2018 SPRING INDUSTRY ADVISORY BOARD MEETING - CHOT
Transcript of 2018 SPRING INDUSTRY ADVISORY BOARD MEETING - CHOT
2018 SPRING INDUSTRY ADVISORY BOARD MEETING April 12-13 | Houston, Texas
1
2
CHOT 2018 SPRING INDUSTRY ADVISORY BOARD MEETING AGENDA
Thursday, April 12
7:30—8:30 Check-In & Continental Breakfast
8:30—9:30 Welcome & Introductions
Welcome: Dr. Bita Kash, Texas A&M University, CHOT Center Director and Dr. Thomas Ferris, Texas A&M University, CHOT Center Co-Director
State of the Center: Dr. Bita Kash, CHOT Center Director and Dr. Thomas Ferris, CHOT Center Co-Director
LIFE Form Review: Dr. Craig Scott, NSF I/UCRC Evaluator
Meeting Overview: Lauren Irlinger, CHOT Managing Director
9:30—10:00 Keynote Presentation
Dr. M. Michael Shabot, Executive Vice President & System Chief Clinical Officer, Memorial Hermann Health System
10:00—10:15 Break
10:15—11:15 CHOT Research Impact and Insight Presentations from Completed Projects
11:15—11:45 Proposal Presentations: Population Health
Theme Champion: Dr. Conrad Tucker, Pennsylvania State University
Pop 1 Comprehensive Analysis on Impact of Social Determinants to Improve Care Across Populations
Pop 2 Participating in a Community Health Improvement Network
Pop 3 The Effectiveness of Substance Abuse Treatment Services in Combating Opioid Crisis
11:45—12:00 IAB Member Discussant Panel & LIFE Forms
12:00—1:00 Networking Lunch
3
CHOT 2018 SPRING INDUSTRY ADVISORY BOARD MEETING AGENDA
Thursday, April 12
1:00—1:30 Proposal Presentations: Care Coordination
Theme Champion: Dr. Christina Mastrangelo, University of Washington
Care 1 Measuring Patient Experience and the Effects of Community Factors on Value-based Reimbursement across the Continuum of Care
Care 2 Developing a Risk Prediction Model for Hospital Acquired Clostridium Difficile Infection
1:30—1:45 IAB Member Discussant Panel & LIFE Forms
1:45—2:15 Proposal Presentations: Analytics and Innovative Technologies
Theme Champion: Dr. Eva Lee, Georgia Institute of Technology
Tech 1 HIE Project for Chronic Disease and Workflow Management
Tech 2 Leveraging technology to enhance communication in healthcare
Tech 3 Data-driven analytics and machine learning for improving healthcare outcomes
2:15—2:30 IAB Member Discussant Panel & LIFE Forms
2:30—2:45 Break
2:45—3:15 Proposal Presentations: Patient Experience
Theme Champion: Dr. Nancy Borkowski & Robert Weech-Maldonado, University of Alabama at Birmingham
Patient 1 Care Coordination Activities for Individuals with Spinal Cord Injury
Patient 2 Generating Tailored Recommendations Automatically with Explanations via an Interactive Dialog-based system
Patient 3 Embedding Routine, Informal Family Caregiver Assessment of Delirium Superimposed on Dementia into Acute Care
3:15—3:30 IAB Member Discussant Panel & LIFE Forms
3:30—4:30 Education & Engaged Scholarship Program
Dr. Norma Padron
4:30—4:45 Day Debrief
4:45—6:00 Networking Reception
4
CHOT 2018 SPRING INDUSTRY ADVISORY BOARD MEETING AGENDA
Friday, April 13
7:30—9:30 Breakfast and Check-in
8:00—9:00 Closed Door IAB Meeting
9:00—9:15 Break
9:15—9:45 Proposal Presentations: Access to Care
Theme Champion: Dr. Christopher Johnson, University of Louisville
Access 1 Telemedicine in Primary Care and in the Management of Chronic Conditions: Exploring Patient & Provider Perspectives
Access 2 Ask Me 3®: A Home Health Intervention to Address Health Literacy Barriers, Increase Patient Engagement, and Improve Patient Experience and Outcomes
9:45—10:00 IAB Member Discussant Panel & LIFE Forms
10:00—10:30 LIFE Feedback Discussion Panel
Dr. Craig Scott, NSF I/UCRC Evaluator
10:30—10:45 Break
10:45—11:45 Grand Challenge Planning Grant Results & Next Steps
Dr. Bita Kash, CHOT Center Director
Dr. Marc Garbey, Center Director for the Center for Cyber-Physical Systems for the Hospital Operating Room (CyBHOR)
11:45—12:00 Debrief & Adjourn
12:00 Box Lunches Available for Pick-up
5
CHOT 2018 SPRING INDUSTRY ADVISORY BOARD MEETING
Key Meeting Contact Information
If you have any questions or need assistance during the meeting, please do not hesitate to contact one of the following:
Bita Kash Center Director [email protected] (979) 575-6768
Lauren Irlinger Center Managing Director [email protected] (407) 506-9757
6
2018 SPRING IAB MEETING PARTICIPANTS April 12-13 | Houston, Texas
Participant List as of April 4th, 2018
Craig Scott [email protected] NSF - University of Washington
Casey Stallsmith [email protected] NSF - Hosparus Health
Thomas Miller [email protected] American Society of Anesthesiologists
Bob Bernstein [email protected] Avizia (formerly Carena)Rachele Misiti [email protected] Central Texas VADorothy Sanders [email protected] Central Texas VAJamey Gigliotti [email protected] HighmarkDustin Dew [email protected] Lakeshore FoundationNorma Padron [email protected] Main Line Health
Kenneth [email protected]
Opelousas General Health System
Thomas Tracy [email protected] State Health Hershey Medical Center
Christopher Hall [email protected] Philips HealthcareChris Juday [email protected] SanofiChris Juday [email protected] Sanofi
Christian [email protected]
Siemens Healthineers
Steven Brown [email protected] A&M University Health Science Center
J.J. Schmidt [email protected] York Risk Services
Jessica Autrey [email protected] AT&TMarc Garbey [email protected] CyBHORMatthew Gibson [email protected] Erlanger Health System
Michael [email protected]
Memorial Hermann Health System
Jill Bell [email protected] Passport Health Plan
Melissa Boltz [email protected] Penn State Hershey Medical Center
Jerome Jourquin [email protected] Susan G Komen
Bita Kash [email protected] CHOT Center
Lauren Irlinger [email protected] CHOT Center
Ankur Agarwal [email protected] Florida Atlantic University
Eva Lee [email protected] Georgia Institute of Technology
Conrad Tucker [email protected] Penn State University
Thomas Ferris [email protected] Texas A&M University
National Science Foundation
IAB Members
Guests
University Faculty & Administration
7
2018 SPRING IAB MEETING PARTICIPANTS April 12-13 | Houston, Texas
Participant List as of April 4th, 2018
Georges Naufal [email protected] Texas A&M University
Nancy Borkowski [email protected] University of Alabama at Birmingham
Robert Weech-Maldonado [email protected] University of Alabama at Birmingham
J'Aime Jennings [email protected] University of LouisvilleChristopher Johnson [email protected] University of LouisvilleTiffany Robinson [email protected] University of LouisvilleJoseph Heim [email protected] University of WashingtonChristina Mastrangelo [email protected] University of Washington
Guanlin Chen [email protected] Georgia Institute of TechnologyQixuan Hou [email protected] Georgia Institute of TechnologyZhuonan Li [email protected] Georgia Institute of Technology
Di Liu [email protected] Georgia Institute of Technology
Joshua Morgan [email protected] Georgia Institute of Technology
Heather Patrick [email protected] Georgia Institute of TechnologyCody Wang [email protected] Georgia Institute of TechnologyZixing Wang [email protected] Georgia Institute of TechnologyLinxi Xiao [email protected] Georgia Institute of TechnologyPeijue Zhang [email protected] Georgia Institute of TechnologyPavan Thaker [email protected] Georgia Institute of Technology Christian Lopez [email protected] Penn State UniversityPreston Blackburn [email protected] Texas A&M UniversityKesler Brock [email protected] Texas A&M University
Abigail Gonzalez [email protected] Texas A&M University
Johnathan McKenzie [email protected] Texas A&M UniversityHannah Meirink [email protected] Texas A&M UniversitySahinya Susindar [email protected] Texas A&M University
Neeraj Puro [email protected] University of Alabama at Birmingham
Reena Joseph [email protected] University of Alabama at Birmingham
Mohamed Ahmed [email protected] University of Louisville
Molly O'Keefe [email protected] University of Louisville
Larissa Prates Guimaraes Petroianu
[email protected] University of Washington
University Faculty & Administration
Students
8
chotnsf.orgNSF CENTER FOR HEALTH ORGANIZATION TRANSFORMATION
As a National Science Foundation industry-university cooperative research center (I/UCRC), CHOT follows a model of an industry-
Industry Advisory Board (IAB) to conduct research that supports the implementation of evidence-based transformational
Pooled Members
$
NSF Funds
Institutional Support
RESEARCH PROJECTS
Innovations in Healthcare Delivery
INVESTIGATE VALIDATEInnovations and prototypes
IMPLEMENT
Value Created
INDUSTRY ADVISORY BOARD (IAB)
INDUSTRY MEMBERSHIP
= $50,000
CHOT’s research model relies on the knowledge and experience of healthcare leaders to guide academic research to ensure that it is meaningful and applicable to the healthcare industry and provides immediate decision support.
CORE FUNDS &
SUPPLEMENTAL FUNDS
$700,000
CHOT UNIVERSITY SITES:
9
CHOT PRESENT INDUSTRY MEMBERS
chotnsf.org
Central Texas Veterans Health Care System
Last Best Chance
10
NS
F C
ente
r fo
r H
ealt
h O
rgan
izat
ion
Tra
nsf
orm
atio
n
Sta
te o
f th
e C
ente
r S
pri
ng
201
8A
pril
12 &
13,
201
8
Bita
Kas
h, C
ente
r D
irec
tor
Tho
mas
Fer
ris,
Co-
Dir
ecto
r La
uren
Irlin
ger,
Man
agin
g D
irect
or
CU
RR
EN
T C
HO
T U
NIV
ER
SIT
Y S
ITE
S
CH
OT
Lea
ders
hip
Team
201
8
Bita
Kas
h, P
hD, M
BA
, F
AC
HE
CH
OT
Cen
ter
Dir
ecto
rTe
xas
A&
M U
nive
rsity
Laur
en I
rling
er,
MH
AC
HO
T M
anag
ing
Dir
ecto
rTe
xas
A&
M U
nive
rsity
Tho
mas
Fer
ris,
PhD
CH
OT
Cen
ter
Co-
Dir
ecto
rTe
xas
A&
M U
nive
rsity
Nor
ma
Pad
rón
, P
hDIA
B C
hair
pers
onM
aine
Lin
e H
ealth
Kar
an U
ppal
, P
hDIA
B V
ice
Cha
irpe
rson
Em
ory
CH
OT
Lea
ders
hip
Team
201
8
Eva
K.
Lee,
PhD
Site
Dir
ecto
rG
eorg
ia I
nstit
ute
of T
echn
olog
y
Con
rad
Tuc
ker,
PhD
Site
Dir
ecto
rP
enns
ylva
nia
Sta
te U
nive
rsity
Rob
ert
Wee
ch-M
aldo
nado
, PhD
Site
Dir
ecto
rU
nive
rsity
of A
laba
ma
at B
irmin
gham
Nan
cy B
orko
wsk
i, D
BA
Site
Co-
Dir
ecto
rU
nive
rsity
of A
laba
ma
at B
irmin
gham
11
CH
OT
Lea
ders
hip
Team
201
8
Chr
isto
pher
E J
ohns
on,
PhD
Site
Dir
ecto
rU
nive
rsity
of
Loui
svill
eS
ite C
o-D
irec
tor
Uni
vers
ity o
f Lo
uisv
ille
Chr
istin
a M
astr
ange
lo,
PhD
Site
Dir
ecto
rU
nive
rsity
of
Was
hing
ton
Jose
ph H
eim
, PhD
Site
Co
-Dir
ecto
rU
nive
rsity
of
Was
hing
ton
CH
OT
Lea
ders
hip
Team
201
8
Ank
ur A
garw
al,
PhD
Site
Dir
ecto
rF
lori
da A
tlant
ic U
nive
rsity
Ste
phen
Tim
mon
s, P
hDS
ite D
irec
tor
Uni
vers
ity o
f N
ottin
gham
NS
F I/
UC
RC
Intr
oduc
tion
Vid
eo
http
s://
ww
w.n
sf.g
ov/e
ng/ii
p/iu
crc/
iucr
c_vi
deo.
jsp
Ind
ust
ry A
dvi
sory
B
oar
d (
IAB
)P
oo
led
M
emb
er
Fee
s
Un
iver
sity
C
ente
r S
ites
Inve
stm
ents
NS
F F
un
ds
NS
F C
ore
&
Su
pp
lem
enta
l F
un
ds
Val
ue
Cre
ated
Inn
ova
tio
ns
in H
ealt
hca
re D
eliv
ery
RE
SE
AR
CH
P
RO
JEC
TS
12
Hea
lth
Sys
tem
s/P
rovi
der
s (1
2)
Ala
care
Hom
e H
ealth
and
Hos
pic
e
Car
e C
oord
inat
ion
Inst
itute
Gra
dy
Hea
lthS
yste
m
Lake
shor
e F
oun
datio
n
Mai
nLi
ne H
ealth
Mor
ehou
se S
choo
l of M
edic
ine
Ope
lous
as G
ener
al H
ealth
Sys
tem
Pen
nS
tate
Her
shey
Med
ical
Cen
ter
UA
B H
ealth
Sys
tem
Uni
vers
ity o
f Lou
isvi
lle H
ospi
tal
CU
RR
EN
T C
HO
T IN
DU
ST
RY
ME
MB
ER
S (
29 m
emb
ers)
Ass
oci
atio
ns/
Go
vern
men
t (7
)
Am
eric
anS
ocie
ty o
f Ane
sthe
siol
ogis
ts
Cen
tral
Texa
s V
eter
ans
Hea
lth C
are
Sys
tem
Cen
tre
Col
labo
ratio
n fo
r Le
ader
ship
in A
pplie
d H
ealth
Res
earc
h an
d C
are
Eas
t Mid
land
s
Eas
t Mid
land
s P
atie
nt S
afet
y C
olla
bora
tive
Pal
mH
ealth
Fou
ndat
ion
Prim
ary
Car
e D
evel
opm
entC
entr
e
Texa
s A
&M
Uni
vers
ity C
olle
ge o
f Med
icin
e
Ret
ail/T
ech
/Ven
do
rs (
8)
AT
&T
Avi
zia,
Inc
.
Qua
ntum
Inno
vatio
n
Last
Bes
t Cha
nce,
LLC
Phi
lips
Hea
lthca
re (
2 m
embe
rshi
ps)
Res
tore
Med
ical
Sol
utio
ns
San
ofi
Sie
men
s
Insu
rers
/Pay
ers
(2)
Hig
hmar
k
Pas
spor
t Hea
lth P
lan
CH
OT
IN
DU
ST
RY
AD
VIS
OR
Y B
OA
RD
(IA
B)
Ind
ust
ry M
emb
ers
del
egat
e re
pre
sen
tati
ves
to
ser
ve
on
th
e C
HO
T IA
BT
he
IAB
:C
ontr
ibut
es to
the
Adv
ises
on
proj
ects
, ne
w u
nive
rsity
par
tner
s, in
dust
ry
mem
bers
, an
d pr
ojec
t vo
ting
and
sele
ctio
n
Mee
ts t
wic
e p
er y
ear
to c
ondu
ct C
HO
T b
usin
ess
IAB
ele
cts
Ch
airp
erso
ns
for
two
year
ter
mN
orm
a P
adro
n w
as e
lect
ed IA
B C
hair
Fal
l 201
7K
aran
Up
pal
was
ele
cte
d IA
B V
ice
Cha
ir F
all 2
017
CH
OT
AN
NU
AL
RE
SE
AR
CH
CY
CL
E
Fal
l C
HO
T
IAB
Mee
tin
gW
inte
rIA
B C
all
Sp
rin
g C
HO
T
IAB
Mee
tin
gS
um
mer
IAB
Cal
l
Co
mm
un
icat
eC
oo
rdin
ate
Co
llab
ora
teC
reat
e
Res
earc
h pr
ojec
t P
rese
ntat
ion
IAB
mem
bers
sh
are
rese
arch
id
eas
and
dev
elo
p
rese
arch
th
emes
Res
earc
h pr
opos
als
deve
lope
d w
ith I
AB
in
put.
