Impact of EHR design and operational procedures on health ......EHR Data more complete than claims • Analysis of diabetics using claims data vs data extracted from the EHR • Claims

Impact of EHR design and operational procedures on

health care data for research EdgarChou,MD

ChiefValueOfficer/ChiefMedicalInformaticsOfficerDrexelUniversityCollegeofMedicine

September29,2016

Objectives

• UnderstandtheunderlyingdriversforElectronicHealthRecordDesign

• UnderstandthedriversthatimpactdataintegrityintheElectronicHealthRecord

• Understandthelimitationsofdatasets

DISCLOSURE STATEMENT

• Nonetoreport

Drexel University College of Medicine: Drexel Medicine • Multispecialtyambulatorypractice•  250Physicians•  17subspecialties•  500Residents/Fellows•  1000medicalstudents•  1000staff•  300,000outpatientvisitsperyear•  ImplementationofEHRin2007

How did we get to this point?

http://www.hermes-press.com/SP17.jpg,AccessedOctober31,2014

Crossing the Quality Chasm - 2001

• 6AimsofHealthcare:• Shouldbe:

• Safe• Effective• PatientCentered• Timely• Efficient• Equitable

Crossing the Quality Chasm

•  Tobepursuedby:•  HealthCareorganizations•  ProfessionalGroups•  PrivatePurchasers•  PublicPurchasers

• Goals:•  Reducetheburdenofillness,injuryanddisabilitytoimprovethehealthandfunctioningof[ourpopulation]

Crossing the Quality Chasm

• Patientassourceofcontrol•  Sharedknowledgeandfreeflowofinformation•  Evidencebaseddecisionmaking•  Safetyasasystemproperty• Needfortransparency• Anticipationofneeds• Decreaseinwaste• Cooperationamongstclinicians

Slides on US healthcare spending

OECD=OrganizationforEconomicCo-operationandDevelopment(internationalorganizationof34countriesfoundedin1961tostimulateeconomicprogressandworldtrade)

“Once again, U.S. has most expensive, least effective health care system in survey” – The Washington Post, June 16, 2014

http://www.washingtonpost.com/news/to-your-health/wp/2014/06/16/once-again-u-s-has-most-expensive-least-effective-health-care-system-in-survey/,AccessedOctober31,2014

Meaningful Use

CentersforMedicare&Medicaid.2011.Accessedfrom:https://www.cms.gov/Regulations-and-Guidance/Legislation/EHRIncentivePrograms/Downloads/CQM_Webinar_10-25-2011.pdf

Meaningful Use Summary

http://www.advisory.com/~/media/Advisory-com/Research/ITSC/Resources/Posters/2012/2014-Edition-MU-Whiteboard_2%2010%2014.pdf

Adoption of Electronic Health Reords

https://www.healthit.gov/sites/default/files/data-brief/2014HospitalAdoptionDataBrief.pdf,lastaccessedSeptember13,2016

Triple Aimà Quadruple Aim •  “improvingtheexperienceofcare,improvingthehealthofpopulations,andreducingpercapitacostsofhealthcare”–TheInstituteforHealthCareImprovement,2012

HealthoftheProvider

Quality Payment Program: Merit Incentive Payment System (MIPS)

Element Was… Weight FutureWeight

ReportingMethod

Quality PatientQualityReportingSystem(PQRS)

50% 30% Registry/QCDR/WebInterface(248patients)

Resource PhysicianValue-PQRS(PV-PQRS)/ValueModifier(VM)

10% 30% Claims

ClinicalPracticeImprovementActivity(CPIA)

Didn’texistbefore 15% 15% Registry/QCDR

AdvancingCareInformation(ACI)

MeaningfulUse(MU) 25% 25% Registry/QCDR

https://www.gpo.gov/fdsys/pkg/FR-2016-05-09/pdf/2016-10032.pdf,lastaccessedJuly12,2016

Time in the EHR measured

•  Forevery1hourofclinicalcaredelivered,2additionalhoursofclerical/documentationworkintheoffice

• Anadditional1-2hoursofworkoccurathome•  Increasedriskforphysicianburnout•  Increasedriskforlowerqualitycare(anddocumentation…)

http://annals.org/article.aspx?articleid=2546704,lastaccessedSeptember14,2016

Medical Education • OnejustneedstolookathowphysiciansaretrainedinfurtherunderstandingthedatacollectedinEHRs

•  “90%ofdiagnosescomefromadetailedhistory”•  Identifythereasonforvisit(ChiefComplaint)•  Discussthedetailsregardingthecomplaint(e.g.timing,severity,associatedsignsandsymptoms)

•  ContributoryFamily,Medical,SocialHistory,SurgicalHistory•  ReviewofOrganSystemstoidentifyotherrelatedissues•  PhysicalExam•  Assessment(synthesisoftheinformationgathered)•  Plan

Documentation and Billing

• Completedocumentationreflectstheperformanceofthephysicianwiththepatientandfollowsaregimentedsetofguidelines

•  1995EvaluationandManagementguidelines•  ChiefComplaint(CC)•  HistoryofPresentIllness(HPI)•  PastFamily,Medical,SocialHistory•  ReviewofSystems(ROS)•  PhysicalExam•  Assessment•  Plan

https://www.cms.gov/Outreach-and-Education/Medicare-Learning-Network-MLN/MLNProducts/Downloads/eval_mgmt_serv_guide-ICN006764TextOnly.pdf,Lastaccessed8/13/16

Evaluation and Management Guidelines • DocumentationmayeitherbeinHPIorspecificsections:Duplication!

http://www.aafp.org/fpm/2010/0300/p22.html,lastaccessed9/6/16

Electronic Progress Notes

• Benefits:•  Legible•  Betterqualitynotes

•  Shortfalls•  NoteBloat

•  Improperuseof“CutandPaste”withduplicationwithoutupdating•  SingleClicktemplates“Allnormal”

Documentation approaches vary…

FREETEXTGOESHERE

SINGLECLICKOPTIONS

Data Integrity

• Dataintegrityreferstotheoverallcompleteness,accuracyandconsistencyofdata.

