emergency critical care center: a lean journey - .global · 2001 2009 ed critical care visits ed...

DepartmentofEmergencyMedicine

EmergencyCriticalCareCenter:

ALeanJourney

BenBassin,MD,FACEP,EDAC

CemalSozener,MD,M.Eng,FACEP,EDAC


• Customer1st

• Peoplearethemostimportant

resource

• Shopfloorfocus(GoandSee)• Kaizenisawayoflife

LeanThinking- ThingsthatMatter


• A3Thinking

• GoandSee

• RapidImprovementEvents(Kaizen)

KeyLEANTools


A3


BACKGROUND/PROBLEM


• Whyisthisimportant?

• Whynow?

DefiningtheproblemisCRITICAL


CreateaClear,ConciseProblemStatement

“Improveaccesstotimelycriticalcareby

enhancingthecapacityandcapabilitytodeliver

highqualitycriticalcareintheAdultEmergency

DepartmentattheUniversityofMichigan.”


Background


MichiganMedicine


UMHSAdmissionStreams:ED=FrontDoor

16% 30%

54%

FY16UMHSinpt/obs admissions=66,620


AnAlarmingNationalTrend

ICUadmissionsfromtheEDhaveDOUBLEDoverthelastdecade- AdmissionsfromtheED:1.2à 2.2millionfrom2001-2009

- AdmissionratefromtheED:0.9%à 1.6%

- 1/3ofallICUadmissionsspend>6hrsintheED

0

2,000,000

4,000,000

6,000,000

8,000,000

10,000,000

12,000,000

2001 2009

EDCriticalCareVisits

EDCriticalCareHours

Herringetal.Crit CareMed2013

Peryear

10%/yr

24%/yr

Growth


Theeffectivenessofcriticalcare

foracuteillnessandinjuryis

time-sensitivewiththerapeutic

windowsrangingfromminutes

tohours.

AcuteCriticalCareistime-sensitive

1358 CRITICAL CARE Nguyen et al. • CRITICAL CARE IN ED

Figure 2. Predicted mortality presented as mean (%) !standard deviation computed from APACHE II andSAPS II scores in the emergency department (ED) andat intervals in the intensive care unit (ICU). Asterisk (*)indicates significant difference (p " 0.02) in predictedmortality between the corresponding time points.

TABLE 5. Predicted Mortality and Changes in Predicted Mortality (! Pred Mort) of All Patients*

APACHE II

PredictedMortality (%)

! Pred Mort (%),p-value†

SAPS II

PredictedMortality (%)

! Pred Mort (%),p-value†

0 (ED admission) 42.0 ! 20.0 — 44.0 ! 22.0 —6 (ED discharge) 34.0 ! 18.0 #8.0 ! 14.0, <0.001 38.0 ! 23.0 #6.0 ! 14.0, <0.00124 (ICU) 26.0 ! 18.0 #7.0 ! 13.0, <0.001 33.0 ! 24.0 #4.0 ! 16.0, 0.0248 21.0 ! 14.0 #1.0 ! 9.0, 0.47 27.0 ! 19.0 #2.0 ! 11.0, 0.2472 20.0 ! 12.0 #1.0 ! 6.0, 0.18 26.0 ! 18.0 #1.0 ! 8.0, 0.22

*Data are presented as mean (%) ! standard deviation computed from APACHE II and SAPS II scores in the emergencydepartment (ED) and at intervals in the intensive care unit (ICU).†P-value reflects comparison of predicted mortalities using Student’s t-test between the successive time points.

ward were excluded; therefore, all patient datawere analyzed. At ED admission, there was ahigher APACHE II score in nonsurvivors vs sur-vivors (p = 0.04), while there was no significantdifference in SAPS II or MODS scores. TheAPACHE II, SAPS II, and MODS scores were sig-nificantly lower in survivors than nonsurvivors atED discharge and throughout the study period (p" 0.001) (Table 4). General linear mixed modelingwas used to examine hourly rate of change in phys-iologic scores (Fig. 1). The hourly rates of change(decreases) in APACHE II, SAPS II, and MODSscores (#0.55 ! 0.64, #1.02 ! 1.10, and #0.16 !0.43, respectively) were significantly greater dur-ing the ED stay than subsequent periods of hos-pitalization in the survivors (p < 0.05). In the non-survivors, there was a significantly greater hourlyrate of change (decrease) in APACHE II score(#0.29 ! 1.05) compared with subsequent inter-vals during hospitalization (p = 0.02). However,there was no significant difference in the hourlyrate of change (actual increase) in SAPS II and

MODS scores between ED therapy and ICU carein the nonsurvivors. The hourly rate of change (ordecrease) in APACHE II, SAPS II, and MODSscores was uniformly greater at each time intervalin survivors compared with nonsurvivors.

