Five complete FMEA worksheets from real hospital quality-improvement projects: medication administration, patient identification, blood transfusion, ED triage, and discharge reconciliation. Each includes Severity, Occurrence, and Detection scores, RPN calculations, and the corrective actions implemented. Use these as reference when building your own HFMEA or preparing Joint Commission LD.04.04.05 documentation.
For a different set of cases across manufacturing, automotive, software, food processing, and aerospace, see our cross-industry FMEA examples. For methodology, see the complete FMEA guide.
Jump to an example:
- HFMEA vs Traditional FMEA — quick primer
- 1. Medication Administration — ICU Heparin Protocol
- 2. Patient Identification — Wrong-Patient Procedure Prevention
- 3. Blood Transfusion — Type and Crossmatch Process
- 4. Emergency Department Triage — Chest Pain Pathway
- 5. Discharge Medication Reconciliation
- HFMEA Hazard Score vs RPN — side-by-side
- Frequently asked questions
HFMEA vs Traditional FMEA — Quick Primer
Healthcare uses two variants of FMEA. Traditional FMEA uses the standard RPN = S × O × D formula on a 1–10 scale (same as automotive and software). HFMEA — developed by the Veterans Affairs National Center for Patient Safety in 2001 — simplifies the scoring to a Hazard Score Matrix of Severity × Probability on a 1–4 scale, then applies a decision tree to judge whether existing controls already mitigate the risk.
The five case studies below use the traditional 10-point RPN format because it remains the most common in hospital quality-improvement work and translates directly to ISO 14971 risk management documentation used by medical device vendors. A side-by-side comparison of both scales is included at the end of this article.
1 Medication Administration — ICU Heparin Protocol (PFMEA)
Healthcare · Pharmacy & NursingScope: Process FMEA for continuous IV heparin infusion on a medical ICU. Triggered by an Institute for Safe Medication Practices (ISMP) alert on concentration mix-ups. The team analyzed ordering, pharmacy preparation, nurse verification, pump programming, and aPTT-based titration. Heparin is on the ISMP High-Alert Medications list; Severity 9–10 items dominate.
| Process Step | Failure Mode | Effect | S | Cause | O | Controls | D | RPN |
|---|---|---|---|---|---|---|---|---|
| Ordering | Wrong concentration selected (100 U/mL vs 25,000 U/500 mL) | 10× dose delivered; major bleed; possible death | 10 | Look-alike order sets in EMR | 3 | Weight-based order set, pharmacist review | 4 | 120 |
| Pharmacy prep | Concentration labeled incorrectly on IV bag | Wrong infusion rate; bleed or sub-therapy | 9 | Manual labeling after compounding | 3 | Barcode verification; second-pharmacist check | 3 | 81 |
| Nurse verification | Two-nurse independent check skipped | Programming error reaches patient | 9 | Staffing pressure; protocol fatigue | 5 | Policy; unit audits quarterly | 6 | 270 |
| Pump programming | Rate entered in mL/hr instead of units/hr | 25× under- or over-dose | 10 | Pump accepts either unit; no interlock | 3 | Smart-pump drug library (soft limit) | 4 | 120 |
| aPTT titration | Dose adjusted using stale lab (> 8 h old) | Sub- or supra-therapeutic; bleed or clot | 8 | Lab result not reviewed before titration | 4 | Nomogram in EMR; reminder at 6 h | 5 | 160 |
2 Patient Identification — Wrong-Patient Procedure Prevention (PFMEA)
Healthcare · Patient SafetyScope: Process FMEA for bedside procedure verification on an inpatient medical/surgical floor. Driven by a near-miss sentinel event (central line attempted on wrong patient). Aligns with National Patient Safety Goal NPSG.01.01.01 (use at least two patient identifiers). Any S ≥ 9 item is automatically escalated regardless of RPN.
| Process Step | Failure Mode | Effect | S | Cause | O | Controls | D | RPN |
|---|---|---|---|---|---|---|---|---|
| Admission | Wristband printed with wrong MRN | Every downstream ID check fails silently | 10 | Registration clerk mis-typed DOB | 2 | Admit nurse visual check | 6 | 120 |
| Pre-procedure verification | Only name asked; DOB skipped | Two roommates with similar names swapped | 9 | Time pressure; single-identifier habit | 4 | Policy; audit tool during rounds | 5 | 180 |
| EMR chart open | Prior patient’s chart left open; order placed on wrong record | Wrong-patient order; delayed care; potential harm | 8 | Shared workstation; no session timeout | 5 | Manual logout reminder at huddle | 7 | 280 |
| Specimen labeling | Blood tube labeled after leaving bedside | Wrong-patient lab result; wrong treatment decision | 9 | Workflow allows labeling at workstation | 3 | Policy: label at bedside | 6 | 162 |
| Consent | Consent signed for wrong procedure | Unconsented procedure; potential harm | 9 | Copy-paste from another chart | 2 | Read-back of procedure to patient | 4 | 72 |
3 Blood Transfusion — Type and Crossmatch Process (PFMEA)
Healthcare · Transfusion MedicineScope: Process FMEA for the full red-cell transfusion chain from specimen draw to bedside administration. ABO incompatibility is rare (~1 in 40,000) but is a Severity 10 never event. Aligned with AABB (Association for the Advancement of Blood & Biotherapies) standards and FDA 21 CFR 606.
