Approaches to Queries for Cohort Identification

11/6/17

1

ApproachestoQueriesforCohortIdentification

PaulHarris,PhDProfessor,BiomedicalInformatics

Patients

CommunityFaculty/Staff

11/6/17

2

Faculty/Staff

11/6/17

3

OR…

11/6/17

4

291Publications

11/6/17

5

Step 1: Potential volunteers (or their parents/caretakers) self-register to indicate a willingness to be contacted for research studies.

11/6/17

6

Potential volunteers (or their parents/caretakers) self-register to indicate a willingness to be contacted for research studies.

Poweredby

Potential volunteers (or their parents/caretakers) self-register to indicate a willingness to be contacted for research studies.

Poweredby

11/6/17

7

Registered researchers search database for individuals based on study inclusion criteria and geographical location (Only De-Id Information)

Researchers send IRB approved recruitment message to ‘matched’ volunteers.

Volunteers make final choice to share identifiable information for direct contact.

11/6/17

8

Researchers contact interested volunteers and follow normal study consent procedures.

152Publications

11/6/17

9

11/6/17

10

?

It’sEarly… NoFormalEvaluationYetforEnrollment,ButPeopleareClicking

11/6/17

11

InterestedinAdopting/Collaborating?paul.harris@vanderbilt.edu

11/6/17

12

20

• AnInformaticsplatformdesignedtocontinuouslyengagepatientsandofferopportunitiestoparticipateinresearch

• Acohortof~20,000Vanderbiltpatientsthathaveoptedintobecontacted directly bye-mailtoparticipateinresearchortoprovideinputonresearchideas

• Aconvenientandefficientpanelofpatientrepresentativesforwhichwehavemedicalrecorddata

• Uses:survey,clinical,interventionalresearchandstakeholderengagementtoguideresearchefforts

11/6/17

13

MRAVCohortAuthenticated(MHAV)individualscanusethislinktoprovideinformationabouthowtheywishtobecontactedforrecruitment(e.g.e-mail,phone,onlybymydoctor,etc)

0100020003000400050006000

18-29

30-39

40-49

50-59

60-69

70-79

80-89

90-99

MRAVCohortAgeDistribution

MRAVCohortTop5 Conditions

HypertensionNOS

BenignHypertension

DiabetesUncomplTypeII

HyperlipidemiaNEC/NOS

AllergicRhinitisNOS

11/6/17

14

• ObtainIRBapprovalforuseofMRAVrecruitmenttoolandemailcontactlanguage

• SubmitaMyResearchAccessRequest

• Reviewedforparticipantburdenandavailabilityofprogrammers

• Self-ServiceIdentificationthroughRDDiscoverInterface

• - or- submitaResearchDerivativeRequesttoidentifyeligiblepatientsbasedonstudycriteria,ifapplicable(IDASCore,$120perprogramminghour)

• Onceapproved,DataCoordinatingCenterCoreprogrammerssendemailnotificationstoparticipantsincludingapprovedlanguage($82perprogramminghour)

HowtoRecruitPatientsFromTheMRAVCohort

ExpressionofInterest/Pre-ScreeningSurvey

ExpressionofInterest/Pre-ScreeningSurvey

11/6/17

15

• Enablesmoreefficientscreeningandrecruitingofpatientsversustraditionalmethodsofmanuallyreviewingupcomingappointmentlistsandcross-referencingEMR

• Basedonalistofupcomingappointmentsinapredeterminedsetofclinics,SubjectLocatorsearchespatient’sEMRviatheresearchdatawarehouseforcommonlyused,discreteinclusion/exclusioncriteriatosignificantlynarrowdownthenumberofpatientsthatrequirescreening

– CriteriaincludesIC9/10andCPTcodes,demographics,vitals,keywords,medications

Snapshot–PilotStudies

Nephrology• Examined: 2598• Candidates: 96

(reduction- 96%)CleftPalate• Examined: 2490• Candidates: 27

(reduction- 99%)

Cardiology(2studies)(reduction- 95%)

StartingHereàFilteringCriteria

…

ReviewList

ClinicsofInterest

StudyWorkQueue(DailyReview)