CH
OT
site
s fa
cilit
ate
colla
bo
rati
ve
rese
arch
Pre
sent
res
earc
h p
rop
osa
lsIA
B p
rovi
des
feed
back
& r
anks
re
sear
ch p
ropo
sals
CH
OT
site
s co
nd
uct
res
earc
h,
dis
sem
inat
e&
p
ub
lish
2015
-201
6 C
HO
T P
roje
cts
& B
ud
get
sB
ita
2017
-201
8 R
ES
EA
RC
H Y
EA
R
Res
earc
h T
hem
esP
roje
cts
Acc
ess
to C
are
4 C
are
Coo
rdin
atio
n 4
Pop
ulat
ion
Hea
lth
3A
naly
tics
& H
IT
2
2017
-201
8 R
esea
rch
Bu
dg
et:
$1,0
60,0
00
13
2015
-201
6 C
HO
T P
roje
cts
& B
ud
get
s
13 R
esea
rch
Proj
ects
Re
sear
ch B
udge
t $1,06
0,00
0
2017
-201
8 C
HO
T P
roje
ct S
tatu
s &
Fin
ance
s
Tit
le%
Co
mp
lete
d$
Rem
ain
ing
Patie
ntEn
gage
men
t and
Hos
pita
l Rea
dmis
sion
s: T
he R
ole
of H
ealth
Lite
racy
25%
$45,
000
Impa
ctof
Dire
ct-t
o-Co
nsum
er T
elem
edic
ine
on D
owns
trea
m H
ealth
Car
e U
tiliza
tion
and
Cost
s15
%$5
0,00
0
Tele
heal
than
d Re
mot
e Pa
tient
Mon
itorin
g Sy
stem
s to
Impr
ove
Acce
ss a
nd
Prom
ote
Activ
e Pa
tient
Eng
agem
ent i
n Ru
ral C
omm
uniti
es35
%$1
00,0
00
Dev
elop
men
tof a
Mid
dlew
are
Fram
ewor
k fo
r Med
ical
Dev
ice
Inte
grat
ion
for
Tele
med
icin
e0%
$50,
000
Usi
ngCa
re C
oord
inat
ion
to A
ddre
ss C
ost,
Qua
lity,
and
Acc
ess t
o Ca
re A
cros
s Sy
stem
s an
d Po
pula
tions
80%
$250
,000
Effe
cts
of C
are
Coor
dina
tion
on th
e Im
prov
emen
t of Q
ualit
y of
Car
e40
%$6
5,00
0
Effe
cts
of C
are
Coor
dina
tion
on C
are
Tran
sitio
ns25
%$1
00,0
00
AM
obile
Bas
ed C
are
Coor
dina
tion
Syst
em fo
r Crit
ical
Car
e25
%$5
0,00
0
2015
-201
6 C
HO
T P
roje
cts
& B
ud
get
s
13 R
esea
rch
Proj
ects
Re
sear
ch B
udge
t $1,06
0,00
0
2017
-201
8 C
HO
T P
roje
ct S
tatu
s &
Fin
ance
s
Tit
le
% C
om
ple
ted
$
Rem
ain
ing
Inte
grat
ion
of P
opul
atio
n H
ealth
Dat
a an
d D
igita
l Ass
ista
nts t
o Re
duce
Re
adm
issi
on R
isks
10%
$50,
000
Gam
ifica
tion
and
its im
pact
on
the
Popu
latio
n H
ealth
Man
agem
ent o
f Chr
onic
Co
nditi
ons
10%
$50,
000
Impr
ovin
gEm
ploy
ee a
nd P
atie
nt H
ealth
thro
ugh
Popu
latio
n D
ata
Min
ing
10%
$50,
000
Dat
a-D
riven
Pred
ictiv
e An
alyt
ics
to Im
prov
e D
iagn
osis
, Tre
atm
ent,
Care
Co
ordi
natio
n, a
nd R
esou
rce
Util
izatio
n60
%$1
20,0
00
Mac
hine
Lea
rnin
g fo
r Evi
denc
e Ba
sed
Prac
tice,
Reso
urce
Allo
catio
n, a
nd R
isk
Pred
ictio
n60
%$8
0,00
0
CH
OT
Pu
bli
cati
on
s
17
10
1919
14
27
33
20
10
20
112
01
22
01
32
01
42
01
52
01
62
01
7
26
Tota
l Nu
mb
er o
f P
ub
licat
ion
s: 1
65
2017
Stu
den
t &
Fac
ult
y E
ng
agem
ent
Un
der
gra
du
ate
80
Do
cto
ral
36
Po
st -
Do
cto
ral
6
Mas
ter
18
Fac
ult
y 20
14
CH
OT
In
sig
hts
Tom
Fer
ris: C
HO
T T
houg
ht(s
)
(600
sq
ft)
Ble
nd T
AM
U in
volv
emen
t am
ong
Hea
lth S
cien
ces
and
Eng
inee
ring
Con
tinue
the
sta
ndar
d of
exc
elle
nce!
Bui
ld o
n C
HO
T
colla
bora
tions
and
like
-m
inde
d ce
nter
s fo
r la
rge-
scal
e pr
ojec
t gra
nts/
initi
ativ
esH
ealt
h &
H
um
an
S
yste
ms
En
gin
eeri
ng
8 fa
cu
lty,
40-5
0
stu
den
t re
sea
rch
ers
PR
OP
OS
AL
CU
LTIV
AT
ION
PR
OC
ES
S
Ste
p 1
: Id
entif
y re
sear
ch th
emes
from
dis
cuss
ions
with
IAB
mem
bers
at
the
IAB
Fal
l Mee
ting
Ste
p 2
: fa
cilit
ate
colla
bora
tion
acro
ss C
HO
T s
ites
and
amon
g in
dust
ry m
embe
rs
Ste
p 3
: C
ham
pion
s cr
eate
3-4
col
labo
rativ
e pr
ojec
t pro
posa
ls p
er
rese
arch
them
e to
pre
sent
at t
he IA
B S
prin
g M
eetin
g
Ste
p 4
: E
nsur
e th
at c
olla
bora
tive
proj
ect p
ropo
sals
add
ress
all
spec
ific
aim
s id
entif
ied
and
enco
mpa
ss IA
B r
esea
rch
inte
rest
s
All
colla
bora
tive
proj
ect
team
s in
clud
e r
esea
rche
rs fr
om m
ultip
le
CH
OT
uni
vers
ity s
ites
and
enga
ge m
ultip
le IA
B m
embe
rs b
ase
d on
sk
ill s
et,
natu
re o
f co
ntri
butio
n to
pro
ject
aim
s, a
nd c
apac
ity.
2018
20
19 R
esea
rch
The
mes
Pop
ulat
ion
Hea
lth
Car
e C
oord
inat
ion
Acc
ess
to C
are
Ana
lytic
s an
d In
nova
tive
Tech
nolo
gies
Pat
ient
Exp
erie
nce
15
2016
-201
7 P
roje
ct P
rop
osa
l V
etti
ng
Pro
cess
28B
rain
sto
rmin
g id
eas
for
pro
ject
p
rop
osa
ls w
ith
Ind
ust
ry m
emb
ers
(Jan
. 201
8) 13P
roje
ct p
rop
osa
ls v
ia
coo
rdin
ated
vet
tin
g b
y IA
B
Mem
ber
s an
d S
ite
Dir
ecto
rs
(Mar
ch 2
018)
2018
-201
9 P
roje
ct P
rop
osa
l V
etti
ng
Pro
cess
Co
mm
un
icat
eC
oo
rdin
ate
Co
llab
ora
teC
reat
e
2018
20
19 P
roje
ct P
ropo
sals Po
pu
lati
on
Hea
lth
Pro
po
sals
Po
p1:
Com
preh
ensi
ve A
naly
sis
on I
mpa
ct o
f S
ocia
l Det
erm
inan
ts to
Im
pro
ve C
are
Acr
oss
Pop
ulat
ions
Po
p 2
: P
artic
ipat
ing
in a
Com
mu
nity
Hea
lth Im
pro
vem
ent N
etw
ork
Po
p 3
: Th
e E
ffec
tiven
ess
of S
ubst
anc
e A
buse
Tre
atm
ent
Ser
vice
s in
Com
bat
ing
Op
ioid
Cris
is
Car
e C
oo
rdin
atio
n P
rop
osa
lsC
are
1:
Car
e C
oord
inat
ion
and
Pa
tient
Exp
erie
nce
acro
ss t
he C
ont
inuu
m o
f C
are:
A V
alue
Bas
ed
Rei
mbu
rsem
ent P
ersp
ectiv
e
Car
e 2
: D
evel
opin
g a
Ris
k P
redi
ctio
n M
odel
for
Hos
pita
l Acq
uire
d C
lost
ridiu
m D
iffic
ile In
fect
ion
An
alyt
ics
and
Inn
ova
tive
Tec
hn
olo
gie
s P
rop
osa
lsTe
ch 1
: In
tegr
ate
d C
hro
nic
Car
e M
ana
gem
ent
Sys
tem
for
Mo
nito
ring
and
Fac
ilita
ting
Com
preh
ensi
veC
linic
al D
eci
sio
n S
upp
ort
Tech
2:
Leve
ragi
ng T
ech
nol
ogy
to E
nha
nce
Com
mu
nica
tion
in H
ealth
care
Tech
3:
Dat
a-dr
iven
Ana
lytic
s an
d M
ach
ine
Lea
rnin
g fo
r Im
pro
ving
Hea
lthca
re O
utco
mes
2018
20
19 P
roje
ct P
ropo
sals Pat
ien
t E
xper
ien
ce P
rop
osa
lsP
atie
nt
1: C
are
Coo
rdin
atio
n A
ctiv
ities
fo
r In
div
idua
ls w
ith S
pina
l Cor
d In
jury
Pat
ien
t 2
: G
ene
ratin
g Ta
ilore
d R
eco
mm
enda
tions
Au
tom
atic
ally
with
Exp
lan
atio
ns v
ia a
n In
tera
ctiv
eD
ialo
g-ba
sed
Sys
tem
Pat
ien
t 3
: E
mb
eddi
ng R
outin
e In
form
al,
Fa
mily
Car
egiv
er A
sse
ssm
ent
of
De
liriu
m S
upe
rim
pos
ed o
n D
eme
ntia
into
Acu
te C
are
Acc
ess
to C
are
Pro
po
sals
Ac
ces
s 1
:Te
lem
edic
ine
in P
ract
ice:
Mu
ltidi
scip
linar
y U
tiliz
atio
n of
Te
lehe
alth
an
d R
emot
e P
atie
nt
Mon
itori
ng S
yste
ms
Ac
ces
s 2
:An
Inte
rven
tion
to A
ddre
ss H
ealth
Lite
racy
Ba
rrie
rs,
Incr
ease
Pat
ient
En
gage
men
t, a
nd
Impr
ove
Pa
tient
Exp
erie
nce
and
Out
com
es
CH
OT
WE
BS
ITE
ch
otn
sf.
org
16
Mem
ber
s O
nly
Sec
tio
nS
pri
ng
201
6 M
eeti
ng
Att
end
ance
Tota
l Reg
iste
red
Att
ende
es a
s o
f 04
.04.
18
Tota
l Att
end
ees:
61
Sp
rin
g 2
018
Mee
tin
g A
tten
dan
ce
Typ
eN
um
ber
of
Att
end
ees
Un
iver
sity
Sit
e A
dm
inis
trat
or
2
Fac
ult
y / R
esea
rch
er13
NS
F R
epre
sen
tati
ve1
IAB
Mem
ber
16
Gu
ests
6
Gra
du
ate
& U
nd
erg
rad
uat
e S
tud
ent
23
Mee
tin
g A
gen
da
ME
ET
ING
AG
EN
DA
Mee
tin
g A
gen
da
ME
ET
ING
AG
EN
DA
17
Thu
rsda
y, O
ctob
er 1
1 &
Fri
day,
Oct
ober
12,
201
8
Hos
ted
by G
eorg
ia I
nstit
ute
of T
echn
olog
y
at t
he M
ITR
E F
acili
ty in
McL
ean,
VA
Fal
l 201
8 IA
B M
EE
TIN
G
McL
ean
, VA
H.E
. Luc
cock
18
CH
OT
Res
earc
h Im
pac
t &
Insi
gh
t:E
ffect
iven
ess
of W
orkp
lace
Hea
lth
Pro
mot
ion
Pro
gram
sP
roje
ct L
eade
rs: G
eorg
e N
aufa
l
Co-
lead
ers:
Ohb
etC
heon
, Bit
aA
. Kas
h
Stud
ent
Nam
e(s)
: N/A
Enga
ged
IAB
mem
bers
: HEB
Enga
ged
CH
OT
Uni
vers
ity S
ites:
Texa
s A&
M
Uni
vers
ity
Pro
ject
Ove
rvie
w a
nd
Des
crip
tio
n
Ob
ject
ive:
•H
ow e
ffect
ive
is a
wor
kpla
ce h
ealth
pro
mot
ion
prog
ram
•P
rovi
des
empi
rical
evi
denc
e on
initi
ativ
es a
ddre
ssin
g he
alth
org
aniz
atio
n m
anag
emen
t an
d se
rvic
e
•E
valu
ate
diffe
rent
hea
lth p
rom
otio
n in
terv
entio
ns,
larg
e ad
min
istr
ativ
e da
ta
Par
tner
s / R
elev
ance
:
•P
rovi
des
empi
rical
evi
denc
e fr
om a
larg
e da
ta s
et
•In
vite
s ot
her
indu
stry
par
tner
s to
con
side
r si
mila
r in
itiat
ives
OB
JEC
TIV
ES
& P
AR
TN
ER
S: D
escr
iptio
n of
Pro
ject
Pro
ject
Ove
rvie
w a
nd
Des
crip
tio
n
Wh
at p
rob
lem
is
this
pro
ject
see
kin
g t
o
add
ress
?
•C
ost
of in
sura
nce
prem
ium
s an
d em
ploy
ee
com
pens
atio
n cl
aim
s co
ntin
ue t
o ris
e; w
orkp
lace
he
alth
pro
mot
ion
prog
ram
can
cut
dow
n th
ese
cost
s (b
oth
heal
th a
nd m
anag
emen
t co
sts)
•M
ost
full-
time
empl
oyee
s sp
end
mor
e th
an 1
/3 o
f th
eir
time
at th
e w
orkp
lace
OB
JEC
TIV
ES
& P
AR
TN
ER
S: D
esig
n T
hink
ing
App
roac
hP
roje
ct O
verv
iew
an
d D
escr
ipti
on
Res
earc
h P
ersp
ecti
ve
•A
hea
lth p
rogr
am p
rom
otio
n th
at a
ll st
akeh
olde
rs
bene
fit fr
om
–E
mpl
oyee
–be
tter
hea
lth o
utco
mes
and
low
er
heal
th p
rem
ium
s
–P
rovi
der –
heal
thie
r em
ploy
ees
(bet
ter
prod
uctiv
ity, l
ess
abse
ntee
ism
), lo
wer
hea
lth
and
man
agem
ent
cost
s
–H
ealth
sys
tem
–le
ss u
sage
of s
ervi
ces
OB
JEC
TIV
ES
& P
AR
TN
ER
S: D
esig
n T
hink
ing
App
roac
h
19
Pro
ject
Ove
rvie
w a
nd
Des
crip
tio
n
Co
ntr
ibu
tio
n t
o In
du
stry
an
d A
cad
emia
•E
mpi
rical
evi
denc
e fr
om a
larg
e em
ploy
er
•R
ich
data
set w
hich
incl
udes
fou
r bi
omet
ric in
dica
tors
: sy
stol
ic, d
iast
olic
, gl
ucos
e, a
nd c
hole
ster
ol
•E
xam
ine
four
diff
eren
t hea
lth p
rogr
am p
rom
otio
ns
–W
elln
ess
cour
se
–H
eart
hea
lth c
ours
e
–D
iabe
tes
prev
entio
n co
urse
–R
egis
tere
d D
ietit
ian
cons
ulat
ion
CO
NT
RIB
UT
ION
: H
ow is
thi
s di
ffere
nt t
han
rela
ted
rese
arch
?A
pp
roac
h
•W
hat a
ppro
ach
was
take
n?–
Was
not
invo
lved
in th
e de
sign
pha
se
–To
ok in
to c
onsi
dera
tion
of s
elec
tion
into
of
fere
d he
alth
pro
mot
ion
prog
ram
sto
def
ine
trea
tmen
t and
con
trol
gro
ups
–U
se q
uasi
-exp
erim
enta
l app
roac
h to
iden
tify
trea
tmen
t effe
ct o
f the
hea
lth p
rogr
am
AP
PR
OA
CH
: Exp
erim
enta
l App
roac
h
Ap
pro
ach
•O
verv
iew
of r
esul
ts•
Ove
rall,
hea
lth p
rogr
ams
offe
red
have
a
posi
tive
impa
ct o
n th
e st
udie
d he
alth
ou
tcom
es o
f the
em
ploy
ees
•A
lthou
gh ty
pica
lly u
nder
10%
, the
mag
nitu
de
of p
rogr
am e
ffect
s ra
nge
from
2%
to 2
7% in
te
rms
of im
prov
emen
t in
hea
lth o
utco
mes
OV
ER
VIE
W O
F R
ES
ULT
S:
Ap
pro
ach
•O
ffer
and
prom
ote
heal
th p
rogr
ams
at th
e w
orkp
lace
•Im
prov
es h
ealth
and
qua
lity
of li
fe o
f em
ploy
ees
•Im
prov
es p
rodu
ctiv
ity o
f em
ploy
ees
•Lo
wer
ove
rall
cost
s (m
anag
emen
t and
he
alth
)
OV
ER
VIE
W O
F R
ES
ULT
S: B
enef
its to
Indu
stry
20
CH
OT
Res
earc
h Im
pac
t &
Insi
gh
t:Im
pro
ve P
atie
nt
Car
e T
hro
ug
h
Clin
ical
Pro
cess
Op
tim
izat
ion
Pro
ject
Lea
ders
: Ev
a K
Lee
, PhD
.
