1	IFTF^* Training Performing a FMEA Jan S. Krouwer, Ph.D. President, Krouwer Consulting http://KrouwerConsulting.com ^*IFTF Integrated Fault Tree FMEA
2	Outline Aspects of medical errors FMEA and FRACAS / RCA The FMEA process Flowcharts Fault trees FMEAs Severity and probability / Pareto Mitigations Evaluation of FMEA quality Culture: Error reporting / Regulatory
3	Failure Mode Effects Analysis History Late 40s Defense 60s Aerospace 70s Automotive Since 2002 – healthcare ISO 14971 “APPLICATION OF RISK MANAGEMENT TO MEDICAL DEVICES” not useful because it’s a horizontal standard
4	Medical errors as a cause of death
5	What it takes to prevent errors
6	Aspects of Medical Errors
7	Potential vs. Observed Errors
8	FME(C)A and FRACAS / RCA FME(C)A describe process / postulate potential errors propose mitigations to prevent potential errors evaluate – difficult, FMEA is a model (Similar to grading an essay – requires a knowledgeable reviewer) FRACAS / RCA run system / observe errors propose corrective actions to prevent recurrence of errors evaluate – easy, measure error rate (Similar to grading a math exam grade = num. correct / num. questions)
9	FMEA vs. FRACAS With all the preventable errors that occur, FRACAS/RCA might be preferred over FMEA FMEA is more comprehensive and JCAHO requires healthcare organizations to perform one FMEA a year RCA is required for sentinel events But there are pitfalls of using FMEA when there are many observed failure events
10	The uses of FMEA and FRACAS
11	The spectrum of preventable errors
12	The anatomy of an error
13	FMEA Steps Goal: Reduce risk of preventable errors in a process Form a team Select the process Flowchart the process (process mapping) Add process steps (potential errors) to fault tree Classify severity and probability of error Fill out rest of FMEAs Decide on ranking scheme Propose mitigation for top events in Pareto Evaluate FMEA
14	FMEA Steps – JCAHO and IFTF
15	JCAHO Sentinel Events 1995-2004^*
16	Select the process – JCAHO Goals Improve the accuracy of patient identification. Improve the effectiveness of communication among caregivers. Improve the safety of using medications. Improve the safety of using infusion pumps Reduce the risk of health care-associated infections Eliminate wrong-site, wrong-patient, wrong-procedure surgery. Accurately and completely reconcile medications across the continuum of care Reduce the risk of surgical fires
17	Select the process New or revised process Process with known problems, including Catastrophic events Near misses Process without observed problems, since FMEA attempts to prevent potential problems Now have to consider processes that affect pay for performance metrics
18	Flowchart the process A flowchart describes the steps required to achieve some end with inputs, process steps, and outputs. A block diagram is a high level flowchart (with many of the steps not shown) Each flowchart process step is a potential error
19	Flowchart the process
20	Flowchart the process
21	Fault tree – Addition to JCAHO FMEA A fault tree is a hierarchical chart of causes for a top level event It is similar to a cause and effect or fishbone diagram but uses “gates” A fault tree helps with severity classification The top level items are effects, the lower level items are causes Template shows why high profile errors deserve more attention
22	Fault tree – Addition to JCAHO FMEA A FMEA (table) contains a largely unstructured list The fault tree is structured and Helps to ensure that the FMEA is complete Allows one to go back and forth from the FT to the FMEA
23	Add process steps to fault tree
24	FMEA Software Comment IFTF is similar to how a word processing program facilitates writing a report compared to pencil and paper Advantage - making changes are easier But, even with the IFTF template, you start with an almost blank screen - just as you have to write the words in a report, you have to go through the FMEA steps Beware of starting with a filled in template
25	Add process steps to fault tree
26	Add to Template Before adding process steps to fault tree, fill in errors Example errors: outlier, patient sample mix-up, wrong test performed, TAT Each error can occur in more than one place on the template Convert performance to attribute errors by using CLSI EP21A
27	Add process steps to fault tree
28	Add process steps to fault tree
29	Fill in Severity and Probability
30	Severity and Frequency – VA* Frequent - Likely to occur immediately or within a short period (may happen several times in one year) Occasional - Probably will occur (may happen several times in 1 to 2 years) Uncommon - Possible to occur (may happen sometime in 2 to 5 years) Remote - Unlikely to occur (may happen sometime in 5 to 30 years) Catastrophic Event - Patient Outcome: Death or major permanent loss of function (sensory, motor, physiologic, or intellectual), suicide, rape, hemolytic transfusion reaction, Surgery/procedure on the wrong patient or wrong body part, infant abduction or infant discharge to the wrong family Major Event - Patient Outcome: Permanent lessening of bodily functioning (sensory, motor, physiologic, or intellectual), disfigurement, surgical intervention required, increased length of stay for 3 or more patients, increased level of care for 3 or more patients Moderate Event - Patient Outcome: Increased length of stay or increased level of care for 1 or 2 patients Minor Event - Patients Outcome: No injury, nor increased length of stay nor increased level of care