| Process Step | Failure Mode | Effect | S | Cause | O | Controls | D | RPN |
|---|---|---|---|---|---|---|---|---|
| Specimen collection | Sample drawn from wrong patient (WBIT — wrong blood in tube) | ABO-incompatible transfusion; fatal hemolytic reaction | 10 | Label not verified at bedside | 3 | Two-identifier check; signed label | 5 | 150 |
| Blood bank typing | Single type recorded (no second historic type on file) | Typing error undetected; wrong unit issued | 10 | Policy allows single-sample issue | 2 | Two-specimen rule for first-time types | 3 | 60 |
| Unit issue | Wrong unit placed in cooler for ward pickup | Wrong patient identifier on unit; potential mismatch | 10 | Manual cooler assembly by tech | 2 | Barcode scan at cooler and at pickup | 2 | 40 |
| Bedside check | Two-nurse verification at bedside skipped | Wrong-patient transfusion not caught | 10 | Remote second nurse over phone | 3 | Policy; infusion pump scan | 4 | 120 |
| Monitoring | Vitals not checked at 15 min | Early transfusion reaction missed; delayed treatment | 8 | Workload; no pump alarm at 15 min | 4 | Policy; nursing documentation audit | 6 | 192 |
4 Emergency Department Triage — Chest Pain Pathway (PFMEA)
Healthcare · Emergency MedicineScope: Process FMEA for ESI (Emergency Severity Index) triage of adult chest pain patients, from walk-in registration to physician evaluation. Trigger: a patient with acute MI was triaged as ESI-3 and waited 48 minutes before EKG. Target: first EKG within 10 minutes per AHA door-to-EKG benchmark.
| Process Step | Failure Mode | Effect | S | Cause | O | Controls | D | RPN |
|---|---|---|---|---|---|---|---|---|
| Registration | Chief complaint not captured accurately (“feeling off” not chest pain) | Acuity under-scored; STEMI missed | 10 | Registrar not clinically trained | 4 | Triage nurse re-asks within 5 min | 5 | 200 |
| ESI assignment | Atypical chest pain scored ESI-3 instead of ESI-2 | Delayed EKG and troponin; missed MI | 10 | Subjective interpretation of presentation | 4 | ESI training; annual competency | 5 | 200 |
| EKG order | EKG delayed > 10 min from arrival | Delayed STEMI diagnosis; worsened outcome | 9 | Tech not co-located with triage | 5 | Department policy; monthly metric | 5 | 225 |
| EKG interpretation | STEMI on EKG not recognized by tech | Cath lab activation delayed | 10 | Tech not trained for preliminary read | 3 | Physician interpretation within 10 min | 4 | 120 |
| Bedding | Patient placed in waiting room after triage | Deterioration unnoticed; cardiac arrest in lobby | 10 | Boarding; no monitored bed available | 5 | Pulse-ox clip in waiting room; rounding every 30 min | 7 | 350 |
5 Discharge Medication Reconciliation (PFMEA)
Healthcare · Care TransitionsScope: Process FMEA for inpatient discharge medication reconciliation on a general medicine unit. A hospital-acquired readmission review identified medication discrepancies as the leading cause of 30-day readmission. Target: zero reconciliation errors on high-risk medication classes (anticoagulants, insulin, opioids, immunosuppressants).
| Process Step | Failure Mode | Effect | S | Cause | O | Controls | D | RPN |
|---|---|---|---|---|---|---|---|---|
| Admission med history | Home medication list incomplete | Drug omitted during stay; disease exacerbation | 7 | Patient cannot recall doses; no pharmacist interview | 6 | Patient self-report; family contact if needed | 6 | 252 |
| Inpatient substitutions | Substitution not reconciled to home med at discharge | Patient given two equivalent drugs; therapeutic duplication | 7 | Formulary substitution not flagged | 5 | EMR flag on substitution | 5 | 175 |
| Anticoagulation | Home warfarin resumed without bridging instruction | Sub-therapeutic INR; thromboembolic event | 9 | Prescriber unaware of INR trajectory at discharge | 4 | Anticoag service consult for INR > 4 or < 2 | 5 | 180 |
| Patient education | Discharge instructions given in wrong language | Patient non-adherence; readmission | 6 | No interpreter at discharge | 5 | Electronic language tag in EMR | 5 | 150 |
| Post-discharge follow-up | No follow-up appointment within 7 days | Medication error uncaught; readmission | 7 | Scheduling after discharge; patient forgets | 6 | Discharge nurse schedules appt pre-discharge | 4 | 168 |
Investigate failures that slipped past your FMEA
When a risk identified in your HFMEA becomes a real event, our free 5 Whys tool guides your team through a structured root-cause analysis — no signup, no downloads. Findings can be fed back into the FMEA to update Occurrence and Detection scores.
