Defining away threats to patient safety: A study of the classification of medication errors in three hospitals

1
Defining away threats to patient safety A
study of the classification of medication errors
in three hospitals

• Michal Tamuz, Ph.D.

AHRQ grant 1PO1HS1154401/3.08
2
Medication Errors

• Preventable medical errors 8th most common cause
  of injury in U.S.
• Errors resulting in adverse drug events
• Common
• Morbidity and Mortality
• Pervasive
• Costly
• (Institute of Medicine, 2000, 2006)

3
Institute of Medicine To Err is Human

• Create a learning environment
• Implement mechanisms offeedback and learning
  from error
• (Institute of Medicine, 2000)

4
Organizational Learning

• Process in which decision makers weigh past
  experience as a basis for changing the routines
  that guide future behavior
• (March, 1999 Levitt March, 1988)

5
Interpreting Experience

• How do organizations and their members
  interpret their past experience?
• Event ClassificationReflects and encodes
  the interpretation of experience

6
Research Questions

• Aviation Safety Surveillance Systems
• How does event classification influence
  information gathering?
• Do individuals and organizations define away
  dangers?
• How does event classification affect the capacity
  for organizational learning?
• Tamuz (2001a, 2001b)

7
How does event classification influence
information gathering?

• Individuals and organizations
• define away dangers.
• Decision makers avoid reporting potentially
  dangerous events by classifying them in
  non-reportable categories.

8
Event Classification

• Decision making
• Organizational memory(Bowker Star,1999)
• Available information (Arrow, 1973)
• Routines
• Classification triggers alternative
  organizational routines (March Simon. 1958
  Dutton Jackson, 1987)

9
Qualitative Research Methods Overview

• Data Collection
• Conducted extensive semi-structured interviews
  with healthcare providers and administrators
• Supplemented interviews
• Observing routine activities
• Gathering documents
• Data Analysis
• Analyzed interview transcripts for themes
• Research questions guided data analysis
• Analysis identified emerging themes
• Identified exemplars that illustrate themes

10
Qualitative Research Methods Overview

• Goals
• Build conceptual framework
• Offer practitioners new perspectives
• Raise practical and policy questions

11
Qualitative Research Methods Sample

• Longitudinal study Part 1
• 3 tertiary care hospitals (no CPOE)
• Purposeful random sampling (Patton, 2002)
• 341 healthcare providers administrators
• 141 Pharmacists
• 141 Nurses and support staff
• 28 Physicians
• 31 administrators

12
Results Exemplars

• Pharmacy Error Definition
• NursingExternal Review Committee

13
First Exemplar

• Pharmacy
• Error Definition

14
Pharmacy Error Definition

• If caught outside hospital pharmacy defined as
  reportable incident
• If caught within pharmacy non-event
• (Tamuz, Thomas, and Franchois, 2004)

15
Pharmacy Error DefinitionDefining away dangers

• It makes a difference where it was detected,
  because if it was in the pharmacy, typically,
  its considered a near miss. Its not
  reported. Its fixed. Its when it leaves the
  pharmacy that it starts becoming an error.
• (Pharmacy management team member)

16
Pharmacy Error Definition Defining away dangers

• Sensible for maintaining bureaucratic
  accountability
• Problematic for identifying sources of potential
  problems

17
Pharmacy Error Definition Event classification
influences organizational learning

• Problematic for organizational learning on
  hospital level
• Underreporting Loss of data to hospital
• Underestimate errors
• Reduce opportunities for early detection

18
Pharmacy Error Definition Event classification
influences organizational learning

• Data used for organizational learning within
  pharmacy
• Reduced stigma
• Formal Research projects
• Informal Pharmacists send group emails

19
Second Exemplar

• NursingExternal Review Committee

20
Nursing External Review Committee

• External Review Committee (ERC)
• Individual accountabilityNurses reported to ERC
  subject to disciplinary measures
• Specific definitionErrors 1 major, 3 minor
• Mandatory reporting

21
Nursing ERC

• Because I don't think there's anything
  really clearly spelled out. I mean, there's not a
  policy that says, "If this happens, this is a
  minor, you must do this if this happens, it's a
  major, you must do this.
• So it's kind of left up to you as a
  leadership team to decide how areyou gonna
  handle these situationsand what are you gonna do?

22
Nursing ERC

• It depends on the circumstances surrounding
  it. It depends on the significance of the
  error.It depends on a lot of things, a lot of
  factors.
• (Nurse management team member)

23
Nursing ERC
System-based Guideline Could anybody do this?
Could it happen again? I saw that as a system
problem, not an individual problem and I
responded to it systematically. (Nurse management
team member)
24
Nursing ERC Defining Away Dangers

• Classifying events as non-reportable
• Implications
• Nurse management teams adhere to reporting
  requirements
• Nurses not reported to ERC
• Buffers nurses from misguided ERC policy

25
Nursing ERC Event classification influences
organizational learning

• Unit Information available among team members
  and learning possible
• Among nursing unitsNo systematic sharing
• Hospital Precludes organizational learning

26
National Health Policy Implications

• Standardization of definitions (Institute of
  Medicine 2003)
• Patient Safety Organizations
• Obstacles to pooling data across hospitalsLocal
  interpretation

27
Policy Questions

• Financial Disincentives for Never Events
• Centers for Medicare and Medicaid Services (CMS)
  
• Financial disincentives Withhold payment for
  never events
• Never events
• Clearly identifiable
• Measurable

28
Policy Questions Future Research

• Clearly identifiable and measurable categories
• Linked to financial disincentives
• Hypothesis Defining away dangers