Ask a room full of physicians to grade their electronic health record the way they’d grade a piece of consumer software, and the result is not close. When researchers actually did this — applying a standardized usability instrument used across hundreds of industries — physician-rated EHR usability came back with a score in the “F” range, worse than roughly 90 percent of software ever benchmarked with the same tool. That is not a complaint buried in a survey’s write-in comments; it is a quantified, peer-reviewed finding, and it sits at the center of a growing body of research connecting EHR usability to two things health systems cannot treat separately: clinician burnout and patient safety.

EHR usability is not a soft, subjective preference about interface aesthetics. It is a measurable property of a system — how efficiently, effectively, and comfortably a clinician can complete real clinical tasks in it — and a decade of adoption data now shows that measurable property has consistently underperformed expectations, even as certification rules and vendor design efforts have tried to close the gap.

What Does “EHR Usability” Actually Mean?

Usability, in the formal human-factors sense, is not a single yes-or-no attribute. The International Organization for Standardization defines it around three components: effectiveness (can the user accomplish the task accurately and completely), efficiency (how much effort and time it takes), and satisfaction (how the user feels about the experience). Applied to an EHR, that means usability covers everything from how many clicks it takes to order a medication, to whether a critical lab value is visually distinguishable from a routine one, to whether the system’s alerts help a clinician catch a problem or simply train them to click through warnings without reading them.

This matters because “usability” gets used loosely to mean almost anything clinicians dislike about their EHR — slow servers, unwanted regulatory fields, or an unfamiliar interface after a conversion. Formal usability evaluation tries to isolate the design and interaction layer specifically: the human-computer interface itself, as distinct from policy requirements or hardware performance, even though in practice these factors interact and are hard to fully separate.

How Is EHR Usability Actually Measured?

The System Usability Scale

The most widely used general-purpose instrument is the System Usability Scale (SUS), a ten-item questionnaire developed in the 1980s and validated across a very wide span of software categories, from consumer apps to industrial control systems. It produces a single score from 0 to 100, and because it has been used in well over a thousand published studies, a given score can be benchmarked against a large distribution rather than judged in isolation. An average SUS score across industries sits around 68; scores below roughly 50 are generally considered to fall in the “not acceptable” range.

A national survey of United States physicians analyzed by researchers including those at Yale and the Mayo Clinic applied the SUS to EHR use specifically. The mean physician-rated score came back at 45.9, which the study’s authors placed in the bottom decile of scores recorded across previous SUS research and squarely in the “F,” not-acceptable range. That figure has become one of the most frequently cited data points in health IT usability discussions, because it converts a vague sense of clinician frustration into a number that can be tracked and compared across systems.

The KLAS Arch Collaborative approach

A second, EHR-specific measurement effort comes from the Arch Collaborative, an initiative organized by the health IT research firm KLAS in which member health systems administer a common survey to their own clinicians and pool results for benchmarking. Rather than a general usability instrument, the Arch Collaborative survey asks clinicians about training, EHR personalization, perceived efficiency, and overall satisfaction, rolling responses into a composite Net EHR Experience Score that organizations can compare against a shared national dataset.

The Arch Collaborative’s early published findings pointed to two things that matter for how health systems interpret their own usability problems. First, satisfaction varies enormously by organization even on the same underlying EHR product, suggesting configuration, training, and governance choices made locally can shift the clinician experience nearly as much as the software itself. Second, satisfaction varies by specialty — reflecting that a single EHR configuration is asked to serve emergency medicine, primary care, and surgical specialties simultaneously, each with different workflows and different tolerance for a one-size-fits-all interface.

Together, the SUS research and the Arch Collaborative data represent two complementary measurement traditions: one benchmarks the software against a broad usability standard, the other benchmarks the local implementation and support environment against peer organizations. Neither alone tells the full story, and researchers have been candid that usability measurement in health IT is still maturing — there is no single, universally agreed instrument.

Why Does EHR Usability Score So Poorly?

Several structural factors recur across the usability literature, and they are largely independent of any single vendor’s product decisions.