Start 5 Whys Analysis →HFMEA Hazard Score vs Traditional RPN
Both methods produce a prioritized list of risks; they differ in scoring granularity and the presence of a decision tree.
| Attribute | Traditional FMEA (10-point RPN) | HFMEA (Hazard Score Matrix) |
|---|---|---|
| Severity scale | 1–10 | 1–4 (Minor / Moderate / Major / Catastrophic) |
| Probability scale | 1–10 (Occurrence) | 1–4 (Remote / Uncommon / Occasional / Frequent) |
| Detection | 1–10 (separate axis) | Not scored — handled in Decision Tree instead |
| Risk output | RPN = S × O × D (range 1–1,000) | Hazard Score = S × P (range 1–16) |
| Action threshold | Typically RPN ≥ 100–200, plus any S ≥ 9 | Hazard Score ≥ 8, and Decision Tree result = “Proceed” |
| Best for | Medical device design (ISO 14971), high-granularity risk scoring | Hospital process risk assessment; Joint Commission LD.04.04.05 |
| Developed by | U.S. Military (MIL-P-1629, 1949); AIAG/VDA (automotive) | VA National Center for Patient Safety (2001) |
Frequently asked questions
What is HFMEA and how is it different from FMEA?
HFMEA (Healthcare Failure Mode and Effects Analysis) is a simplified version developed by the VA National Center for Patient Safety in 2001. It uses a 4-point Severity × Probability matrix instead of three 1–10 scales and adds a Decision Tree to evaluate whether existing controls mitigate the hazard. It is recommended by The Joint Commission for proactive risk assessment.
Is FMEA required in healthcare?
Yes, in most accredited facilities. The Joint Commission Leadership Standard LD.04.04.05 requires at least one proactive risk assessment every 18 months. CMS Conditions of Participation require a QAPI program that includes prospective risk analysis. Medical device manufacturers must also apply FMEA under ISO 14971 for design risk management.
Which healthcare processes should be analyzed with FMEA first?
Start with high-severity, human-factor-heavy processes: medication administration, patient identification, blood product transfusion, surgical site marking and prep, handoff communication, and discharge medication reconciliation. These account for most of the sentinel events reported to The Joint Commission.
What is a sentinel event and how does FMEA prevent it?
A sentinel event is any unexpected occurrence involving death, serious physical or psychological injury, or the risk thereof. FMEA prevents sentinel events by mapping every failure mode, scoring its risk, and mandating action on Severity ≥ 9–10 items regardless of probability.
How do you score Severity for patient safety in FMEA?
In 10-point FMEA: Severity 10 is death or permanent disability, 9 is major injury, 8 is temporary hospitalization, 6–7 is minor injury, 4–5 is inconvenience without harm. In HFMEA: 4 is Catastrophic (death or permanent loss), 3 is Major, 2 is Moderate, 1 is Minor.
Can HFMEA be applied to medication error analysis?
Yes — medication-use FMEA is one of the most common applications. The Institute for Safe Medication Practices (ISMP) publishes an FMEA template for the five-step medication-use process. Separate FMEAs are typically run for high-risk drug classes: opioids, anticoagulants, insulin, and chemotherapy.
Who should participate in a healthcare FMEA?
A multidisciplinary team of 4–8 people with direct knowledge of the process. For a medication FMEA: pharmacist, nurse, physician, pharmacy technician, quality-improvement nurse, IT/informatics. Include frontline staff who perform the process daily — their workaround knowledge reveals undocumented failure modes.
Related resources
- FMEA: The Complete Guide — RPN formula, rating scales, 7-step process, DFMEA vs PFMEA
- FMEA Examples: 6 Real-World Cases Across Industries — manufacturing, automotive, software, food, aerospace
- Free FMEA Template — Excel & Google Sheets
- 5 Whys in Healthcare: Patient Safety RCA Guide — reactive companion to proactive HFMEA
- RCA Tools Compared: 5 Whys, Fishbone, FTA, FMEA & Pareto
- Free 5 Whys Tool — investigate failures that your FMEA did not prevent