10 years after "To Err is Human": An RCA of Patient Safety Research? (Text Version)

Slide presentation from the AHRQ 2008 conference showcasing Agency research and projects.

Slide Presentation from the AHRQ 2008 Annual Conference

On September 9, 2008, Peter Pronovost, made this presentation at the 2008 Annual Conference. Select to access the PowerPoint® presentation (1.8 MB; Plugin Software Help).

Slide 1

10 years after "To Err is Human": An RCA [root cause analysis] of Patient Safety Research?

Peter Pronovost, MD, PhD

Slide 2

Objectives

  • To reflect on some of the barriers to patient safety research
  • To consider an overview for training in patient research

Slide 3

Bilateral cued finger movements

The slide shows three Magnetic Resonance Imaging (MRI) scans of a patient's brain.

Notes:

  • This pre-operative functional MRI (fMRI) was performed to help plan surgery. Results show fMRI signal changes elicited by bilateral finger movements. Some of the signal changes directly overly shortest path to tumor. Surgeon therefore elected to come in via dorsal para-sagittal approach (from high just-to-left of midline and go down towards mass).
  • Patient awoke without deficit, and has had no evidence of recurrence in the two years since surgery. Now a third year med student.

Slide 4

The slide shows an x-ray of a patient's torso with two arrows pointing to a retained lap sponge and retracted ureteral stent.

Slide 5

System Failures Slowing Progress in Patient Safety

The slide shows an image of a red arrow flowing through slices of Swiss cheese. Each hole the arrow flows through represents the following:

  • Insufficient capacity to train researchers
  • Insufficiently robust research
  • Failure to view the delivery of care as a science*
  • Insufficient partnerships between academic and quality communities
  • Patients continue to suffer preventable harm
  • Reason model
  • Note: *Highlighted text

Slide 6:

Translation Superhighway

The slide presents a "Transitional Research Model."

  • Understanding Disease Biology
  • T1: Translating to Humans
    • Formulating, Analyzing, and Testing Pre-Clinical Models
  • Identifying and Comparing Effective Therapies
  • T2: Translating to Practice
    • Summarizing evidence and understanding if and how these therapies work in practice.
  • Implementing, Disseminating and Sustaining Research, Monitoring Outcomes
  • Improved Health Outcomes

Slide 7

The slide shows a black and white photograph of a toddler playing in the sand.

Slide 8

System Failures Slowing Progress in Patient Safety

The slide repeats the diagram from slide 5. Each hole the arrow flows through represents the following:

  • Insufficient capacity to train researchers
  • Insufficiently robust research*
  • Failure to view the delivery of care as a science
  • Focus on differences rather than similarities with other types of research
  • Patients continue to suffer preventable harm
  • Reason model
  • Note: *Highlighted text

Slide 9

The slide presents a diagram composed of two lines intersecting and forming a cross. Where the lines intersect is a red circle. In the open space of the upper right hand quadrant of the cross, is a red X. The top of the cross reads, "Central Mandate;" on the right, "Feasible;" on the bottom, "Local Wisdom;" and on the left, "Scientifically Sound."

Notes:

  • Safety efforts are where the X is. We need to migrate to be more scientifically sound and tap into local wisdom.

Slide 10

Exercise

  • Please answer each question with a score of 1 to 5.
    1 is below average, 3 is average and 5 is above average.
    • How smart am I?
    • How hard do I work?
    • How kind am I?
    • How tall am I?
    • How good is the quality of care we provide?

Slide 11

Improving Sepsis Care (n= 19 intensive care units [ICUs])

The slide shows two, separate bar graphs presenting the results for "Mortality" and "ICU LOS [Length of Stay]."

  • Mortality:
    • Oct-Dec 2003: 41.8%
    • Mar-May 2004: 21.9%
    • July-Sept 2004: 13.1%
  • Note: 69% Reduction (p <0.001)
  • ICU LOS:
    • Oct-Dec 2003: 10.0 days
    • Mar-May 2004: 7.6 days
    • July-Sept 2004: 6.2 days
      • Note: 36% Reduction (NS)

Notes:

  • Data for quality improvement

Slide 12

Improving Sepsis Care (n= 19 ICUs)

The slide presents a large, red "X" within a circle over top of the duplicate bar graphs from the previous slide.

