5.2 Proactive Risk Assessment (FMEA) vs Reactive Root Cause Analysis (RCA2)

5.2 Proactive Risk Assessment (FMEA) vs Reactive Root Cause Analysis (RCA2)

---

Proactive Risk Assessment: Failure Mode and Effects Analysis (FMEA)

Originally developed in the aerospace and defense sectors, Failure Mode and Effects Analysis (FMEA) is a prospective risk assessment method. The Veterans Health Administration's National Center for Patient Safety adapted it for health care as Healthcare FMEA (HFMEA), which scores each failure mode with a hazard score (severity × probability, each on a four-point scale) and a decision tree rather than the three-factor RPN described below. The Joint Commission expects hospitals to select at least one high-risk process for proactive risk assessment every 18 months (NPG.02.03.01). Rather than waiting for a patient to suffer injury, an FMEA systematically dissects a proposed clinical process, technology deployment, or high-risk workflow to identify where, how, and with what consequence the process might fail.

The Step-by-Step FMEA Methodology

+-------------------------------------------------------------------------+
|                         THE 5-STEP FMEA PROCESS                         |
|                                                                         |
|  [Step 1: Select High-Risk Process & Form Multidisciplinary Team]       |
|           Include frontline RNs, pharmacy, MDs, biomed, human factors  |
|                                    v                                    |
|  [Step 2: Map the Process Completely]                                   |
|           Flowchart every sequential step and sub-step micro-activity   |
|                                    v                                    |
|  [Step 3: Brainstorm Failure Modes & Causes]                            |
|           For each step: How could this fail? Why would it fail?        |
|                                    v                                    |
|  [Step 4: Score RPN (Risk Priority Number)]                             |
|           RPN = Severity (S) x Occurrence (O) x Detection (D)           |
|                                    v                                    |
|  [Step 5: Redesign Process & Re-evaluate Post-Mitigation RPN]           |
|           Implement strong forcing functions; measure new RPN score     |
+-------------------------------------------------------------------------+

1. Select a High-Risk Clinical Process & Assemble Team: Choose a process involving substantial vulnerability—such as the enterprise-wide transition to a new smart infusion pump platform, automated medication dispensing cabinets, pediatric chemotherapy administration, or surgical specimen handling. The CNO ensures the team includes frontline bedside nurses, pharmacists, physicians, biomedical engineers, risk managers, and human factors specialists.
2. Process Mapping (Flowcharting): Construct an exhaustive, granular process map breaking the overall workflow into numbered sequential steps and sub-steps. Bypassing granular steps blinds the team to latent failure points.
3. Identify Potential Failure Modes, Causes, and Effects: For every individual sub-step, the team answers three questions:
   • Failure Mode: In what way could this specific step fail to achieve its intended function? (e.g., smart pump library not selected; concentration incorrectly programmed; primary/secondary clamp left closed).
   • Failure Cause: What latent condition, cognitive trap, or environmental defect causes this failure? (e.g., drug library search screen requires 8 clicks; ambient glare obscures pump LCD display; alarm tone identical to bed exit monitor).
   • Failure Effect: What is the ultimate clinical consequence to the patient if this failure mode reaches them? (e.g., massive vasoactive overdose, cardiovascular collapse, prolonged ICU length of stay, death).
4. Quantify the Risk Priority Number (RPN): Each failure mode is assigned a numerical score (typically on a 1 to 10 scale) across three variables:
   • Severity (S): The clinical impact of the failure mode on the patient ($1 = \text{negligible harm/inconvenience}$; $10 = \text{catastrophic injury or death}$). Note: Severity cannot be reduced unless the fundamental clinical intervention itself changes.
   • Occurrence (O): The likelihood or estimated frequency that the failure cause will take place ($1 = \text{virtually impossible / extremely rare}$; $10 = \text{almost inevitable / occurs daily}$). Redesigning workflows directly lowers Occurrence.
   • Detection (D): The probability that the system defenses will intercept and detect the failure mode before it reaches the patient ($1 = \text{detection is guaranteed by automated forcing function}$; $10 = \text{stealth failure; zero detection capability prior to patient impact}$). Adding automated verifications directly lowers Detection scores.

