning modules. Ongoing education is essential given frequent EHR and workflow updates. Continuous training helps clinicians learn time-saving best practices and adapt to new capabilities.
2. Promote peer champions and mentorship. Identify and leverage peer “super-user” champions who can serve as first-line support. These champions model efficient EHR use in daily clinical settings. Consider pairing clinicians who struggle with EHR efficiency with high-performing peers for practical, workflow-based coaching.
3. Use data to target training where it’s needed most. Employ EHR usage analytics to deliver role-specific and personalized training. For example, if data show a provider rarely uses a particular tool and consistently spends longer on a certain process than peers, that individual would benefit from targeted training on those tasks. This ensures training time is focused on areas of highest impact for each clinician.
4. Build dedicated training tracks for AI-enabled tools. As you introduce new technologies like ambient documentation, AI-assisted clinical decision support (CDS), or other AI tools, provide specialized training on how to review and use AI-generated outputs. Emphasize how to interpret and validate AI suggestions and reinforce that clinicians remain accountable for final decisions and documentation accuracy.
5. Reward and recognize training engagement. Encourage participation in ongoing training by providing incentives such as protected time for training, continuing education credits, formal recognition, or small stipends for those who lead peer training sessions.

Poorly designed CDS can overwhelm clinicians with low-value alerts and often fails to influence care decisions in a meaningful way. Many organizations have deployed a high volume of alerts and reminders — most of which are overridden or ignored. Furthermore, clinicians still spend substantial time on manual, repetitive tasks that could be streamlined or automated. If not carefully managed, adding more CDS or automation could increase cognitive burden and noise rather than improving efficiency or outcomes.

1. Evaluate existing alerts to reduce noise. Conduct a thorough review of existing CDS alerts and notifications. Retire or consolidate alerts that have low clinical value or high override rates. Reducing the total volume of alerts is often a necessary first step before introducing any new decision support, since alert fatigue is one of the most common EHR complaints.
2. Deploy predictive analytics that work in the background. Use predictive models (e.g., sepsis early warnings, readmission risk scores, or advanced diagnostic suggestions) that monitor patient data behind the scenes. These tools should only alert the care team when a significant risk or recommendation is identified, rather than interrupting clinicians with constant low-value pop-ups.
3. Let clinicians personalize their alerts. Where appropriate, give providers some control over which non-critical alerts they see. User-customizable settings can help fit decision support into individual workflows and rebuild trust in CDS, since clinicians can tailor it to their preferences.
4. Establish governance for CDS design and management. Effective CDS requires ongoing oversight to ensure alerts and automated actions remain up to date and truly helpful. Without governance, even well-intentioned CDS can become outdated, irrelevant, or excessive over time. Implement a multidisciplinary governance committee to review and manage CDS rules, address provider feedback, and oversee the introduction of any new decision support.

Patient messages and other inbox tasks have become a growing source of workload and after-hours effort for physicians. Many patient communications do not require a physician’s level of expertise yet still end up in doctors’ inboxes due to unclear workflows or inadequate triage rules. As patient portal use has increased, organizations have struggled to manage rising message volumes efficiently, contributing to physician burnout and reducing time available for direct patient care. Effective inbox management requires routing routine requests to appropriately trained team members while reserving physician involvement for clinically complex issues.

1. Adopt a top-of-license message management model. Route routine patient inquiries to the appropriate non-physician team member, reserving physician involvement for issues that require clinical judgment, diagnosis, or treatment decisions. Develop clear protocols defining what nurses, pharmacists, front-desk staff, and physicians can address independently, and provide template responses and standing orders so non-physician team members can resolve common questions. Use well-defined escalation criteria to identify when a message needs physician input. Organizations can further support this model with AI-driven triage tools that categorize incoming messages or draft responses for staff review.
2. Even without a call center, optimize how in-basket tasks are handled within clinics. For instance, assign dedicated “inbox time” blocks or rotate inbox duty among team members so that responding to patient messages doesn’t continuously interrupt face-to-face care and doesn’t spill over into personal time.
3. Use analytics (from Tactic 1) to track inbox volumes and rebalance workload. Monitor inbox volumes and response times for each provider and team member. Redistribute tasks if one person is overloaded. Ensure response times remain within acceptable ranges by tracking metrics like inbox volume per clinician, average response lag, and after-hours message counts.

