Using Analytics to Drive Change in Health Professions Education Marc M. Triola, MD NYU School of Medicine @marctriola
Using&Analytics&to&Drive&Change&in&Health&Professions&Education
Marc%M.%Triola,%MD
NYU%School%of%Medicine
@marctriola
Why%Now?
• Analytics%Driving%Changes%in%Care:
– Changes%in%health%care%delivery%
– Explosion%of%scientific%knowledge%
– Demand%for%a%data%literate%physician%workforce
• Analytics%Driving%Changes%in%Education:
– Personalized%education
– CompetencyKgated%pathways
– Increasing%amounts/capacity%for%data%collection
– New%learning%analytics%methods%to%derive%actionable%
knowledge
"I%know%more%about%my%11KyearKold%son's%sixthK
grade%basketball%team%than%the%average%college%
faculty%member%knows%about%their%incoming%
class,%in%terms%of%key%variables%that%are%going%to%
make%them%successful%or%not%successful”
»Mark%Milliron,%Bill%&%Melinda%Gates%Foundation
The&Education&Data&Continuum
Curriculum
USMLE%
Exams%and%Evaluations
MCAT%/%
Admissions
Competency%Portfolios
Clinical% patient%outcomes
CME%Performance
Medical%School Training Clinical%Practice
NRMP ACGME
Boards%and%licensure
Warehouse%of%longitudinally% integrated%learner%data
Institution%A Institution%B Institution%C
The&Education&Data&Continuum
Med%
SchoolClinical% Practice
Residency
College
Density%of%Education%Data
Density%of%Patient%Data%and%Outcomes
Inspired%by%Dr.%Tom%Aretz,%Harvard%Medical%School
How%to%Drive%Change?
• Our%ability%to%collect%data%has%vastly%outpaced%
our%ability%do%gain%insight%and%action%from%it.
• Most%data%and%systems%are%not%integrated%nor%
interoperable.
• Local%resources%are%needed,%as%is%expertise%
not%normally%a%part%of%the%education%mission.
• An%ongoing%strong%partnership%between%
Informatics/IT,%faculty%and%administration%is%
key%– it’s%a%culture%change.
Education&Data&Warehouse
Education%Data%
Warehouse
Education%Data%
Warehouse
LMS
Curricular%
Content
AdmissionsExams
ePortfolioPatient%
Log
Learning%
ModulesEvaluations
SIS
Reporting&and&AnalyticsData&Marts&for&
Education&Research
Simulation
Clinical%Data%
Warehouse
Clinical%Data%
Warehouse
NYU%
Practice%
Network
Epic%EMR
Division%of%Educational%
Informatics
Institute%Structure
Curriculum%Mapping
Multimedia%and%Medical%
Visualization
Educational%Technology
Division%of%Education%Quality%
and%Analytics
Instructional%Design
Education%Analytics%for%
UMEKGME%Continuum
Benchmarking%and%
Outcomes
UME/GME%Assessment%
Program
Digital%Press
Faculty%iBook% Publishing
Commercialization% and%
Licensing
UMEKGME%
Continuum
Education%Data%
Warehouse
Learning%analytics%refers%to%the%interpretation%of%a%
wide%range%of%data%produced%by%and%gathered%on%
behalf%of%students%in%order%to%assess%academic%
progress,%predict%future%performance,%and%spot%
potential%issues
KU.S.%Department%of%Education%2012
Analytics
• Descriptive:&&How%did%I%do%on%the%assessment%items?%
• Diagnostic:&&Did%I%do%better%on%application%or%recall%questions?
• Predictive:&Based%on%my%prior%data,%what%is%my%estimated%USMLE%score?%
• Prescriptive:&My%learning%data%indicates%I%should%practice%genomics%and%pharmacology%using%
application%questions.
Analytics
• Early&warning&system
• Early&reward&system
• Anecdote&to&epidemiology
Individual%Dashboards%for%Learners
Individual%Dashboards%for%Faculty
Dashboards%for%Edu%Leadership%and%Mentors
Dashboards%for%Learning%Modules
Clinical%Data%and%the%Education%
Continuum
• Use%of%EMR%data%in%the%medical%school%curriculum%has%
tremendous%potential
• Can%our%clinical%data%drive%our%curricular%content?
• Which%learner%and%patient%outcomes%that%can%be%used%
to%monitor%the%impact%of%curriculum%on%clinical%
effectiveness?
Predictive%Analytics
• 7,859%students%in%36%classes%entering%Jefferson%Medical%College%between%1970%and%2005
• Could%MCAT%predict:%performance%in%medical%school,%USMLE%scores,%and%ratings%of%clinical%competence%in%residency
MCAT%as%Predictor
• MCAT%scores%moderately%correlated%with%
medical%school%performance%and%with%scores%
on%USMLE%(up%to%4%years%later).
• Predictive%validity%coefficients%for%women%were%
consistently%higher%than%those%for%men.
• MCAT%performance%did%not%predict%clinical%
competence%or%professionalism.
Predicting%USMLE
• Base%Model:%Gender,%MCAT,%College%GPA
• Exams:%Preclinical%exam%scores
• Knowledge%domains:%%Performance%on%topics,%
weighted%by%correlations%to%prior%student%
USMLE
Base%Model Exams Knowledge%Domains
Domains%+%Base
Within&10&points 55% 58% 52% 65%
Within&15&points 76% 86% 82% 86%
Within&20&points 85% 95% 95% 97%
Pusic et.%al.,%unpublished% data
Driving%Change
• Educational%technologies%and%analytics%should:
– drive%the%move%towards%individualized% learning%and%%
personalized%progression
– accelerate%the%transformation% towards%health%
education% that%is%competencyKdriven,%affordable,%and%
accessible%to%each%learner
• Faculty%need%new%skills%and%expertise%in%the%
effective%use%of%educational%technologies%and%
learning%analytics%to%enhance%the%teachingK
learning%process.
Driving%Change
• Shared%goals,%shared%pressures,%robust%data,%
and%analytics%can%bridge%the%gap%between%our%
educational%and%clinical%missions.
• Insights%from%the%overlap%of%education%and%
clinical%data%can%enhance%the%value%of%
teaching%and%learning%to%the%healthcare%
delivery%system
Driving%Change
• Educational%technologies%should%be%
interoperable%with%respect%to%both%content%and%
outcomes%data.
• Learners%and%faculty%should%own%their%data.
• Analytics%and%data%integration%should%be%easy.
• Impact%of%educational%innovations%will%be%
maximized%by%sharing%across%platforms,%
teams,%schools,%institutions,%and%the%public.
• Organizations%like%MedBiquitous are%key!