There was a significant decrease in APACHE IIand SAPS II predicted mortality during the EDstay and at 24 hours in the ICU (p " 0.02) (Table5). After 24 hours there was no significant changein predicted mortality with both scoring systems.The APACHE II and SAPS II predicted mortalityapproaches actual in-hospital mortality at approx-imately 12 hours and 36 hours after ED admission(in the ICU), respectively (Fig. 2).

Mortality predictions at each time point werecompared using ROC curve analysis (Fig. 3). Therewas a stepwise increase in mortality prediction(sensitivity and specificity) as represented by theAUC from ED admission to ED discharge, and at48, 72, and 24 hours in the ICU using APACHE II.The AUC for mortality prediction using SAPS IIwas progressively increased from ED admission toED discharge, and at 24, 48, and 72 hours.

DISCUSSION

The increased provision of critical care in theED1–4 is a result of increasing patient volumes,limited ICU bed availability, increasing levels ofpatient acuity, and the use of the ED as the firstcontact for primary care. In addition, the applica-tion of therapeutic1 and monitoring23,24 technolo-gies once considered under the domain of the ICUhas made the delivery of critical care more prac-tical in the ED. Studies have quantitated that from1541 to 1864 critical-patient-days per year are pro-vided in the urban ED setting. Our data indicated464.4 critical-patient-days per year (38.7 critical-patient-days per month or 30.5 critical-patient-hours per day) were delivered at this institution.

If the findings of this study are seen in similarEDs and generalized nationally, a substantialamount of critical care is delivered in this setting.Professional organizations have recognized andare currently addressing these issues.25 The Soci-ety of Critical Care Medicine currently provides a

TheFIRST6HOURS

Mostrapidchangeinphysiology

Nguyen2000


ED-ICUInterface

EDLocation ICULocation

Time0to24hr

CriticalCareNeedsBeingAddressed?

EDTeam ICUTeamNoMan’s

Land?


CurrentState/Analysis


CurrentSituation– Wherearewenow?


Currentstate2012– ICUAdmissions

• FY12 – 1,886 ICU Admits

• 5.2 per day

• ED LOS 6 hr

• Projected 10% growth


Patient(Home/EMS/OSH)

ED

Admission

ICU

Floor

Discharge

DeclinedED-EDTransferRequests(750/yr)– 25%ICU

ShuntingofResourcesfrom

Non-CriticalPatients

Short-StayICUAdmissions<24hr

(440/yr)

EmergentTransfertoICUin<24hrs

(200/yr)

EC3CC

Boarding(6hrs)

CCD/CfromED?

CCinnon-CCarea

(450/1286ICUadmits–

35%)


• IncreasingEDvisits

• IncreasingICUdemand(EDandtransfers)

• IncreasingEDLOSforourICUpatients

• Outcomes?

Summarytodate

(wherewewerein2012)


FutureState/Goals


A3


FutureState


ParadigmShift


ApplyLEANprinciplestoproblem

identificationandsolution


• Engagefrontlineworkerstocreateoptimal

workflowsandeliminatewaste

• Architectsanddesignerstransformprocess

mapsintoadesign

• Focusonremovingphysicalbarriersfrom

theworkflows

• Paradigmshiftofarchitectsfromproject

leadstoteammemberandfacilitators

LeanFacilityDesign


• Purpose– fixpatients’problemsdefinitively

• Process– fundamentallyredesignthe

processes(notjusttighteninguptheold

way)

• People– atruemultidisciplinaryteam

LeanPatientFlowTransformation


• Assumptions:

– Currentvolume

– EC3LOS– 12hrs

– 50%ofpatientstreatedinresuscitationbayswouldpassthroughEC3

– 10-20%ofpatientswouldnolongerneedICU

• ModelOutput:– Require5resuscitationbaysand7EC3rooms

QueueingModel


GrowthProjection


ProposedFutureState

5


CreatingaMultidisciplinaryTeam


Manyteammeetings2yearsprior


• A3Thinking

• GoandSee

• RapidImprovementEvents(Kaizen)