$\text{Risk Priority Number (RPN)} = \text{Severity (S)} \times \text{Occurrence (O)} \times \text{Detection (D)}$

With 10-point scales, RPN values range from 1 to 1,000. Critical failure modes exhibiting high RPN scores (e.g., scores exceeding institutional thresholds like 200–300, or any failure mode with a Severity score of 9 or 10 regardless of RPN) mandate immediate engineering redesign.

1. Process Redesign and Post-Mitigation RPN Calculation: The team implements systemic safeguards—prioritizing physical forcing functions and automated hard stops—and recalculates the projected post-mitigation RPN to confirm that residual risk has been driven to an acceptable level.

---

Reactive Investigation: Root Cause Analysis and Action ($RCA^2$)

When a catastrophic adverse clinical event, sentinel event, or high-consequence near-miss breaches existing defenses, healthcare systems execute a reactive investigation. Historically known as Root Cause Analysis (RCA), the process was fundamentally revolutionized by the National Patient Safety Foundation (NPSF) and the Institute for Healthcare Improvement (IHI) with the publication of the $RCA^2$: Improving Root Cause Analyses and Actions to Prevent Harm guidelines.

The critical addition in $RCA^2$ is the deliberate inclusion of Action ($A^2$). Traditional RCAs spent 90% of their operational energy dissecting what occurred and only 10% generating corrective actions—which often defaulted to weak, administrative platitudes like "retrain the nurse" or "revise the hospital policy." $RCA^2$ mandates that an investigation is incomplete without rigorous, sustainable, high-impact systemic action plans.

Core Tenets of the $RCA^2$ Methodology

• Timely Completion: The $RCA^2$ guidance recommends finishing the analysis and action plan within about 30 to 45 days. For sentinel events, the Joint Commission requires a comprehensive systematic analysis and corrective action plan within 45 business days.
• Multidisciplinary Team with "Outside Eyes": The panel must include clinical subject-matter experts, frontline staff from the involved service line, and critically, at least one clinician or engineer completely detached from the department (an "outsider" whose fresh perspective unearths blind spots and normalized deviances that insiders overlook).
• Exclusion of Involved Staff from the Investigative Panel: Staff directly involved in the event must be interviewed thoroughly and treated with deep Just Culture empathy and psychological support (Care for the Caregiver / Second Victim support). However, they must never sit on the formal $RCA^2$ review committee, as personal guilt, defensiveness, or trauma compromises objective analysis.
• Focus on Latent System Failures vs. Active Human Slips: $RCA^2$ operates on James Reason's Swiss Cheese Model of System Accidents. Active failures (the proximal human slip, such as pushing the wrong button) are merely the final hole in the cheese. The true investigation focuses on the latent conditions: institutional management decisions, staffing algorithms, equipment purchasing choices, software architectures, and cultural norms that aligned the holes in the first place.

+-------------------------------------------------------------------------+
|                    JAMES REASON'S SWISS CHEESE MODEL                   |
|                                                                         |
|   [Latent Hazard]                                                       |
|        |                                                                |
|        v   [Organizational Deficiencies]  -> Flawed budget/staffing     |
|        O                                                                |
|        |   [Supervisory Deficiencies]     -> Inadequate preceptor train |
|        v                                                                |
|        O                                                                |
|        |   [Preconditions for Unsafe Acts]-> Extreme fatigue / clutter  |
|        v                                                                |
|        O                                                                |
|        |   [Active Human Failures]        -> Nurse selects wrong vial   |
|        v                                                                |
|   [PATIENT HARM / SENTINEL EVENT]                                       |
+-------------------------------------------------------------------------+

Analytical Tools: 5 Whys and the Ishikawa (Fishbone) Diagram

Within the $RCA^2$ process, two primary diagnostic tools assist teams in moving beyond superficial explanations:

1. The 5 Whys: An interrogative technique that originated with Sakichi Toyoda and was popularized by Taiichi Ohno in the Toyota Production System. The team explores a defect by asking "Why?" at least five successive times. Each answer forms the basis of the next inquiry, driving past active clinician slips to uncover systemic root causes.
   • Defect: A post-operative patient received a 10-fold overdose of IV hydromorphone.
   • Why 1? The nurse selected a 10 mg/mL vial instead of a 1 mg/mL vial from the automated dispensing cabinet.
   • Why 2? Both concentrations were co-located in adjacent, unlocked matrix bins in the medication room.
   • Why 3? Pharmacy stocked high-concentration hydromorphone on the general surgical floor following an unexpected central supply shortage of pre-filled 1 mg syringes.
   • Why 4? The supply chain purchasing committee made a bulk substitute purchase without notifying pharmacy clinical specialists or nursing leadership.
   • Why 5 (Root Cause)? The health system lacks an enterprise-wide policy and electronic cross-functional review mechanism requiring multidisciplinary safety evaluation before purchasing high-alert medication substitutes.
2. Ishikawa (Fishbone / Cause-and-Effect) Diagram: Categorizes contributing factors into structured enterprise branches: People (training, fatigue, communication), Methods/Processes (policies, clinical pathways), Machines/Equipment (EHR, pumps, user interfaces), Materials/Supplies (medication packaging, tubing), Environment (lighting, acoustic noise, interruptions), and Management/Governance (staffing models, production pressure).

---

The NPSF Hierarchy of Actions

The defining executive deliverable of $RCA^2$ is the Hierarchy of Actions, categorized by systemic strength and durability. Executive nurse leaders must evaluate every proposed safety remediation against this hierarchy, ruthlessly rejecting reliance on weak interventions.

+-------------------------------------------------------------------------+
|                   RCA2 HIERARCHY OF ACTIONS (NPSF)                      |
|                                                                         |
|  [STRONGER ACTIONS]                                                     |
|  * Architectural or physical plant changes                             |
|  * Forcing functions (ENFit connectors, interlocks, hard stops)         |
|  * Simplifying the process; standardizing equipment or process         |
|  * Tangible involvement by leadership                                  |
|                                                                         |
|  [INTERMEDIATE ACTIONS]                                                 |
|  * Redundancy; increased staffing or decreased workload                |
|  * Software enhancements; eliminating look- and sound-alikes           |
|  * Eliminating distractions; checklists and cognitive aids             |
|  * Standardized communication tools; simulation-based training         |
|                                                                         |
|  [WEAKER ACTIONS]                                                       |
|  * Double checks; warnings and signs                                   |
|  * New procedures, memoranda or policies                               |
|  * Training alone; additional study or analysis                        |
+-------------------------------------------------------------------------+

Evaluating Action Strength and Executive Decision-Making

---

Comparison: Proactive FMEA vs. Reactive $RCA^2$

---

Action Sustainability: Process vs. Outcome Measurement

A common failure in executive risk oversight is assuming that once an $RCA^2$ action plan is approved, the risk has been mitigated. Executive nursing governance requires rigorous longitudinal tracking of both process measures and outcome measures:

• Process Measures (Fidelity & Compliance): Measures whether the redesigned workflow is actually being executed as intended by clinical staff. For example, if a new independent double-check workflow was instituted for high-alert insulin infusions, the process measure tracks the percentage of infusions that have documented dual-nurse biometric verification within the electronic medical record (auditing compliance target: $\ge 95\%$).
• Outcome Measures (Clinical Harm Eradication): Measures whether the systemic intervention achieved the ultimate goal of eliminating patient harm. For the insulin example, the outcome measure tracks the incidence of severe inpatient hypoglycemic events (blood glucose $< 40\text{ mg/dL}$) per 1,000 patient days over the subsequent 6 to 12 months.