Clin
ical
Lea
ders
: A
mel
ia A
. Lan
gsto
n, M
D, S
agar
Lon
ial,
MD
, Ed
mun
d K
. Wal
ler,
MD
, PhD
, FAC
P.
Eng
aged
IAB
Mem
bers
: Em
ory
Uni
vers
ity S
choo
l of
Med
icin
e, C
hild
ren’
s H
ealth
care
of A
tlant
a
Eng
aged
CH
OT
Uni
vers
ity
Site
s:
Geo
rgia
Inst
itute
of T
echn
olog
y
Stud
ent
Nam
e(s)
: T
ess E
. Bee
ler,
Bin
gy B
ao, R
ache
l B. D
efili
pp, R
ober
ts
Rod
rigu
ez, J
onat
han
C. Y
. Pan
g, A
bhin
av B
hard
waj
, Yifa
n W
ang
Pro
ject
Ove
rvie
w a
nd
Des
crip
tio
n
Ob
ject
ive:
•Im
prov
e pa
tient
exp
erie
nce,
sch
edul
ing,
utili
zatio
n, c
linic
al p
roce
sses
, and
prov
ider
s’ m
oral
e.
Par
tner
s / R
elev
ance
:
•O
verc
ome
the
inef
ficie
ncy
in r
esou
rces
and
staf
f util
izat
ion
to im
prov
e pa
tient
expe
rienc
e.
OB
JEC
TIV
ES
& P
AR
TN
ER
S:
Des
crip
tion
of P
roje
ct
Pro
ject
Ove
rvie
w a
nd
Des
crip
tio
n
Wh
at p
rob
lem
is t
his
pro
ject
see
kin
g t
o
add
ress
?•
The
over
all
auto
logo
ushe
mat
opoi
etic
stem
cell
tran
spla
ntat
ion
proc
ess
can
bech
alle
ngin
gan
dst
renu
ous
for
man
ypa
tient
sto
goth
roug
h
•P
roce
dure
ssu
chas
hem
aphe
resi
s(s
tem
cell
colle
ctio
n)ca
ncr
eate
bottl
enec
kdu
eto
patie
ntva
riabi
lity
•V
aria
bilit
yca
nca
use
inef
ficie
ncy
inre
sour
cean
dst
affu
tiliz
atio
n
OB
JEC
TIV
ES
& P
AR
TN
ER
S:
Des
ign
Thi
nkin
g A
ppro
ach
Pro
ject
Ove
rvie
w a
nd
Des
crip
tio
n
Res
earc
h P
ersp
ecti
ve
•P
atie
nts:
Red
uce
back
logs
, wai
ting
time.
•P
rovi
ders
: Im
prov
e ef
ficie
ncy
for
iden
tifie
dob
ject
ives
.
•H
ealth
sys
tem
s: Im
prov
e ov
eral
l res
ourc
eut
iliza
tion
and
allo
catio
n.
OB
JEC
TIV
ES
& P
AR
TN
ER
S:
Des
ign
Thi
nkin
g A
ppro
ach
21
Pro
ject
Ove
rvie
w a
nd
Des
crip
tio
n
Co
ntr
ibu
tio
n t
o In
du
stry
an
d A
cad
emia
A fr
amew
ork
for
the
heal
thca
re t
oef
ficie
ntly
allo
cate
re
sour
ces
and
staf
f bas
ed o
n th
e va
riabi
lity
of p
atie
nts.
CO
NT
RIB
UT
ION
: How
is th
is d
iffer
ent
than
rel
ated
res
earc
h?A
pp
roac
h
•O
ur a
ppro
ach:
1.In
terv
iew
rele
vant
stak
ehol
ders
inth
ebo
nem
arro
wtr
ansp
lant
proc
ess.
2.P
erfo
rmtim
e-m
otio
nst
udie
san
dob
serv
atio
nsto
esta
blis
hth
eB
MT
proc
ess
map
s.
3.A
naly
zeob
serv
atio
nal,
time-
mot
ion,
sche
dulin
g,ap
poin
tmen
tan
dse
rvic
edu
ratio
n,an
dpa
tient
data
toes
tabl
ish
patie
ntan
dse
rvic
ech
arac
teris
tics
and
syst
emic
inte
rdep
ende
ncie
s.
4.M
odel
and
optim
ize
sche
dule
san
db
edca
paci
ties
toal
ign
dem
and
and
supp
ly,e
qual
ize
utili
zatio
nan
dre
duce
over
time.
5.Im
plem
enta
ndev
alua
teac
hiev
edre
sults
.
AP
PR
OA
CH
: E
xper
imen
tal A
ppro
ach
Ap
pro
ach
•P
revi
ous
vs O
ptim
ized
Res
ults
•P
revi
ous
obse
rvat
ions
•W
eekl
y ut
iliza
tion
of 5
6% le
aves
muc
h ro
om f
or o
ptim
izat
ion
•O
nly
67.5
% o
f pa
tient
s ar
e ab
le to
fin
ish
colle
ctio
n in
one
day
•S
ched
ulin
g fo
r pa
tient
s w
ho n
eed
mul
tiple
day
s to
col
lect
targ
et s
tem
cel
l qu
antit
ies
is n
ot d
one
in a
dvan
ce
•C
onsi
dera
ble
stan
dard
dev
iatio
n be
twee
n ut
iliza
tion
on w
eekd
ays
resu
lts in
ei
ther
res
ourc
e/st
aff
shor
tage
or
exce
ss
•T
he n
eed
for
Sat
urda
y co
llect
ions
res
ults
in u
nnec
essa
ry s
taff
over
time
•O
ur o
ptim
ized
res
ults
•P
atie
nt S
tem
Cel
l Col
lect
ion
Dat
a ac
quire
d fo
r al
l pat
ient
s be
fore
and
afte
r im
plem
enta
tion
of t
he o
ptim
ized
sch
edul
ing
mod
el
•A
vera
ge u
tiliz
atio
n in
crea
sed
from
56%
to
92%
•N
o ov
ertim
e st
affin
g on
Sat
urda
y
•30
% in
crea
sed
patie
nt t
hrou
ghpu
t pe
rmitt
ed b
y ne
w s
ched
ulin
g m
odel
OV
ER
VIE
W O
F R
ES
ULT
S:
Ap
pro
ach
OV
ER
VIE
W O
F R
ES
ULT
S:
0
0.2
0.4
0.6
0.81
1.2
Mon
Tue
sW
edT
hur
Fri
Sat
Ori
gin
al U
tiliz
atio
n
00.
10.
20.
30.
40.
50.
60.
70.
80.
91
Mon
Tue
sW
edT
hur
Fri
Sat
Th
eore
tica
l Uti
lizat
ion
0
0.2
0.4
0.6
0.81
1.2
Mon
Tue
sW
edT
hur
Fri
Sat
Imp
lem
ente
d U
tiliz
atio
n
Orig
inal
util
izat
ion
vs th
eore
tical
(op
timiz
ed)
vs im
plem
ente
d re
sults
22
Ap
pro
ach
•A
fram
ewor
k th
at b
uild
s pr
oces
s m
aps,
ana
lyze
s da
taan
dop
timiz
es s
ched
ules
to:
–Im
prov
e sc
hedu
ling
capa
city
.
–R
educ
e pa
tient
bac
klog
s an
d w
ait-
time.
–In
crea
se r
esou
rces
util
izat
ion.
–A
void
unn
eces
sary
sch
edul
es.
•B
oth
patie
nts
and
heal
thca
re p
rovi
ders
wou
ld b
enef
it fr
om th
e im
prov
emen
t of
the
serv
ice
deliv
ery
effic
ienc
y.
OV
ER
VIE
W O
F R
ES
ULT
S:
Ben
efits
to In
dust
ry
23
24
An
Un
sup
ervi
sed
Mac
hin
e L
earn
ing
M
eth
od
fo
r D
isco
veri
ng
Pat
ien
t C
lust
ers
Bas
ed o
n G
enet
ic S
ign
atu
res
Pro
ject
Lea
ders
: Con
rad
Tuck
er, P
h.D
.
Co-
Lead
ers:
Chr
isto
pher
DeF
litch
, MD
; Gre
g Le
wis
, Ph.
D.
Stud
ent
Nam
e(s)
: Chr
istia
n Lo
pez,
Scot
t Tuc
ker, T
arik
Sal
ameh
Enga
ged
IAB
mem
bers
: Sie
men
s, H
ighm
ark,
Her
shey
Med
ical
Cen
ter,
AT
&T
Enga
ged
CH
OT
Uni
vers
ity S
ites:
Penn
Sta
te U
nive
rsity
Pro
ject
Ove
rvie
w a
nd
Des
crip
tio
n
Ob
ject
ive:
•D
evel
op n
ovel
mac
hine
lear
ning
te
chni
ques
for
iden
tifyi
ng a
ctio
nabl
e ge
nom
ic s
imila
ritie
s am
ong
patie
nts
with
chr
onic
imm
une
dise
ases
.
Par
tner
s / R
elev
ance
:
•R
educ
e tr
ial-a
nd-e
rror
trea
tmen
t co
sts
and
impr
ove
patie
nt o
utco
mes
.
OB
JEC
TIV
ES
& P
AR
TN
ER
S: D
escr
iptio
n of
Pro
ject
Pro
ject
Ove
rvie
w a
nd
Des
crip
tio
nW
hat
pro
ble
m i
s th
is p
roje
ct s
eeki
ng
to
ad
dre
ss?
•C
hron
ic im
mun
e di
sord
ers
man
ifest
diff
eren
tly
from
pat
ient
to
patie
nt,
yet
have
a g
enet
ic
etio
logy
.
•E
nhan
ced
unde
rsta
ndin
g of
the
gen
etic
un
derp
inni
ngs
can
impr
ove
dise
ase
trea
tmen
t.
OB
JEC
TIV
ES
& P
AR
TN
ER
S: D
esig
n T
hink
ing
App
roac
h
un
Pro
ject
Ove
rvie
w a
nd
Des
crip
tio
nR
esea
rch
Per
spec
tive
•P
atie
nt –
enha
nced
car
e
•P
rovi
der
–in
form
ed tr
eatm
ent d
ecis
ions
•H
ealth
sys
tem
–im
prov
ed p
atie
nt
outc
omes
•P
ayer
–ef
ficie
nt p
atie
nt-b
ased
trea
tmen
t
OB
JEC
TIV
ES
& P
AR
TN
ER
S: D
esig
n T
hink
ing
App
roac
h
25
Pro
ject
Ove
rvie
w a
nd
Des
crip
tio
n
Co
ntr
ibu
tio
n t
o In
du
stry
an
d A
cad
emia
•A
pre
limin
ary
step
to
iden
tifyi
ng g
enet
ical
ly-
dist
inct
pat
ient
coh
orts
with
in t
he p
opul
atio
n su
fferin
g fr
om c
hron
ic im
mun
e di
sord
ers.
CO
NT
RIB
UT
ION
: H
ow is
thi
s di
ffere
nt t
han
rela
ted
rese
arch
?A
pp
roac
h•
Hie
rarc
hica
l uns
uper
vise
d m
achi
ne le
arni
ng is
em
ploy
ed t
o im
mun
ochi
p da
ta c
olle
cted
fro
m a
M
ultip
le S
cler
osis
pat
ient
coh
ort.
•S
igni
fican
t ge
netic
clu
ster
s ar
e de
fined
by
the
algo
rith
m a
nd p
atie
nts
are
assi
gned
to
thes
e cl
uste
rs b
ased
on
thei
r ge
netic
mak
eup.
AP
PR
OA
CH
: Exp
erim
enta
l App
roac
h
Ap
pro
ach
OV
ER
VIE
W O
F R
ES
ULT
S:
•2
dist
inct
clu
ster
s ar
ise
from
the
pat
ient
coh
ort.
•G
ene
Pat
hway
Ana
lysi
s re
veal
s si
gnifi
cant
di
ffere
nces
bet
wee
n th
e cl
uste
rs in
cel
lula
r ad
hesi
on,
cyto
kine
res
pons
e, a
nd im
mun
e pr
oces
s pa
thw
ays.
Ap
pro
ach
•C
lust
ers
iden
tifie
d in
the
patie
nt
coho
rt c
an b
e as
sess
ed fo
r lin
ks
to c
linic
al o
utco
mes
suc
h as
:–
Age
of
onse
t of
dis
ease
–R
espo
nse
to t
reat
men
t
–S
peci
fic d
isea
se c
ours
e
•A
ccur
ate
pred
ictio
n of
thes
e ou
tcom
es w
ill im
prov
e ov
eral
l di
seas
e ou
tcom
es a
nd id
entif
y ap
prop
riate
trea
tmen
t str
ateg
ies.
OV
ER
VIE
W O
F R
ES
ULT
S: B
enef
its to
Indu
stry
26
Pro
gra
ms
to R
edu
ce M
edic
atio
n-
Rel
ated
Rea
dm
issi
on
sP
roje
ct L
eade
rs:
Nat
han
Car
roll,
Ph.D
., M
HA
Stud
ent
Nam
e(s)
: R
eena
Jose
ph, M
HA
Nee
raj P
uro,
MH
A
Eng
aged
IAB
mem
bers
: H
ealth
Sout
h
Eng
aged
CH
OT
Uni
vers
ity
Site
s: U
nive
rsity
of A
laba
ma
at B
irm
ingh
am
Pro
ject
Ove
rvie
w a
nd
Des
crip
tio
n
Ob
ject
ive:
•To
iden
tify
inte
rven
tions
tha
t ha
ve b
een
succ
essf
ul in
red
ucin
g m
edic
atio
n-re
late
d re
adm
issi
ons
Par
tner
s / R
elev
ance
:
•E
stim
ated
13-
20%
of
read
mis
sion
s ar
e ca
used
by
med
icat
ion-
rela
ted
fact
ors
•R
educ
ing
read
mis
sion
s im
prov
es t
he q
ualit
y of
ca
re a
nd r
educ
es p
rovi
der
pena
lties
OB
JEC
TIV
ES
& P
AR
TN
ER
S: D
escr
iptio
n of
Pro
ject
Pro
ject
Ove
rvie
w a
nd
Des
crip
tio
n
Wh
at p
rob
lem
is t
his
pro
ject
see
kin
g t
o
add
ress
?•
Fro
m 2
007-
2015
rea
dmis
sion
rat
es d
eclin
ed
from
22%
to 1
8%–
Man
y pr
ovid
ers
are
impl
emen
ting
prac
tices
to
redu
ce r
eadm
issi
ons
–R
eadm
issi
on r
ates
are
stil
l rel
ativ
ely
high
OB
JEC
TIV
ES
& P
AR
TN
ER
S: D
esig
n T
hink
ing
App
roac
h
Pro
ject
Ove
rvie
w a
nd
Des
crip
tio
n
Res
earc
h P
ersp
ecti
ve
•R
esea
rch
focu
ses
on in
terv
entio
ns
impl
emen
ted
by h
ospi
tals
•P
atie
nt p
ersp
ectiv
e is
impo
rtan
t but
isn’
t ad
dres
sed
in m
ost o
f the
stu
dies
we
foun
d
OB
JEC
TIV
ES
& P
AR
TN
ER
S: D
esig
n T
hink
ing
App
roac
h
27
Pro
ject
Ove
rvie
w a
nd
Des
crip
tio
n
Co
ntr
ibu
tio
n t
o In
du
stry
an
d A
cad
emia
•F
ocus
on
med
icat
ion-
rela
ted
caus
es o
f re
adm
issi
ons
•D
isse
min
ate
wha
t aca
dem
ics
have
lear
ned
abou
t rea
dmis
sion
inte
rven
tions
to m
anag
ers
•Id
entif
y ba
rrie
rs to
usi
ng a
cade
mic
res
earc
h in
pra
ctic
e
CO
NT
RIB
UT
ION
: H
ow is
thi
s di
ffere
nt t
han
rela
ted
rese
arch
?
Ap
pro
ach
•Li
tera
ture
rev
iew
-22
stu
dies
iden
tifie
d–
Pee
r-re
view
ed e
mpi
rical
stu
dies
fro
m a
cade
mic
da
taba
ses
(Sco
pus,
Pub
Med
, C
INA
HL)
pu
blis
hed
sinc
e 20
00
–C
onta
ins
MeS
Hte
rms
rela
ted
to
•“M
edic
atio
n ad
here
nce”
or
“dru
g-re
late
d si
de e
ffect
s”
AN
D “
read
mis
sion
s”
–In
clus
ion
crite
ria•
The
stu
dy d
escr
ibes
an
inte
rven
tion
•Ta
rget
s m
edic
atio
n-re
late
d fa
ctor
s le
adin
g to
re
adm
issi
ons
AP
PR
OA
CH
: Exp
erim
enta
l App
roac
h
Ap
pro
ach
•7
inte
rven
tions
res
ulte
d in
a s
igni
fican
t re
duct
ion
in r
eadm
issi
ons
•M
any
used
com
bina
tion
of c
ompo
nent
s (d
/c
plan
ning
, ed
ucat
ion,
pos
t d/
c ph
one
calls
, pa
tient
cou
nsel
ling,
med
rec
onci
liatio
n)
•S
ever
al o
nly
used
a c
are
coor
dina
tor
assi
gned
to
impr
ove
coor
dina
tion
and
com
mun
icat
ion
betw
een
prog
ram
co
mpo
nent
s an
d cl
inic
ians
•C
are
coor
dina
tion
seem
s ef
fect
ive
OV
ER
VIE
W O
F R
ES
ULT
S:
Ap
pro
ach
•D
eliv
erab
le:
Writ
ten
repo
rt th
at:
–D
etai
ls e
valu
ated
effo
rts
to r
educ
e m
edic
atio
n-re
late
d re
adm
issi
ons
–S
ynth
esiz
es w
hat
we
know
abo
ut s
ucce
ssfu
l in
terv
entio
ns
•Im
port
ance
: –
Man
y pr
ovid
ers
have
suc
cess
fully
red
uced
re
adm
issi
ons
usin
g a
varie
ty o
f st
rate
gies
–In
cent
ives
to
redu
ce t
hem
fur
ther
rem
ain
larg
e (H
RR
P, b
undl
ed p
aym
ent)
OV
ER
VIE
W O
F R
ES
ULT
S: B
enef
its to
Indu
stry
28
Med
ical
Imag
ing
Lo
gfi
les:
Big
Dat
a A
nal
ysis
& F
ind
ing
sP
roje
ct L
eade
rs:
Chr
istin
a M
astr
ange
lo
Co-
lead
ers:
Chr
isto
pher
Hal
l –Ph
ilips
Hea
lthca
reR
ebec
ca J.
