High reliability in healthcare, the executive's view
Facts checked against nursingworld.org, codeofethics.ana.org and web.archive.org · Last checked Nov 6, 2026 · Not affiliated with ANCC, ANA, AONL, NLN or AACN.
High reliability in healthcare describes an organization that runs hazardous work with very few failures because it hunts for small ones all the time. NEA-BC domain II, 32 scored questions, asks how an executive builds that: the culture, the analysis tools, survey readiness, benchmarks and emergency plans.
Weick and Sutcliffe (Managing the Unexpected) describe high reliability organizations by five principles. The first three are about anticipating trouble, the last two about containing it:
Preoccupation with failure: near misses are read as data about the system, not as luck.
Reluctance to simplify: leaders distrust the first tidy explanation.
Sensitivity to operations: leaders know what is happening at the front line now, through rounding, safety huddles and live data.
Commitment to resilience: errors are expected and recovery is rehearsed.
Deference to expertise: in a fast-moving problem, authority moves to whoever knows the situation best.
Chassin and Loeb (Milbank Quarterly, 2013), writing for The Joint Commission, add the executive's frame: leadership commitment to zero harm, a culture of safety, and robust process improvement.
18 questions on safety at system scale
Question 1 of 18 · Quality and Safety · 26%
During a system-wide ED surge, the CNO lets the house supervisor who has tracked every bed and pending discharge tonight make bed placement decisions rather than routing them through the senior executive on call. Which high reliability principle does this illustrate?
Preoccupation with failure means treating small lapses and near misses as warning signs; it does not describe shifting decision rights.
Commitment to resilience is about recovering and adapting after something has already gone wrong, not about who makes the call.
Sensitivity to operations does not fit, because the CNO's move changes who decides, not what leaders can see.
Letting the supervisor with the most current knowledge of beds and discharges decide, regardless of rank, is deference to expertise.
Question 2 of 18 · Quality and Safety · 26%
Before opening a new outpatient chemotherapy infusion center, the CNO wants to identify how the medication process could fail and prioritize the riskiest steps. Which method should the team use?
A rapid PDSA cycle tests one change on a small scale and does not systematically list and rank possible failures.
Apparent cause analysis is a limited look back at a low-harm event that already happened, not a forward look at a new process.
FMEA is prospective: before launch it lists how each step could fail and ranks the failures by risk, which is what a new infusion center needs.
DMAIC reduces variation in an existing process with data; it is not the standard tool for predicting failures in a process that does not exist yet.
Question 3 of 18 · Quality and Safety · 26%
A hospital's restraint policy is being revised. Under the CMS Conditions of Participation, what is required when restraint is used to manage violent or self-destructive behavior?
There is no 8-hour exemption; the evaluation is required whenever restraint is used for violent or self-destructive behavior.
CMS requires a face-to-face evaluation within 1 hour of starting restraint or seclusion for violent or self-destructive behavior.
A 24-hour window is far too late; for violent or self-destructive behavior the CMS face-to-face evaluation is due within 1 hour.
Four hours is the renewal limit for an adult's restraint order, not the deadline for the face-to-face evaluation.
Question 4 of 18 · Quality and Safety · 26%
Stratifying 30-day readmission rates by race, ethnicity, and language shows that patients with limited English proficiency are readmitted at twice the overall rate. Which IOM quality aim does closing this gap address?
Equitable care means quality does not vary by personal characteristics such as language or ethnicity, so closing this gap addresses equity.
Efficient care means avoiding waste of resources; a readmission gap tied to language is an equity problem first.
Patient-centered care means respecting individual preferences and values, while this gap is about outcomes that differ by group.
Effective care means matching care to evidence for the whole population, while the finding here is unequal results across groups.
Question 5 of 18 · Quality and Safety · 26%
A hospital's average length of stay (LOS) rose 0.4 days this year, while its case mix index (CMI) rose from 1.45 to 1.62. The board asks whether efficiency has worsened. What should the CNO do before answering?
Comparing actual LOS with the CMI-adjusted expected LOS separates sicker patients from true inefficiency before the board gets an answer.
Discharge-before-noon rates are a unit process measure and cannot show whether case mix explains the longer stays.
Benchmarking unadjusted LOS against national averages repeats the same error of ignoring case mix.
Calling the LOS rise a throughput failure ignores that sicker patients (higher CMI) are expected to stay longer.
Question 6 of 18 · Quality and Safety · 26%
A Joint Commission tracer finds transfer summaries missing required elements in 6 of 10 records reviewed across three units. Which corrective approach will best sustain compliance?
Auditing 100% of summaries each month catches errors after the fact and adds burden without preventing them.