Clinical documentation remains one of the most time-intensive and burdensome aspects of physician work. Much of the work involved in writing visit notes is manual and repetitive, which encroaches on time that could be spent on direct patient care or other higher-value activities. Documentation support tools — including ambient documentation, medical scribes, AI-assisted drafting, templates, and other workflow enhancements — can help reduce this burden when they are integrated into clinical workflows and adopted consistently by clinicians.

1. Deploy ambient clinical documentation tools and integrate them into workflows. These AI-driven systems listen to the patient–clinician conversation and produce a draft clinical note, which the clinician then reviews and finalizes. It’s crucial to integrate ambient documentation into existing processes (such as in-room scribing or immediate post-visit note review) rather than layering it on top of current routines.
2. Identify best-fit scenarios and scale gradually. Begin in specialties or settings that have high documentation burdens, such as primary care or urgent care, where each day is packed with brief visits. These areas often see the greatest return on investment from ambient documentation or other support tools. Start with willing early adopters who can champion technology and help refine workflows. Use small pilots with robust training, feedback loops, and leadership support before scaling up organization wide.
3. Build trust through training, governance, and oversight. Educate clinicians on what the ambient documentation AI does and doesn’t do, emphasizing the need to review the AI-generated notes for accuracy. Establish governance to monitor note quality and safety, including periodic audits and an easy way for clinicians to report errors in AI-generated notes. This oversight will help ensure any issues are caught early and that clinicians maintain confidence in the tool.

Care coordination workflows — such as managing referrals, handoffs at discharge, or follow-up processes — are often fragmented and manual. This fragmentation leads to delays, dropped handoffs, redundant work, and added burden on both clinicians and staff. While automation can accelerate individual tasks, orchestration ensures that multi-step processes move forward, dependencies are managed, and unresolved work is escalated when necessary. If the underlying workflow is unclear or key data cannot move across systems, adding automation will simply scale the inefficiency.

1. Establish clear ownership, handoffs, and escalation paths across care coordination workflows. For key processes such as referrals, discharge transitions, and follow-up care, define responsibilities at each step, standardize handoff protocols, and create escalation mechanisms when tasks stall. Prioritize interoperability and standardized referral data so information can move across teams and systems without manual rework.
2. Use AI to automate referral processing, triage, and scheduling. Intelligent automation can interpret unstructured referral information, route urgent cases immediately, and even initiate scheduling of appointments without manual intervention.
3. Focus on creating closed-loop workflows so referrals, transitions, and follow-ups are completed, not just initiated. Redesign care coordination processes so every task has a defined endpoint and confirmation of completion (e.g., referral scheduled, patient seen, or follow-up conducted). Track task status across team and system boundaries and automatically escalate unresolved or delayed tasks to the appropriate team member. Pay special attention to hospital-to-clinic transitions and post-acute follow-ups, where incomplete handoffs most often lead to readmissions or patient frustration.
4. Use AI to orchestrate multistep workflows across teams and systems. Consider using AI agents or automation to coordinate tasks across clinical, administrative, and operational teams. An AI “orchestrator” can sequence tasks, assign work to the right person at the right time, and ensure prerequisites are met. Ideally, these AI systems should both act (execute routine tasks) and alert (flag exceptions or delays to the team) as needed.

¹ Berg S. Physician burnout rate continues to decline, falling to nearly 42%. American Medical Association. April 16, 2026.

² Guerra A. EHR frustration major factor in doc burnout, but optimization can have big effect, KLAS finds. Health System CIO. July 24, 2025.

³ Mahnke CB. Small steps add up to much less “pajama time” for physicians. American Medical Association. June 26, 2024.

⁴ Rebholz T. Redesigning clinician training to improve EHR readiness and efficiency. EpicShare. August 14, 2023.

⁵ Levick D. Reduce medication alert fatigue with patient-focused clinical decision support. FDB (First Databank). April 18, 2023.