Ventilator-Associated Pneumonia Getting to the Bundle and Getting

Ventilator-Associated Pneumonia Getting to the Bundle (and Getting Beyond the Bundle? ) At an Academic Medical Center MICHAEL D. HOWELL, MD MPH DIRECTOR, CRITICAL CARE QUALITY ASSOCIATE DIRECTOR, MEDICAL CRITICAL CARE BETH ISRAEL DEACONESS MEDICAL CENTER HARVARD MEDICAL SCHOOL JEAN GILLIS, RN Clinical Nurse Specialist Patient Care Services Beth Israel Deaconess Medical Center BETH ISRAEL DEACONESS MEDICAL CENTER

These slides are only meant to Note illustrate the discussion; they aren’t a discussion of the topic in and of themselves.

Key Lessons Learned at BIDMC Sell the problem, not the solution. “Common, lethal, expensive, preventable. ” If you sell the problem, the clinicians will help you find the solutions. Definitions are inadequately explicit for real work Defining “VAP” is subjective, variable, and expensive. Vent bundle definitions are inadequately explicit, but can be really useful. The head-of-bed angle is like an onion. Having a kit matters for oral care.

Defining “VAP” PRE-CALL QUESTION : “HOW ARE PEOPLE DEFINING/DIAGNOSING A VAP? EVEN USING THE CDC DEFINITION, WE FIND THERE IS ROOM FOR INTERPRETATION. IN THE END, WE CAN DEFINE IT FOR OUR ORGANIZATION AND JUST FOCUS ON IMPROVING OUR RATE. HOWEVER, MANY FOLKS ARE REPORTING RATES OF ZERO AND PAYERS MAY REFUSE TO PAY FOR THE CARE ASSOCIATED W/ VAP…. SO THE DEFINITION BECOMES MORE IMPORTANT. “ 1

The CDC definitions are complicated and subjective. (This is just part of PNU-1. ) “What do you mean by ‘VAP symptoms’? ” 2

CDC PNU 1 VAP “Symptoms, ” as best we can figure out… 3

Which definition you use affects the ‘answer’ you get. “Depending on the definition evaluated, criteria were met for a diagnosis of VAP from as low as 4% of patients by the Johanson definition to as high as 48% of patients by the CDC definition. ” 4

Our conclusions about defining “VAP” It’s hard, time-consuming, subjective, and expensive We use somewhere between 0. 25 – 0. 5 FTE of experienced critical care nurse time to screen four of our nine ICUs. We review all CXRs with three critical care MDs to arbitrate the final “rate. ” This is about 15 -20 hours of physician time per month screened. We therefore only do it for 3 – 5 months per year. The rest of the time, we work on process. It makes sense to follow your own, internal “VAP rate” – if you do it the same way, time after time. Be very wary about comparing “rates” among hospitals. 5

Implementing Change THE BUNDLE Pre-call Question : “To improve, we find measuring the process more helpful than following the rate (though the outcome is certainly important!). We have questions about what and how people are measuring the process. Are folks just reviewing compliance w/ the bundle? once a day? Once a shift? Etc. ? “ 6

Surrogate Team Function Metric? Measure Head of bed Stress ulcer proph DVT proph Daily wake-up RSBI / SBT Disciplines Required RN / RT MD / RN / Pharmacy RN / RT / (MD) RT / RN / MD 7

Approach: Unit Champions Selection of local nurse and respiratory therapy leaders Incorporation of them into data gathering phase Three snapshots per week Distributed across shifts around the clock Metric is therefore approximately proportional to ventilator days Two-week feedback cycles, using unit champions to disseminate change Supported by Clinical Nurse Specialists (0. 5 - 1. 0 FTE) 8

Four Useful Lessons (among many!) Documentation ≠ Reality! Corollary: The head of the bed is like an onion. The Bundle definitions are not explicit enough for our needs. Oral care kits make implementation easier. Corollary: People like gizmos. Issue: Our experience was that not all oral care kits are the same. We provided three things that (we think) helped accelerate change Data that is trusted by providers Very frequent data feedback to each ICU (q 2 weeks while improving) 18 times a month! Actionable analysis of the q 2 week data, when needed 9

The head of the bed is like an onion. “We hypothesized that head of bed angles would be at or above 30 degrees among mechanically ventilated patients throughout the day due to a hospital-wide initiative on ventilator-associated pneumonia prevention and standardized electronic order entry system to keep head of bed at an angle of 30 degrees or greater. . Contrary to our hypothesis, all patients had head of bed angles less than 30 degrees. ” 10

The head of the bed is like an onion Our beds’ electronic Fowler’s angle: wrong (randomly, by random magnitudes) 10

We needed more explicit bundle definitions. 11

Oral Care: Gizmos make implementation easier, but not all gizmos are the same. 12

Results

Process Measure

13

14 Note: We use a “temporal bundle” for oral care. If you miss one oral care chance during a 24 -hour period, you get a “ 0” for the day. (We just made this up, though).

Outcome Measure

Outcomes No VAP rate data) 15

16

17

Resource Utilization Measure

No vent day data 18

Throughput Impacts CAVEAT: LOTS OF OTHER THINGS WERE GOING ON!

19

20

Summary

No data (ICU Throughput) Better (Three-ICU Sample) Better Ventilator Days Better VAP Cases (Three-ICU Sample) Better Ventilator Bundle Compliance Long-Stay ICU Patients Better Number of ICU Patients (% of patients with ICU LOS > 10 days) 21

Conclusions

Conclusions Defining “VAP” is complicated and challenging. but may be useful to follow internal. The Ventilator Bundle is really useful. but documentation may not equate to reality! Our providers responded to Selling that a real problem exists Data they trust (and help collect) Frequent data feedback (q 2 weeks), with actionable analysis Delivered by a respected Clinical Nurse Specialist When Ventilator Bundle (and oral care) compliance improve, VAP goes down.

Thank you. QUESTIONS? PLEASE EMAIL MHOWELL@BIDMC. HARVARD. EDU