Notes:

  • Measurement error as large an issue as selection bias

Slide 13

Framework for Patient Safety Research and Practice

  • Measuring Patient Safety
  • Translating Evidence Intro Practice (TRIP)
  • Identifying and Mitigating hazards
  • Improving Culture and Communication
  • Building Capacity and Organizing for Safety
  • Reducing Diagnostic Errors
  • Note: Pronovost Circulation in press

Notes:

  • Study design must be appropriate for question

Slide 14

  • Translating Evidence Into Practice:
    • Envision the problem within the larger health care system
    • Engage collaborative multi-disciplinary teams centrally (stages 1,2, and 3) and locally (stage 4)
  1. Summarize the Evidence:
    • Identify Interventions associated with improved outcomes
    • Select interventions with the largest benefit and lowest barriers to use
    • Convert interventions to behaviors
  2. Identify local barriers to implementation: understand the process and context of work
    • Observe staff performing the interventions
    • "Walk the process" to identify defects in each step of intervention implementation
    • Enlist all stakeholders to share concerns and identify potential gains/losses associated with intervention implementation
  3. Measure Performance:
    • Select Measures (process and/or outcome)
    • Develop and pilot test measures
    • Measure Baseline Performance
  4. Ensure all patients receive the interventions:
    • Engage: Explain why the interventions are important
    • Educate: Share the evidence supporting the interventions
    • Execute: Design an intervention "toolkit" targeted to barriers employing standardization, independent checks and reminders, and learning from mistakes
    • Evaluate: Regularly assess performance measures
  • Note:  Pronovost BMJ in press

Slide 15

The slide shows a photograph of plastic respirator tubes.

Slide 16

Patient Safety Learning Communities

The diagram shows three gears representing "Industry Level," "Unit Level," and "Hospital or Trust Level." Arrows show "Industry and Unit Level" rotating clockwise, whereas "Hospital or Trust Level" is rotating counter clockwise. Around the gears is a large circle. At the top of the circle a box reads, "1. Identify Hazards;" on the right, "2. Analyze and Prioritize Hazards;" on the bottom, "3. Mitigate Risks;" and on the left, "Evaluate Effectiveness of Risk Reduction."

  • Patient safety learning communities relate to each other in a gear like fashion: as the identified hazards require stronger levels of intervention to achieve mitigation, the next learning community is engaged in action, eventually feeding back to the group that provided the initial thrust. Each group (unit, hospital, industry) follows the same four- step process, but they engage unique matrices of stakeholders to mitigate hazards that are within their locus of control.

Slide 17

System Failures Slowing Progress

The slide repeats the diagram from slide 5. Each hole the arrow flows through represents the following:

  • Insufficient capacity to train researchers
  • Insufficiently robust research
  • Failure to view the delivery of care as a science
  • Focus on differences rather than similarities with other types of research*
  • Patients continue to suffer preventable harm
  • Reason model
  • Note: *Highlighted text

Notes:

  • Data on quality improvement; what does that mean?
  • Context mechanism and outcomes; context becomes mechanism. Must unite.
  • Recent Wall Street Journal article on context of voting is now mechanism. James Fowler political scientist from University of California, San Diego, studies 1082 identical and fraternal twins. He found that whether you run for office, donated to a candidate, attend a rally, or join a political organization were heritable. Without genetics you missed half the story.
  • Recent advances in cancer biology. Found two types of cancer, pancreatic and glioblastoma ; 38,000 people will develop pancreatic cancer this year and fewer than 5% will be alive in 5 years. Another 20,000 will develop gliobloastoma with similar or worse prognosis. Researchers form 18 centers, discovered 83 genes in pancreatic cells and 42 in glioblastoma cells with mutations, making them likely candidates for turning on uncontrolled cell growth. What is novel is that they found that, rather than acting alone, these mutated genes acted in concert orchestrating pathways that allow unfettered growth to occur. This changes the games for researchers looking for therapies; rather than tarteting individual genes, they can tartet the pathways. It is likened to stopping traffic by raising a drawbridge than stopping individual cars. Even more interesting connect in glioblastoma: 3 pathways covered 75% of cancers. Defect in gene IDH1 only in 12% of cancers, but half of those younger than 35. These patients live longer.
  • Without the human genone sequencing, this would not be possible. We need the human geneome project linking, researchers doing safety work.