Interfaces built around data capture and billing, not clinical workflow

Many EHR systems evolved from billing and transaction-processing software rather than from tools designed around how a clinician actually thinks through a patient encounter. The result, described repeatedly in usability research, is an interface organized around discrete data fields and coded elements rather than around the clinical narrative — which forces extra navigation, clicking, and context-switching into tasks that used to take a single line in a paper chart.

Alert fatigue and interruptive design

Clinical decision support — drug interaction warnings, allergy alerts, reminder pop-ups — is meant to improve safety, but usability researchers have documented that when a system fires too many low-value alerts, clinicians learn to dismiss them reflexively. This phenomenon, generally called alert fatigue, is as much a usability failure as a clinical one: a well-designed alert system should make the truly dangerous warning visually and behaviorally distinct from routine noise, and many current systems do not.

One system serving many masters

An EHR configured for a large health system typically has to satisfy regulatory reporting requirements, billing and coding rules, quality-measure capture, legal documentation standards, and the clinical workflows of dozens of specialties — all within the same base platform. Usability researchers and the Office of the National Coordinator for Health Information Technology (ONC) have both pointed to this as a core tension: a system tuned to satisfy every external requirement simultaneously is unlikely to feel simple to any one user completing a routine task.

Local configuration and training gaps

Because most EHRs are highly configurable, two hospitals running the identical vendor product can produce very different clinician experiences depending on how order sets, templates, and shortcuts were built, and how well clinicians were trained to use them. This is the variance the Arch Collaborative data has highlighted, and it means usability complaints are not always resolvable by the vendor alone — organizational governance, informatics staffing, and training investment all shape the day-to-day experience layered on top of the underlying software.

Usability is not only a satisfaction issue. The Agency for Healthcare Research and Quality (AHRQ) has funded and published research using a sociotechnical framework to analyze how health IT — including EHR interface design — contributes to patient safety events, alongside factors like workflow, training, and organizational policy. That framework treats the human-computer interface as one of several interacting elements that can either support or undermine safe care, rather than as a standalone technical feature.

Patient-safety event reports have documented recurring usability-linked hazards: look-alike medication or dosing fields positioned close together on screen, wrong-patient errors when multiple charts are open simultaneously, and clinically important information that is easy to overlook because of poor visual hierarchy. None of this means EHRs are, on net, less safe than paper records — the broader evidence on EHR adoption points to real safety gains from legible orders and automated interaction checking. But it does mean usability defects are not merely an efficiency problem; they sit on the causal pathway to specific, documented safety incidents, which is why usability and patient safety are usually discussed together in the health IT literature.

How Does EHR Usability Connect to Clinician Burnout?

The same Mayo Clinic Proceedings study that produced the SUS finding of 45.9 also tested the relationship between usability and physician burnout directly, using the Maslach Burnout Inventory alongside the usability survey. After adjusting for factors such as specialty, practice setting, and hours worked, the researchers found that each one-point improvement in a physician’s SUS rating was independently associated with a measurable reduction in the odds of burnout. That is a dose-response relationship, not just a correlation between two things that happen to move together — it suggests usability is doing real explanatory work rather than standing in for some other variable like workload.

That finding sits alongside a related body of research on time burden. A widely cited 2016 time-and-motion study published in the Annals of Internal Medicine found that physicians in ambulatory practice spent roughly two hours on EHR and administrative desk work for every one hour of direct face time with patients during the clinic day, with additional charting frequently extending into evenings — the pattern clinicians call “pajama time.” Poor usability compounds that burden directly: a workflow requiring excess clicks, redundant data entry, or difficult navigation does not just cost seconds per task, it accumulates across a day of encounters into hours of additional, often uncompensated work.

Burnout researchers note that documentation burden and usability are not the sole cause of physician burnout — workload, staffing ratios, loss of autonomy, and regulatory demands all contribute independently. But among contributors that are structurally addressable by health systems and vendors, usability is one of the few with both a validated measurement tool and a demonstrated, quantifiable relationship to the outcome.

What Has Been Done to Improve EHR Usability?

Certification requirements for safety-enhanced design

Since the 2015 Edition of ONC’s health IT certification criteria, EHR vendors seeking certification for certain capabilities have had to complete a formal usability process: adopting a recognized user-centered design methodology during development, then conducting summative usability testing — structured evaluation with real representative users completing defined clinical tasks — on specific safety-critical functions such as medication ordering and clinical decision support. Vendors must document the process and results, giving purchasers and regulators at least some visibility into how a product was usability-tested before reaching the certification threshold required for federal incentive programs.