Mie
losz
yk –
Phili
ps H
ealth
care
Stud
ent
Nam
e:
Lari
ssa
P. G
. Pet
roia
nu
Eng
aged
IAB
mem
bers
: Ph
ilips
Hea
lthca
re
Eng
aged
CH
OT
Uni
vers
ity
Site
s: U
nive
rsity
of W
ashi
ngto
n
Pro
ject
Ove
rvie
w a
nd
Des
crip
tio
n
Bac
kgro
un
d a
nd
Rel
evan
ce:
•M
RI
mac
hine
s ar
e im
port
ant
for
dete
ctin
g or
mon
itorin
g sp
ecifi
c he
alth
co
nditi
ons.
•M
RI
is a
hig
h co
st p
roce
dure
: ab
out
$600
/hou
r in
the
US
•W
ides
prea
d ef
fect
: •
39 M
RI m
achi
nes
per
1,00
0,00
0 in
habi
tant
s1~
12,
519
mac
hine
s
•11
7.8
MR
I exa
ms
per
1,00
0,00
0 in
habi
tant
s pe
r ye
ar2
~ 3
7,81
3.8
exam
s/ye
ar
•D
iffer
ent
fact
ors
can
influ
ence
in w
aste
d sc
anne
r tim
e, fo
r ex
ampl
e:–
Poo
r sc
hedu
ling,
faile
d/lo
w v
alue
seq
uenc
es, o
r id
lene
ss
–H
uman
fact
ors
, suc
h a
s tr
ans
por
t, m
ove
me
nt o
f pat
ient
s an
d te
chni
cian
s
•F
ind
a w
ay
to d
ecre
ase
the
exam
tim
e an
d its
var
iabi
lity
will
–re
duce
the
cost
s of
the
exa
m
–im
prov
e th
e pa
tient
’s e
xper
ienc
e
•D
evel
op a
met
hodo
logy
tha
t can
gen
eral
ize
to o
ther
site
s.
OB
JEC
TIV
ES
& P
AR
TN
ER
S: D
escr
iptio
n of
Pro
ject
1 ht
tps:
//dat
a.oe
cd.o
rg/h
ealth
eqt/m
agne
tic-r
eson
ance
-imag
ing
-mri-
units
.htm
-03
/26/
2018
2 ht
tps:
//dat
a.oe
cd.o
rg/h
ealth
care
/mag
netic
-res
onan
ce-im
agin
g-m
ri-ex
ams.
htm
-03
/26/
2018
Pro
ject
Ove
rvie
w a
nd
Des
crip
tio
n
Wh
at p
rob
lem
is t
his
pro
ject
see
kin
g t
o
add
ress
?•
Hig
h va
riabi
lity
in t
ime
and
quan
tity
of s
eque
nces
for
si
mila
r ex
ams
whi
ch a
ffect
s th
e pa
tient
exp
erie
nce
and
cost
of M
RI p
roce
dure
s.
•O
bje
ctiv
es:
–Id
entif
y an
d m
easu
re “
was
ted”
tim
e in
MR
I exa
ms.
–A
scer
tain
the
root
cau
ses
of th
e no
n-pr
oduc
tive
time.
–Id
entif
y ke
y m
easu
res.
–U
se lo
g fil
es a
s th
e so
urce
of d
ata
to a
naly
ze d
urat
ion
and
varia
bilit
y.
OB
JEC
TIV
ES
& P
AR
TN
ER
S: D
esig
n T
hink
ing
App
roac
hP
roje
ct O
verv
iew
an
d D
escr
ipti
on
Res
earc
h P
ersp
ecti
ve
OB
JEC
TIV
ES
& P
AR
TN
ER
S: D
esig
n T
hink
ing
App
roac
h
Pat
ient
Re
du
ctio
nin
du
ratio
n o
f th
e
exa
m
Kn
ow
led
ge o
f th
e e
xpe
cte
d
du
ratio
n
Pro
vide
rR
ed
uct
ion
of
cost
Re
du
ctio
n o
f tim
eB
ett
er
sch
ed
ule
p
lan
nin
g
Hea
lth
syst
emL
ess
va
riab
ility
Re
du
ctio
n o
f tim
e
Pay
erR
ed
uct
ion
of
cost
29
Pro
ject
Ove
rvie
w a
nd
Des
crip
tio
n
Co
ntr
ibu
tio
n t
o In
du
stry
an
d A
cad
emia
•U
nder
stan
d ho
w t
o an
alyz
e in
form
atio
n ex
istin
g in
MR
log
files
.
•Im
prov
e pa
tient
’s e
xper
ienc
e.
•Id
entif
y ca
uses
of
varia
bilit
y.
•Id
entif
y im
prov
emen
ts t
o re
duce
“w
aste
d” t
ime
and
incr
ease
pro
cess
qua
lity.
•A
pply
to
othe
r ra
diol
ogy
area
s, s
uch
as C
T
CO
NT
RIB
UT
ION
: H
ow is
thi
s di
ffere
nt t
han
rela
ted
rese
arch
?A
pp
roac
hA
PP
RO
AC
H: E
xper
imen
tal A
ppro
ach
Und
erst
and
the
log
files
Und
erst
and
the
proc
ess
Pre
-pr
oces
sing
th
e da
ta
Cor
rela
tion
anal
ysis
Out
lier
dete
ctio
nP
redi
ctio
n m
odel
s
Site
s co
mpa
rison
Iden
tify
ing
Co
rrel
ates
OV
ER
VIE
W O
F R
ES
ULT
S
•V
aria
ble
s re
late
d t
o p
atie
nt’
s ch
arac
teri
stic
s an
d d
ura
tio
ns
•U
nex
pec
ted
lac
k o
f co
rrel
atio
n b
etw
een
pat
ien
t’s
char
acte
rist
ics
and
du
rati
on
•E
xpec
ted
co
rrel
atio
n b
etw
een
rep
eate
d s
can
s an
d c
on
tras
t w
ith
exa
m d
ura
tio
n
Exa
m D
ura
tio
n a
nd
Rat
io o
f R
epea
ted
Seq
uen
ceb
y A
nat
om
y an
d S
ite
Sho
rtes
tLo
nges
t
•L
arg
e va
riab
ility
exi
sts
bet
wee
n s
ites
.•
Sit
e 3
has
bet
ter
met
rics
an
d w
ill b
e an
alyz
ed a
s a
ben
chm
ark.
30
Fra
ctio
n o
f R
epea
ted
Seq
uen
ces
by
An
ato
my
OV
ER
VIE
W O
F R
ES
ULT
S
•‘B
rain
’ an
d ‘
Hea
d’ h
ave
mo
re e
xam
s, h
ow
ever
are
mo
re c
on
sist
ent
wit
h f
ew r
epea
ted
sc
an s
equ
ence
s.•
‘Ab
do
men
’ an
d ‘
Liv
er’ h
ave
hig
h r
atio
of
rep
etit
ion
an
d i
s an
op
po
rtu
nit
y fo
r im
pro
vem
en
t.
Ap
pro
ach
•D
efin
ed 3
key
met
rics
to a
naly
ze w
aste
:–
% o
f rep
eate
d se
quen
ces
(sca
ns)
–E
xam
dur
atio
n
–Id
le ti
me
•P
atie
nt p
hysi
olog
ical
cor
rela
tes
are
cont
rary
to w
hat
was
ex
pect
ed;
they
do
not s
igni
fican
tly a
ffect
dur
atio
ns.
•S
igni
fican
t va
riabi
lity
exis
ts w
ithin
the
sam
e ex
am c
ard
sequ
ence
s be
twee
n th
e di
ffere
nt s
ites.
•D
iffer
ent p
redi
ctiv
e m
odel
s ha
ve b
een
expl
ored
to
anal
yze
the
data
, but
no
sign
ifica
nt r
esul
ts y
et.
•T
he b
igge
st o
ppor
tuni
ty t
o re
duce
exa
m d
urat
ion
is b
y re
duci
ng t
he n
umbe
r of
rep
eate
d sc
ans.
OV
ER
VIE
W O
F R
ES
ULT
S: B
enef
its to
Indu
stry
31
32
RESEARCH PROPOSALS
Research Theme #1:
Population Health
Research Theme #2:
Care Coordination
Research Theme #3:
Analytics and Innovative Technologies
Research Theme #4:
Patient Experience
Research Theme #5:
Access to Care
33
34
Research Theme #1:
POPULATION HEALTH
Theme Champion: Dr. Conrad Tucker, Pennsylvania State University
Pop 1 Comprehensive Analysis on Impact of Social Determinants to Improve Care Across Populations Pop 2 Participating in a Community Health Improvement Network Pop 3 The Effectiveness of Substance Abuse Treatment Services in Combating Opioid Crisis
35
36
PI:
Comprehensive Analysis on Impact of Social Determinants to Improve Care Across Populations
Pop1 Borkowski, Mastrangelo, Ferris, Kash
Population Health
$150,000 YES 1
Social and environmental determinants, such as economic stability, housing and physical environment, food access, communitysupport, and availability of health care systems impact health outcomes and the satisfaction of patients and care providers. Initiativesare under way to improve our understanding of how these social factors can influence the organization and delivery of care to patientcommunities, targeting the broad population as well as focused efforts on those with certain medical conditions. However, only limitedresearch has focused on the relationship between social determinants and healthcare for specific sub-populations, such as militaryveterans, pediatric populations, and those with histories of readmission or repeated ED visits. Additionally, there is value in morecompletely understanding how these social factors interact with health conditions in regard to patient satisfaction and HCAHPSscores. This collaborative project seeks to address these aims and broaden the knowledge base concerning the effects of socialdeterminants on mental and physical health and the utilization of available healthcare.
This proposed research focuses on identifying the social factors that contribute to the mental and physical health of individuals fromtarget populations (veterans, children, recurrent ED visitors). The understanding of these factors can inform ways of improving theeffectiveness of care and the patient experience. This research will support efforts to extend patient care beyond clinical parameters,to understand the effects of social demographics on patient experience through HCHAPS, and to orient clinical practice for improvedpatient experience and reduction in the cost of care.
We will obtain data from the UAB enterprise data warehouse (EDW) for all patients who were discharged between 2016 - 2017.Notably, these data will include Social Determinants of Health (SDH). Data will also be obtained from patients enrolled in the CentralTexas Veteran's Health Care system that are enrolled in their integrated health program. The compiled database will be analyzed viamixed methods to determine significant direct and interacting social and medical factors impacting healthcare utilization. This analysiswill inform the development of a predictive model that can be evaluated in simulation.
Month 1-2: Obtain IRB approvalMonth 3-6: Distinguish relationships among social determinants and health variables, obtain/merge/develop databases that includeboth mental and physical health flags, process data and produce descriptive analyticsMonth 7-8: Develop predictive model and evaluate in simulation, iteratively refine model.Month 9-12: Write report to include results for each target population
1) Predictive models for readmission, which incorporate socialand medical factors, can be used to inform and guide careplans.2) Access to databases for selected patient groups.
1) Publication of results of the analysis, including the identificationof social determinants affecting care need and readmissions.2) Publication of the developed predictive model to assesslikelihood of repeat ED visits and readmissions.3) Access to developed databases that integrates both mentalhealth and clinical factors for selected patient groups.
37
Com
preh
ensi
ve A
naly
sis
on Im
pact
of
Soc
ial D
eter
min
ants
to Im
prov
e C
are
Acr
oss
Popu
latio
nsP
roje
ct L
eade
r: T
hom
as F
erri
s, P
hD
Co-
lead
ers:
Chr
isti
na M
astr
ange
lo, P
hD, M
SN
ancy
Bor
kow
ski,
DB
AB
ita
Kas
h, P
hD, M
BA
, FA
CH
ESt
uden
t N
ame(
s):
Enga
ged
IAB
mem
bers
: Vet
eran
s Adm
inis
trat
ion
Enga
ged
CH
OT
Uni
vers
ity S
ites:
Cen
tral
Tex
as
Vete
rans
Hea
lthca
re, S
eatt
le C
hild
ren’
s H
ospi
tal,
Mai
n Li
ne H
ealth
Proj
ect O
verv
iew
and
Des
crip
tion
Obj
ectiv
es:
Dev
elop
pre
dict
ive
mod
els
to e
xam
ine
the
rela
tions
hip
betw
een
soci
al fa
ctor
s an
d th
e lik
elih
ood
of e
mer
genc
y ro
om
visi
ts a
nd re
peat
hos
pita
lizat
ions
.
Dev
elop
a d
atab
ase
that
incl
udes
bot
h m
enta
l and
phy
sica
l hea
lth fa
ctor
s fo
r ta
rget
pop
ulat
ions
.
OB
JEC
TIV
ES
& P
AR
TNE
RS
: Des
crip
tion
of P
roje
ct
Part
ners
/ R
elev
ance
:Th
e m
odel
s w
ill pr
ovid
e in
sigh
t int
o se
rvic
e ne
eds
that
mig
ht n
ot o
ther
wis
e be
iden
tifie
d fo
r vet
eran
and
ped
iatri
c po
pula
tions
.
The
deve
lope
d da
taba
se w
ill su
ppor
t tra
ckin
g pa
tient
pro
gres
s an
d an
alys
is
of s
ocia
l fac
tors
.
Proj
ect O
verv
iew
and
Des
crip
tion
Wha
t pro
blem
doe
s th
is p
roje
ct a
ddre
ss?
•Fe
w h
ealth
car
e m
odel
s in
corp
orat
e m
edic
al a
nd s
ocia
lfac
tors
in p
redi
ctin
g un
met
pat
ient
car
e ne
eds
•N
eed
to c
onsi
der b
oth
phys
ical
and
men
tal c
are
need
s•
Dev
elop
ed m
odel
s w
ill su
ppor
t ana
lysi
s of
com
plex
rela
tions
hips
am
ong
soci
al d
emog
raph
ics,
phy
sica
l and
men
tal h
ealth
, and
hea
lthca
re u
tiliz
atio
n
OB
JEC
TIV
ES
& P
AR
TNE
RS
: Des
ign
Thin
king
App
roac
h
Hea
lthca
re U
tiliz
atio
n
Proj
ect O
verv
iew
and
Des
crip
tion
Res
earc
h Pe
rspe
ctiv
eTh
e m
odel
s w
ill he
lp c
are
prov
ider
s id
entif
y pa
tient
s w
ho
have
gre
ater
nee
d fo
r car
e se
rvic
es, b
ased
on
soci
al a
nd
med
ical
fact
ors
OB
JEC
TIV
ES
& P
AR
TNE
RS
: Des
ign
Thin
king
App
roac
h
The
pred
ictiv
e m
odel
s w
ill al
so
be u
sed
to e
valu
ate
way
s to
re
duce
cos
ts, p
reve
nt
read
mis
sion
s, a
nd in
crea
se
patie
nt s
atis
fact
ion
38
Proj
ect O
verv
iew
and
Des
crip
tion
Con
trib
utio
n to
Indu
stry
and
Aca
dem
ia•
Con
tribu
te to
und
erst
andi
ng o
f com
plex
re
latio
nshi
p am
ong
inte
ract
ing
soci
al a
nd
med
ical
fact
ors
in h
ealth
care
util
izat
ion
•P
redi
ctiv
e m
odel
s of
hea
lth o
utco
mes
, lik
elih
ood
of re
adm
issi
on c
an b
e us
ed to
in
form
and
gui
de c
are
plan
s
CO
NTR
IBU
TIO
N: H
ow is
this
diff
eren
t tha
n re
late
d re
sear
ch?