KeyLEANTools


Roadtrips


SCHEMATICDESIGN


EC3v1.0


FullScaleMock-up


UtilizeActualEquipment


MultidisciplinaryTeams


Egress


ImmovableBarriers


IdeaBoards


• Thisisareallygoodidea

• NoprecedentatMM

• Empowering/Inclusive/Unifying

• SignificantdesignchangesBEFORE

construction• Nosignificantchangeorders

• Underbudget

• Aheadofschedule

MockupLessonsLearned


EmergencyCriticalCareCenter

99

98 97 96 95

92

9394

EDCBA

91

EmergencyCriticalCareCenter(EC3)(91-99)

EDResuscitationBays(A-E) AmbulanceEntrance


BringintheMuscle…

September2014


ImplementationPlan


JoyceandDonMasseyFamilyFoundation

EmergencyCriticalCareCenter(EC3)

GrandopeningFebruary16,2015


LEANDESIGNFEATURES


• Cold,CATscanin

zone

• Nocrowdcontrol

• Patientsand

familiesnotin

ideal

environmentat

theirgreatest

timeofneed.

Before– 3bedResuscitationBay


OldResus Bay

Cluttered,old,

poorlydesigned

boomsystems.

ERfleamarketwith

tonsofequipment

pushedtotheback.

Can’tfindanything

quickly


DirectAccess

toEC3

Multiple

Video

Displays

Camerasfor

QAand

Teaching


Dedicated

Nurse

Charting

Station

Dedicated

Physician

OrderEntry


Abilitytocarefor

multiplepatients

Power,Data

andGas-

360˚


In-situ

teaching

technology


CustomCarts

toMaximize

Spaceand

Efficiency


Transaction

Height

Counters

KeepCarts

outof

Corridor


LineofSight

toResus Bays


OpenLineof

SightAcross

Unit


Ambient

LightHOB

ClearBariatric

Lifts

Dialysis

Boxes

Negative

Pressure


PatientCareSpaces

FamilyZone PatientZone CaregiverZone


EARLYRESULTSANDTARGETMETRICS


Patient(Home/EMS/OSH)

ED

Admission

ICU

Floor

Discharge

DeclinedED-EDTransferRequests(750/yr – 25%ICU)

ShuntingofResourcesfrom

Non-CriticalPatients

Short-StayICUAdmissions<24hr

(440/yr)

EmergentTransfertoICUin<24hrs

(200/yr)

EC3CCinnon-CC

area(450/1886ICUadmits–

24%)

CCBoarding(6hrs)

CCD/CfromED?

DeclinedED-EDTransferRequests(597/yr – 9%ICU)

14dedicatedbedsforCC

EC3– 0hrsIfnoEC3–2.5hrs

Decreased43%

IncreasedICU-ICUAcceptedTransfer32%incr FY16

0%change=safestrategy

9%(460pts)

CCinnon-CCarea

(229/1816ICUadmits–

12%)


• Improveaccesstotimelycriticalcareby

enhancingthecapacityandcapabilityto

deliverhighqualitycriticalcareintheAdult

EmergencyDepartmentattheUniversityof

Michigan.

• Inreality,themodelhascompletely

changedtheED&ICUhealthcaredelivery

paradigm

Summary/Conclusion


Thankyou


Questions?


RAPIDIMPROVEMENTEVENTS(KAIZEN)


ExcessEquipment


ProceduralLineCarts


NewLineCart


Airwaycartredesign

17 1


Airwaycartredesign


AppliedPrinciples


Content Title

• Add content here• Add content here• Add content here




99

98 97 96 95

92

9394

EDCBA

91



ED:EC3:Floorv.Pre-EC3Baseline

1 2 3 4 5 6 7 8-14 15+ ED:EC3:Floorv.Pre-EC3Baseline -350.42 -2,276.61 -2,498.15 -2,695.02 -2,882.40 -2,697.81 -2,886.92 -2,736.51 -2,230.10

-3,500.00

-3,000.00

-2,500.00

-2,000.00

-1,500.00

-1,000.00

-500.00

0.00

Cost

LOS(days)


• Patientmix Tertiary/quaternarycare

• Currentvolume 67,014/year

• Admitrate 35.4%

• ICUadmitrate 10%ofadmissions

• TransfersfromoutsideED 3440/year~300/month

• Transfersdeclined 750/year(25%ICUlevel)

• LWBSRate 3%

• Volumeprojections 3-4%overall,10%criticalcare

• ICUadmitsw/LOS< 24hrs ~440/yr

• FloortoICUTxfr < 24hrs ~200/yr

CurrentState- AdultEmergencyServicesFY12


TimeinEDwaitingforICUbed…

0

1

2

3

4

5

Reference Chalfin2007 Rincon2010 Singer2011 Hung2014 Cha2015

OddsRatiofordeath

>6hr

>5hr

>12hr

>4hr >6hr


• Average7patientsperday• MedianEC3LOSICUadmit=7.2hr• MedianEC3LOSNon-ICUadmit=12hr