A policy memo is a weak action that relies on memory and does not change the workflow where the omission happens.
Counseling individual nurses treats a cross-unit design problem as personal failure.
Building the required elements into the EHR transfer workflow makes completion the default, and auditing confirms the fix holds.
Question 7 of 18 · Quality and Safety · 26%
A sepsis pathway achieved 90% bundle compliance within three months of go-live. The project team is being disbanded. Which action BEST prevents compliance from drifting back to baseline?
Auditing for 30 days and then closing the project ends monitoring just when drift starts.
Publishing the pathway in the policy manual is a weak action that does not sustain behavior.
Naming a process owner who reviews compliance monthly gives the gain lasting accountability after the project team disbands.
Recognizing top units motivates but does not create ongoing ownership.
Question 8 of 18 · Quality and Safety · 26%
Nurses tell the patient safety officer they avoid writing detailed near-miss analyses because they fear the documents will be used in malpractice lawsuits. Which action BEST addresses this concern?
Physician peer review protection varies by state and covers medical staff review, not nursing near-miss analyses in general.
Under the Patient Safety and Quality Improvement Act, work reported to a federally listed PSO is privileged and confidential.
HIPAA protects patient information but does not shield event analyses from discovery in a lawsuit.
Attorney-client privilege covers legal advice, not routine safety analyses, and overusing it does not create a learning system.
Question 9 of 18 · Quality and Safety · 26%
During a respiratory surge, a hospital activates crisis standards of care for scarce ventilators. Which design element is most important to a fair allocation process?
Deciding case by case without written criteria invites inconsistency and bias, the opposite of a transparent process.
First-come, first-served ignores the chance of benefit and is not a fair way to allocate scarce ventilators in a crisis.
A triage team separate from the treating clinicians applies the criteria consistently and frees bedside teams to advocate for patients.
Leaving allocation to each attending physician puts the treating clinician in a conflict between advocacy and rationing.
Question 10 of 18 · Quality and Safety · 26%
A CNO wants to benchmark a 30-bed adult step-down unit's falls-with-injury rate through NDNQI and present the comparison to the quality committee. Which comparison group gives the most valid benchmark?
A single best-performing unit is an outlier target, not a benchmark distribution that shows where the unit stands.
Step-down units in hospitals of similar size and type form a like-for-like peer group, which makes the benchmark valid.
All adult inpatient units mix very different populations and acuity, so the comparison is not like for like.
The hospital's own med-surg units differ in acuity and unit type, and internal comparison is not a peer benchmark.
Question 11 of 18 · Quality and Safety · 26%
At 0700 the command dashboard shows 99% inpatient occupancy, 20 admitted patients boarding in the ED, no available float staff, and 14 expected discharges not yet written. Which action should the nurse executive take first?
Activating the pre-approved surge plan sets off its capacity and staffing triggers at once, which addresses the boarding crisis.
Diverting all ambulance traffic is a last resort that shifts risk to the community and does not create beds.
A committee meeting about next week's schedule does nothing for today's full house and ED boarding.
Opening an overflow unit by pulling day-shift staff creates gaps elsewhere and skips the coordinated surge plan.
Question 12 of 18 · Quality and Safety · 26%
A root cause analysis confirms that a patient's death was caused by a defibrillator malfunction. Under the Safe Medical Devices Act, which reporting must the nurse executive ensure?
A sentinel event report to The Joint Commission is voluntary and does not satisfy the federal device reporting rule.
A 24-hour state report is not the federal device rule, and it leaves out both the FDA and the manufacturer.
The Safe Medical Devices Act requires user facilities to report a device-related death to both the FDA and the manufacturer within 10 work days.
Reporting only to the manufacturer within 10 work days is the rule for serious injuries, but a death must also go to the FDA.
Question 13 of 18 · Quality and Safety · 26%
A CNO is building a scorecard for a systemwide falls-prevention program. Which measure is a process measure that serves as a leading indicator?
Mobility-related pressure injuries are an outcome of a different program, and they lag behind the prevention work.
Completing fall-risk plans within 8 hours of admission measures whether prevention is being done, which makes it a leading process measure.
Falls with injury is an outcome measure that shows harm after it happens, so it is a lagging indicator.
RN hours per patient day describes resources, which is a structure measure, not a process.
Question 14 of 18 · Quality and Safety · 26%
During a staffing crisis with a temporary 16-hour shift waiver, a nurse in her 15th hour gives a double dose of heparin after misreading a pump screen and reports the error immediately. How should the CNO direct leaders to respond?
Retraining every nurse on the pump treats a fatigue-related slip as a knowledge gap across the whole staff.
