EHR Optimization Strategies for the Post-Go-Live Era

Go-live day gets the champagne. The eighteen months after go-live get the burnout surveys.

That gap is the reason “EHR optimization” has become its own discipline rather than an afterthought tacked onto implementation contracts. Standing up an electronic health record is a project with a deadline; making that EHR work well for the clinicians who touch it every day is an ongoing operational commitment with no natural end point. Organizations that treat go-live as the finish line tend to see early efficiency gains erode within a year, as workarounds accumulate, order sets drift out of date, and notes grow longer without becoming more useful.

This piece looks at what EHR optimization actually involves in practice: personalizing the system to specialty and role, keeping order sets and clinical decision support under active governance, coaching users toward more efficient habits, controlling note bloat, and measuring whether any of it is working.

Why optimization is a distinct phase, not a cleanup task

Most EHR implementations are scoped, staffed, and budgeted as one-time projects. The build team configures the system to a “safe minimum” standard designed to get every department live without introducing patient safety risk, then moves to the next go-live. That approach is defensible — it would be reckless to spend months perfecting a cardiology order set before the hospital has any EHR at all — but it means the system clinicians inherit on day one is intentionally generic.

Optimization is the work of closing the gap between that generic baseline and a system tuned to how a given specialty, role, or individual actually practices. Because clinical workflows, staffing, and regulatory requirements keep changing after go-live, this isn’t a project with an end date. It’s closer to ongoing facilities maintenance than a renovation.

The Office of the National Coordinator for Health Information Technology (ONC) has documented usability and clinician burden as persistent, systemic issues in health IT rather than problems solved once and left alone, noting that configuration and implementation decisions made well after initial deployment continue to shape how usable a system is in practice.¹ That is the core argument for a standing optimization function: usability is not fixed at go-live, it’s negotiated continuously.

What does EHR personalization actually mean?

“Personalization” in this context is not a UI preference toggle. It typically covers a handful of concrete capabilities:

• Specialty-specific note templates and SmartPhrases/dot phrases that reflect how a given specialty documents (an orthopedic surgeon’s post-op note has little in common with a psychiatrist’s intake note, yet both may start from the same generic template if no one has customized it).
• Personalized order sets and preference lists scoped to an individual physician’s typical practice patterns, subject to the same safety review as system-wide order sets.
• Custom result flowsheets and dashboards that surface the handful of data points a given role checks most often, rather than a one-size-fits-all view.
• Role-based navigation so that a nurse, a scheduler, and a physician are not all working through the same generic menu structure.

The KLAS Arch Collaborative — a research initiative that gathers clinician EHR satisfaction data across hundreds of health systems — has published findings tying personalization directly to satisfaction and burnout outcomes. Their work describes personalization as one of the pillars of what they call the “EHR House of Success,” alongside governance and ongoing education, and finds organizations get the most benefit when personalization content is built into both initial and ongoing training rather than offered as an optional, self-service add-on that clinicians have to discover and pursue on their own.² In practice, that means embedding a personalization module in every specialty’s training curriculum and revisiting it at defined intervals, not just mentioning that the feature exists during a one-time orientation session.

It’s worth being honest about the limits here: personalization tools reduce friction, but they can’t compensate for a fundamentally poor underlying workflow design, and over-customization at the individual level can create its own maintenance burden and safety inconsistency across a department. Most governance frameworks try to strike a balance — standardizing the clinical content (what the order set contains) while allowing personalization of the interface (how it’s presented and accessed).

Order-set governance: cleaning up what accumulates

Order sets and clinical decision support (CDS) rules tend to multiply after go-live. New specialties request new sets, guidelines change, drug shortages force one-off workarounds that never get retired, and duplicate or near-duplicate order sets pile up because it was faster to build a new one than track down and revise an existing one.

Published governance frameworks for CDS consistently recommend a standing, multidisciplinary committee — typically drawing representation from medicine, nursing, pharmacy, and informatics — with clearly defined responsibilities that include review of new requests, periodic review of existing content against current clinical guidelines, and retirement of order sets that show low or no usage.³ A commonly cited structure separates governance into three functions:

1. Intake and prioritization — a defined process for who can request new order sets or CDS content, and a committee that vets requests against clinical need rather than individual preference.
2. Content review — checking that an order set or alert reflects current evidence and organizational standards, not what was correct when it was originally built.
3. Usage monitoring and retirement — pulling utilization data on existing order sets and CDS alerts so that low-value or unused content can be sunset rather than accumulating indefinitely.

That third function is frequently the weakest link. It’s easier to get a committee to approve new content than to get the same committee to decommission old content, but analytics teams that have looked closely at alert fatigue and order-set sprawl generally find that a small fraction of order sets account for the large majority of use, while a long tail sees little to no uptake and mostly adds clutter and maintenance risk.

Efficiency coaching and workflow redesign

Where personalization changes the system, efficiency coaching changes the user. Many health systems now run some version of a physician efficiency program — sometimes staffed by “physician builders” or clinical informaticists, sometimes contracted through the EHR vendor — that provides one-on-one or small-group coaching on keyboard shortcuts, templates, in-basket management, and documentation habits.

The Arch Collaborative’s research on this front points to ongoing, role-specific education as one of the strongest predictors of EHR satisfaction, more so than raw years of experience with the system. Their data suggests that clinicians who receive continued, personalized training well after go-live report meaningfully higher satisfaction than those who received only initial go-live training and were left to learn the rest on their own.² That has pushed many organizations toward “at the elbow” support models — brief, informal coaching sessions embedded in clinical time rather than requiring clinicians to attend a separate class.

