~90% Value = Benefit/Cost. Benefit = low as possible rate of the 8 major VLBW infant morbidities. Are Low Morbidity NICUs

Transcription

1 Are Low Morbidity NICUs (proficiency) also Low Cost NICUs (efficiency)? Who Provides Value? Joe Kaempf, MD Providence St. Vincent Medical Center Women and Children s Services Portland, OR June 24, 2018 I have no conflicts of interests nor commercial bias to report. The Vermont Oxford Network played no role in the design, conduct, analysis, interpretation, or reporting of this presentation. The views, conclusions, and opinions expressed are solely those of the presenter and do not represent the Vermont Oxford Network. Improve Individual Care Experience The Quadruple Aim Improve Population Health Improve Affordability Enhance Provider Fulfillment and Growth Is this lofty target realistic? Value = Benefit/Cost Value = Population Health + Care Experience + Provider Fulfillment Affordability VLBW infant discharged home without CLD, Grade 3 4 IVH, Stage 3 4 ROP, PVL, NEC, FIP, nosocomial infection, or a D/C weight <10 th percentile. Likelihood of good long term health and neurodevelopment? ~90% Benefit = low as possible rate of the 8 major VLBW infant morbidities CLD Grade 3 4 IVH PVL Stage 3 4 ROP Any Late Infection D/C Wt <10 th % NEC GI Perforation 1

2 Risk Adjusters Gestational Age Birth weight/sga Inborn v. Outborn Major Birth Defect Apgar 1 Minute Vaginal vs. C/S Gender Singleton vs. Multiple Mortality Deduction MEED Why are some NICUs more proficient than others? Arch Dis Child FNN Kaempf et al, 2017 Value = Benefit/Cost Benefit = Rate of the 8 major VLBW infant morbidities risk adjusted Cost = Mean Total Hospital Length of Stay in Survivors JAMA Pediatrics, 2015 Arch Dis Child FNN, Why am I speaking today? Total Length of Stay Surviving VLBWs Prov STV LOS VON LOS Inter Q d 66 d 58, 72 d d 71 d 59, 76 d Why is length of stay increasing? Democracy is the worst form of government, except for all the other forms that have been tried. Winston Churchill House of Commons, 1947 <32 weeks weeks weeks USA births 2% 9% 89% NICU Admits 15 20% 60% 25% Bed Days 35 40% 40 50% 20% PMA at D/C weeks* wks wks *The younger the GA, the greater and more variable the PMA is at discharge 2

3 Locale Position Length of Stay Discussions Yes No University NICUs Salary X Kaiser Permanente HMO Managed Care Salary X Private Practice Neonatologist owned Fee for Service X* Private Practice Employee National Firm Salary with Incentives X* Providence CQI Director for W&CS Health Plan X Getting older has some advantages,.i have worked 35 years in five distinct settings. Where is length of stay discussed? ** What is important to us? Census LOS. Benefit Metric Collaborative ,618 Inborn VLBW Survivors Morbidity Count LOS vs. Number of Morbidities Patient level data Younger GA means more morbidities and is associated with increased LOS. Calculating NICU expenditures is complex and imperfect. Major morbidities are difficult to compare. Resource utilization comparisons between NICUs are inadequate. Hospital length of stay might over simplify cost and value calculations. Paralysis by Analysis? Suggestion: Every premature infant should carry in their EHR a concise, accurate, risk adjusted morbidity count and a total hospital length of stay year 1 tally. Value Metric = Benefit Metric normalized to the total length of stay in survivors 2 arrows are Providence NICUs, is this astonishing variance acceptable? What distinguishes the 16 more proficient and efficient NICUs from the other 23 NICUs? Value Metric from 39 NICUs in North America NICU Length of Stay Predictors Demographic risk factors difficult to influence Modifiable risk factors we can influence Risk adjustment is essential for valid comparisons and meaningful CQI and research LOS literature is messy analyses blur demographic, socioeconomic, and modifiable risk factors,. is BPD or NEC a risk factor the way gestational age is,.maternal education level,.apgar scores,..? Risk adjustment Take Home Message Gestational Age Birth Weight accts for ~30 45% LOS variance Sex Apgar score 1 /5 ups it to ~45 50% LOS variance Severity of Illness score C/S vs Vaginal Inborn vs Outborn Maternal socio economic factors Geography Isaiah Berlin Oxford,