Data input: If you’ve seen one physician… •  you’veseenonephysician… Coronary Artery Disease

CAD Heart Failure Heart Disease HF CHF Congestive Heart Failure Low Ejection Fraction ASCVD Myocardial infarction MI MI s/p stent MI s/p PTCA MI s/p balloon angioplasty CABG Coronary Artery Bypass Graft CHF with EF<55% CHFpEF Ischemic Cardiomyopathy

ALLSCRIPTS ENTERPRISE V11.2.3

LABCORPINDEPENDENT LAB VENDOR

HUH: CERNPATHPATHOLOGYInpatient and

Outpatient results

QUESTINDEPENDENT LAB VENDOR

HUH: HPF: OPERATIVE REPORTS

HUH: PACSWEB LINK: URL

HUH: IDXRADRADIOLOGY

Outpatient radiology results only

HUH: SUNQUEST LABS

Oupatient lab results only

Bidirectional

Bidirectional

UnidirectionalConnect R

Unidirectional Ensemble

UnidirectionalConnectR

HUH: HPF: DISCHARGE SUMMARIES

(Future goal: Task PCP)

Unidirectional (Future Goal: Bidirectional)Open Link

IDX (Centricity)Practice Management System

V5.0Unidirectional to Allscripts: Demographics, Visit Type,

Appointment Status, Preferred method of communication

Charges from Allscripts

HUH: PBARREGISTRATION

SYSTEM

PBAR Thru Connectr

AnyEHR

Most organizations are not like Kaiser Permanente… • Datasourcescomefromdisparatesources

•  Payers•  Hospitals•  Otherhealthcaresystems•  AncillaryServices

•  HomeHealth•  NursingHomes•  SkilledNursingFacilities•  LabVendors

Data Integrity •  InaccurateReporting

•  Sensitivityofreportingrangingfrom46-98%permeasure

1AnnInternalMed.2013,158:77-83.

EHR Data more complete than claims

• AnalysisofdiabeticsusingclaimsdatavsdataextractedfromtheEHR• Claimsdataonlyidentified75%ofexistingdiabetics•  EHRdataidentified97%ofexistingdiabetics

JAmMedInformAssoc.2007Jan-Feb;14(1):10-5.Epub2006Oct26

83yo with PMHx CAD, atrial fibrillation, MR s/p MVR, DM, HTN, CHF with EF=45%, urinary incontinence presents for followup.

83yo with PMHx CAD, atrial fibrillation, MR s/p MVR, DM, HTN, CHF with EF=45%, urinary incontinence presents for followup.

SUMMARY:34problemsLimitof10submissionsforMedicareClaims(usedtobe4)Diagnosesneedtobesubmittedannuallyforpatients(impactsriskscores)

Documented Plan

Charge Capture

Additional means of data capture

Improving the Health of a Population: Data and Patient Quality Reporting System (PQRS)

•  CaseStudy:•  PQRSMeasure#7:BetablockertherapyforPriorMyocardialInfarctionorLeftVentricularSystolicDysfunction(LVEF<40%)

•  Howwouldyoufindthisinformation?•  PQRSreliesonbillingcodes(ICD-10-formerlyICD-9)

•  Additionalbackground:•  LeftVentricularSystolicDysfunctiondoesnothaveabillingcode

•  Requiresadditionalbillingcode:G8694•  PatientQualityReportingSystemresultsreportednationallyin2014forconsumerstoallowcomparisonwithotherorganizations

•  ValueBasedModifierprogram–Paymenttotheorganizationwillbebasedonperformanceonqualitymeasures

• Whatconcernswouldyouhaveforyourorganization?•  Potentialriskfor1)pooroutcomesfalselyattributedsolelybasedonextractiontechnique;2)lowerreimbursements

Data and Predictive Analytics

https://upload.wikimedia.org/wikipedia/commons/f/f9/Poland._Trash_002.JPG;lastaccessed9/14/16

http://www.ghymcablog.org/blog1-wp/wp-content/uploads/2013/05/lemonade.jpg;lastaccessed9/18/16

What’s happening?

http://www.cleveland.com/healthfit/index.ssf/2014/10/cleveland_clinic_ibm_to_begin.html,AccessedNovember1,2014

How to improve?

• Understandthedatasourcesandmethodsofdataentrytounderstandthelimitationsofthedatarequested

•  Lessfocusonprogramsthatdonotnecessarilyimprovequality

Additional Sources of Data and Future Applications • HealthShareExchanges• PrecisionMedicine• AdvancementsinArtificialIntelligence