1. Assessing Scheduled Support
of Medical Equipment
North Central Biomedical Association Seminar
September 5, 2008
Tim Ritter Senior
Project Engineer .

2. Who is ECRI Institute?
Nonprofit, international health services research agency
Seeking to promote the highest standards of safety, quality and cost effectiveness in healthcare since 1968
Resources include
Membership-based access to Web resources, databases, research reports, guides, directories
Consultation services and special projects
Information used by
Hospitals, health systems, health plans, and insurers
Government agencies, legal and regulatory professionals
©2008 ECRI Institute  

3. Alternative Presentation Titles . . .
To PM or Not to PM
How Often is Enough?
Recommendations or Requirements?
© 2008 ECRI.
©2008 ECRI Institute  

4. How safe is “safe”?
© 2008 ECRI.
©2008 ECRI Institute  

5. In the Beginning . . . Emergency Care Research Institute – 1968
“1,200 US hospital electrocutions/yr” – 1969
5,000/yr according to Ralph Nader by 1971
The electrical safety “scare” gives birth to clinical engineering
Health Devices Volume 1 Number 1 addresses the isolated power controversy
Monthly inspections of defibrillators
Quarterly inspections of monitors
© 2008 ECRI.
©2008 ECRI Institute  

6. In the Beginning . . .
Technology was not reliable and possibly dangerous
Vacuum tubes
High leakage currents
Metal chassis with minimal fluid protection
Analog meters
Heated stylus recorders
Lousy batteries
Reusable pressure transducers
© 2008 ECRI.
©2008 ECRI Institute  

7. Over the Past 40 Years Technology has become more reliable
Better designed and frequently double insulated
Devices have self diagnostics and error/event logs
Significant failures rarely identified during inspections
The need for preventive maintenance is declining
ECRI Institute has no evidence of injury or death attributable to neglected inspection or preventive maintenance
© 2008 ECRI.
©2008 ECRI Institute  

8. Initial Observations
Many “PM” programs are outdated ─ electrical safety and equipment risk levels are overemphasized
Manufacturer recommendations ─ the same for all units under all operating conditions
Risk managers and regulatory agencies frequently interpret those recommendations to be requirements
© 2008 ECRI.
©2008 ECRI Institute  

9. Initial Observations
Equipment management is essentially risk management.
Nothing can be 100% safe or 100% reliable!
Inspect something today and it can fail tomorrow . . .
© 2008 ECRI.
©2008 ECRI Institute  

10. What is Scheduled Support?
Inspection – verification of performance and safety
Preventive maintenance – periodic procedures to minimize risk of failure and to ensure continued proper operation
“PM” is frequently used incorrectly to mean inspection
Scheduled support = inspection and/or preventive maintenance
Relatively few devices require true preventive maintenance
The focus of this presentation: Scheduled Inspections
© 2008 ECRI.
©2008 ECRI Institute  

11. performing scheduled support?
What reasons are given for
performing scheduled support?
To reduce the risk of injury (to patients, staff, visitors)
To reduce the risk of significant adverse impact on patient care (e.g., due to downtime)
To comply with codes, standards, and regulations
© 2008 ECRI.
©2008 ECRI Institute  

12. Initial Questions for Any Technology
Is there value in performing inspections?
If so, do what? How often?
How to determine, document and update the decision?
Are manufacturer recommendations for performance verifications justified?
© 2008 ECRI.
©2008 ECRI Institute  

13. What is Risk?
Risk
“ Combination of the probability of occurrence of harm and the severity of that harm.”
ANSI/AAMI/ISO 14971:2000
© 2008 ECRI.
©2008 ECRI Institute  

14. What is Risk?
Risk
“ the chance of something happening that will have an impact upon objectives. It is measured in terms of consequences and likelihood.”
AS/NZS 4360:1999
© 2008 ECRI.
©2008 ECRI Institute  

15. Assessing Equipment Risk
High-risk devices
Medium-risk devices
Low-risk devices
© 2008 ECRI.
©2008 ECRI Institute  

16. High-risk devices Ventilators Defibrillators Anesthesia units
Life-support, key resuscitation, critical monitoring and other likely devices whose failure or misuse is reasonably likely to seriously injure patients or staff
Ventilators
Defibrillators
Anesthesia units
Patient lifts
© 2008 ECRI.
©2008 ECRI Institute  

17. Medium-risk devices Ultrasound scanners Electrocardiographs
Devices, including many diagnostic instruments, whose misuse, failure or absence (e.g. out of service with no replacement available) would have a significant impact on patient care, but would not be likely to cause direct serious injury
Clinical laboratory equipment
Ultrasound scanners
Electrocardiographs
© 2008 ECRI.
©2008 ECRI Institute  

18. Low-risk devices Electronic thermometers Cast cutters
Devices whose failure or misuse is unlikely to result in serious consequences
Ophthalmoscopes
Electronic thermometers
Cast cutters
© 2008 ECRI.
©2008 ECRI Institute  

