Michigan’s Keystone ICU Project:

1. Michigan’s Keystone ICU Project: An exemplar Chris Goeschel RN MPS MPS Johns Hopkins Quality and Safety Research Group

2. How will we know ? Consider in the end at the beginning:Are the citizens of Michigan less likely to harmed?

3. State wide effort to improve ICU care in Michigan Funded by AHRQ

4. Michigan: Facts • Total Population: 10,120,860 (8). • 2000 percent population 18 and over: 73.9; • 65 and over: 12.3; median age: 35.5. • Major Industries - car manufacturing, farming (corn, soybeans, wheat), timber, fishing • 10,083 inland lakes and 3,288 mi of Great Lakes shoreline (most registered boaters in the US) • 138 acute care hospitals (not all with ICU’s) • 3 beds to 1500 beds

5. Keystone ICU The aim was to use evidence-based tools to improve quality and patient safety in Michigan intensive care units.

6. Goals of Keystone ICU • Reduce harm: BSI and VAP • Ensure 90% of patients receive EB interventions for preventing VAP, • Learn from one defect per month • Improve culture of safety 20% (SAQ) • Improve quality improvement

7. Collaborative Process • Written Commitment to Participate & Provide Resources to do the work • Senior Leader as part of ICU Team • Bi-weekly or Monthly Calls: Collaborative Leaders, Teams, Hopkins • Content, Coaching and Team Sharing • Monthly Standardized Web based Data Collection • Transparency at local level • “Harm is Untenable”

8. Comprehensive Unit-based Safety Program (CUSP) • Evaluate culture of safety • Educate staff on science of safety http://www.jhsph.edu/ctlt/training/patient_safety.html • Identify defects • Assign executive to partner with the unit • Learn from one defect per month and implement teamwork tools; daily goals, a.m. briefing, culture checkup • Evaluate culture www.safetyresearch.jhu.edu Pronovost J, Patient Safety, 2005

9. Interventions to prevent Central Line Blood Stream Infections: 5 Key Behaviors • Remove Unnecessary Lines • Wash Hands Prior to Procedure • Use Maximal Barrier Precautions • Clean Skin with Chlorhexidine • Avoid Femoral Lines MMWR. 2002;51:RR-10

10. Safety Tips: Label devices that work together to complete a procedureRule: stock together devices need to complete a task CASE IN POINT: An African American male ≥ 65 years of age was admitted to a cardiac surgical ICU in the early morning hours. The patient was status-post cardiac surgery and on dialysis at the time of the incident. Within 2 hours of admission to the ICU it was clear that the patient needed a transvenous pacing wire. The wire was Threaded using an IJ Cordis sheath, which is a stocked item in the ICU and standard for PA caths, but not the right size for a transvenous pacing wire. The sheath that Matched the pacing wire was not stocked in this ICU since transvenous pacing wires are used infrequently. The wire was threaded and placed in the ventricle and staff Soon realized that the sheath did not properly seal over the wire, thus introducing risk of an air embolus. Since the wire was pacing the patient at 100%, there was no Possibility for removal at that time. To reduce the patient’s risk of embolus, the bedside nurse and resident sealed the sheath using gauze and tape. SYSTEM FAILURES: OPPORTUNITIES for IMPROVEMENT: Knowledge, skills & competence.Care providers lacked the knowledge needed to match a transvenous pacing wire with appropriate sized sheath. Regular training and education, even if infrequently used, of all devices and equipment. Unit Environment: availability of device. The appropriate size sheath for a transvenous pacing wire was not a stocked device. Pacing wires and matching sheathes packages separately… increases complexity. Infrequently used equipment/devices should still be stocked in the ICU. Devices that must work together to complete a procedure should be packaged together. Medical Equipment/Device. There was apparently no label or mechanism for warning the staff that the IJ Cordis sheath was too big for the transvenous pacing wire. Label wires and sheaths noting the appropriate partner for this device. ACTIONS TAKEN TO PREVENT HARM IN THIS CASE The bedside nurse taped together the correct size catheter and wire that were stored in the supply cabinet. In addition, she contacted central supply and requested that pacing wires and matching sheaths be packaged together.

11. Slowing our progress

12. Insufficient Leadership Support

13. Leading Change • One of most common leadership mistakes is expecting technical solutions to solve adaptive problems…. • Ron Heifetz “Leadership without Easy Answers

14. Creating Reliable Health care

15. JHU Toolkits to Assist Teams • Engage (local work) • Opportunity calculator, stories of harm • Educate (central work) • Original papers, fact sheet, slides • Execute (local work) • Standardize, create independent checks, learn • Evaluate (central work) • Web based data reports

16. Safety Scorecard

17. 80% Reduction in BSI in One Year from 103 ICU Data from 100 ICUs Analysis: multilevel GLLAMM

18. Safety Climate Across Michigan ICUs % of respondents within an ICU reporting good safety climate

19. No BSI = 6 months or more w/ zero No BSI 21% No BSI 44% No BSI 31% Teamwork Climate Across Michigan ICUs The strongest predictor of clinical excellence: caregivers feel comfortable speaking up if they perceive a problem with patient care % of respondents within an ICU reporting good teamwork climate

20. Develop an Eye for the Unexpected

21. # RNs who left the ICU 1 RN Turnover and Teamwork Climate: 26 Keystone ICUs reporting r=-.650, <.001

22. Impression: In Hindsight, the Successful KICU Project Looks Easy

23. FACT: Participants Say These Results Never Would Have Been Achieved Without the Johns Hopkins Keystone ICU Collaborative Why is That??

24. CLIP

25. Our Experience: Factors for Success • Use evidence-based tools • Pilot – Input from frontline staff is key • Make sure tools are practical • Treat the project like a clinical trial • Involve frontline staff in the initiative– ownership AND provide feedback

26. Our Experience: Factors for Success • Project goals must drive measurement • Care most about patient level goals; others are predictor variables • Design data collection and management plan at outset • Reduce bias in data collection • Give up on quantity not quality of data • Central Development/ local implementation • Strive for scientifically sound, feasible, useable

27. Our Experience: Factors for Success • Adaptive lessons • Commit that harm is untenable; make harm visible • What are CLABSI rates? Do all clinical caregivers know them? • Ohana • How have you shared what you are learning with others? Administrators, clinicians, teams, facilities? • Local modification of execution • Have you adapted the implementation in light of your organizational culture?

28. Our Experience: Factors for Success • Leadership Engagement • Regional Collaborative Leaders • Hospital Executive/Administration • Clinicians • Ownership • The teams and staff must own the project • Collaborative “Virtual Learning Community” • OHANA