31	Analytical Performance vs. FMEA Goals FMEA - Attribute error frequency from VA may happen sometime in 5 to 30 years Glucose goal from ISO 15197 standard applied to all lab assays (95% of results must meet medical requirements) implies …. number of results reported that fail to meet medical requirements must be < 50,000 per year* This is 875,000 times more lenient than VA
32	Severity and Probability
33	Severity and Probability Severity Independent of probability Often refers to the effect of the event, rather than the event itself Severity can’t be changed, but its frequency of occurrence can be reduced
34	Severity and Probability Probability – Classify BASIC events Use “(c+d)/a” cell
35	Fill out initial FMEA
36	QSEs Addition to JCAHO FMEA process
37	The Ranking Problem – Transplant process
38	Choose ranking options
39	Choose ranking options
40	Can also view Risk Map
41	Can also view Risk Chart
42	Propose mitigations for top Pareto events Mitigation - consider: error occurrence reduction error detection improvement chance detection implies an unsafe situation error recovery improvement Focus on systems (not people), using Ralph Evans’s advice: “Make it easy to do the right thing and hard to do the wrong thing”
43	Evaluate cost benefit of mitigations
44	Evaluate FMEA Quality Beware of before and after RPNs Easy to measure reduction in observed errors Difficult to measure reduction in risk of potential errors RPN won’t change for a high severity lowest probability error – but the mitigation may have great benefit. Evaluate the FMEA process – how can the next FMEA be improved
45	Mitigations are Design Changes Design changes require a new FMEA Change may not be effective Change may cause new problems Example: Cedars-Sinai MC in LA abandoned 34 million dollar CPOE system after 3 months
46	Other issues More thoughts on web site at: http://krouwerconsulting.com/Essays/ListofEssays.htm Focus on systems not people Why detection shouldn’t be ranked Recovery – neglected but important Preventability – what does it mean, how is it measured, what is the effect of regulation Pay for performance – the missing metric
47	Other IFTF Features Fault trees and FMEAs from a different project or organization may be loaded into a fault tree / FMEA project Complete lists of QSEs in manual for healthcare organizations and hospital labs No additional software is needed for IFTF. IFTF uses an Access database for fault trees / FMEAs and xml files for flowcharts. IFTF Reports can be exported to PDF, Word, and Excel. These files require appropriate viewers.
48	Tools - Summary Even with the right culture, good tools are still needed
49	Cultural Issues – Reporting Errors The tools won’t work without the right culture A policy should encourage error reporting
50	Reporting errors – blame-free? “To collect productive investigative data, we must promote a culture in which employees are willing to come forward in the interests of system safety. Yet, no one can afford to offer a “blame-free” system in which any conduct can be reported with impunity”
51	Reporting errors – blame-free?
52	Classification of human errors Human error - Missing a turnoff on the highway Negligence - should have known High risk behavior – aware of the risk, did it anyway Knowingly violates rules Classifications are not mutually exclusive
53	Disciplinary action (punishment) Can be based on: error negligence rule violations high risk behavior outcomes or some combination of the above
54	Blame-free^*: Except for Premeditated or intentional acts of violence against people or damage to equipment/property; Actions or decisions involving a reckless disregard toward the safety of our customers, our fellow employees, or significant economic harm to the company; or Failure to report safety incidents or risk exposures as required by standard operating procedures and/or this policy.
55	The lack of a quality culture Many reasons for lack of quality culture “Working harder” often wins out over “working smarter”¹ Quality loses in most funding competitions Not what many people signed up for Management not committed Some quality activities perceived as adversarial Result: Quality activities are mandated by regulation
56	The regulatory issue Often, the FMEA effort will be just enough to pass inspection but will limit quality improvement. Regulatory quality goal - pass inspection, with the least amount of effort. Inspections don’t protect: - the other MGH.¹ Extremely rare for a hospital to lose accreditation “Horizontal” standards aren’t useful. Too much reliance on documentation.² Same for ISO 15189 Inspections should place more emphasis on measurements against goals
57	The regulatory issue
58	Just Do It
59	Thank you