App
roac
h•
UA
BH
S E
nter
pris
e D
ata
War
ehou
se (c
linic
al,
soci
al, a
nd e
cono
mic
var
iabl
es) f
or 2
016
and
2017
•M
ixed
ana
lytic
al m
etho
ds to
unc
over
re
latio
nshi
ps a
mon
g so
cial
det
erm
inan
ts a
nd
med
ical
fact
ors
in h
ealth
care
util
izat
ion
•D
evel
op a
nony
miz
ed p
atie
nt-le
vel d
atab
ases
to
supp
ort p
redi
ctiv
e m
odel
ing
•D
evel
op p
redi
ctiv
e m
odel
of r
eadm
issi
on/ E
D
visi
t fre
quen
cy a
nd e
valu
ate
in s
imul
atio
n
AP
PR
OA
CH
: Exp
erim
enta
l App
roac
h
App
roac
hO
VE
RV
IEW
OF
RE
SU
LTS
: Exp
ecte
d M
ilest
ones
Mon
ths
1-2
Obt
ain
IRB
ap
prov
al
Mon
ths
3-6
Dis
tingu
ish
rela
tions
hips
, de
velo
p da
taba
ses,
pr
oces
s da
ta
Mon
ths
7-8
Dev
elop
and
ev
alua
te
pred
ictiv
e m
odel
Mon
ths
9-12
W
rite
repo
rts
incl
udin
g re
sults
for e
ach
targ
et
popu
latio
n
App
roac
h
•P
ublic
atio
n(s)
and
pre
sent
atio
n(s)
:–
Res
ults
of a
naly
sis,
iden
tific
atio
n of
sig
nific
ant
soci
al fa
ctor
s as
soci
ated
with
car
e ne
ed a
nd
read
mis
sion
beh
avio
r–
Rep
ort o
n de
velo
pmen
t and
eva
luat
ion
of
pred
ictiv
e m
odel
for l
ikel
ihoo
d of
repe
at
hosp
italiz
atio
ns•
Acc
ess
to d
atab
ases
that
inte
grat
e bo
th
men
tal h
ealth
and
clin
ical
fact
ors
for
sele
cted
pat
ient
s
OV
ER
VIE
W O
F R
ES
ULT
S: B
enef
its to
Indu
stry
39
40
PI:
Participating in a Community Health Improvement Network
Pop2 Judah Thornewill
Population Health
$90,000 NO 1
The project participants will participate in one or more innovative "consumer-directed" community health information sharing networksdesigned to improve ability for people and organizations to more easily share health and well-being data to improve care coordinationand support research across the community. The project will explore implementation and use of several innovations: 1) amulti-stakeholder, community-based governance and oversight structure; 2) consumer-directed health information exchange leveragingHIPAA individual right of access mechanisms; 3) a technology platform supporting safe, secure, encrypted data sharing among people,providers, plans, apps, AI firms, and researchers; 4) a sustainable business model tied to reducing costs and improve quality andoutcomes; and 5) potential to generate millions in new private-sector led investments from sales of equity and a blockchain baseddigital currency.
1. Organizing meeting2. 3-4 documents describing best practices and lessons learned.3. Formal presentation in month 9.
The CHIN project will research a new paradigm of health information sharing that puts the person (patient/consumer) in the center ofinformation sharing about them, and also uses next generation security technologies, including block chain, to protect and enablepersonal health information sharing. The CHIN will engage public and private-sector organizations to advance this emerging form ofsecure "consumer-directed exchange." The project will focus on use cases with potential for high impact and ROI, emphasizingpopulations with type-II diabetes and opiate addiction, but also looking at other chronically ill or under-served populations with carecoordination challenges.
CHOT members will be invited to actively participate in the CHIN, working alongside embedded "action researchers" from UofL.Action-oriented research will focus on developing best practice recommendations for: 1) goverance of the CHIN; 2) organizationalparticipation agreements with the CHIN; 3) engaging "connectors" at providers, plans, social agencies and other offices to helppatients/consumers sign-up with the system; 4) obtaining informed consent to access and re-share data with members of the CHIN; 5)conducting searches of data, starting with type II diabetes; and 6) potential to expand nationwide - in a network of CHINs, potentiallyincluding other CHOT sites.
Month 2: CHOT members invited to participate in CHIN process. Getting organized.Month 6-9: First "connectors" trained and certified. Taught how to use system.Month 6-9: "Dummy" and real patients enrolled; data moved. Experience evaluated.Month 6-9: Encrypted data-search tested with selected sources.Month 6-9: Presentation of results presented to sponsor(s).Month 9-12: Write-up best practice recommendations.
1. Advance knowledge of best practices for consumer-directedexchange (CDEx).2. Advance knowledge of best practices for encrypted searchwith high security3. Models for profitable CDEx-driven sharing in systems,communities, states.
41
Part
icip
atin
g in
a C
omm
unity
H
ealth
Impr
ovem
ent N
etw
ork
Pro
ject
Lea
ders
: Jud
ah T
horn
ewill
, PhD
Co-
lead
ers:
Paul
Win
drum
, PhD
Stud
ent
Nam
e(s)
: Moh
amed
Ahm
ed, M
olly
O’K
eefe
Enga
ged
IAB
mem
bers
: Pas
spor
t Hea
lth P
lan,
Sa
nofi,
Uni
vers
ity o
f Lou
isvi
lle H
ospi
tal
Enga
ged
CH
OT
Uni
vers
ity S
ites:
Uni
vers
ity o
f Lo
uisv
ille,
Uni
vers
ity o
f Not
tingh
am
Proj
ect O
verv
iew
and
Des
crip
tion
OB
JEC
TIV
ES
& P
AR
TNE
RS
: Des
crip
tion
of P
roje
ct
Obj
ectiv
e:D
evel
op a
"co
nsum
er-d
irect
ed"
com
mun
ity h
ealth
info
rmat
ion
shar
ing
netw
ork
desi
gned
to im
prov
e ab
ility
for
peop
le a
nd o
rgan
izat
ions
to m
ore
easi
ly s
hare
hea
lth a
nd w
ellb
eing
dat
a to
impr
ove
care
coo
rdin
atio
n an
d su
ppor
t res
earc
h ac
ross
the
com
mun
ityPa
rtne
rs /
Rel
evan
ce:
Cre
ates
a n
ew p
arad
igm
of h
ealth
in
form
atio
n sh
arin
g th
at p
uts
the
pers
on
(pat
ient
/con
sum
er) i
n th
e ce
nter
of
info
rmat
ion
shar
ing
abou
t the
m, a
nd
also
use
s ne
xt g
ener
atio
n se
curit
y te
chno
logi
es, i
nclu
ding
blo
ckch
ain,
to
prot
ect a
nd e
nabl
e pe
rson
al h
ealth
in
form
atio
n sh
arin
g
and
Des
crip
tion
and
Des
can
dD
esc
Pt
/Rl
Proj
ect O
verv
iew
and
Des
crip
tion
Prob
lem
: Po
or h
ealth
out
com
es a
nd th
e ne
ed fo
r a b
ette
r way
to s
hare
hea
lth
impr
ovem
ent d
ata
for c
are
and
rese
arch
in
clud
ing
cons
umer
s an
d th
eir f
amili
es
OB
JEC
TIV
ES
& P
AR
TNE
RS
: Des
ign
Thin
king
App
roac
h
•C
onsu
mer
s–
Enr
ollm
ent d
elay
s, p
oor c
are
coor
dina
tion,
hig
h co
sts
•Pr
ovid
ers
–hi
gh c
osts
, lac
k of
acc
ess
to q
ualit
y da
ta o
ut o
f ne
twor
k, q
ualit
y/ou
tcom
e sc
ores
•Pl
ans
–hi
gh c
osts
enr
ollin
g an
d en
gagi
ng p
atie
nts;
clin
ical
er
rors
and
cha
lleng
es
•Pu
rcha
sers
–hi
gh c
osts
•D
igita
l hea
lth fi
rms
–ch
alle
nges
acc
ess
data
nee
ded
to
inno
vate
and
mak
e a
diffe
renc
e
•R
esea
rche
rs–
high
cos
ts o
f acc
essi
ng d
ata
for c
linic
al tr
ials
, re
sear
ch, p
op. h
ealth
•D
iabe
tes
initi
al fo
cus
•M
any
new
type
s of
dat
a•
Soc
ial.
Hom
e. S
urve
ys.
App
s.
Gen
omic
s. B
eyon
d E
MR
.
•Fr
ustra
ted
peop
le /
fam
ilies
•E
nrol
lmen
t cha
lleng
es
•C
are
coor
dina
tion
chal
leng
es
•D
uplic
ate
test
s an
d pr
oced
ures
•M
edic
al e
rror
s
•Av
oida
ble
spen
ding
Proj
ect O
verv
iew
and
Des
crip
tion
OB
JEC
TIV
ES
& P
AR
TNE
RS
: Des
ign
Thin
king
App
roac
h
B-C
-B –
New
Par
adig
m
42
Proj
ect O
verv
iew
and
Des
crip
tion
CO
NTR
IBU
TIO
N: H
ow is
this
diff
eren
t tha
n re
late
d re
sear
ch?
Sour
ces
• • • • • • • •
Acce
ss
Perm
issio
ns
•IR
A /
othe
r•
FHIR
CCDA
HIE
Grap
h
Loca
l En
cryp
tions
• •
PP
P
P P8.
Dat
a us
ers r
ecei
ve
serv
ices
and
pay
with
ca
sh/d
igita
l coi
ns
(Rxc
oin)
Cons
umer
Nav
igat
or
5. C
onsu
mer
s req
uest
ac
cess
to d
ata
Trus
t Aut
horit
ies
• • • • •
P P
X
P
4. U
nifie
d He
alth
Re
cord
s con
nect
to
netw
ork
X
P
RRR
R
PR
XP
AAR
R
A
Why
?
Kent
ucky
Hea
lth Im
prov
emen
t Net
wor
k
Rx
Rx
Rx
7. M
essa
ges a
nd A
lert
sM
M M
2. D
ata
sour
ces a
cces
s res
ourc
es a
nd
conn
ect
P
9. S
ourc
es g
et p
aid
1.Da
ta u
sers
requ
est b
ette
r dat
a &
se
rvic
es
3. C
loud
Se
rvic
es• • • • • • • • •
6. C
onse
nt e
ngin
eT
T
P
App
roac
hA
ctio
n-or
ient
ed re
sear
ch w
ill fo
cus
on
deve
lopi
ng b
est p
ract
ice
reco
mm
enda
tions
for:
1.
CH
IN g
over
nanc
e2.
Org
aniz
atio
nal p
artic
ipat
ion
agre
emen
ts w
ith th
e C
HIN
3.E
ngag
ing
"con
nect
ors"
at p
rovi
ders
, pla
ns, s
ocia
l ag
enci
es a
nd o
ther
offi
ces
to h
elp
patie
nts/
cons
umer
s si
gn-u
p w
ith th
e sy
stem
4.O
btai
ning
info
rmed
con
sent
to a
cces
s an
d re
-sha
re d
ata
with
mem
bers
of t
he C
HIN
5.C
ondu
ctin
g se
arch
es o
f dat
a, s
tarti
ng w
ith ty
pe II
di
abet
es
AP
PR
OA
CH
: Exp
erim
enta
l App
roac
h
App
roac
hO
VE
RV
IEW
OF
RE
SU
LTS
: Exp
ecte
d M
ilest
ones
Mon
th
2
Mon
ths
6-9
Mon
ths
9-12
•C
HO
T m
embe
rs in
vite
d to
par
ticip
ate
in
CH
IN p
roce
ss•
Get
ting
orga
nize
d
•Fi
rst c
onne
ctor
s tra
ined
and
cer
tifie
d•
Dum
my
and
real
pat
ient
enr
olle
d; d
ata
mov
ed;
expe
rienc
e ev
alua
ted
•E
ncry
pted
dat
a-se
arch
test
ed w
ith s
elec
ted
sour
ces
•P
rese
ntat
ion
of re
sults
pre
sent
ed to
spo
nsor
s
•W
rite
up b
est p
ract
ice
reco
mm
enda
tions
App
roac
hO
VE
RV
IEW
OF
RE
SU
LTS
: Ben
efits
to In
dust
ry
Ben
efits
to In
dust
ry:
1.Ad
vanc
e kn
owle
dge
of b
est
prac
tices
for c
onsu
mer
-di
rect
ed e
xcha
nge
(CD
Ex)
2.A
dvan
ce k
now
ledg
e of
bes
t pr
actic
es fo
r enc
rypt
ed s
earc
h w
ith h
igh
secu
rity
3.M
odel
s fo
r pro
fitab
le C
DE
x-dr
iven
sha
ring
in s
yste
ms,
co
mm
uniti
es, s
tate
s
Expe
cted
Del
iver
able
s:1.
Org
aniz
ing
mee
ting
2.D
ocum
ents
des
crib
ing
best
pr
actic
es a
nd le
sson
s le
arne
d
3.Fo
rmal
pre
sent
atio
n in
mon
th
nine
43
44
PI:
The Effectiveness of Substance Abuse Treatment Services in Combating Opioid Crisis
Pop3 Hui Zhao
Population Health
$50,000 NO 1
Each day, more than 115 Americans die due to overdosing on opioid. Addiction to opioid (including heroin and fentanyl) becomes aserious national crisis that devastates public health. To combat the opioid crisis, the Substance Abuse Treatment Services (SATS)facilities across the country provide opioid addicts professional counseling and treatments. The objective of this project is to evaluatethe effectiveness of the different opioid addiction treatment programs provided by these SATS facilities, by using sophisticatedeconometric models to analyze the national survey data on SATS facilities, the epidemic data on opioid abusers, and other relateddata.
While the opioid crisis has attracted much research recently, most of the research focuses on the supply side of the problem andinvestigate ways to control the supply and prescription of opioid such that only for legitimate reasons, patients can have access toreasonable amount of opioid. Much less research investigates the demand side of the problem. As we know, opioid abusers, whotypically have developed addictions, often seek illegal access to opioid or other alternatives. Our study takes this angle and looks fromthe demand side of the opioid problem by investigating the effectiveness of the different treatments and services used to recoveropioid addicts. While some medical papers have examined the effectiveness of SATS from individual patient perspective, our studyevaluates SATS from a population perspective, from which we expect to provide policy guidance of managing such programs.
The experimental plan includes the following tasks:1) Acquire and clean data from the National Survey of Substance Abuse Treatment Services from 2006-2016.2) Acquire metrics such as total deaths caused by opioid overdosing from Centers for Disease Control and Prevention (CDC)WONDER Database from 2006-2016.3) Collect demographic information from census data.4) Build an econometric model that evaluates the effectiveness of SATS in reducing opioid overdosing related deaths.5) If possible, collect individual opioid addicts information from a few SATS facilities to validate finding.6) Report and disseminate research findings.
The expected milestones of this project include:1) Collection and compilation of data from National Survey of Substance Abuse Treatment Services.1) Collection and compilation of data on opioid overdosing related deaths from CDC WONDER.3) Integration of different data sources into a master datafile.4) An econometric model that evaluates the effectiveness of SATS.5) A manuscript documenting the research.
1) Evaluate the effectiveness of SATS in curbing opioidoverdosing related deaths from a population level perspective.2) Recommend effective programs (e.g. individual counseling,group counseling) to recover opioid abusers.3) Combat the opioid crisis from the demand-side
1) A master datafile on opioid overdosing related deaths andSATS facility across the country.2) An econometric model that predicts the effectiveness of SATSprogram.3) Policy recommendation for how to manage SATS programs tobest curb opioid overdosing death.
45
The
Effe
ctiv
enes
s of
Sub
stan
ce
Abu
se T
reat
men
t Ser
vice
s in
C
omba
ting
the
Opi
oid
Cris
isP
roje
ct L
eade
rs:
Hui
Zha
o, S
mea
lCol
lege
of B
usin
ess,
Supp
ly C
hain
Co-
lead
ers:
Chr
isto
pher
DeF
litch
Stud
ent
Nam
e(s)
: TB
D
Enga
ged
IAB
mem
bers
: Sie
men
s, H
ersh
ey M
edic
al C
ente
r, H
ighm
ark,
AT
&T
Enga
ged
CH
OT
Uni
vers
ity S
ites:
Penn
Sta
te
Proj
ect O
verv
iew
and
Des
crip
tion
Obj
ectiv
e:E
valu
ate
the
effe
ctiv
enes
s of
the
diffe
rent
opi
oid
addi
ctio
n pr
ogra
ms
prov
ided
by
Sub
stan
ce A
buse
Tr
eatm
ent S
ervi
ces
(SAT
S).
OB
JEC
TIV
ES
& P
AR
TNE
RS
: Des
crip
tion
of P
roje
ct
Part
ners
/ R
elev
ance
:In
form
hea
lthca
re p
rovi
ders
abo
ut th
e ef
fect
iven
ess
of c
urre
nt tr
eatm
ent
prog
ram
s us
ed in
com
batti
ng th
e op
ioid
cris
is
Are
sub
stan
ce
abus
e tre
atm
ent
serv
ices
ef
fect
ive
in
treat
ing
opio
id
addi
tion?
Wha
t pro
blem
is
this
pro
ject
see
king
to
add
ress
?G
athe
r and
qua
ntify
de
man
d si
de d
ata
usin
g so
phis
ticat
ed
econ
omet
ric m
odel
s to
ana
lyze
the
natio
nal s
urve
y da
ta
on S
ATS
faci
litie
s, th
e ep
idem
ic d
ata
on
opio
id a
buse
rs, a
nd
othe
r rel
ated
dat
a.
OB
JEC
TIV
ES
& P
AR
TNE
RS
: Des
ign
Thin
king
App
roac
hPr
ojec
t Ove
rvie
w a
nd D
escr
iptio
nR
esea
rch
Pers
pect
ive
The
econ
omic
an
d so
ciet
al
impa
cts
of th
e op
ioid
epi
dem
ic
are
stag
gerin
g.
