Patient Safety

1
Patient Safety Usability of Medical
DevicesPart I
Gill Ginsburg, M.A.Sc Human Factors Biomedical
Engineer Trillium Health Centre
Erin Barkel, B.A.Sc Patient Safety/Risk
Management Specialist Niagara Health System

• 2004 Fall CESO Conference

2
Outline Part I

• Intro to usability
• Intro to Human Factors Engineering
• Why do users make mistakes?
• Intro to patient safety medical error
• Canadian Adverse Events Study
• Examples

3
Introduction to Usability
Mikes New CarMonsters, Inc.
4
Introduction to Usability

• Usability issues with Mikes new car
• Complex dashboard
• Too many buttons / switches
• Functions are not obvious
• No logical grouping
• Hood is too high for Mike
• Sully doesnt fit
• New exciting features are too complicated to
  useMike wants his old car back!

5
Introduction to Usability
www.baddesigns.com
6
Introduction to Usability
www.baddesigns.com
7
Introduction to Usability
X
8
Introduction to Usability
www.baddesigns.com
9
Introduction to Usability
Other Usability Examples
???
10
Introduction to HFE

• Human Factors Engineering (HFE) ensures that
  systems are easy-to-use
• Multidisciplinary engineering, medicine,
  psychology, computing, statisticsetc.
• Design of systems according to Human Factors
  Principlesiterative process incorporating user
  feedback
• Evaluation of systems for usability, safety,
  efficiency effectiveness

11
HFE Principles

• Easy-to-use systems incorporate these Human
  Factors Principles

• Good error messages
• Prevent errors
• Clear closure
• Reversible actions
• Use users language
• Users in control
• Help documentation

• Visibility of system status
• Consistency standards
• Match between system world
• Minimalist design
• Minimize memory load
• Informative feedback
• Flexibility efficiency

12
Illustration of HFE Principles
13
An Easy-to-Use System is

• Effective
• Task completed, users goals met
• Efficient
• Task completed quickly without undue cognitive
  effort
• Easy-to-learn
• System is predictable and consistent

• Engaging
• User experiences pleasant interaction with the
  system
• User satisfied with how system supports
  completion of task
• Error tolerant
• System prevents errors and assists in error
  recovery

14
HFE Techniques to Ensure Usability of Systems

• Heuristic evaluation
• How does the system violate the HFE principles?
• What is the severity of the violations?
• User testing
• Real users
• Realistic tasks
• What mistakes are made?
• What is the severity of the mistakes?
• Other performance measures task completion time,
  mental workload, user preference

15
HFE Techniques to Ensure Usability of Systems

• Observations
• Task analysis
• Work domain analysis
• Questionnaires
• Surveys
• Interviews
• Focus groups

16
Why do users make errors?
Device Use
human error
patient injury or death
Adapted from Kaye Crowley, 2000
17
Examples of Medical Error

• Incorrectly sterilizing equipment
• Administering wrong medication
• Administering wrong dose
• Administering wrong blood type
• Wrong site surgery
• Making an incorrect diagnosis
• Burning a patient

18

• Computers allow us to make mistakes faster than
  any other invention in history
• -Unknown

19
Canadian Adverse Events Study

• Principal Investigators Ross Baker and Peter
  Norton
• Released May 2004
• Based on a review of 3,700 charts from 20 acute
  care facilities
• Year 2000 data

20
Methodology

• Nurses reviewed the charts looking for any of the
  18 triggers that might indicate that an AE had
  occurred
• 40.8 of charts had at least one trigger
• Charts were then reviewed by Doctors
• Looking for evidence that an injury that caused
  disability, death or a prolonged LOS was present
• Injury caused by health care management

21
Findings

• 1 in 13 patients will experience an AE
• 255 of these AEs required an additional 1521 days
  in hospital
• About 1 million bed days nation wide
• 5 of AEs resulted in permanent disability
• 16,500 deaths

22
Recommendations

• Near Miss/Close Catch Reporting
• Accident Ratio Study
• Incident Reporting
• Renewed efforts to promote incident reporting
• Using Root Cause Analysis to investigate
  incidents
• Ask why 5x

