FMEA training for Healthcare - Sample

Failure Mode & Effect Analysis (FMEA)

Training for Healthcare

© 2013 Workflow Diagnostics, Inc., unless otherwise noted. Any redistribution or commercial use of this presentation without permission

from Workflow Diagnostics, Inc., is expressly forbidden.

Class objectives

• Understand the purpose of the FMEA

• Understand the steps of the FMEA process

• Understand how to use FMEA

• Complete an exercise and actually create an FMEA to begin feeling comfortable with the process

2

FMEA

Tra

inin

g –

Ob

ject

ives

What is Failure Mode & Effect Analysis?

• A team-based systematic and proactive approach for identifying the ways that a process can fail, why it might fail, the effects of that failure, and how it can be made safer.

• The goal is to eliminate or minimize the potential for failures, to stop failures before harm reaches the patient, or to minimize the consequences of the failure.

3

*Institute for Safe Medication Practices Canada (ISMP Canada)

FMEA

Tra

inin

g –

FMEA

Exp

lain

ed

Why Use FMEA?

• Aimed at preventing a tragedy, not simply responding to it

• Doesn’t require previous bad experience or close call (“near-miss”)

• Makes a system more fail-proof

• Fault-tolerant

*VA National Center for Patient Safety

4

FMEA

Tra

inin

g –

FMEA

Exp

lain

ed

Why me? Why you?

• Practitioners in the systems know the specific vulnerabilities and failure points

• Professional & moral obligation to “first do no harm”

• Increased expectation that we create safe systems

5

*ISMP Canada

FMEA

Tra

inin

g –

FMEA

Exp

lain

ed

Rationale for FMEA in Healthcare

Historically…..

• Accident prevention has not been a primary focus of hospital medicine

• Misguided reliance on “faultless” performance by healthcare professionals

• Hospital systems were not designed to prevent errors; they just reactively changed and were not typically proactive.

6

*VA National Center for Patient Safety

FMEA

Tra

inin

g –

FMEA

Exp

lain

ed

• Identify and prioritize high-risk processes

• Annually: • Select at least one high-risk process

• Identify potential “failure modes”

• For each “failure mode,” identify the possible effects

• For the most critical effects, conduct a root-cause analysis

• Redesign the process to minimize the risk of that failure mode or to protect patients from its effects

• Test and implement the redesigned process

• Identify and implement measures of effectiveness

• Implement a strategy for maintaining the effectiveness of the redesigned process over time

7

Joint Commission Requirement (Standard LD.5.2 effective July 2001)

*VA National Center for Patient Safety

FMEA

Tra

inin

g –

FMEA

Exp

lain

ed

Healthcare Applications

• Specimen identification

• Hospital-acquired conditions – pressure ulcers, patient falls, VAP, surgical site infections, wrong-site surgery, etc.

• Medication safety and dispensing

• Fall prevention

• Tests – delays and results

• Infection control

• Facility or new process design 8

FMEA

Tra

inin

g –

FMEA

Exp

lain

ed

How does FMEA work?

• To narrow in on key failures to address, assign each failure mode three ratings:

• Severity

• Occurrence

• Detectability

• Rate each failure mode as 1-10 for each of the three categories. (Some people use 1-5.)

9

You do this to get to the ultimate goal: Reduce/eliminate risk to the patient

FMEA

Tra

inin

g –

FMEA

Exp

lain

ed

Risk Priority Number (RPN)

Multiply the three ratings together to get the Risk Priority Number or RPN:

• RPN = Severity Occurrence Detectability

10

FMEA

Tra

inin

g –

FMEA

Exp

lain

ed

FMEA Process Steps 1 2

Select a High-Risk Clinical Process

Assemble the team

Map the Process

11

3

4 5

Brainstorm potential failure

modes

Identify effects of each failure mode (Severity; Occurrence;

Detectability)

Prioritize the failure modes

(RPN)

6

7 8

Develop mitigation strategies and

redesign process

Implement and evaluate the

redesign

Monitor effectiveness of new processes.

9

FMEA

Tra

inin

g –

Co

nst

ruct

ion

Select a High-Risk Process

• Select processes with high potential for having an adverse impact on the safety of individuals served.