Results:EC3OperationalCharacteristics

BiPap/Intubation/Vent 595 11.5%DKA 276 5.3%EndofLife 61 1.2%GIBleed 421 8.1%PostCardiacArrest 84 1.6%Sepsis 794 15.4%ShortnessofBreath 514 9.9%StatusEpilepticus 57 1.1%SubarachnoidHemorrhage 133 2.6%Undifferentiated 2,546 49.2%Unknown 391 7.6%

EC3PathwayMultipleperpatientpossible Count %


Pre-EC3(744d)

Post-EC3(744d)

RelativeChange

OverallED Visits 147,030 157,190 6.9%

HospitalAdmissions 51,451 55,912 8.7%

ICUAdmissions 3,742 3,279 -12.4%

Results:EDVisitsandAdmissions


Total EDVisits

ICUAdmission

ICUAdmission

Rate

95%CI

Pre-EC3 147,030 3,742 2.54% 2.4-2.6%

Post-EC3 157,190 3,279 2.08% 2.0-2.2%

EC3AssociatedwithDecreased

ICUAdmissionRatefromED

Relative Risk Reduction = 18% [95%CI:11-23%]Number Needed To Treat = 218 [95%CI:174-361]


TotalAdmissions

TransfertoICU

≤24hrsafteradmittoward

Rate

Pre-EC3 51,451 377 0.8%

Post-EC3 55,912 400 0.8%

EC3NotAssociatedwithIncreased

TransferstoICUwithin24hoursof

GeneralWardAdmission


TotalEDAdmissions

Death≤48hrsafter

admission

MortalityRate

Pre-EC3 51,451 280 0.54%

Post-EC3 55,912 281 0.50%

EC3NotAssociatedwithIncreased

ShortTerm(≤48hrs)Mortality


ICUBedDaysSaved

MinimumICUBedDaysSaved1 1,188AverageperMonth2 51.4AverageperDay2 1.7MedianICUBedDaysSaved 3,326AverageperMonth2 143.3AverageperDay2 4.7


Reductionin“ShortStay”ICU

Admissions


ReducingMedicalICUAdmissions


DiseaseSpecificOrderSets

Undifferen

tiatedSepsis

BiPap/Intu

bation/

Vent

Shortness

ofBreath

Feb-16*

GIBleed UnknownDKAAug-

15*

Subarachn

oid

Hemorrha

ge

Post

Cardiac

Arrest

EndofLife

Dec-15*

Status

Epilepticus

EC3Pathway 49.2% 15.4% 11.5% 9.9% 8.1% 7.6% 5.3% 2.6% 1.6% 1.2% 1.1%

2,546

794

595514

421 391276

133 84 61 57

-5.0%

5.0%

15.0%

25.0%

35.0%

45.0%

55.0% EC3PathwaysUtilized

Multipleperpatientpossible2/16/2015- 2/28/2017

n=5,872

%ofT

otalEC3

StatusP

atients

*Dateorderset started


DiseaseSpecificOrderSets


LengthofStay&CostAnalysis


LengthofStay

MedianHospitalLOS MedianICULOSPreEC3 PostEC3

EDtoMICU 6.00

ED toEC3 toMICU 7.00

ED toEC3 toFloor 4.00

PreEC3 PostEC3

ED toMICU 2.00

ED toEC3 toMICU 3.00

0.00

0.50

1.00

1.50

2.00

2.50

3.00

3.50

ED:MICU ED:EC3:MICU

Med

ianICULO

S

PatientPathway

PreEC3

PostEC3

0.00

1.00

2.00

3.00

4.00

5.00

6.00

7.00

8.00

ED:MICU ED:EC3:MICU ED:EC3:Floor

Med

ianHo

spita

lLOS

PatientPathway

PreEC3

PostEC3


CostComponents

Cost

Radiology

X-ray

CT

MRI

Ultrasound

Pharmacy Nursing LabTestsAncillarySupport

RT

PT

OT

SpeechLanguagePathology

OtherTests

EKG

EEG

Echo


EC3CostComparison

$0.00

$1,000.00

$2,000.00

$3,000.00

$4,000.00

$5,000.00

$6,000.00

$7,000.00

$8,000.00

1 2 3 4 5 6 7 8 9

Cost

LOS(days)