11/6/17

16

11/6/17

17

BioVUØ LinksDNAextractedfromdiscardedblood

samplestode-identifiedEMRØ ~177kDNAsamples,over20k pediatricØ 74 BioVUprojectsapprovedtodate

SyntheticDerivativeØ Researchtooltoenablestudieswith

de-identifiedclinicaldataØ Contains2.5millionrecords;highlydetailed

longitudinalclinicaldatafor~1million

ResearchDerivativeØ IdentifiedclinicaldatawarehouseØ Tools(e.g.SubjectLocator)Ø Services(FeeforService)Ø NewREDCap extractiontoolkit

11/6/17

18

11/6/17

19

•Recommendations for clinical data representation to support phenotyping• 1. Structure clinical data into queryable forms.• 2. Recommend use of a common data model, but also support customization for

the variability and availability of EHR data among sites.

•Recommendations for phenotype representation models• 3. Support both human-readable and computable representations.• 4. Implement set operations and relational algebra.• 5. Represent phenotype criteria with structured rules.• 6. Support defining temporal relations between events.• 7. Use standardized terminologies, ontologies, and facilitate reuse of value sets.• 8. Define representations for text searching and natural language processing.• 9. Provide interfaces for external software algorithms.• 10. Maintain backward compatibility.

11/6/17

20

PARTNERS

Visionandpurpose

Ourgoalistopositivelyimpacthumanhealthbyimprovingparticipantenrollmentandretentioninmulti-centerclinicaltrials.

Achievingthisgoalwillrequiresophisticatedinformatics-basedrecruitmenttools andnovelengagementapproachestoacceleraterecruitmentandretention.

Participant Trial

11/6/17

21

KeyPrinciples• RespectingCTSAautonomyanddiversity• Afocusonminorityandunderservedpopulations• Makingthemostofelectronichealthrecords• Preservingadiseaseneutralapproach• Focusoncostefficiency• Respectingandreturningvaluetoparticipants• Buildonbestpractice(avoidreinventingthewheel)• Evidencebased….Whatworks?(testbed)• Finiteresources– scalability/tools• Homeforrecruitmentexperts(across+beyondCTSA)

EHR-Based Site Feasibility

11/6/17

22

EHR-Based Feasibility

11/6/17

23

EHR-Based Feasibility

11/6/17

24

Same,butDifferent

EHR-Based Feasibility

EHR-Based Feasibility

11/6/17

25

11/6/17

26

11/6/17

27

11/6/17

28

11/6/17

29

11/6/17

30

Non-Fed Fed

EHR-Based Feasibility: Federated

11/6/17

31

RelatedResultsFederated Non-Federated

Counts TriNetXCountsDifferencebetweencounts

TriNetX(CTSASites)

Differencebetweencounts

Count1 237,770 172,500

Count2 233,310 2.28% 168,710 2.19%

Count3 214,630 8.00% 157,650 6.56%

Count4 181,630 15.0% 129,660 17.75%

Vanderbilt Differencebetweencounts RIC Differencebetween

counts

9,731 39,571

9,505 2.32% 38,319 3.16%

9,106 4.20% 37,132 3.10

7,302 19.81% 7,302 16.76%

ThechangeinthecountsisvirtuallythesamebetweenFederatedandNon-Federated

+NeedFederatedtodosophisticatedsensitivityanalysisofinclusion/exclusionrules

11/6/17

32

We’reworkingonjoining…

We’reworkingonleveraging…

11/6/17

33

We’reworkingonjoining…

We’reworkingonleveraging…

11/6/17

34

11/6/17

35

11/6/17

36

11/6/17

37

11/6/17

38

THREE Immediate Opportunities To Collaborate

Sharematerials,tools,bestpractices,… www.trialinnovationnetwork.orgàà Toolkit

1

11/6/17

39

Recentofferings(videoavailable)

ShowcaseYourWork--- www.trialinnovationnetwork.orgGiveaWebinar(logistics,setup,archivalsupportedbyRIC)

2

InterestedinAdopting/Collaborating?paul.harris@vanderbilt.edu

3Collaborate–TrialsToday- Local

11/6/17

40