Failure Mode and Effects Analysis Part II

1
Failure Mode and Effects AnalysisPart II

• Acknowledgements
• Dr. Michael Maddox
• Dr. Janey Barnes
• Dr. Kristine Delano
• HFES 2004

2
Setting the Stage for a UFMEA

• Question Why would we use FMEA method instead of
  some other
• risk analysis technique?
• Regulatory Basis There isnt one at least
  there isnt a requirement for a specific risk
  analysis technique.
• 21 CFR Part 820 Quality System Regulation,
  Subpart C Design Controls
• FDA Publication 1497 (2000) Medical Device Use
  Safety Incorporating Human Factors Engineering
  into Risk Management
• JCAHO Hospital Accreditation Standards 2004,
  Standard PI.3.20 An ongoing, proactive program
  for identifying and reducing unanticipated
  adverse events and safety risks to patients is
  defined and implemented.
• Question What is the difference between a
  usability FMEA and a component FMEA?

3
Setting the Stage for a UFMEA

• Dept. of Veterans Affairs (VA) Healthcare FMEA
  (HFMEA)http//www.patientsafety.gov/SafetyTopics
  /CogAids/HFMEA/http//www.fda.gov/cdrh/CDRHHHC/0
  91202-presentation-purcell.html
• Hybrid approach FMEA, RCA, HACCP
• Five-step process, including testing
• Imports diagramming and failure modes
  identification from FMEA
• Radically changes definitions of likelihood,
  severity, and mitigation
• Develops fixes and assigns responsibility
• Tests changes to verify effectiveness
• Aimed specifically at hospital processes
• Includes forms and subprocesses

4
Setting the Stage for a UFMEA

• Planning for a UFMEA
• Form analysis team.
• Define anchors for nominal rating categories for
  likelihood, severity, and mitigation.
• Decide on action breakpoints (optional).
• Set up documentation typically a spreadsheet.

5
Setting the Stage for a UFMEA

• Analysis Team
• Human Factors
• Medical
• Hardware Engineering
• Software Development
• Industrial Design
• Marketing/Sales
• Technical Support
• Others?

6
Setting the Stage for a UFMEA

• Define General Usage Considerations
• User population description
• Training and experience
• Usage environment
• Performance shaping factors
• Overall UI complexity

7
Define Likelihood Anchors

• Define Likelihood Anchors
• Rating Category Definition
• 1 Extremely Unlikely Not likely to be
  encountered more than a few times in the
  normal operational life of a particular
  device. Error seen in less than or equal to
  5 of participants in usability testing.
• 2 Unlikely Not likely to be
  encountered more than a few times in a year
  of operation for a particular device. Error
  seen in between 5- 10 of participants in
  usability testing.
• 3 Possible Not likely to be
  encountered more than once a month in the
  operational use for a particular device.
  Error seen in between 10-30 of participants
  in usability testing.
• 4 Likely Likely to be encountered
  about once or twice a week during operation
  of a particular device. Error seen in between
  30- 50 of participants in usability testing.
• 5 Extremely Likely Likely to be
  encountered once, or more, during each day of
  operational use. Error seen in 50, or more,
  of participants in usability testing.

8
Define Severity Anchors

• Rating Category Definition
• 1 Negligible No injury or lengthened
  treatment
• 2 Marginal Minor injury or slightly
  lengthened treatment
• 3 Significant Moderate injury,
  slight temporary loss of
• function, or moderate lengthened treatment
• 4 Critical Serious injury, moderate
  temporary loss of
• function, or prolonged lengthened treatment
• 5 Catastrophic Serious injury, death,
  permanent loss of one,
• or more, function, extensive prolongation of
  
• treatment

9
Define Mitigation Anchors

• Rating Category Definition
• 1 Definite Self-mitigating - no
  user action required
• 2 Extremely Likely Immediately
  detected and corrected
• (t lt 5 seconds)
• 3 Likely Quickly detected and
  corrected
• (5 sec lt t lt 30 seconds)
• 4 Unlikely Will propagate for a
  significant time before detected and
  corrected
• (30 seconds lt t lt 5 minutes)
• 5 Extremely Unlikely Will propagate
  without being detected or
• corrected ( t gt 5 minutes)

10
Define Action Breakpoints

• RPN Category Definition
• N/A Single Failure Criticality will become
  High Risk if a
• Condition component fails or a software
  error occurs.
• 1-8 Low Risk No further design or evaluation
  efforts required.
• 9-27 Moderately Low Unlikely that significant
  design changes will be required.
• Risk Failure mode should be further
  evaluated and then monitored during usability
  testing.
• 28-64 Moderately High Likely that significant
  design, training, or procedural element will be
  Risk required. Failure mode should be further
  evaluated and specifically addressed during
  usability testing.
• 65-125 High Risk Significant design,
  training, or procedural change is required before
  the product goes to production. Failure mode
  should be specifically addressed during
  usability testing (after a fix is made).
• What is RPN? Risk Priority Number Likelihood
  X Severity X Mitigation
• Hazard Index Likelihood x severity (Might be
  better to leave out mitigation becuas mitigation
  could lower score)

11
Example
12
Conduct UFMEA

• Break into teams of three
• Usability FMEA of blood sugar testing device

13
FMEA Software

• FMEA Information Center
• http//www.fmeainfocentre.com/software.htm
• Reliasoft
• http//www.reliasoft.com/xfmea/
• Kinetic
• http//www.fmeca.com/ffsoft/ffhome.htm
• Relex
• http//www.relexsoftware.com/products/fmeafmeca.as
  p
• Dyadem
• http//www.dyadem.com/products/fmea/
• Reliability Center
• http//www.reliability.com/proact/leap_software.ht
  m