• Processes that: • have variable input

• are complex

• non-standardized

• heavily dependent on human intervention

• performed under tight or loose time constraints

• tightly coupled and hierarchical (not team-oriented)

are all candidates for consideration. 12

FMEA

Tra

inin

g –

Co

nst

ruct

ion

*ISMP Canada

High-Risk Processes – Examples

• Medication administration

• Surgery

• Transfusions

• Restraints

• Isolation

• Emergency or resuscitative care

• High-risk populations

13

FMEA

Tra

inin

g –

Co

nst

ruct

ion

*ISMP Canada

Assemble a team

• Leader

• Facilitator

• Scribe/Recorder

• Process experts

• Include all areas involved in the process

• “Outsider” – objective, “naïve”

• 6-10 optimal size

14

FMEA

Tra

inin

g –

Co

nst

ruct

ion

*ISMP Canada

Map the Process

• Pick a manageable portion of the process

• Make sure the topic is narrow enough of a focus

(don’t try to cure world hunger)

15

FMEA

Tra

inin

g –

Co

nst

ruct

ion

*ISMP Canada

Map the Process

• Define beginning and end of the process

• If process is complex, identify the area to focus on.

• Chart the process as it is normally done, using the collective process knowledge of the team.

• Number each step

16

FMEA

Tra

inin

g –

Co

nst

ruct

ion

*ISMP Canada

For each step in the process, list all the failure modes –all the ways the process could break down or go wrong

Could have multiple failures for each process step.

jProcess Step

Failure Mode #1

Failure Mode #2

Failure Mode #3

Failure Mode #4

17

Brainstorm Potential Failure Modes

FMEA

Tra

inin

g –

Co

nst

ruct

ion

Identify Effects

• Review each failure mode and identify the potential Effect(s) for each one

• Try not to overlook any Effects -> results will impact the risk ratings done later

If failure mode occurs, then what are the consequences?

18

FMEA

Tra

inin

g –

Co

nst

ruct

ion

*ISMP Canada

Severity • The seriousness and Severity of the Effect (to the process or

patient) of a failure if it should occur.

19

Process step

Potential failure mode

Potential failure effects

S

E

V

Potential causes

O

C

C

Current process controls

D

E

T

R

P

N

What is the

step?

In what ways can the step go wrong?

What is the impact on the

customer if the failure

mode is not prevented or corrected?

N

What causes the step

to go wrong?

(i.e., How could the failure

mode occur?)

N

What are the existing controls

that either prevent the failure mode

from occurring or detect it

should it occur?

N N

FMEA

Tra

inin

g –

Co

nst

ruct

ion

Scoring Scale

• Need to have a scale to reference

• Want to have a consistent scale throughout your organization

• Make it meaningful to your organization

• Start with 1 and end with 10 (don’t use zero)

• 1 is best and 10 is worst

• Relate 1 through 10 to each of Severity, Occurrence, and Detectability

• Customize the scales to your organization 20

FMEA

Tra

inin

g –

Co

nst

ruct

ion

Occurrence • Also known as Frequency, it is the likelihood or number of times a specific

failure (mode) could occur.

• Before we assign ratings for the probability of occurrence , we’ll list the possible causes for each failure mode.

• We’ll assign occurrence ratings to each cause of the failure.

• 1 means it almost never happens-10 means it always happens

21

Process step

Potential failure mode

Potential failure effects

S

E

V

Potential causes

O

C

C

Current process controls

D

E

T

R

P

N

What is the step?

In what ways can the step go wrong?

What is the impact on the

customer if the failure mode is not prevented or corrected?

N

What causes the step to go

wrong? (i.e., How could the failure mode

occur?)

N

What are the existing controls

that either prevent the failure

mode from occurring or

detect it should it occur?

N N

FMEA

Tra

inin

g –

Co

nst

ruct

ion

Occurrence:

In the Titanic example, we would rate the probability of an "iceberg that could potentially cause a hull failure” and not simply the “probability of a hull failure”

22

FMEA

Tra

inin

g –

Co

nst

ruct

ion

Detectability • The likelihood of detecting a failure or effect of a failure BEFORE it is felt

by the patient.

• Need to understand “Current Process Controls”

• On a scale of 1-10, a “1” means the control is absolutely certain to detect the problem; “10” means the control has no chance to detect the problem or no control exists.

23

Process step

Potential failure mode

Potential failure effects

S

E

V

Potential causes

O

C

C

Current process controls

D

E

T

R

P

N

What is the step?

In what ways can the step go wrong?

What is the impact on the

customer if the failure mode is not prevented or corrected?

N

What causes the step to go

wrong? (i.e., How could the failure mode

occur?)

N

What are the existing controls

that either prevent the failure

mode from occurring or

detect it should it occur?

N N

FMEA

Tra

inin

g –

Co

nst

ruct

ion

Detectability • “Detectability” means detecting the issue while we

can still do something about it (before it reaches the customer/patient).

• In the Titanic example, we’re not about detecting that the watertight compartments are flooded. We’re about detecting the iceberg and avoiding it.