A few patterns show up repeatedly in efficiency initiatives that report success:

• Peer-led coaching rather than IT-led training, since physician coaches tend to have more credibility with physician colleagues about workflow trade-offs.
• Data-informed targeting — using EHR usage metrics (time in system, after-hours documentation, click counts) to identify which individuals or departments would benefit most from coaching, rather than offering generic refreshers to everyone equally.
• Short, workflow-embedded sessions over long classroom retraining, on the theory that clinicians retain more when coaching happens in the context of their actual patient load.

How do you reduce note bloat without cutting corners on documentation?

Note bloat — the tendency of clinical notes to grow longer and more repetitive over time — is one of the more measurable downsides of EHR-era documentation. A widely cited 2022 analysis of more than 100 million EHR notes found that just over half of all text in a given note was duplicated from earlier documentation on the same patient, largely through copy-forward and copy-paste functions.⁴

The patient safety literature is not simply anti-copy-paste; it’s more precise than that. AHRQ’s Patient Safety Network has published guidance acknowledging that copy-paste and templated auto-population can genuinely support documentation efficiency, but the same functions make it easy to propagate outdated or incorrect information forward indefinitely and to bury clinically important findings inside dense, repetitive text.⁵ One frequently cited illustration involves an abbreviation (“PE”) that was copied forward across multiple notes with the wrong meaning attached, contributing to an unnecessary diagnostic workup — a small documentation habit compounding into a real clinical and safety issue.

Approaches that organizations have used to address note bloat include:

• Auditing copy-paste usage at the note level to identify outlier documentation patterns, rather than banning the function outright (which is rarely practical and can push clinicians toward slower, more error-prone manual re-entry).
• Redesigning templates so that required fields don’t force redundant restatement of information already captured elsewhere in the chart.
• Note-bloat-specific metrics, such as average note length by service line or percentage of duplicated text, tracked over time rather than as a one-time audit.
• Governance sign-off on new templates so that template proliferation itself doesn’t become a second source of bloat.

None of this is a solved problem as of this writing. Note length and documentation burden remain active areas of research, and organizations should treat any single intervention — including scribes, voice recognition, or template redesign — as a partial mitigation rather than a fix.

Measuring whether optimization is working

Optimization efforts are hard to sustain politically if there’s no way to show they’re producing results. Programs that report durable support from clinical leadership tend to track a consistent set of metrics before and after each optimization initiative, rather than relying on anecdotal feedback alone. Commonly tracked measures include:

• Time in system per patient encounter and time spent in the EHR outside scheduled clinical hours (“after-hours” or “pajama time” documentation).
• Clinician-reported satisfaction, often gathered through validated survey instruments rather than informal feedback, to allow comparison across departments and over time.
• Order-set and CDS alert utilization rates, used both to guide governance decisions and to demonstrate that low-value content is being actively managed rather than ignored.
• Note length and duplication metrics, tracked by service line to identify where documentation habits diverge from organizational norms.
• Turnover and burnout indicators in departments with high EHR burden, recognizing that documentation friction is one contributing factor among several, not a sole cause.

A meaningful caveat applies to all of these: EHR usage metrics are proxies, not direct measures of quality of care or clinician wellbeing, and they can be confounded by factors like patient complexity, staffing ratios, or seasonal volume. Organizations that rely on a single metric — time in system, for instance — risk optimizing for the metric rather than the underlying experience. Most mature programs pair quantitative EHR usage data with periodic qualitative feedback (surveys, focus groups, at-the-elbow observation) to avoid that trap.

Building a sustainable optimization program

Taken together, the research and case studies on this topic point to a few consistent structural elements behind programs that sustain improvement rather than losing momentum after an initial push:

• Standing governance, not ad hoc committees convened only when something breaks.
• Ongoing, role-specific training that continues well past go-live, rather than a single onboarding curriculum.
• Personalization built into training, so clinicians learn to tailor the system as part of normal education rather than as a separate, optional pursuit.
• Usage-based prioritization, using EHR analytics to direct limited optimization resources toward the areas with the greatest documented burden.
• Realistic timelines, since the evidence suggests meaningful efficiency and satisfaction gains typically unfold over many months of sustained effort rather than a single optimization sprint.

None of this is a substitute for sound clinical judgment or organizational change management, and the specifics of any optimization program should be adapted to a given organization’s specialty mix, patient population, and existing governance structures.

This article is for general informational purposes and does not constitute medical, legal, or clinical advice. Healthcare organizations should consult qualified clinical informatics, compliance, and IT professionals when designing or modifying EHR optimization programs.

Related reading

• EHR Implementation Best Practices: A Field Guide
• From Meaningful Use to Promoting Interoperability

Frequently Asked Questions

What is EHR optimization?

EHR optimization is the ongoing process of improving an already-implemented electronic health record system — refining workflows, personalizing templates and order sets, cleaning up outdated clinical decision support, and reducing documentation burden — rather than a one-time project completed at go-live.

How is EHR optimization different from EHR implementation?

Implementation is the finite project of configuring and deploying a new EHR safely across an organization. Optimization begins after go-live and continues indefinitely, adapting the system to specialty workflows, retiring outdated content, and coaching users as needs evolve.

What causes note bloat in electronic health records?

Note bloat largely stems from copy-forward and copy-paste documentation habits, along with templates that require redundant restatement of information. A widely cited 2022 study found roughly half the text in a large sample of EHR notes was duplicated from earlier documentation.

Who should be involved in order-set governance?

Effective order-set and clinical decision support governance typically involves a standing multidisciplinary committee with representation from medicine, nursing, pharmacy, and clinical informatics, tasked with reviewing new requests, auditing existing content, and retiring low-use order sets.

How do organizations measure EHR optimization success?

Common metrics include time spent in the EHR per encounter, after-hours documentation time, clinician satisfaction surveys, order-set utilization rates, and note-length or duplication metrics — ideally paired with qualitative feedback rather than relied on in isolation.