EC3CostComparison

ED:EC3:MICU

ED:EC3:Floor

Pre-EC3:Baseline


ED

ICU WARDWithin24hours

- 4xlongertogettoICU

- O.R.Death3.07

Molina2014

WARDàICUtransferisalsoassociatedwithincreasedmortality

AnOpportunity


• 7.5AttendingFTEs

• 7PAFTEs

• 40NursingFTEs

• DedicatedRT

IncrementalFTEs


• Transitionzonebetweenacuteresuscitation

andongoingcriticalcaredelivery

• Needs-basedtreatmentvsgeography

–RightCareRightNow

• Timesensitive,nottimelimited

ED-ICUinterface


Scope


• Internalaudit– Needsbasedassessmentofcurrentstate

• CriticalCareAssessment

– Siloed

– Specificspaceforspecializedcare

• EmergencyMedicineAssessment

– EmergencyCriticalCareemergingspecialty(care/education/research)

– Currentstaffingmodel

– Currentphysicalplant

Roadmap


StaffingModel

Attending1 Attending2

PA/Fellow/ResidentA PA/Fellow/ResidentB

PA/Fellow/ResidentC PA/Fellow/ResidentD

EC3BedsideRNx4(2:1)

EC3TeamLeadRNx1(1:1)

RespiratoryTherapy

Pharmacy

MedicalProvider

Nursing

Ancillary

24Hours

Laboratory


ImplementationPlan


EC3:LogisticsandPatientFlow

EMS/Triage

ResusBay

EC3

DefinitiveCare ICU Ward Discharge

PalliativeCare

MainEDPatientSelection

Pushvs.Pull

90-120min

LOS11hrs (2-50hrs)

CollaborativeDisposition


RibbonCutting– Feb2015


• Defects

• Overproduction

• Waiting

• ExcessProcessing

• Transporting

• Inventory

• Motion

• NotUsingTalent

EightWastesofHealthcare

Grunden andHagood.Lean-LedHospitalDesign.CRCPress;2012:13-14


CostImpactofLEAN

SavingsOverTime

HigherUpFront

Investin

Personnel/LEAN

Leadership

Process

Improvement

Reduces

Cost/Increases

ValueOver

Time

Returns


IntensityofEDServicesIncreasing


• CCadvisorygroupformedin2012–– MedicaldirectorsandunitrepresentativesfromallinpatientadultICUs

• Agreeduponbasictreatmentprotocol/strategyformostcommonadmissions– Sepsis

– Cardiacarrest

– RespiratoryFailure

– GIBleed

– Anticoagulationreversal

– SAH

CollaborativeModel


• IntroductiontoMichiganMedicine

• DefineLEAN

• LEANTools

• PressuresAffectingEmergencyCare

• DemonstrativeCasePresentation

• Summary/Wrap-Up

Outline


• Engageallworkers

– Frontline,MiddleManagement,AdminLeadership

– AllJobFamiliesAffectingPatientExperience

• Empowers

• FostersTeamworkAcrossJobFamilies

• FrequentlyUncoverLargerSystemsIssues

• Leaders’primaryjobistogrowmoreleaders

People


Quality• ValuenotVolume

• HospitalAcquiredInfections

• MedicalErrors

Flow• IncreasingVolumes

• SickerPatients

• LackofPCPAccess

Financial• RisingCosts

• DecreasingReimbursement

• CapitatedPayments

• AccountableCare

• RepurposingExistingSpace

EmergencyCareSystemStressors


PhysicalSpaceInsufficienttoMeetCurrent&

ForecastedDemand


FullScaleMock-up


MultidisciplinaryTeams


•ExecutionoftheUMHSStrategicPlantodouble adulthighcomplexitymarketshare (from6%- to12%)willamplifythe

demandforemergencycriticalcareinoursystem

•Theemergencycaresystem,init'scurrentstructure,isnotpreparedtorespondtothesechallenges.

HealthSystemStrategicPlan


Roadtrips


Centralized,commonlyused

items

Rarelyusedbutcritical

proceduralitems


MichiganMedicine- Facts&Figures


GoldenTickets


LessonsLearned


StorageSpace/Technology


FullScaleMock-up


ValueStreamMapping


• file:///.file/id=6571367.12173219


Fiberoptic video

imagecanbe

displayedonmultiple

HDscreensaround

RESUSbay

emergency critical care center: a lean journey - .global · 2001 2009 ed critical care visits ed...

Documents