Hea
lthca
re
prov
ider
s, p
olic
y m
aker
s, a
nd th
e co
mm
unity
nee
d to
und
erst
and
wha
t tre
atm
ents
ar
e m
ost e
ffect
ive.
OB
JEC
TIV
ES
& P
AR
TNE
RS
: Des
ign
Thin
king
App
roac
h
46
Proj
ect O
verv
iew
and
Des
crip
tion
Con
trib
utio
ns to
Indu
stry
and
Aca
dem
ia•
Cur
rent
rese
arch
prim
arily
focu
ses
on s
uppl
y si
de p
robl
ems
and
how
to c
ontro
l the
sup
ply
and
pres
crip
tion
of o
pioi
ds. T
his
rese
arch
w
ill in
vest
igat
e th
e de
man
d si
de.
•M
edic
al p
aper
s ex
ist o
n th
e ef
fect
iven
ess
of
SAT
S fr
om th
e pe
rspe
ctiv
e of
indi
vidu
al
patie
nts.
Thi
s re
sear
ch w
ill ev
alua
te S
ATS
fro
m a
pop
ulat
ion
pers
pect
ive.
•R
esul
ting
rese
arch
is e
xpec
ted
to p
rovi
de
polic
y gu
idan
ce fo
r tre
atm
ent p
rogr
am
man
agem
ent.
CO
NTR
IBU
TIO
N: H
ow is
this
diff
eren
t tha
n re
late
d re
sear
ch?
App
roac
hW
hat d
o yo
u pl
an o
n do
ing?
1.A
cqui
re a
nd c
lean
dat
a fro
m th
e N
atio
nal S
urve
y of
Sub
stan
ce A
buse
Tr
eatm
ent S
ervi
ces
from
200
6 -2
016
2.A
cqui
re m
etric
s su
ch a
s to
tal d
eath
s ca
used
by
opio
id o
verd
osin
g fro
m th
e C
DC
WO
ND
ER
Dat
abas
e fro
m 2
006
–20
163.
Col
lect
dem
ogra
phic
info
rmat
ion
from
cen
sus
data
4.Bu
ild a
n ec
onom
etric
mod
el th
at e
valu
ates
SAT
S e
ffect
iven
ess
in re
duci
ng
opio
id d
eath
s5.
Valid
ate
findi
ngs
usin
g da
ta fr
om re
al w
orld
opi
oid
addi
cts
6.R
epor
t and
dis
sem
inat
e fin
ding
s
AP
PR
OA
CH
: Exp
erim
enta
l App
roac
h
App
roac
hO
VE
RV
IEW
OF
RE
SU
LTS
: Exp
ecte
d M
ilest
ones
Expe
cted
Mile
ston
es1.
Col
lect
ion
and
com
pila
tion
of d
ata
from
Nat
iona
l S
urve
y of
Sub
stan
ce
Abu
se T
reat
men
t Ser
vice
s2.
Col
lect
ion
and
com
pila
tion
of d
ata
on o
pioi
d ov
erdo
sing
rela
ted
deat
hs
from
CD
C W
ON
DE
R3.
Dev
elop
an
econ
omet
ric
mod
el th
at e
valu
ates
the
effe
ctiv
enes
s of
SAT
S4.
Pub
lish
man
uscr
ipts
that
do
cum
ent r
esea
rch
findi
ngs
5.C
reat
e a
heal
thca
re
deci
sion
sup
port
tool
that
vi
sual
izes
the
resu
lts
App
roac
h
Ben
efits
to In
dust
ry:
1.E
valu
ate
the
effe
ctiv
enes
s of
SAT
S in
cur
bing
opi
oid
over
dosi
ng re
late
d de
aths
from
a p
opul
atio
n le
vel p
ersp
ectiv
e2.
Rec
omm
end
effe
ctiv
e pr
ogra
ms
for o
pioi
d ab
user
s3.
Com
bat t
he o
pioi
d cr
isis
from
the
dem
and-
side
OV
ER
VIE
W O
F R
ES
ULT
S: B
enef
its to
Indu
stry
Expe
cted
D
eliv
erab
les:
1.
A m
aste
r dat
a fil
e on
op
ioid
ove
rdos
ing
rela
ted
deat
hs a
nd
SAT
S fa
cilit
ies
acro
ss
the
US
2.A
n ec
onom
etric
mod
el
that
pre
dict
s ef
fect
iven
ess
of S
ATS
pr
ogra
ms
3.P
olic
y re
com
men
datio
n on
how
to m
anag
e S
ATS
pro
gram
s to
bes
t cu
rb o
pioi
d ov
erdo
sing
de
ath.
47
48
Research Theme #2:
CARE COORDINATION
Theme Champion: Dr. Christina Mastrangelo, University of Washington
Care 1 Measuring Patient Experience and the Effects of Community Factors on Value-based Reimbursement across the Continuum of Care Care 2 Developing a Risk Prediction Model for Hospital Acquired Clostridium Difficile Infection
49
50
PI:
Care Coordination & Patient Experience across the Continuum of Care: A Value-Based Reimbursement Perspective
Value-based reimbursement (VBR) in health care has resulted in an increased focus on care coordination and patient experience(CAHPS) across the continuum of care, including hospitals, home health, and ambulatory care. This project has three aims: 1) Identifyand pilot test a survey instrument that can be used to assess patient experience across the continuum of care from acute care topost-acute care; 2) incorporate VBR concepts into medical curricula and adapt practices to support patient experience in a VBR-basedsystem; and 3) develop a HIPAA compliant messaging platform to ensure timely delivery of all messages to a care team with a criticalpatient information attached with each message.
Care1 Agarwal, Ferris, and Weech-Maldonado
Care Coordination
$100,000 YES 1
Aim 1: a) Review the literature on patient experience measurement across the continuum of care from acute care to post-acute care;b) Identify and adapt measures that can be used to assess patient experience across the continuum of care; and c) Pilot test thesurvey instrument identified under b.
Aim 2: a) Evaluate current platforms for care coordination; b) Develop a mobile application framework for secured asynchronousmessaging system; c) Implement the framework for centralized monitoring and integration of care services. The current proposal willaddress the phase-3 of the project.
Months 1-2: Obtain IRB approvalMonths 3-4: Literature review and evaluation of current modelsMonths 5-6: Focus groups and stakeholder interviewsMonths 7-8: Develop pilot studyMonths 9-10: Conduct pilot studyMonths 11-12: Analyze results from pilot study and write final report
This research contributes to the areas of measurement, training and development, and technology implementation with the goal ofimproving care coordination and patient experience across the continuum of care. First, there has been a focus on assessing patientexperience on separate components of the continuum of care care, or a silo approach; however, less research has been conductedon assessing patient experience across the continuum of care as patients transition from one setting to another. Second, there is aneed for medical curricula to incorporate principles associated with value-based reimbursement. Finally, while several new mobilehealthcare messaging applications are available, such as HIPPA compliant WhatsApp, they ultimately create more silos.
With the rise of integrated delivery systems, such as theaccountable care organizations, it is important for health careorganizations and payers to have tools that can be used tocoordinate care assess patient experience and across thecontinuum of care.
A final report outlining the findings of the study.
51
Car
e C
oord
inat
ion
and
Patie
nt
Expe
rienc
e ac
ross
the
Con
tinuu
m
of C
are:
A V
alue
-Bas
ed
Rei
mbu
rsem
ent P
ersp
ectiv
eP
roje
ct L
eade
rs:
Ank
urA
garw
al, T
hom
as F
erri
s, an
d R
ober
t Wee
ch-M
aldo
nado
Enga
ged
IAB
mem
bers
: Te
xas A
&M
Uni
vers
ity C
olle
ge o
f Med
icin
e, U
AB
Hea
lth S
yste
m
Enga
ged
CH
OT
Uni
vers
ity S
ites:
UA
B/TA
MU
/FA
U
Proj
ect O
verv
iew
and
Des
crip
tion
Obj
ectiv
es:
•Id
entif
y an
d pi
lot t
est a
sur
vey
inst
rum
ent t
o as
sess
pat
ient
exp
erie
nce
acro
ss th
e co
ntin
uum
of
car
e fro
m a
cute
car
e to
pos
t-acu
te c
are
•In
corp
orat
e va
lue-
base
d re
imbu
rsem
ent (
VB
R)
conc
epts
into
med
ical
cur
ricul
a an
d ad
apt
prac
tices
to s
uppo
rt pa
tient
exp
erie
nce
in a
VB
R-
base
d sy
stem
•D
evel
op a
HIP
AA
com
plia
nt m
essa
ging
pla
tform
to
ensu
re ti
mel
y de
liver
y of
all
mes
sage
s to
a c
are
team
with
a c
ritic
al p
atie
nt in
form
atio
n at
tach
ed
with
eac
h m
essa
ge
OB
JEC
TIV
ES
& P
AR
TNE
RS
: Des
crip
tion
of P
roje
ct
Proj
ect O
verv
iew
and
Des
crip
tion
Prob
lem
:•
Valu
e-ba
sed
reim
burs
emen
t (V
BR
) has
resu
lted
in a
n in
crea
sed
focu
s on
car
e co
ordi
natio
n an
d pa
tient
exp
erie
nce
(CA
HP
S) a
cros
s th
e co
ntin
uum
of c
are
•M
easu
rem
ent
–Fo
cus
has
been
on
asse
ssin
g pa
tient
exp
erie
nce
on s
epar
ate
com
pone
nts
of th
e co
ntin
uum
of c
are,
or a
silo
app
roac
h.
–Le
ss re
sear
ch o
n as
sess
ing
patie
nt e
xper
ienc
e ac
ross
the
cont
inuu
m o
f ca
re
•Tr
aini
ng a
nd D
evel
opm
ent
–N
eed
for m
edic
al c
urric
ula
to in
corp
orat
e pr
inci
ples
ass
ocia
ted
with
va
lue-
base
d re
imbu
rsem
ent.
•Te
chno
logy
–S
ever
al n
ew m
obile
hea
lthca
re m
essa
ging
app
licat
ions
are
ava
ilabl
e;
how
ever
, the
y ar
e ba
sica
lly H
IPA
A co
mpl
iant
text
mes
sagi
ng a
mon
g do
ctor
s -i
.e. H
IPA
A co
mpl
iant
wha
tsap
p, a
nd e
ffect
ivel
y cr
eate
mor
e,
not l
ess,
silo
s.
OB
JEC
TIV
ES
& P
AR
TNE
RS
: Des
ign
Thin
king
App
roac
hPr
ojec
t Ove
rvie
w a
nd D
escr
iptio
n
Res
earc
h Pe
rspe
ctiv
e•
Pro
vide
rs: H
ospi
tals
, hea
lth s
yste
ms
•P
ayer
s: M
edic
are,
insu
ranc
e co
mpa
nies
•P
atie
nts:
Acu
te to
pos
t-acu
te p
atie
nts
OB
JEC
TIV
ES
& P
AR
TNE
RS
: Des
ign
Thin
king
App
roac
h
52
Proj
ect O
verv
iew
and
Des
crip
tion
Con
trib
utio
n to
Indu
stry
and
Aca
dem
ia•
With
the
rise
of v
alue
-bas
ed
reim
burs
emen
t mod
els,
suc
h as
bun
dled
pa
ymen
ts a
nd a
ccou
ntab
le c
are
orga
niza
tions
, it i
s im
porta
nt fo
r hea
lth
care
org
aniz
atio
ns a
nd in
sure
rs to
hav
e to
ols
that
can
be
used
to c
oord
inat
e ca
re
and
asse
ss p
atie
nt e
xper
ienc
e ac
ross
the
cont
inuu
m o
f car
e.
CO
NTR
IBU
TIO
N: H
ow is
this
diff
eren
t tha
n re
late
d re
sear
ch?
App
roac
h•
Aim
1:
–R
evie
w th
e lit
erat
ure
on p
atie
nt e
xper
ienc
e m
easu
rem
ent a
cros
s th
e co
ntin
uum
of
car
e fro
m a
cute
car
e to
pos
t-acu
te c
are
–Id
entif
y an
d ad
apt m
easu
res
that
can
be
used
to a
sses
s pa
tient
exp
erie
nce
acro
ss th
e co
ntin
uum
of c
are
–P
ilot t
est t
he s
urve
y in
stru
men
t ide
ntifi
ed u
nder
ste
p 2
•A
im 2
: –
Pha
se 1
: Eva
luat
e cu
rren
t pla
tform
s fo
r car
e co
ordi
natio
n–
Pha
se 2
: Dev
elop
a m
obile
app
licat
ion
fram
ewor
k fo
r sec
ured
asy
nchr
onou
s m
essa
ging
sys
tem
–P
hase
3: I
mpl
emen
t the
fram
ewor
k fo
r cen
traliz
ed m
onito
ring
and
inte
grat
ion
of
care
ser
vice
s. (T
he c
urre
nt p
ropo
sal w
ill a
ddre
ss th
e ph
ase-
3 of
the
proj
ect)
AP
PR
OA
CH
: Exp
erim
enta
l App
roac
h
App
roac
h
•M
onth
s 1-
2:
Obt
ain
IRB
app
rova
l•
Mon
ths
3-4:
L
itera
ture
revi
ew a
nd
eval
uatio
n of
cur
rent
mod
els
•M
onth
s 5-
6:
Foc
us g
roup
s an
d st
akeh
olde
r int
ervi
ews
•M
onth
s 7-
8:
Dev
elop
pilo
t stu
dy•
Mon
ths
9-10
: C
ondu
ct p
ilot s
tudy
•M
onth
s 11
-12:
Ana
lyze
resu
lts fr
om p
ilot
stud
y an
d w
rite
final
repo
rt
OV
ER
VIE
W O
F R
ES
ULT
S: E
xpec
ted
Mile
ston
esA
ppro
ach
OV
ER
VIE
W O
F R
ES
ULT
S: B
enef
its to
Indu
stry
•D
eliv
erab
les
•S
urve
y in
stru
men
t tha
t can
be
used
to
asse
ss p
atie
nt e
xper
ienc
e ac
ross
the
cont
inuu
m o
f car
e•
Med
ical
cur
ricul
a th
at in
corp
orat
es V
BR
pr
inci
ples
•A
mob
ile/ m
essa
ging
app
licat
ion
to im
prov
e co
mm
unic
atio
n am
ong
prov
ider
s of
a c
are
team
53
54
PI:
Developing a Risk Prediction Model for Hospital Acquired Clostridium Difficile Infection
Care2 Midge Ray, Ferhat Zengul
Care Coordination
$42,000 NO 2
Hospital Acquired Infections (HAI), infections acquired after admission to the hospital, consume about $25 to $31 billion. In 2002, theCenters for Disease Control and Prevention estimated the rate for HAIs to be about 1.7 million, with 99,000 related to deaths duringhospitalization. Hence, HAI affect quality and cost of healthcare. In 2008, Medicare initiated a plan to reduce payments to hospitals forcomplications that occur during the hospital stay, including HAI. One such infection is the Clostridium difficile infection (CDI), which isthe most common cause of infectious diarrhea occurring in the hospital. Treatments costs per patient is approximately $8,911 to$30,049 in the U.S. In Phase I, we identified predictors of hospital acquired CDI. In Phase II, we will conduct a retrospective studyusing UAB i2b2 dataset to develop a risk prediction model for CDI. The main goal of this study is to develop a CDI risk predictionmodel that allows categorizing patients into high, medium, and low risk categories, which will allow for more targeted strategies.
Even though there have been studies exploring the predictors of CDI, there has not been a study that develops a CDI risk predictionmodel that allows categorizing patients into high, medium, and low risk categories. These risk categories would allow the developmentof more targeted testing strategies for CDI infection. Given that CDI testings tend to be expensive and they are not reimbursed by thepayers, it is crucial for hospitals to be able to develop more targeted CDI testing strategies.
Phase I: Literature Review - CompletedPhase II: Initial UAB i2b2 exploration generated 20,325 patients who were tested for CDI. Currently, we don't know how many of thesepatients had positive test results. We will be able to know this information after acquiring the IRB approval. If a patient was tested forCDI, there should be some clinical reason. Having both positive and negative tested patients would allow us to develop models thatwould better differentiate these two categories of patients.Phase III: Testing the findings by using industry partner's data set and writing a grant proposal for the development of a clinicaldecision support system, integrated into electronic health records, that would use findings from machine learning algorithms to provideprobability of risk for each patient for present-on-admission CDI or development of CDI during hospital stay.
A risk prediction model of CDI will be useful in screening patientsat risk for the infection at the time of admission, leading to moretargeted strategies in testing CDI, reducing costs. Identifyingpredictors of CDI would also allow development of potentialinterventions to reduce the rate of CDI by (1) administeringproactive treatment to patients at risk (2) diagnosing patients whohave pre-existing or HAI CDI resulting in quicker treatment, and(3) potentially reducing the length of stay, also reducing costs.
Year 1: Final report on risk prediction model of CDIYear 2: (1) Publish a manuscript based upon the findings and (2)Poster presentation by PhD student at professional meeting.