Suspending the nurse is discipline for an unintended slip and discourages others from reporting errors.
A written warning punishes human error, and blaming the decision to work the extended shift ignores that leadership authorized it.
A 15th-hour misread that was promptly reported is human error, so the response is to console the nurse and fix the system factors, including shift length.
Question 15 of 18 · Quality and Safety · 26%
Two years after launching a rapid response team (RRT), the board asks whether it has improved patient outcomes. Which measure BEST answers the board's question?
Non-ICU cardiac arrests per 1,000 discharges is the outcome the RRT exists to prevent, so it shows whether patients are better off.
ICU mortality measures patients after transfer, while the RRT is designed to rescue deteriorating patients before they reach the ICU.
Nurse satisfaction with the team is a perception measure, not a patient outcome.
Monthly call volume is a process measure that shows use, not whether outcomes improved.
Question 16 of 18 · Quality and Safety · 26%
A hurricane has damaged a hospital's power plant and incident command is running operations. The board chair asks to receive hourly updates and to sit in on incident command briefings. How should the CNO, as incident commander, respond?
Bringing the board chair into command briefings blurs governance and operations and pulls the command team off task.
Hourly operational reports give the board detail it cannot act on and add reporting work in the middle of the crisis.
The public information officer handles external messaging; briefing the board for oversight belongs with the incident commander.
Scheduled strategic briefings keep the board informed for its oversight role while operational decisions stay inside the command structure.
Question 17 of 18 · Quality and Safety · 26%
An ED redesign cuts average door-to-provider time from 70 to 25 minutes without changing staffing costs. Which IOM quality aim does this change MOST directly advance?
Patient-centered care concerns preferences and values; shorter waits help, but timeliness is the direct aim.
Cutting the wait to see a provider from 70 to 25 minutes reduces harmful delay, which is the timely aim.
Effective care is about using evidence to choose the right treatment, not about how long patients wait.
Efficient care means less waste for the same output; with costs unchanged, the direct gain is in timeliness.
Question 18 of 18 · Quality and Safety · 26%
An RCA of a wrong-site procedure finds that the time-out was rushed, the consent form listed the wrong side, and site marking was skipped because skin markers had been out of stock for weeks after a supply contract change. In Reason's Swiss cheese model, how is the prolonged marker shortage best classified?
An active failure is the unsafe act at the sharp end, such as skipping the mark or rushing the time-out, not the upstream shortage.
A procedural violation is a person deviating from a rule, while the marker shortage is a condition created by the organization.
At-risk behavior is a Just Culture category for an individual's choice, not a description of a system-level weakness.
A supply problem that persisted for weeks after a contract change is an upstream system weakness, which Reason calls a latent condition.
Your score shows up after the last answer.
RCA or FMEA: read the clock
Root cause analysis works backward from something that already happened, a sentinel event or a cluster of harm, and asks which system conditions let it through. The Joint Commission's Sentinel Event Policy expects a comprehensive systematic analysis and a corrective action plan.
Failure mode and effects analysis works ahead of a process, before it starts or changes. The team maps the process, names the ways it might break down, and rates every failure mode for severity, likelihood and the chance of catching it; multiplying the three gives a risk priority number that orders the fixes.
Two traps follow. The first is the clock: harm already done points to RCA, a process not yet running points to FMEA. The second is the output: an analysis that ends with "retrain the nurse" has stopped at the person. The RCA² guidance of the National Patient Safety Foundation (2015) ranks forcing functions, standardization and physical redesign as stronger actions than training or a new policy.
Mistakes domain II punishes
Judging by outcome. Just culture (David Marx) sorts the behavior: an inadvertent slip, risky drift whose danger the person no longer sees, or reckless disregard. The response follows the behavior, not the amount of harm.
Counts in a benchmark. Compare rates that share one definition and one denominator, never raw totals. A percentile is a position, not a target.
Survey vocabulary. CMS cites deficiencies against the Conditions of Participation; The Joint Commission issues Requirements for Improvement. The two bodies use different terms and paperwork.
Emergency phases. Mitigation, preparedness, response and recovery are separate phases. The CMS emergency preparedness rule asks for a risk-based plan, policies and procedures, a communication plan, and training and testing.
EBP, QI and research. Same continuum, different aims; owned by the NPD-BC page linked above.
No. High reliability assumes failures will happen and builds early detection and recovery around them. Zero harm is a goal some systems set on top of that.
Can a near miss get a root cause analysis?
Yes. A close call exposes the same system weakness without the harm, which makes it a good RCA candidate. FMEA remains the tool for a process that is not yet running.