A Comprehensive Approach to Improving the Quality of Neonatal Intensive Care Dan Ellsbury MD, Director of Continuous Quality Improvement The Center for Research, Education, and Quality Pediatrix Medical Group Disclosures Dr. Ellsbury is an employee of Pediatrix Medical Group. Off label use of medications will be discouraged in this presentation. Institute of Medicine Report: Crossing the Quality Chasm 2001 IOM Core Conclusions There are serious problems in quality.. Between the health care we have and the care we could have lies not just a gap but a chasm. Examples of the Chasm Overused, Underused, or Misused Examples of the Chasm Antenatal Steroids Breast Milk CABSI Reduction Bundles Delivery Room Thermal Management Postnatal Steroids H-2 Blockers, Metoclopramide Antibiotic Overuse, Cefotaxime Oxygen Many more. 1
Bridging the Quality Chasm IOM Core Conclusion: We have a problem:? The problems come from poor systems, not bad people! The NICU system is very complex Current s Potential s Probability of Performing Perfectly Steps (each with 95% success rate) Overall Success of the Process 1 95% 25 28% 40 12% 100 0.06% Overall success rate drops and errors increase as more steps are added to a process Quality is a system property!! Central Law of Improvement: Every system is perfectly designed to achieve exactly the results it achieves F Morriss, Neoreviews 2008, 9: e8-e23 2
Change the system Change the results Patient Neuro ID GI Pulmonary Renal Organ Systems We take care of babies each day. Each organ system and disease state gets attention Your NICU is your patient too Patient Organ Systems NICU IVH, BPD, NEC, PDA, etc The NICU is its own MICROSYSTEM Change the NICU system Change the NICU results Example of Large Scale System Change Comprehensive Oxygen Management for Prevention of Retinopathy of Prematurity (COMP-ROP) Started as small scale projects in multiple Pediatrix NICUs, focused on the implementation of practical oxygen management, based on the work of Chow (Pediatrics 2003:111(2):339-45 Structure and Redesign Model Clin Perinatol 37 (2010) 203 215 3
Problems 5/11/2011 Progression Improvement!!??? Lets Think Bigger But are we just playing a sad game of whack a mole Improve one problem, another pops up, go after that one and the original pops up How can we improve the multiple bad problems in the NICU??? How can we design our NICU system to improve multiple problems simultaneously???? Clin Perinatol 37 (2010) 203 215 Pareto Chart 80% of the problem is due to 20% of the causes 100 90 80 70 60 50 40 30 20 10 0 Vital Few Trivial Many A B C D E F Causes Pediatrix 100,000 Babies Campaign Re-engineering the NICU care of 100,000 Babies to Optimize Target Areas: The Vital Few 4
Increase breast milk use Early provision of IV protein in small infants Use of a customized feeding protocol Central Use a central line insertion bundle Use a central line maintenance bundle ilator Use a customized weaning and extubation protocol Optimize nasal CPAP technique Medications Increase/optimize use of surfactant, caffeine, antenatal steroids Decrease use of H-2 blockers, postnatal steroids, cefotaxime, heparin Optimize use of supplemental oxygen (COMP-ROP) Why these processes? They are the drivers of important outcomes, they are modifiable and generally under the control of the clinician Neurologic Infection NEC Length of stay Mortality BPD Blindness Why these processes? They are the drivers of important outcomes, they are modifiable and generally under the control of the clinician Neurologic Infection NEC Length of stay Mortality BPD Blindness Admission erature Use of plastic wraps, heated mattresses, radiant warmer, and/or heated room Monitoring for hypothermia and hyperthermia Growth IVH and PVL Growth IVH and PVL Why these processes? They are the drivers of important outcomes, they are modifiable and generally under the control of the clinician Neurologic Infection Growth NEC Length of stay Mortality BPD Blindness IVH and PVL System Evaluation Phase Assessment tool provided for system evaluation NICU Staff Survey to determine attitudes, beliefs, and knowledge gaps Review individualized NICU database review, to determine current general outcomes and features of clinical practice System Analysis analysis of each NICU s system design relevant to the target areas s discuss survey findings with their staff Prioritization Matrix - Develop consensus regarding initial problem areas to target for improvement System Evaluation Phase 5
Breast Milk CQI Council: permanent group that selects and oversees projects TPN Feeding Protocol Line Insertion Central Line Maint CQI Council & CPAP CPAP Days COMP ROP Other DR s: Many short-term groups that works on one specific project each Prioritization Matrix Complete this table, based on the current systems, practices, environment and outcomes of your NICU How Important? 1 to 10 scale (1 = very important, 10 = not important) In General In Your NCU How Difficult? 1 to 10 scale (1 = very easy, 10 = very difficult) In General Increasing breast milk use 1 3 Providing early IV protein 4 1 Standardized feeding protocol 2 5 Standardized central line insertion 2 3 Standardized central line maintenance 1 5 Decrease contaminated blood cultures 5 2 Minimize ventilator duration 2 5 Optimize nasal CPAP 2 6 Optimize use of antenatal steroids, caffeine, and surfactant 1 3 Optimize antibiotic choice and exposure 3 5 Decrease use of cefotaxime, metoclopramide, H-2 blockers, spironolactone, erythropoietin, postnatal 3 2 steroids Eliminate use of heparin flush in peripheral IVs 6 1 Standardize oxygen management 1 5 Optimize initial thermal management 1 2 In Your NICU Key Issues for Success Simplicity, simplicity, simplicity Automate measurement if possible, and minimize manual measurement Adapt general change packages to the specific, local NICU environment Use a system based approach, not an effort based approach Don t go after one problem and ignore other important areas/processes Status Over 100 NICUs participating thus far Preliminary results are encouraging, with network wide improvement in target process areas, and very encouraging results in morbidities and mortality 35 30 25 20 15 10 5 0 2003 2004 2005 2006 2007 2008 * Ellsbury PAS 2010 Percent of babies 501-1500 g receiving the medication Metoclopramide Cefotaxime H2-blockers Erythropoietin Spironolactone 2004: QI intervention started 2005-2007: Intervention and med reporting refined and expanded *p<0.01 for 2008 35 compared to 2003 Mortality in Extremely Preterm Infants Infants 24 through 27 6/7 weeks gestation, inborn, with no major anomalies (n=13,532) 20% 18% 16% 14% 12% 10% 8% 6% 4% 2% 0% 2005 2006 2007 2008 2009 Ellsbury PAS 2011 6