Six Sigma Approach to Reduction of Infections Lois Yingling, RNC, MSN, CPHQ, Black Belt Florida Hospital Orlando, Florida Lois.Yingling@flhosp.org
Objectives At the conclusion of the presentation participants will: List the 5 steps of Six Sigma Identify components of the IHI central line bundle Appreciate the value of a systematic approach to process improvement
Overview Who is Florida Hospital Bloodstream infections Five steps of Six Sigma Define Measure Analyze Improve Control Lessons learned with CDT
Who is Florida Hospital? Founded in 1908 by Adventist Church Oldest & largest healthcare system in Central Florida Seven campuses in 3 counties Licensed for over 1800 beds Third largest employer in Central Florida Largest Medicare population in the nation Recognized as one of America s Best Hospitals in U.S. News & World Report for the seventh year in a row HealthGrades 2005 Award for Excellence in Patient Safety
DMAICMAIC Define
Why Bloodstream Infection (BSI) Published mortality rates as high as 35% Baseline CVC related BSI: 13% Additional therapy costs $56,000 Baseline CVC related BSI: $16,699 variable cost Increased length of stay Baseline CVC related BSI: 20.6 additional days per case
National Interest Institute for Healthcare Quality (IHI) Central line bundle Hand hygiene Maximal barrier precautions Chlorhexadine skin antisepsis Appropriate care of site and line system No routine replacement Center for Disease Control (CDC) Guidelines
DMAIC Measure
In Scope: Scope Inpatients system-wide >17 y/o Positive blood culture within 48 hours of admission (2 weeks re- admission) Confirmed based on CDC definition CVC Out of Scope PICC lines Tunneled, port, dialysis, peripherals
Project Description/ Problem Statement Based on 2003 & annualized 2004 data: 43% of BSIs were secondary to CVCs LOS is increased by 20.6 days per case Variable treatment cost is increased by $16,699 per case Goal: Decrease the number of CVC related cases by 10%, a decrease of 16 cases per year
SIPOC High Level Process Map Supplier Input Process Output Customer Start = line Insertion Physician Referral Select device No BSI Patient Line Select site BSI Family Protective Garb Don full barrier garb Tray Prep site Insert line Care for line & dressing End = line removal
Baseline Process in control, no special cause variation
Gauge Repeatability Patient ICP Initial Surveillance Determination Surveillance ICP Second Surveillance Determination Surveillance result Agree Disagree result BSI Surveillance Hospital Acquired Hospital Acquired BSI Surveillance Hospital Acquired Hospital Acquired BSI Surveillance Hospital Acquired Hospital Acquired BSI Surveillance Hospital Acquired Hospital Acquired UTI Surveillance Hospital Acquired Hospital Acquired UTI Surveillance Hospital Acquired Hospital Acquired UTI Surveillance Hospital Acquired Hospital Acquired CDT Surveillance Hospital Acquired Hospital Acquired CDT Surveillance Hospital Acquired Hospital Acquired CDT Surveillance Hospital Acquired Hospital Acquired Total 10 0 100% One person repeatedly measures same unit
Gauge Reproducibility ICP #1 ICP#2 Agree Disagree Patient Surveillance result Surveillance result BSI Surveillance Hospital Acquired Hospital Acquired BSI Surveillance Hospital Acquired Hospital Acquired BSI Surveillance Hospital Acquired Hospital Acquired BSI Surveillance Hospital Acquired Hospital Acquired UTI Surveillance Hospital Acquired Hospital Acquired UTI Surveillance Hospital Acquired Hospital Acquired UTI Surveillance Hospital Acquired Not Hospital Acquired CDT Surveillance Hospital Acquired Hospital Acquired CDT Surveillance Hospital Acquired Hospital Acquired CDT Surveillance Hospital Acquired Hospital Acquired Total 9 1 90% Two or more persons measure the same unit
DMAIC Analyze
Process Capability Y1 All BSI Overall Z.USL -1.39 Sigma 0
Vital Xs CVC related blood stream infections cause & effect fishbone Patient/Visitor Equipment Technique Hand hygiene Handling catheter Handling drsg Visitors Patient Physician Staff Antimicrobial catheter Non-antimicrobial catheter Cost Stabilizer Contamination Hub care Dressing A septic techniue Line maintenance Hand hygiene Hand hygiene Sterile barrier Sk ill level Prep Contamination BSI PICC Candidate Femoral Subclavien Jugular Education Catheter Care Site
DMAIC Improve
Interventions & Results Nail P&P Chloraprep Staff BSI Education Began conversion to antimicrobial catheters in custom trays with sterile garb in all trays except Anesthesia Trays Hand Hygiene Campaign
Statistical Significance Two-Sample T-Test and CI: Historical VS New Mean Two-sample T for Rate C7 N Mean StDev SE Mean 1 11 0.658 0.154 0.047 (Jan 2003 - November 2003) 2 14 0.355 0.131 0.035 (Nov 2003 - January 2005) Difference = mu (1) - mu (2) Estimate for difference: 0.303182 95% CI for difference: (0.181309, 0.425054) T-Test of difference = 0 (vs not =): T-Value = 5.21 P-Value = 0.000 DF = 19 Difference between historical & new mean is statistically significant
Error Proofing Trays Custom Trays: Anesthesia Trays - no sterile garb ED & Unit Trays sterile garb Issue: Anesthesia trays without sterile garb distributed to units Error Proof: All custom trays include sterile garb and antimicrobial catheter
DMAIC Control
Reliable Measurements What to Measure Total Blood Stream Infections Clarify Data Collection Goals Type of Measure Y=BSI Rate Type of Data Continuous Data Operational Definition Procedure & What How X 1 X Data Form What Where When Positive Blood culture after 48 hours od admission or readmission within 7 days for S&S if BSI Query Medmined for positive blood cultures Develop Operational Definitions and Procedures Other Conditions to Record Line Type: CVC Swan Ganz, PICC, tunneled Collecting and Recording Infection Control Survelance Criteria for Center for disease Control Definition BSI Sampling Plan How Many Systemwide Monthly!00% X=number of infections secondary to CVC Discrete data Same as above Same as above Extarpolate CVC lines Confirmed BSI secondary to CVC line Review records for accuracy Systemwide Monthly!00% Gage R&R for all new emplyees after 90 days & for all staff annually. Gage R&R may be done more frequently if indicated. BSI Rate is based on CDC definition. CVC related BSI extrapolated from total BSI cases.