Slide 18

Context become Mechanism

The slide shows "Mechanism," "Context," and "Outcome" with arrows that point both ways in between each word.

Notes:

  • Need to advance the basic science of quality. Invest in tools and measures.

Slide 19

System Failures Slowing Progress in Patient Safety

The slide repeats the diagram from slide 5. Each hole the arrow flows through represents the following:

  • Insufficient capacity to train researchers*
  • Insufficiently robust research
  • Failure to view the delivery of care as a science
  • Focus on differences rather than similarities with other types of research*
  • Patients continue to suffer preventable harm
  • Reason model
  • Note:  *Highlighted text

Slide 20

Simple Rules for Producing Researchers

  • Obtain formal degree
  • Identify willing and capable mentor
  • Obtain protected time to participate in research project

Slide 21

Core Skills for Patient Safety Researchers

  • Epidemiology
  • Biostatistics
  • Health services
  • Economics
  • Sociology
  • Psychology
  • Informatics
  • Systems analysis
  • Qualitative
  • Leadership
  • Change management
  • Project management

Notes:

  • Clinicians in safety need to be excellent methodologist. Need to know fatal flaws, and balance scientifically sound and feasible. But cannot be expert so must be part of an interdisciplinary team.

Slide 22

Quality and Safety Research Group Mixing Bowl

The slide shows an empty table cross referencing "EPI/Stats," "Psych/Soc," "HSR," and "Econ" with "Critical Care," "Surgery," "Pediatrics," and "Medicine."

Slide 23

Improving Patient Safety in Michigan ICUs. Funded by AHRQ.

Slide 24

2 year results from 103 ICUs

The table presents the results for "Median CRBSI rate" and "Incidence rate ratio" for various "Time periods."

  • Baseline: 2.7; 1
  • Peri intervention: 1.6; 076
  • 0-3 months: 0; 0.62
  • 4-6 months: 0; 0.56
  • 7-9 months: 0; 0.47
  • 10-12 months: 0; 0.42
  • 13-15 months: 0; 0.37
  • 16-18 months: 0; 0.34
  • Note: Pronovost, NEJM 2006

Notes:

  • From over 103 ICUs, we reduced the bsi rate to 0 for nearly two years after the interventions.

Slide 25

"Needs Improvement" Statewide Michigan Comprehensive Unit-Based Safety Program (CUSP) ICU Results

The slide shows a bar graph presenting the percentages for "Safety Climate" and "Teamwork Climate" for 2004 and 2007. The results show:

  • Safety Climate:
    • 2004: 84%
    • 2007: 23%
  • Teamwork Climate:
    • 2004: 82%
    • 2007:; 22%
  • Less than 60% of respondents reporting good safety climate = "needs improvement."
    • Statewide in 2004 84% needed improvement, in 2006 41%
    • Non-teaching and Faith-based ICUs improved the most
    • Safety Climate item that drives improvement: "I am encouraged by my colleagues to report any patient safety concerns I may have"

Notes:

  • We do not know exactly what the goal for safety culture should be. We typically set a goal of 80% of staff reporting positive safety culture. We also recognize that if safety culture is below 60% it is associated with worse clinical and economic outcomes. We reduced the percent of teams scoring below 60% by nearly 50%.

Slide 26

Keystone ICU Safety Dashboard

  • How often did we harm (BSI)?
    • 2004: 2.8/1000
    • 2006: 0
  • How often do we do what we should?
    • 2004: 66%
    • 2006: 95%
  • How often did we learn from mistakes?
    • 2004: 30%
    • 2006: 100%
  • Percent needs improvement in safety climate?
    • 2004: 84%
    • 2006: 43%
  • Teamwork climate?
    • 2004: 82%
    • 2006: 42%

Slide 27

Focus and Execute

The slide shows a black and white photograph of a white porcelain wash basin.

Slide 28

The slide shows a black and white photograph of a toddler playing in the sand.

Current as of February 2009
Internet Citation: 10 years after "To Err is Human": An RCA of Patient Safety Research? (Text Version). February 2009. Agency for Healthcare Research and Quality, Rockville, MD. http://www.ahrq.gov/news/events/conference/2008/Pronovost.html