• That’s why we write “Process Controls,” “Current Controls” or “Current Process Controls” on our FMEA forms.

24

FMEA

Tra

inin

g –

Co

nst

ruct

ion

Calculate Risk Priority Number (RPN)

• Rates impact of failure on patient / system based on Severity, Occurrence and Detectability

• Multiply three scores to obtain RPN Severity – consequence of failure if it occurs

Occurrence – probability or actual frequency of failure

Detectability – probability of the failure being detected or prevented before the effect is realized

• Consider assigning priority to high Severity score even if RPN is low

25

FMEA

Tra

inin

g –

Co

nst

ruct

ion

*ISMP Canada

Prioritizing RPN’s

• Prioritize your efforts and improvement resources according to RPN.

• High RPNs are more serious, should be addressed first, and deserve more effort and resources.

• Note that functions with low RPN’s might often have “none” as the recommended action unless the action were particularly easy and low cost.

26

Note: May use a threshold for action (RPN or severity score)

FMEA

Tra

inin

g –

Co

nst

ruct

ion

Determine Root Cause(s)

Use appropriate root-cause analysis tools: • Fishbone (Ishikawa, Cause and Effect)

• 5-Why

• Pareto

• Scatter plots

27 Important: Determine true root cause of each potential Failure Mode before determining the mitigation strategies, if it is not readily apparent.

May be considered a separate step

in FMEA

FMEA

Tra

inin

g –

Co

nst

ruct

ion

Reduce Severity Examples to reduce Severity:

• Protection – gloves, masks, face shields

• Emergency shut-offs, fail-safe operation

• Sprinkler systems, fire doors

• Patient positioning

• Alternative materials, e.g., safety glass, Pyrex

• Warnings and messages

• Backup and redundant systems

• Patient and family awareness and education

• Expanding supplier base, multiple sources

• Shared design with vendors 28

Also consider the impact on Occurrence and Detectability

FMEA

Tra

inin

g –

Co

nst

ruct

ion

Prevent Occurrence Examples to Prevent Occurrence

• Continual improvement, problem-solving teams

• Increasing process performance (capability)

• Address multiple causes

• Move checks earlier in the process

• Staff education and training

• Error-proofing (poka-yoke)

• Better data collection, publish data

• Protective storage, inventory management

• Supplier evaluation and monitoring

29

FMEA

Tra

inin

g –

Co

nst

ruct

ion

Improve Detectability

Ways to improve Detectability (and Prevention)

• In-process checks instead of post-process

• Automated checks and early warnings

• Barcodes, wristbands, asking name/DOB

• Better measuring devices, calibration checks

• Verification and double-checks

• Error-proofing (poka-yoke)

• Use colors, shapes to identify materials

• Statistical process control (SPC)

• Equipment and process validations

• Audits, system testing and monitoring 30

FMEA

Tra

inin

g –

Co

nst

ruct

ion

• Important: Assign a person responsible for each action item, and a due date for completion

• Follow-up on assigned actions

• Verify actions taken have intended results

31

Implement & Evaluate the New Process – Key Steps

FMEA

Tra

inin

g –

Co

nst

ruct

ion

Evaluate New Process

After implementation:

• Determine what actions were taken – different than proposed?

• Collect data on the new process

• Calculate new RPNs based on actions taken

• Reassess RPNs and determine next actions (on same item if RPN is still high, or new high RPN)

• At regular intervals, re-assess to ensure the new process remains in place and effective

32

FMEA

Tra

inin

g –

Co

nst

ruct

ion

SECTION 3

FMEA TAKEAWAYS

33

Hea

lth

care

FM

EA –

Take

away

s

Tips for Success

• Start small, aim for early success

• Narrow … Narrow … Narrow

• Use different team members from same dept. for different parts of the process (compare to RCA – not able to do that)

34

Hea

lth

care

FM

EA –

Take

away

s

*ISMP Canada

FMEA Pitfalls • Incomplete flowchart of the process

• Confusing Effects (symptoms) with Causes

• Not determining the true root causes

• Mixing Severity, Occurrence, Detectability

• Inconsistent scoring of Sev, Occ, Det

• Lack of RPN resolution (not a full 10-point scale)

• Not devoting enough time to the effort

• Not having a facilitator to keep process moving

• Not gathering input from all roles in the process

• Not having the right team members

• Low accountability for conducting

35

Hea

lth

care

FM

EA –

Take

away

s

Benefits of Implementing FMEA

• Safety-minded culture

• Proactive problem resolution

• Prevention of failures vs. rework and damage control

• Failure-proof approach vs. punitive

• Sense of control and ownership

36

Hea

lth

care

FM

EA –

Take

away

s

*ISMP Canada