| Audience: | CIO · CMIO/Chief Clinical Informatics Officer · CTO/VP Digital Health |
| Primary Sectors: | Healthcare Providers · Public Sector Health Systems |
| Decision Horizon: | 0-6 months |
Executive Summary
Most health systems are treating the documentation burden as a staffing and tooling problem. The evidence says the constraint is more basic: clinicians are burning cognitive capacity on navigation, fragmentation, and poorly organized data, not on care. Clinicians can spend one-third to one-half of their day in the EHR; median usability scores are poor (45.9/100) and even small usability declines correlate with higher burnout risk.
Verdict: Pilot a targeted redesign and don’t scale with new add-ons yet. Over the next 90 days, focus investment on measurable UX and data-usability fixes in the highest-volume workflows. Treat AI scribes as a bounded experiment only after baseline interaction friction is reduced.
Our Analysis
To have an effective EHR, you need two levers: system usability, how hard it is to operate the EHR, and data usability, whether the EHR's data is complete, reliable, and contextually meaningful. System usability reduces extraneous cognitive load and wasted effort, while higher data usability supports the mental work you want physicians doing for diagnosis (germane cognitive load). It can also reduce perceived overload.
The Narrative vs. The Reality
EHR developers propose that if you improve documentation capture using dictation and ambient AI then the clinician cognitive burden will drop. Research shows that this does not track cleanly:
- EHR work is fragmented at the micro-level. Clinicians average ~1.4 task switches per minute, driven largely by data viewing, data entry and order work.
- The “click and screen tax” is not rhetorical. Single documentation tasks can take hundreds of clicks (346) and dozens of screens (43). Deep hierarchies also correlate with wrong-field entry events, which is an important patient safety signal to address.
- Workflow misalignment creates shadow systems where clinicians resort to off-system notes and workarounds. This misalignment can extend workdays by ~90 minutes on average in reported settings.
- Better data alone can backfire if the interface remains hard to use because cognitive overload is driven more by how data is accessed and navigated than by the underlying data architecture.
The Signal in the Noise
The simplest wins are human-factors fixes like: view consolidation, interrupt reduction, use of progressive disclosure, and aligning task sequences to real clinical flow.
Why This Matters Now
Cost pressure and clinician capacity are tightening, while EHR usability remains a measurable operational drag. Poor usability scores add to the evidence that small usability declines are associated with higher burnout risk. At the same time, adding more data will increase the cognitive demands on clinicians unless healthcare organizations pair it with better interaction design and data governance. This is an operating-model problem, because if you don’t redesign the work, you’ll just finance more workarounds and burnout.
Recommended Actions
Do This
- Declare a Top 10 workflow redesign sprint. For example, focus on notes, inbox, orders, meds, discharge, etc. Set a gate that no workflow is done unless it hits a ~50% reduction in clicks per screens from observed baseline examples and passes clinical safety review.
- Separate data usability work from UI usability work with distinct owners. Use KPIs like data completeness or quality vs. navigation burden or interruptions.
- Ringfence AI scribe and automation spend as R&D and establish a gate of not scaling beyond the pilot until after-hours EHR time drops by at least 15% without increasing note rework or corrections.
Avoid This
- Purchasing enterprise-wide licenses for documentation tools before reducing navigation and workflow friction.
- Template proliferation. Adding more fields and more alerts without progressive disclosure and signal prioritization.
- Letting vendors define efficiency metrics for you. Use your own workflow telemetry and clinician time outcomes to determine what is best for your clinicians.
Bottom Line
The EHR burden is a navigation and cognition problem. Spend less energy capturing notes and more on reducing interaction friction. Work to make data usable in context. If you are not addressing click per screen reduction in the top workflows, you’re not fixing the cognitive burden, you’re funding new workarounds and burnout.