19. Defining versus Predicting
Risk level definitions don’t predict probability or nature of device or system failure
High risk devices should be and usually are very reliable
Risk level should be used for prioritizing the completion of scheduled inspections
© 2008 ECRI.
©2008 ECRI Institute  

20. Case Study: Infusion Pumps
They are usually considered to be high-risk devices
They have mechanical parts and are used for many years
A comparatively high number of adverse incidents are associated with pumps
© 2008 ECRI.
©2008 ECRI Institute  

21. Why pumps may not require periodic testing!
Flow accuracy does not significantly deteriorate over time (5-10 years)
No reports of insidious or preventable failures
When they do fail, they fail “safe” (stop and alarm rather than over- or under-infusing)
Event logs show primary cause of adverse incidents is
operator error
Survey: some have stopped scheduling inspection or gone to every other year; some inspect a small sample
© 2008 ECRI.
©2008 ECRI Institute  

22. Inspecting Infusion Pumps:
How Often Is Enough?
“Many hospitals have a large number of general-purpose infusion pumps in their inventory. Therefore, the frequency at which these facilities schedule routine inspection of these pumps can have a major impact on their workforce utilization and costs. . .
ECRI believes that, for most pumps in most facilities, inspection need not be scheduled for more than once a year, and that in many cases even this frequency is unnecessary.” (Health Devices 1998)
© 2008 ECRI.
©2008 ECRI Institute  

23. Low-risk devices
If failure of a device is unlikely to result in serious consequences...
then there is little value in inspecting low risk devices
Some exceptions: ultrasound therapy units and warming cabinets
© 2008 ECRI.
©2008 ECRI Institute  

24. Joint Commission OK not to schedule IPM – but document decision
OK for different schedules for same device based on differences in use (hospital versus ambulance defibs)
Hospitals may modify manufacturer protocols based on service experience
Support decisions with records that identify issues related to failure and misuse
© 2008 ECRI.
©2008 ECRI Institute  

25. Is Scheduled Support Required?
Acceptance Inspection
Scheduled Support
Required? No
Yes
Document Decision
and Interval
Document Decision
Change in Use
Review Annual
Repair Data
Identify/Develop
IPM Procedure
Perform IPM(s)
Review Annual
IPM/Repair Data
© 2008 ECRI.
©2008 ECRI Institute  

26. Annual Review of IPM / Repair Data
Review Annual
IPM/Repair Data
Any Problems Minimized/Prevented by IPM? No
Yes
Decrease/Eliminate Scheduled Support
Maintain/Increase Scheduled Support
© 2008 ECRI.
©2008 ECRI Institute  

27. Determining and Documenting Scheduled Support Decisions
Need to switch emphasis from equipment “risk level” to failure mode analysis
Need to identify differences in use/environment
Need to document these issues have been considered
© 2008 ECRI.
©2008 ECRI Institute  

28. Scheduled Support Assessment Form
ECRI Institute’s
Scheduled Support Assessment Form
Equipment history – failure data
Use issues
portable device/battery power?
heavy or infrequent use?
Routine user performance verification or pre-use check?
© 2008 ECRI.
©2008 ECRI Institute  

29. Numeric/Formulaic Assessment Tools
These tools tend to emphasize failure effects and ignore failure mode data
Can these tools produce an objective determination?
Can they produce consistent results?
© 2008 ECRI.
©2008 ECRI Institute  

30. Hypothetical Scheduled Support Assessment Scheme
Consequence of Failure (1–4)
multiplied by
Likelihood of Failure (1–4) =
“Risk” Score
© 2008 ECRI.
©2008 ECRI Institute  

31. Hypothetical Scheduled Support Assessment Scheme
Risk Score ≥8: device is likely to need inspection
How can factoring 2 subjective decisions
identifying a worst case failure result
estimating the likelihood of such a failure
produce an objective, definitive determination???
© 2008 ECRI.
©2008 ECRI Institute  

32. Beef up
© 2008 ECRI.
©2008 ECRI Institute  

33. © 2008 ECRI.
©2008 ECRI Institute  

34. © 2008 ECRI.
©2008 ECRI Institute  

35. An Interesting Study
Global Failure Rate: A Promising Medical Equipment Management Outcome Benchmark
Journal of Clinical Engineering July/September 2006
Binseng Wang et al. present failure data from three independent service organizations
Data is presented as Failures/Device/Year
Failure = # completed repair work orders
© 2008 ECRI.
©2008 ECRI Institute  

36. What is a Failure? “The device doesn’t operate”
“Repair or calibration had to be performed”
“Equivalent to a work order”
“Shouldn’t include cosmetic repairs”
“Shouldn’t include work orders for user abuse or error”
© 2008 ECRI.
©2008 ECRI Institute  

37. Failure: Proposed Definition
“The condition of not meeting intended function or safety requirements and or a breach of physical integrity.
A failure is corrected by repair and/or calibration.”
A Universal Medical Technology Service Nomenclature
Health Devices - June 2007
© 2008 ECRI.
©2008 ECRI Institute  