4 Which NICU LOS risk factors can we modify? CLD, NEC, nosocomial infection, and surgery reduction associated with shorter LOS With select morbidities factored in, ~75% of LOS variance can be explained,..what accounts for the other 25%? Quantum Mechanics Time is relative Matter is energy Multiverses Disentanglement Particles are waves All VLBW infants 2000 to NICUs Green high proficiency BM Yellow average Red lower proficiency BM Median Length of NICU Stay 2 high proficiency NICUs as ranked by the Benefit Metric 25 week survivors have a 32 day difference in median LOS,.why? Other factors besides morbidities affect LOS 40 NICUs Benefit Metric in 21,423 VLBW infants LOS decreases as the morbidity score improves LOS increases as mortality decreases LOS increase seen with decreasing mortality is blunted by proficiency, i.e., a better morbidity score LOS days Mortality (MEED) Risk adjusted Composite Morbidity Score A Culture of Excellence Classic Newtonian physics is 100% comprehensible, but only approximates truth. Quantum mechanics is 100% true, but is incomprehensible Quality outcome variance is true, but our explanatory models are Newtonian approximations. Quantum CQI,..we know some NICUs outperform others,..but we fundamentally don t know why! Morbidities are intertwined with decreasing GA and BWT,. horse or cart as we study value? Mortality rate may or may not affect LOS and total bed days,.ditto for transfer rates (CPQCC data). Decreasing morbidities in VLBW infants reduces mortality which may actually increase total bed days as more ELGANs survive. LOS is 4 dimensional, it s quantum,.non linear interactions in the complex adaptive system of NICUs and culture and incentives and payers. 4

5 We have a problem, a BIG PROBLEM. Neurodevelopmental outcomes in premature infants <28 weeks are NOT improving,...and might be declining. Twilhaar, JAMA Pediatrics, 2018 Spittle, Pediatrics, 2018 Cheong, Pediatrics, 2017 Adams Chapman, Pediatrics, 2018 The 10 POD NICUs 20 years of formal VON collaboration 23 to 27 weekers N = 4,709 Mean age 25.8+/ 1.4 wks POD has better composite morbidity scores than VON How? Why? Worst POD morbidity score Best POD morbidity score Morbidity reduction correlates with LOS reduction. Win Win! Benefit Metric is better in the green NICU. 30 day difference in LOS. ~$100,000 per baby Obstetric and NICU strategies that likely reduce NICU LOS safely and effectively Reduce prematurity and optimize prenatal care appropriate transfer, ACS, placental transfusion, HIE prevention, skilled delivery, reduce C/S rate Explicit Periviability Guidelines decision aids, informed consent (NICU vs Comfort Care) SAFETY Decrease errors and iatrogenic events Reduce Morbidities BPD, NEC, nosocomial infection, IVH, ROP, surgery Family Centered Care parent empowerment and active engagement in care and the transition to home 5

6 Obstetric and NICU strategies that likely reduce NICU LOS safely and effectively Explicit care guidelines, milestones to home expectations, use of discharge coordinators Feeding and nippling guidelines (infant driven feeding) Domiciliary care home health visits Home NG feeding and/or G tube, judicious home monitor use Reduce medication use to the essentials only Improved NAS guidelines (Yale), and reduce R/O Sepsis work ups (Kaiser calculator) Strategies that perhaps improve LOS Clear apnea/bradycardia/cyanosis definitions and monitor guidelines Caffeine stoppage at <35 weeks Developmental Care Kangaroo, NIDCAP, massage, reduce pain/light/noise/movement Osteopathic Manipulative Therapy Wean from incubator protocol Strategies that perhaps improve LOS Post discharge herniorrhaphy Early NICU vaccinations Liberal saturation guidelines for TTN, post C/S respiratory distress Flexible, sensible car seat guidelines Appropriate home monitor use Modify re imbursement strategies for physicians and hospitals Strategies that perhaps improve LOS Co bedding twins Physician and RN continuity staffing models Music therapy Incentivizing families with $$$ Myriad of surgical techniques for gastroschisis, bowel atresias, etc. Opinion Alert! 1) Modifying physician and hospital perverse $$$ incentives (fee forvalue rather than fee for service) would be LOS enlightening. 2) Sensory Processing Disorder Vinall et al U Brit Columbia Ask any colleague >55 y/o who regularly does long term VLBW infant follow up: Are nippling issues in premies the same, better, or worse in 2018 compared to the 1980 s and 90 s? Dysfunctional, aversive nippling and feeding in premature infants has grown to epidemic proportions. Prematurity is thousands of painful procedures, most of them unnecessary and wasteful. Sensory processing disorder is the premie brain trying to interpret, integrate, and respond to unnatural stimuli. Skillful nippling is a fantastic concert of 6 cranial nerves, the limbic system, the prefrontal cortex. If we figure out how to reduce the scourge of SPD, NICU LOS will improve. 6

7 CMGCHUKR Thank you to Lian Wang, MS and Gary Grunkemeier, PhD Providence St. Vincent Medical Center biostatisticians 7