Month 1-2: Acquiring IRB approvalMonth 3-6: Preprocessing the data and running descriptive statisticsMonth 7-8: Predictive AnalysesMonth 9-12: Write report on the results of risk prediction model of CDI
55
Dev
elop
ing
a R
isk
Pred
ictio
n M
odel
for
Hos
pita
l Acq
uire
d C
lost
ridiu
m D
iffic
ile
Infe
ctio
n
Pro
ject
Lea
ders
: M
idge
N. R
ay;
Fer
hat
D. Z
engu
l
Co-
lead
ers:
Ken
Coc
hran
, Jea
nine
Tho
mas
Stud
ent
Nam
e(s)
:N
eera
j A. P
uro;
Ree
na Jo
seph
, Rac
hel V
irgi
nia W
isni
ewsk
i
Enga
ged
IAB
mem
bers
: O
PELO
USA
S G
ENER
AL
HEA
LTH
SYS
TEM
Enga
ged
CH
OT
Uni
vers
ity S
ites:
The
Uni
vers
ity o
f Ala
bam
a at
Bir
min
gham
Prob
lem
and
Obj
ectiv
e:•
Clo
strid
ium
diff
icile
infe
ctio
ns (C
DIs
) are
one
of t
he m
ost
com
mon
nos
ocom
ial i
nfec
tions
that
incr
ease
pat
ient
s’
mor
bidi
ty &
mor
talit
y, a
nd a
re a
ssoc
iate
d w
ith a
nnua
l co
sts
in e
xces
s of
$4.
8 bi
llion
•P
urpo
se o
f thi
s st
udy
is to
dev
elop
a C
DI r
isk
pred
ictio
n m
odel
that
cat
egor
izes
pat
ient
s in
to h
igh,
med
ium
and
lo
w ri
skPa
rtne
rs /
Rel
evan
ce:
•H
ealth
care
org
aniz
atio
ns w
ould
be
able
to d
evel
op m
ore
targ
eted
test
ing
stra
tegi
es fo
r CD
I•
HC
Os
may
be
able
to re
duce
unn
eces
sary
test
ing
and
asso
ciat
ed c
osts
OB
JEC
TIV
ES
& P
AR
TNE
RS
: Des
crip
tion
of P
roje
ct
Proj
ect O
verv
iew
and
Des
crip
tion
•P
er p
atie
nt c
ost a
ssoc
iate
d w
ith tr
eatin
g C
DIs
rang
e fro
m a
ppro
xim
atel
y $8
,911
to $
30,0
49
•Tr
eatin
g C
DIs
can
resu
lt in
long
er h
ospi
tal s
tays
and
hi
gher
cos
ts to
the
heal
th c
are
orga
niza
tion
•Te
stin
g al
l pat
ient
s fo
r CD
I is
not f
easi
ble
sinc
e te
stin
g is
co
stly
and
hos
pita
ls a
re o
ften
not r
eim
burs
ed fo
r sc
reen
ing
OB
JEC
TIV
ES
& P
AR
TNE
RS
: Des
ign
Thin
king
App
roac
h
Res
earc
h Pe
rspe
ctiv
e
•C
DIs
acq
uire
d du
ring
hosp
ital s
tay
are
pena
lized
by
CM
S u
nder
the
Hos
pita
l-Acq
uire
d C
ondi
tion
Red
uctio
n P
rogr
am (H
AC
RP)
•C
DI r
isk
pred
ictio
n m
odel
is im
porta
nt fo
r pro
vide
rs to
aid
w
ith e
arly
iden
tific
atio
n of
hig
h ris
k pa
tient
s
•P
atie
nts
may
ben
efit
from
bet
ter i
nfor
med
and
co
ordi
nate
d cl
inic
al c
are
thro
ugh
CD
I ris
k pr
edic
tion
mod
el
OB
JEC
TIV
ES
& P
AR
TNE
RS
: Des
ign
Thin
king
App
roac
h
56
Con
trib
utio
n to
Indu
stry
and
Aca
dem
ia
•C
DI r
isk
pred
ictio
n m
odel
can
aid
in p
oten
tial
inte
rven
tions
suc
h as
: –
early
dia
gnos
is o
f pat
ient
s at
risk
for d
evel
opin
g C
DIs
–pr
oact
ive
treat
men
t to
at-r
isk
patie
nts,
ther
eby
min
imiz
ing
likel
ihoo
d of
HA
I pen
alty
for h
ospi
tals
–po
tent
ially
redu
cing
the
leng
th o
f sta
y an
d as
soci
ated
ho
spita
l cos
ts
CO
NTR
IBU
TIO
N: H
ow is
this
diff
eren
t tha
n re
late
d re
sear
ch?
App
roac
h•
Usi
ng th
e pr
edic
tors
of C
DIs
that
wer
e pr
evio
usly
id
entif
ied
in P
hase
I, w
e w
ill co
nduc
t a re
trosp
ectiv
e st
udy
usin
g th
e U
AB
i2b2
dat
aset
to d
evel
op a
risk
pre
dict
ion
mod
el•
This
dat
aset
incl
udes
:
•P
redi
ctio
n m
odel
will
allo
w fo
r dev
elop
men
t of t
arge
ted
test
ing
stra
tegi
es
•W
ill us
e st
anda
rd s
tatis
tics
(i.e.
, lin
ear r
egre
ssio
n) a
nd
mac
hine
lear
ning
app
roac
hes
AP
PR
OA
CH
: Exp
erim
enta
l App
roac
h
Posi
tive
Neg
ativ
eTo
tal #
of C
DI
Test
s Pe
rfor
med
4,81
336
,677
41,4
90
Prop
osed
Tim
elin
e
•M
onth
1-2
: Acq
uirin
g IR
B a
ppro
val
•M
onth
3-6
: Pre
proc
essi
ng th
e da
ta a
nd ru
nnin
g de
scrip
tive
stat
istic
s•
Mon
th 7
-8: P
redi
ctiv
e A
naly
ses
•M
onth
9-1
2: W
rite
repo
rt on
the
resu
lts o
f ris
k pr
edic
tion
mod
el o
f CD
I
OV
ER
VIE
W O
F R
ES
ULT
S: E
xpec
ted
Mile
ston
esA
ppro
ach
•A
CD
I ris
k pr
edic
tion
mod
el th
at c
an b
e ex
pand
ed a
nd im
prov
ed b
y ut
ilizi
ng la
rger
dat
a so
urce
s an
d ul
timat
ely
inte
grat
e in
to th
e E
HR
of
heal
thca
re p
rovi
ders
, whi
ch w
ould
impr
ove
care
an
d po
tent
ially
redu
ce c
ost
•P
ublis
hed
man
uscr
ipt b
ased
upo
n th
e fin
ding
s
•P
oste
r pre
sent
atio
n at
pro
fess
iona
l mee
ting
and/
or a
cade
mic
con
fere
nce
OV
ER
VIE
W O
F R
ES
ULT
S: B
enef
its to
Indu
stry
57
58
Research Theme #3:
ANALYTICS AND INNOVATIVE TECHNOLOGIES
Theme Champion: Dr. Eva Lee, Georgia Institute of Technology
Tech 1 HIE Project for Chronic Disease and Workflow Management Tech 2 Leveraging technology to enhance communication in healthcare Tech 3 Data-driven analytics and machine learning for improving healthcare outcomes
59
60
PI:
HIE Project for Chronic Disease and Workflow Management
There is a need in the healthcare industry to unify disparate sources of patient data to provide better care at lower costs. Theintegration of administrative, clinical, environmental and personal data sources is an opportunity to reconcile patients across all of theirdata records and the failure to do so can impede interoperability, leading to patient safety risks, revenue loss and decreased providerefficiency. The goal of this project is to explore the opportunities and challenges of integrated patient health data by characterizing thedata, identifying the issues, determining relevant algorithms and models and applying the algorithms and models to healthcaredelivery. The aims of this project are to apply this knowledge to three areas: 1) understanding the effects of integrated patient healthdata on workflows, 2) developing an architecture for a chronic care management system and 3) predicting re-admissions for a chronicdisease.
Tech1 Mastrangelo, Weech-Maldonado, Borkowski, Agarwal
Analytics & Innovative Technologies
$150,000 YES 1
A system architecture, a working prototype, a white paper onintegration, a journal paper on using ML in workflow analysis.
This project is novel for using machine learning methods in new healthcare applications and using the resulting analytics for visual,patient-centered information to support meaningful and actionable decisions regarding chronic condition and potential hospitalre-admission.
Conducting literature and market search.Identify preliminary workflows.Identify and obtain data sources.Develop methods to complete definition (feature engineering) and construction of models.Develop and implement ML algorithms and decision support rules.Evaluate the models and metrics and integrate.
8/31/ 2018 Literature and market research5/31/2018 Conceptual Modeling7/31/2018 Construction of datasets11/31/2018 Implementation of algorithms2/28/2019 Evaluation of the models and metrics4/30/2019 Integration
Looking at healthcare delivery workflows from the perspectiveof integrated data analytics and visualization will result in thedevelopment of procedures to improve patient flow, to providetimely treatment of chronic conditions, to maximize utilization ofavailable resources and reduce re-admissions.
61
HIE
Pro
ject
for C
hron
ic D
isea
se
and
Wor
kflo
w M
anag
emen
tP
roje
ct L
eade
rs:
Ank
urA
garw
al, C
hris
tina
Mas
tran
gelo
, Be
n O
zayd
in
Co-
lead
ers:
Ferh
atZ
engu
l, Et
a Be
rner
, Joe
Hei
m
Stud
ent
Nam
e(s)
: TBD
Enga
ged
IAB
mem
bers
: Edi
fecs
Enga
ged
CH
OT
Uni
vers
ity S
ites:
FAU
, UA
B, U
W
OB
JEC
TIV
ES
& P
AR
TNE
RS
: Des
crip
tion
of P
roje
ct
•B
ackg
roun
d•
Hea
lth In
form
atio
n E
xcha
nge
(HIE
) he
lps
mee
t the
nee
d in
the
heal
thca
re in
dust
ry to
uni
fy
disp
arat
e so
urce
s of
pat
ient
dat
a to
pr
ovid
e be
tter c
are
at lo
wer
cos
ts.
•Th
e in
tegr
atio
n of
thes
e da
ta
sour
ces
is a
n op
portu
nity
to
reco
ncile
pat
ient
s ac
ross
all
of th
eir
data
reco
rds.
•R
esul
ts in
incr
ease
d in
tero
pera
bilit
y w
hich
lead
s to
redu
ced
to p
atie
nt
safe
ty ri
sks,
incr
ease
d re
venu
e an
d in
crea
sed
prov
ider
effi
cien
cy.
•O
bjec
tive:
To
expl
ore
new
op
portu
nitie
s an
d ch
alle
nges
as
soci
ated
with
HIE
thro
ugh
3 ai
ms.
Proj
ect O
verv
iew
and
Des
crip
tion
Aim
1:
•E
xpan
d H
IE b
y in
tegr
atin
g da
ta fr
om h
ome
med
ical
dev
ices
use
d by
pa
tient
s w
ith c
hron
ic c
ondi
tions
by
–U
sing
the
initi
al p
hase
to d
esig
n, b
uild
, and
test
an
inte
grat
ed c
hron
ic c
are
man
agem
ent s
yste
m to
cap
ture
, ana
lyze
and
pre
sent
real
-tim
e, c
onsu
mab
le
heal
th in
form
atio
n–
Util
izin
g an
alyt
ics
and
visu
aliz
atio
n to
ols
to p
rese
nt a
ggre
gate
d in
form
atio
n th
at
is m
eani
ngfu
l and
act
iona
ble
for t
he p
rovi
ders
–M
aint
aini
ng in
tero
pera
bilit
y w
ith o
ther
info
rmat
ion
syst
ems
in th
e en
viro
nmen
t an
d co
mpl
ianc
e w
ith p
rivac
y an
d se
curit
y pr
otoc
ols
Rel
evan
ce fo
r Par
tner
s:•
Sys
tem
will
allo
w p
rovi
ders
to m
anag
e pa
tient
s w
ith m
ultip
le c
hron
ic
cond
ition
s re
mot
ely,
redu
cing
cos
t by
avoi
ding
unn
eces
sary
vis
its.
OB
JEC
TIV
ES
& P
AR
TNE
RS
: Des
crip
tion
of P
roje
ctPr
ojec
t Ove
rvie
w a
nd D
escr
iptio
n
Aim
2:
•D
evel
op a
n ar
chite
ctur
e fo
r int
egra
ting
unst
ruct
ured
dat
a el
emen
ts
that
can
be
used
for p
redi
ctiv
e m
odel
s in
a s
yste
m c
hara
cter
ized
by
–A
web
-bas
ed p
roto
type
for p
redi
ctin
g ho
spita
l rea
dmis
sion
s fo
r pa
tient
s w
ith C
hron
ic O
bstru
ctiv
e P
ulm
onar
y D
isea
se (C
OP
D)
–P
redi
ctiv
e m
odel
s th
at a
lso
utili
ze lo
cal/e
nviro
nmen
tal f
acto
rs
Rel
evan
ce fo
r Par
tner
s:•
Sys
tem
will
redu
ce th
e ra
te o
f rea
dmis
sion
by
acco
untin
g fo
r fac
tors
th
at v
ary
by lo
catio
n w
hich
will
, in
turn
, im
prov
e ca
re c
oord
inat
ion
and
hom
e he
alth
care
.
OB
JEC
TIV
ES
& P
AR
TNE
RS
: Des
crip
tion
of P
roje
ct
62
OB
JEC
TIV
ES
& P
AR
TNE
RS
: Aim
2 C
once
ptua
l Arc
hite
ctur
ePr
ojec
t Ove
rvie
w a
nd D
escr
iptio
n
Aim
3:
•U
nder
stan
d th
e im
plem
enta
tion
of H
IE b
y st
udyi
ng th
e ef
fect
of
inte
grat
ed d
ata
on w
orkf
low
s by
–Id
entif
ying
whe
re d
ata
disp
ariti
es m
ay o
ccur
in a
wor
kflo
w b
y m
odel
ing
the
wor
kflo
w p
roce
sses
–D
evel
opin
g M
L al
gorit
hms
for t
he c
hara
cter
izat
ion
of s
uch
disp
ariti
es, s
uch
as
proc
ess
bottl
enec
ks–
Eva
luat
ing
the
mod
els
and
met
rics
for a
repr
esen
tativ
e he
alth
care
reco
ncili
atio
n so
lutio
n
Rel
evan
ce fo
r Par
tner
s :
•G
iven
the
pres
sure
s to
con
tain
cos
ts, i
t is
criti
cal f
or h
ospi
tals
, hea
lth
care
sys
tem
s, a
nd p
ayer
s to
dev
elop
hig
hly
effic
ient
sys
tem
s w
hich
ca
n re
conc
ile d
ata
disp
ariti
es a
cros
s th
e sy
stem
nod
es q
uick
ly a
nd
accu
rate
ly.
OB
JEC
TIV
ES
& P
AR
TNE
RS
: Des
crip
tion
of P
roje
ct
Proj
ect O
verv
iew
and
Des
crip
tion
Res
earc
h Pe
rspe
ctiv
e
OB
JEC
TIV
ES
& P
AR
TNE
RS
: Des
ign
Thin
king
App
roac
h
Pat
ient
Impr
oved
and
ta
ilore
d ca
reR
educ
ed R
isk
Pro
vide
rR
educ
ed c
osts
Incr
ease
d ac
cess
to
patie
nt d
ata
Hea
lth
syst
emIn
crea
sed
effic
ienc
yR
educ
ed re
-ad
mis
sion
s
Pay
erR
educ
es c
ost
Proj
ect O
verv
iew
and
Des
crip
tion
Con
trib
utio
n to
Indu
stry
and
Aca
dem
ia•
Ana
lytic
s an
d vi
sual
izat
ion
tool
s to
pre
sent
agg
rega
te
info
rmat
ion
that
is m
eani
ngfu
l and
act
iona
ble
•A
n un
ders
tand
ing
of th
e sc
ienc
e be
hind
the
deve
lopm
ent
of d
ecis
ion
supp
ort s
yste
m w
hich
is k
ey in
co
mm
erci
aliz
ing
new
pro
duct
s an
d ap
plic
atio
ns•
An
early
look
at h
ealth
care
del
iver
y w
orkf
low
s fro
m th
e pe
rspe
ctiv
e of
inte
grat
ed a
dmin
istra
tive
and
clin
ical
dat
a re
sulti
ng in
new
pro
cedu
res
to im
prov
e pa
tient
flow
CO
NTR
IBU
TIO
N: H
ow is
this
diff
eren
t tha
n re
late
d re
sear
ch?