23
Niagara Health System

• Last of the HSRC amalgamations, and the largest
• 7 sites
• 6 municipalities
• Population based of approximately 450,000

24
The Challenge

• Regionalization
• 7 Distinct Site Cultures
• Different levels of awareness of patient safety
• Different attitudes towards reporting
• Different methods of reporting
• Need to standardize reporting
• Consistent data set
• Consistent, conscientious reporting

25
Standardize Data Collection

• In June 2004, 3 of 7 sites were using the Encon
  Incident Reporting system
• The remaining 4 were using homemade forms
• Inservice sessions were run at the remaining 4
  sites
• As of September, all NHS sites are using Encon

26
Continuing Efforts

• Need for continuous inservicing
• Maintain staff awareness
• Develop awareness of Near Miss/Close Catch
  situations
• Increase visibility of Risk Management
  initiatives and demonstrate accountability
• Address staff fear (e.g. that reporting is
  punitive)

27
Cautionary Note

• Increased volume is not reflective of a higher
  error rate
• Incidents are presently under reported at most
  facilities
• Education of staff will lead to an increase in
  reporting

28
Medication Safety Committee

• Part of our Service Excellence Initiative
• Reporting to the Inspiring Excellence Council
• Representatives from Risk Management, Pharmacy,
  Nursing, Human Resources and Finance

29
Medication Safety Committee

• First Year Goals
• Increase incident reporting
• Complete/Revise the Regional Medication
  Administration Policy
• Provide education to frontline staff on the
  policy and the importance of reporting
• Work on developing the framework for a Just
  Culture (Marx, 2001)
• Creating a list of Look-a-like, Sound-a-like
  drugs in our facilities
• Implement a education strategy to reduce errors
  associated with these drugs

30
Other Projects

• Best-of-Breed
• Joint effort by Finance, Information Technology
  and Biomedical Departments
• Standardize purchasing only the best products,
  that are well supported and are usable, will be
  purchased

31
Projects at Trillium Health Centre

• Infusion pump selection
• Usability of bed alarms
• Usability of diagnostic imaging systems
• Incorporating human factors specifications into
  Request for Proposal process

32
IV Pump Selection

• Background
• Over 500 general-purpose IV pumps in hospital
• Existing contract expiring
• Need for smart features for patient safety
• Dose-error reduction
• Automated programming
• Need for standard pump across hospital

33
IV Pump Selection

• 3 pumps after RFP
• Similar functionality features
• Initial selection process not successful
• Used HFE to evaluate usability of pumps to
• Choose best pump for end users
• Enhance patient safety

34
IV Pump Selection

• Heuristic Evaluation
• Based on Human Factors principles
• Revealed usability issues
• Revealed information about causes of errors
• User testing
• 5 clinical areas, 14 nurses 3 anaesthetists
• Realistic scenarios
• Observed recorded of errors severity
• Usability errors
• Critical usability errors
• Critical undetected usability errors

35
IV Pump Selection
Total Number of Usability Errors
36
IV Pump Selection
Number of Critical Usability Errors
37
IV Pump Selection
Number of Undetected Critical Usability Errors
38
IV Pump Selection
Total of Errors Across Clinical Areas
39
IV Pump Selection
Usability Characteristic Participants who preferred Participants who preferred Participants who preferred
Usability Characteristic Pump A Pump B Pump C
Easiest to program a basic infusion 4 5 12
Easiest to program from a drug library 8 3 5
Easiest to program from a drug calc 10 5 4
Easiest to loading a set 8 5 7
Easiest to transport 12 2 6
Most user-friendly prompts 6 5 5
Most user-friendly keypad 5 6 9
Most user-friendly display 6 6 9
Overall preference 6 5 8
40
IV Pump Selection

• Benefits of using HFE to evaluate usability
• Structured objective approach
• User involvement
• Feedback to vendors
• Customize user training
• User familiarity preference not always an
  indicator of device usability

41
Thank you!
Questions?

• Gill Ginsburg
• gginsburg_at_thc.on.ca
• 905-848-7580 x 3016

Erin Barkel EBarkel_at_niagarahealth.on.ca 905-684-72
71 x 4420