38. Definition Challenges
It may be difficult to determine when user abuse or user error is the cause of a failure.
Similarly, environmental conditions (e.g., temperature, electrostatic discharge, line voltage spikes) beyond manufacturer specifications would ideally be excluded but are often not easily identified.
© 2008 ECRI.
©2008 ECRI Institute  

39. Failure Data: Key Questions for Optimizing Scheduled Support Activities
Review findings from inspections and repairs:
were any failures not detected by users?
were any failures preventable?
were any failures due to user abuse?
© 2008 ECRI.
©2008 ECRI Institute  

40. Sources of Information/Evidence
Internal
Maintenance management system – Failure Data!
Use/environment considerations
External
Equipment manufacturers
Official bodies (FDA, Health Canada)
Independent bodies (ECRI Institute)
Other equipment users
© 2008 ECRI.
©2008 ECRI Institute  

41. Assessing the Need for Inspection
Start with manufacturer recommendations but consider that they were developed:
before reliability data was available
for the device in every type of facility, anywhere over its expected life (i.e., a worst-case use scenario)
And, manufacturers sell . . .
service contracts!
© 2008 ECRI.
©2008 ECRI Institute  

42. Assessing the Need for Inspection
Consider manufacturer recommendations but also equipment service experience and the use environment
Determine appropriate inspection procedure and inspection interval
Eliminate unnecessary inspections particularly on low-risk devices
De-emphasize electrical safety testing
© 2008 ECRI.
©2008 ECRI Institute  

43. Another Look at Risk © 2008 ECRI.
“ Combination of the probability of occurrence of harm and the severity of that harm.” In terms of scheduled support, the probability of occurrence of harm translates to quantifying the probability and types of equipment failure
© 2008 ECRI.
©2008 ECRI Institute  

44. Assessing the Need for Inspection
Document decisions on a support assessment form
Present conclusions to hospital’s safety/quality committee for approval
Modify inspection intervals based on failure data from ongoing repairs and changes in device use
© 2008 ECRI.
©2008 ECRI Institute  

45. Failure Data: The Key to Scheduled Support
Failure data is needed to assess the value of scheduled support
Determine and track any failures
- that were not/would not have been recognized by clinicians
- that could have been prevented
Step 1: Collect and analyze CMMS data
Step 2: Compare it to data from other service groups
© 2008 ECRI.
©2008 ECRI Institute  

46. Strategies for Change
Process needs to become driven by relevant failure data ─ start by reviewing experience with pumps and monitors
Adopt the proposed definition for failure
Collect data for each model to share/compare with other groups
ECRI Institute will be facilitating a model-specific failure database
When in doubt, perform inspections on a sample
© 2008 ECRI.
©2008 ECRI Institute  

47. Strategies for Change
Lobby manufacturers to be more flexible with maintenance recommendations ( . . .“or in accord with hospital experience”)
Request scheduled support requirements in equipment RFPs
Forward replies of excessive requirements to ECRI Institute
Educate governmental agencies and accreditation groups by demonstrating relevant failure data
© 2008 ECRI.
©2008 ECRI Institute  

48. What is Benchmarking?
The continuous, systematic search for, and implementation of better practices that lead to improved performance (
The process of comparing business practices and performance levels to gain new insights and to identify opportunities for making improvements (Cohen et. al. 1995)

49. Benchmarking and Clinical Engineering
Indicators or metrics – numbers that represent level of performance of significant CE activities
Annual service cost/device
Average response time
Multiple indicators for contrasting cost with service quality
Indicators or metrics – numbers that characterize equipment performance
Failure rate
Downtime

50. Evolution of a Product Key new feature: Support Assessment database
Health Devices Inspection and Preventive Maintenance System™
becomes BiomedicalBenchmark™
Key new feature: Support Assessment database

51. BiomedicalBenchmark™ ECRI Institute data
BENEFIT
Acquisition Cost
Better budgeting; comparison of Service Cost to Acquisition Cost (a common denominator)
Expected Life
Improved planning/budgeting
Service Contract Cost
Compare the cost of service options; negotiate better terms
IPM Intervals
Estimate workload; suggestions for reducing inspection frequency
© 2008 ECRI.

52. BiomedicalBenchmark™ User-supplied data
BENEFITS
CE Department Demographics
Compare staffing levels / work space / non-IPM and repair activities
Inspection Intervals
External support for your findings
Acquisition Cost
Budgeting and negotiating better pricing
Equipment Failure Rates
Objective basis for scheduled support decisions; also useful for equipment selection
Average IPM / Repair Times
Estimate / compare work load by model
Model-specific IPM Procedures
Reduce time to develop new procedures
© 2008 ECRI.

53. Questions / Comments? tritter@ecri.org 610-825-6000 x5168
ECRI Institute Web site:
© 2008 ECRI.
©2008 ECRI Institute