Documentation Burden at Mercy Ridge: Informatics, Education, or Staffing?
Mercy Ridge is a 240-bed community hospital. A time-and-motion study found that medical-surgical nurses spend a mean of 46 minutes of every 12-hour shift in EHR documentation, and a build audit shows 31% of nursing flowsheet fields are never viewed by any clinician. Three of six med-surg units report chronic end-of-shift overtime. At the operations meeting the CNO says, 'Documentation is out of control - I need you to fix the EHR.' The CIO offers a vendor documentation-efficiency module priced at $180,000. Unit managers believe the real problem is that two RN positions were cut last year. You are the sole nurse informaticist, four months into the role, with no dedicated analyst support and one fiscal quarter before budget decisions.
1. Translate the CNO's request into a scoped, measurable problem statement, and explain how you would determine whether this is an informatics problem, an education problem, or a staffing problem.
Read the worked answer
Scoped statement: Reduce non-value-adding documentation time for med-surg RNs by 20% within two quarters without reducing required documentation or the data quality, regulatory, and billing consumers rely on. Note what this excludes: it does not promise to eliminate overtime, which may have other drivers.
Then resist starting with the EHR. Triangulate three sources:
- System data: audit logs by screen and field, time-stamped, stratified by unit and by tenure of the user.
- Observation: where the 46 minutes actually goes - login and re-authentication, searching, duplicate entry between flowsheet and care plan, handoff tools, device scanning.
- Consumer interviews: every never-viewed field has an owner (quality abstraction, regulatory reporting, billing, research). The build audit tells you usage, not necessity.
Discriminating the three hypotheses:
- If burden concentrates in redundant build, duplicate fields, and navigation, it is an informatics problem.
- If charting time per patient is stable but overtime tracks patient ratios and persists in weeks with low census, it is a staffing problem wearing a documentation costume.
- If the pattern is confined to new graduates and travelers, or free-text duplication of existing structured fields, it is education.
The answer will likely be mixed, and the honest deliverable is a weighted attribution rather than a single label.
2. The CIO wants to purchase the $180,000 module immediately. What evidence would you gather, and how would you frame your recommendation so it is defensible whether or not the purchase proceeds?
Read the worked answer
Evidence to gather before any capital decision:
- The mechanism. Ask the vendor to state in writing which of your measured time sinks the module removes. A module that improves note authoring does nothing for a login or a duplicated intake form.
- Reference sites running your EHR version and module, with their measured before/after and their staffing context.
- Your own baseline, locked before the trial, so post-implementation change can be attributed rather than asserted.
- Compliance and quality sign-off that hiding or removing fields will not break a reported measure.
- Total cost: license plus analyst build hours, training, and the opportunity cost of that analyst time against the free optimizations you could do instead.
- Change-fatigue risk given any recent upgrade.
Framing: Present a staged decision rather than yes/no. Set the criteria first (for example: if a 60-day optimization sprint recovers at least 15 of the 46 minutes, defer purchase), then run the sprint on the highest-yield, lowest-risk items. This is defensible in either direction because the decision rule is public before the result is known, and because it protects the CIO from a purchase that later cannot be shown to have worked. If the sprint recovers little and the module demonstrably addresses the residual gap, recommend the purchase with confidence and with a measurement plan attached.
3. Walk the raw findings through the data–information–knowledge–wisdom continuum and describe how you would present the final recommendation to the executive team.
Read the worked answer
Data: 46 minutes per shift, 31% of fields never viewed, overtime hours by unit.
Information: organized and contextualized - minutes per screen and per field, ranked by yield; field usage cross-referenced against the name of the consumer who relies on it; unit-level trends against census.
Knowledge: patterns and relationships - duplication between flowsheet and care plan accounts for a large share; several high-burden fields exist only because of historical habit and have no downstream consumer, while a few low-burden fields feed a publicly reported measure. Knowledge is the synthesis, not any single number.
Wisdom: the judgement layer, and where the executive conversation actually happens - weighing nurse time against the residual risk of removing a field, deciding what to stop measuring, acknowledging that the two cut positions may be part of the truth, and choosing a sequence under budget constraint. Wisdom also includes saying plainly what you do not know.
Presentation: one page. The problem in a sentence, the attribution with confidence (informatics / education / staffing shares), the three highest-yield actions with owner and date, the decision requested, and the metric that will tell the board in 90 days whether it worked. Do not lead with the tool; lead with the attribution.