This certification requirement has real limits. It applies to specific certified capabilities rather than the entire product experience, testing occurs before widespread field deployment rather than continuously afterward, and a product can pass its required usability testing while still scoring poorly once deployed inside a specific organization’s configuration — part of why certified systems still show the wide field-usability variation documented by the Arch Collaborative.

Federal attention to regulatory and documentation burden

Recognizing that usability complaints are entangled with regulatory documentation requirements, ONC — working with the Centers for Medicare & Medicaid Services under a mandate from the 21st Century Cures Act — released a draft strategy in late 2018 aimed at reducing regulatory and administrative burden associated with health IT and EHR use. It identified reducing the time clinicians spend recording information, reducing the burden of regulatory reporting, and improving the functionality and intuitiveness of EHRs as its central goals, treating usability and administrative-burden reduction as parts of the same problem.

Vendor-side and organization-side design work

Independent of regulation, some EHR vendors have invested in dedicated user-experience teams, iterative usability testing with practicing clinicians, and configurable role-based views intended to reduce on-screen clutter for a given specialty. At the health-system level, the Arch Collaborative’s own published guidance points toward organizational levers — governance that gives clinicians real input into build decisions, deeper personalization training, and ongoing optimization rather than a one-time go-live — as being at least as influential on the clinician’s ultimate experience as the underlying platform.

What Should Health Systems and Clinicians Take Away From This?

The available evidence supports a few reasonably firm conclusions, while leaving open questions for the field to keep working through. EHR usability, measured with standardized instruments, has consistently scored below general software benchmarks. That poor usability is statistically linked, independent of other factors, to physician burnout, and separately to specific categories of patient-safety events. And usability outcomes are shaped by both the underlying software and the local choices an organization makes about configuration, training, and ongoing optimization — meaning no single actor, vendor or health system, fully controls the outcome alone.

What the evidence does not yet support is a simple, universal fix. Usability testing requirements, burden-reduction policy, and organizational optimization programs are incremental interventions layered onto a genuinely difficult design problem: building a single system that satisfies billing, regulatory, legal, and clinical-workflow demands at once, for users across dozens of specialties, without asking any of them to tolerate excessive friction. Progress so far has been real but partial, and the usability gap that shows up in SUS scores and Arch Collaborative benchmarks remains one of the more consequential unsolved problems in health IT.

Frequently Asked Questions

What is a good SUS score for an EHR?

Across industries, the average System Usability Scale score is roughly 68, with scores in the 80s generally considered excellent. Published research on physician-rated EHR usability has found average scores well below that general benchmark, often in ranges usability researchers classify as “poor” or “not acceptable,” though scores vary meaningfully by vendor, specialty, and local implementation.

Is EHR usability the same thing as EHR satisfaction?

They are related but distinct. Usability refers to specific, task-level measures of effectiveness, efficiency, and ease of use, typically captured with structured instruments like the SUS. Satisfaction is a broader attitude that also reflects training quality, system reliability, workload, and organizational support — which is why two clinicians on the identical software can report different satisfaction levels.

Does poor EHR usability cause physician burnout?

Research has found a statistically independent association between lower usability scores and higher odds of burnout, even after adjusting for factors like specialty and hours worked. However, burnout is multi-causal — workload, staffing, and administrative demands also contribute — so usability should be understood as one significant, addressable factor rather than the sole cause.

Are certified EHRs required to be tested for usability?

Since the 2015 Edition of ONC’s certification criteria, EHR products seeking certification for certain capabilities must undergo a user-centered design process and summative usability testing on specific safety-critical functions. This testing occurs before widespread deployment and applies to defined capabilities rather than the full product experience, which is one reason certified systems can still show wide usability variation once deployed in real clinical settings.

Can better training improve a clinician’s EHR experience without changing the software?

Yes, according to benchmarking research from initiatives like the KLAS Arch Collaborative, which has found meaningful satisfaction differences between organizations using the identical EHR product. Governance, personalization, and ongoing training investment appear to shape the clinician experience substantially, alongside — not instead of — the underlying software’s design.