Current Status I Chart of CVC BSI Rate Historical VS New Mean Data Source: AICE 1.2 1 2 Per 1000 Patient Days 1.0 0.8 0.6 0.4 0.2 UCL=0.733 _ X=0.383 0.0 LCL=0.033 Marc h June Septe mbe r De cember Marc h Jun e Septe mber De cembe r Ma rc h Jun e January 2003 through July 2005 Process is in Control
Target: 5 or Less/Month
Process Capability Process Capability of Rate P rocess D ata LS L * Target * U S L 0.42000 S am ple M ean 0.38000 Sample N 18 S td ev (Within) 0.12359 S td ev (O v erall) 0.13445 USL Within Overall P ote ntial (W ithin) C apability Z.Bench 0.32 Z.LS L * Z.U S L 0.32 Cpk 0.11 CCpk 0.11 O v erall C apability Z.Bench 0.30 Z.LS L * Z.U S L 0.30 Ppk 0.10 Cpm * 0.1 0.2 0.3 0.4 0.5 0.6 0.7 O bserv ed P erform ance PPM < LSL * P P M > U S L 388888.89 P P M Total 388888.89 Exp. Within Performance PPM < LSL * P P M > U S L 373103.19 P P M Total 373103.19 E xp. O v erall P erform ance PPM < LSL * P P M > U S L 383038.47 P P M Total 383038.47 Y1 All BSI overall Z.USL 0.30 current Sigma 1.8
Owner Accountability What Who When Data collection Process Confirmed with IC Director & Manager Director Monthly beginning June 2005 Monthly report of CVC BSI Cases & LOS by Campus to Esmond Chan Variable cost/capacity adjustment Director Financial Analyst Monthly beginning June 2005 January 2005 & monthly
Results Capacity YTD April Actual 296 Days Target 110 Days Variance 186 Days Dollar Savings YTD April Actual $207,196 Target $77,233 Variance $129,963
CDT: Lessons Learned
Scope: Containment In Scope: Inpatients system-wide >17 y/o Diarrhea with confirmed assay diagnosis of CDT Out of Scope Outpatients Inpatients without diarrhea & confirmed assay diagnosis of CDT
Scope: Prevention In Scope: Inpatients system-wide, except Campus 3, >17 y/o with a history of a surgical procedure on the SIP list Diarrhea with confirmed assay diagnosis CDT Out of Scope All patients admitted to Campus 3 All medical patients and all surgical patients not on SIP list
CDT Baseline Out of Control I Chart of CDT Rate 5.5 1 Individual Value 5.0 4.5 4.0 3.5 3.0 1 1 UCL=4.530 _ X=3.564 2.5 1 1 LCL=2.597 2.0 March June 1 S eptem ber December March June September Decem ber Marc h Month Baseline 2003 through June 2004 June
CDT Rate I Chart of CDT Rate Pre-Assay VS Post Assay Data Source: AICE Per 1000 Patient Days 5.5 5.0 4.5 4.0 3.5 3.0 2.5 1 2 1 1 1 UCL=5.117 _ X=4.195 LCL=3.273 2.0 March June S eptem ber Dec ember March January 2003 through May 2004 June Septem ber Decem ber Marc h June June 2004 through June 2005 100% Assay Testing increased Case Finding
Containment Bleach April 2005 Terminal Cleans with bleach for rooms of CDT patients May 2005 Error Proofing Terminal Cleans for all rooms July Pilot Campus 6 New non-bleach product Kills spores No damage to furniture
CDT Rate
CDT Cases/Month Terminal bleach clean CDT rooms Terminal bleach clean all Rooms 2004 - June 2005: Target 152 or less/month
Prevention Right Antibiotic Right time Within one hour of incision Right duration Discontinue within 24 hours for prophylaxis Document if treating infection
Business Case Improved clinical quality (absence of infection) Capacity opportunity of 1639 days Financial opportunity of $1,298,484
Summary Six Sigma: Well defined methodology Systematic approach Robust Data driven Directional Statistical application for other initiatives
Alice came to a fork in the road. Which road do I take? she asked. Where do you want to go? responded the Cheshire cat? I don t know. Alice answered. Then said the cat, it doesn t matter. From Alice in Wonderland by Lewis Carroll