63
App
roac
h•
Con
duct
lite
ratu
re a
nd m
arke
t sea
rch
•Id
entif
y pr
elim
inar
y w
orkf
low
s•
Iden
tify
and
obta
in d
ata
sour
ces
•D
evel
op m
etho
ds to
com
plet
e de
finiti
on (f
eatu
re
engi
neer
ing)
and
con
stru
ctio
n of
mod
els
•D
evel
op a
nd im
plem
ent M
L al
gorit
hms
and
deci
sion
sup
port
rule
s•
Eva
luat
e th
e m
odel
s an
d m
etric
s an
d in
tegr
ate
AP
PR
OA
CH
: Exp
erim
enta
l App
roac
hA
ppro
ach
•O
verv
iew
of r
esul
ts: U
sing
mac
hine
lear
ning
met
hods
in n
ew
heal
thca
re a
pplic
atio
ns a
nd u
sing
the
resu
lting
ana
lytic
s fo
r vis
ual,
patie
nt-c
ente
red
info
rmat
ion
to s
uppo
rt m
eani
ngfu
l and
act
iona
ble
deci
sion
s re
gard
ing
chro
nic
cond
ition
and
pot
entia
l hos
pita
l re-
adm
issi
on•
Lite
ratu
re a
nd m
arke
t res
earc
h8/
31/2
018
•C
once
ptua
l Mod
elin
g5/
31/2
018
•C
onst
ruct
ion
of d
atas
ets
7/31
/201
8•
Impl
emen
tatio
n of
alg
orith
ms
11/3
1/20
18•
Eva
luat
ion
of th
e m
odel
s an
d m
etric
s2/
28/2
019
•In
tegr
atio
n4/
30/2
019
OV
ER
VIE
W O
F R
ES
ULT
S: E
xpec
ted
Mile
ston
es
App
roac
h
•A
syst
em a
rchi
tect
ure
•A
web
-bas
ed p
redi
ctio
n pr
otot
ype
•A
whi
te p
aper
on
inte
grat
ion
•A
jour
nal p
aper
on
usin
g M
L in
wor
kflo
w
anal
ysis
OV
ER
VIE
W O
F R
ES
ULT
S: D
eliv
erab
les
64
PI:
Leveraging technology to enhance communication in healthcare
Tech2 Eva K Lee
Analytics & Innovative Technologies
TBD NO 1
With continuous advancements in technology, care providers have access to more tools than ever to combat breakdowns incommunication with referring physicians and to ultimately play a greater role in improved patient care. Often overwhelmed with heavyworkloads, care communication may suffer. For example, radiologists may be hesitant to assume additional responsibilities related toconveying test results and ensuring proper follow-up with patients. Certain symptoms discovered during surgical procedures bysurgeons may be conveyed ineffectively to intensivists and bed-side teams. Yet those activities can play an important role in not onlycarefully interpreting images or making recommendations but also acting as a safe, patient-centered back-up system and ensuringthat actionable results are not overlooked. In a similar manner, non-English speaking patients may require enhanced carecoordination plan to ensure that they understand the discharged and home care process.
This study is the first study which utilizes machine learning, text mining, and deep learning techniques to hospital discharge notes tobuild an accurate automatic translation system which will facilitate discharge and home care process design, particularly fornon-English speaking patients. In addition, it incorporates system design and human-device interaction technologies to offer real-timedecision support providers.
We will first conduct literature review on existing machine translation systems and explore gaps in discharge and home care process.Various sites with lack of medical translation experts will be identified and their needs will be assessed. Machine learning and naturallanguage processing techniques will be applied to build automatic translation systems using discharge notes. We will use the BLEUalgorithm to evaluate the translation quality and further refine the system. In addition, we will design chatbox and virtual messaging toenhance family engagement and facilitate knowledge dissemination across sites.
Aim 1: Improve communication, compliance and quality of carethrough automated machine translation. Initial focus will be thedischarge procedureAim 2: Design chatbox and virtual messaging to enhance familyengagement, knowledge dissemination (e.g., feeding plan,compliance and awareness of hospital acquired infection, childhealth).
1) Systematic literature review of a) existing machine translation algorithms and accuracies, b) indicators of lack of medical translationexperts and where the most needs for them are.2) Develop a machine learning / natural language processing framework using hospital discharge notes that can automatically andaccurately translate clinical documents to foreign languages.3) Evaluation and refinement of the translation system4) Design and implementation of chatbox and virtual messaging systems
-Enhanced care coordination plan-Reduced workforce requirements for translation experts-Improved discharged and home care process-Enhanced family engagement-Improved patient compliance and treatment outcome-Reduced staff time and cost for treatment
65
Leve
ragi
ng te
chno
logy
to
enha
nce
com
mun
icat
ion
in
heal
thca
re
Pro
ject
Lea
ders
: Eva
Lee
Stud
ent
Nam
e(s)
: Cod
y W
ang
Enga
ged
IAB
mem
bers
: Gra
dy, C
hild
ren'
s: H
ealth
care
of A
tlant
a, M
oreh
ouse
Sch
ool o
f M
edic
ine,
Res
tore
Med
ical
Sol
utio
ns
Enga
ged
CH
OT
Uni
vers
ity S
ites:
GIT
Proj
ect O
verv
iew
and
Des
crip
tion
Obj
ectiv
es:
Obj
ectiv
e1:
Impr
ove
com
mun
icat
ion,
com
plia
nce,
and
qual
ityof
care
thro
ugh
auto
mat
edm
achi
netra
nsla
tion.
Initi
alfo
cus
will
beth
edi
scha
rge
proc
edur
e.
OB
JEC
TIV
ES
& P
AR
TNE
RS
: Des
crip
tion
of P
roje
ct
Obj
ectiv
e2:
Des
ign
chat
box
and
virtu
alm
essa
ging
toen
hanc
efa
mily
enga
gem
ent,
know
ledg
edi
ssem
inat
ion
(e.g
.,fe
edin
g pl
an,c
ompl
ianc
e an
d aw
aren
ess
ofho
spita
lacq
uire
din
fect
ion,
child
heal
th).
OB
JEC
TIV
ES
& P
AR
TNE
RS
: Des
crip
tion
of P
roje
ct
Part
ners
/ R
elev
ance
:•
Car
e pr
ovid
ers
will
be
equi
pped
with
tool
s to
com
bat b
reak
dow
ns in
com
mun
icat
ion
with
refe
rrin
g ph
ysic
ians
and
to u
ltim
atel
y pl
ay a
gre
ater
role
in im
prov
ed p
atie
nt c
are
•Im
prov
e pa
tient
exp
erie
nce,
MR
I util
izat
ion,
and
sch
edul
ing
effic
ienc
y
OB
JEC
TIV
ES
& P
AR
TNE
RS
: Des
crip
tion
of P
roje
ct
66
Proj
ect O
verv
iew
and
Des
crip
tion
Wha
t pro
blem
is th
is p
roje
ct s
eeki
ng to
ad
dres
s?•
Util
izes
mac
hine
lear
ning
, tex
t min
ing,
and
dee
p le
arni
ngte
chni
ques
to h
ospi
tal d
isch
arge
not
es to
bui
ld a
nac
cura
te a
utom
atic
tran
slat
ion
syst
em w
hich
will
faci
litat
edi
scha
rge
and
hom
e ca
re p
roce
ss d
esig
n, p
artic
ular
ly fo
rno
n-E
nglis
h sp
eaki
ng p
atie
nts
•P
rovi
de a
saf
e, p
atie
nt-c
ente
red
back
-up
syst
em a
nden
sure
that
act
iona
ble
resu
lts a
re n
ot o
verlo
oked
OB
JEC
TIV
ES
& P
AR
TNE
RS
: Des
ign
Thin
king
App
roac
hPr
ojec
t Ove
rvie
w a
nd D
escr
iptio
n
Res
earc
h Pe
rspe
ctiv
e
OB
JEC
TIV
ES
& P
AR
TNE
RS
: Des
ign
Thin
king
App
roac
h
Pat
ient
Bet
ter
inte
rpre
tatio
n of
re
com
men
datio
ns
and
hom
e ca
re
proc
ess
Red
uced
leng
th
of s
tay
Bet
ter t
reat
men
t ou
tcom
e
Pro
vide
rR
educ
tion
of s
taff
time
Impr
oved
hom
e ca
re p
roce
ssIm
prov
ed c
are
coor
dina
tion
Hea
lth
syst
emIm
prov
ed
effic
ienc
yIm
prov
ed
utiliz
atio
n of
te
chno
logy
Pay
erR
educ
tion
of c
ost
Proj
ect O
verv
iew
and
Des
crip
tion
Con
trib
utio
n to
Indu
stry
and
Aca
dem
ia•
This
stud
y is
the
first
stu
dyw
hich
utili
zes
mac
hine
lear
ning
,tex
tmin
ing,
and
deep
lear
ning
te
chni
ques
toho
spita
l dis
char
ge n
otes
tobu
ild
anac
cura
te a
utom
atic
tran
slat
ion
syst
emw
hich
w
illfa
cilit
ate
disc
harg
ean
dho
me
care
proc
ess
desi
gn,p
artic
ular
ly fo
r non
-Eng
lish
spea
king
pa
tient
s. In
addi
tion,
it in
corp
orat
essy
stem
de
sign
and
hum
an-d
evic
e in
tera
ctio
n te
chno
logi
esto
offe
r rea
l-tim
ede
cisi
onsu
ppor
t to
pro
vide
rs.
CO
NTR
IBU
TIO
N: H
ow is
this
diff
eren
t tha
n re
late
d re
sear
ch?
App
roac
hA
PP
RO
AC
H: E
xper
imen
tal A
ppro
ach
•S
yste
mat
ic li
tera
ture
revi
ewof
a) e
xist
ing
mac
hine
tra
nsla
tion
algo
rithm
san
d ac
cura
cies
, b) i
ndic
ator
sof
la
ck o
f med
ical
tran
slat
ion
expe
rts a
nd w
here
the
mos
t ne
eds
fort
hem
are
•D
evel
op a
mac
hine
lear
ning
/nat
ural
lang
uage
pr
oces
sing
fram
ewor
kus
ing
hosp
itald
isch
arge
not
es
that
can
aut
omat
ical
ly a
nd a
ccur
atel
y tra
nsla
te c
linic
al
docu
men
tsto
fore
ign
lang
uage
s•
Eva
luat
ion
and
refin
emen
t of t
he tr
ansl
atio
n sy
stem
•D
esig
n an
d im
plem
enta
tion
of c
hatb
oxan
dvi
rtual
mes
sagi
ng s
yste
ms
67
App
roac
hO
VE
RV
IEW
OF
RE
SU
LTS
: Exp
ecte
d M
ilest
ones
•Id
entif
icat
ion
and
sum
mar
y of
gap
s in
hom
e ca
repr
oces
s, in
par
ticul
ar th
e di
scha
rge
proc
ess
•D
evel
op a
sec
ure
onlin
e au
tom
atic
tran
slat
ion
syst
emfo
rhos
pita
l dis
char
ge n
otes
•A
n ev
alua
tion
syst
em b
ased
on
BLE
U to
ass
ess
the
qual
ity o
f tra
nsla
ted
docu
men
ts•
Des
ign
and
impl
emen
tatio
n of
virt
ual m
essa
ging
syst
ems
p
•E
nhan
ced
care
coo
rdin
atio
n pl
an•
Enh
ance
wor
kfor
ce c
apab
ility
in tr
ansl
atio
n•
Impr
oved
dis
char
ged
and
hom
e ca
re p
roce
ss•
Enh
ance
d fa
mily
eng
agem
ent
•Im
prov
ed p
atie
nt c
ompl
ianc
e an
d tre
atm
ent
outc
ome
•R
educ
ed s
taff
time
and
cost
for t
reat
men
t•
Impr
ove
patie
nt-p
rovi
der t
rust
and
rela
tions
hip
OV
ER
VIE
W O
F R
ES
ULT
S: B
enef
its to
Indu
stry
68
PI:
Data-driven analytics and machine learning for improving healthcare outcome
Tech3 Lee (GT), Mastrangelo (UW), Tucker (PSU)
Analytics & Innovative Technologies
$150,000 YES 1
Data-driven healthcare has the potential to revolutionize care delivery and trim costs. A major challenge is that providers must siftthrough and analyze mountains of disparate data to materialize the substantial gain. We continue our healthcare innovation throughsystems and data analytics. Utilizing EMR and various procedural and personal health data, along with social and behavioralinformation, we will address all aims - with specific regard to radiologic exam variability. This also had implications in utilizingpredictive models to use at the point-of-care when treating infectious disease.
1) Develop predictive models for KPI’s (Exam Duration, Idle Time,Ratio of Repeated Scans), using logfile variables; 2) Define bettersequence of scans per exam (exam cards); 3) Design pilot projectimproving exam cards; 4) Develop models that predict machineutilization efficiency; 5) Design evidence-based expert decisionsupport system and optimize personalized treatment plans.
This is the first study where 1) large amounts of patient data are extracted unbiased and globally analyzed, 2) automated encryption ofPHI and data integration through terminology mapping is achieved using natural language processing, 3) time series clustering isdone with consideration of disease progression despite sparse and missing data, 4) discriminatory factors that inform key decisionsare systematically selected using machine learning, 5) individual patient conditions are addressed with the design of personalizedevidence-based treatment methods, and 6) this research will be able to be replicated to other cases and sites to improve theirprocess.
1. Conduct & benchmark literature review2. Data collection, extraction, and encryption of PHI3. Data cleaning and integration across types of records and multiple sites4. Build predictive models using machine learning and derive knowledge for personalized treatment, resource utilization andtreatment procedure optimization for best outcome5. Run pilot projects with optimized procedures and analyze preliminary findings to further refine processes
Objective 1: Use MRI log file data to identify variability and "wasted" time opportunities and to develop predictive models of examduration, idle time, and repeated scansObjective 2: Leverage the size and availability of population health data to model and predict machine utilization efficiencyObjective 3: Apply machine learning techniques to electronic Health Records containing patient demographics, labs, medications,procedures, and clinical notes to establish health trends and uncover definitive factors that can predict treatment outcome and optimalpatient care characteristicsObjective 4: Design evidence-based expert decision support system to facilitate early diagnosis, optimize and personalize treatment,and ensure safety and reduce errors to provide high quality outcome
1) Industry practitioners can make more informed decisionsand achieve care that is personalized, timely, evidence-based,and appropriate; 2) Optimize the usage of hospital resources,treatment process, and outcome; 3) Reduce waste, risk, andcost associated with procedures and operations; 4) Thisresearch can be replicated to other cases and sites to improveprocess.
69
Dat
a-d
rive
n a
nal
ytic
s an
d
mac
hin
e le
arn
ing
fo
r im
pro
vin
g
hea
lth
care
ou
tco
me
Pro
ject
Lea
ders
: Eva
Lee
Co-
lead
ers:
Con
rad
Tuc
ker,
Chr
isti
na M
astr
ange
lo
Stud
ent
Nam
e(s)
: TB
D
Enga
ged
IAB
mem
bers
: Col
labo
rativ
e
Enga
ged
CH
OT
Uni
vers
ity S
ites:
Col
labo
rativ
e
Pro
ject
Ove
rvie
w a
nd
Des
crip
tio
nO
bje
ctiv
es:
Obj
ectiv
e 1:
Use
MR
I log
file
dat
a to
iden
tify
varia
bilit
y an
d “w
aste
d” ti
me
oppo
rtun
ities
and
to d
evel
op p
redi
ctiv
e m
odel
s of
exa
m d
urat
ion
, idl
e tim
e, a
nd
repe
ated
sca
ns
OB
JEC
TIV
ES
& P
AR
TN
ER
S:
Des
crip
tion
of P
roje
ct
p
Pro
ject
Ove
rvie
w a
nd
Des
crip
tio
n
Obj
ectiv
e 2:
Lev
erag
e th
e si
ze a
nd a
vaila
bilit
y of
pop
ulat
ion
hea
lth d
ata
to m
odel
an
d pr
edic
t mac
hine
util
izat
ion
effic
ienc
y
OB
JEC
TIV
ES
& P
AR
TN
ER
S:
Des
crip
tion
of P
roje
ctP
roje
ct O
verv
iew
an
d D
escr
ipti
on
Obj
ectiv
e 3:
App
ly m
achi
ne le
arni
ng te
chni
ques
to e
lect
roni
c H
eal
th R
ecor
ds
cont
aini
ng p
atie
nt d
emog
raph
ics,
labs
, med
icat
ions
, pro
cedu
res,
and
clin
ical
no
tes
to e
stab
lish
heal
th tr
ends
and
unc
over
def
initi
ve fa
ctor
s th
at c
an p
redi
ct
trea
tmen
t out
com
e an
dop
timal
pat
ient
car
e ch
arac
teris
tics
OB
JEC
TIV
ES
& P
AR
TN
ER
S:
Des
crip
tion
of P
roje
ct
70
Pro
ject
Ove
rvie
w a
nd
Des
crip
tio
n
Obj
ectiv
e 4:
Des
ign
evid
ence
-bas
ed e
xper
t dec
isio
n su
ppor
t sys
tem
to fa
cilit
ate
early
dia
gnos
is, o
ptim
ize
and
pers
onal
ize
trea
tmen
t, an
d en
sure
saf
ety
and
redu
ce e
rror
s to
pro
vide
hig
h qu
ality
out
com
e
OB
JEC
TIV
ES
& P
AR
TN
ER
S:
Des
crip
tion
of P
roje
ctP
roje
ct O
verv
iew
an
d D
escr
ipti
on
Par
tner
s / R
elev
ance
:•
Est
ablis
h po
pula
tion
heal
th m
odel
that
pre
dict
s m
achi
ne u
tiliz
atio
n ef
ficie
ncy
•Im
prov
e pa
tient
exp
erie
nce,
MR
I util
izat
ion,
and
sch
edul
ing
effic
ienc
y
•In
crea
se e
ffici
ency
of r
esou
rce
allo
catio
n an
d re
duce
cos
ts
OB
JEC
TIV
ES
& P
AR
TN
ER
S:
Des
crip
tion
of P
roje
ct
Pro
ject
Ove
rvie
w a
nd
Des
crip
tio
n
Wh
at p
rob
lem
is t
his
pro
ject
see
kin
g t
o
add
ress
?•
Incr
ease
util
izat
ion
of la
rge
amou
nts
of d
ispa
rate
med
ical
data
incl
udin
g M
RI i
mag
ing
logs
, tre
atm
ent
proc
edur
es,
dem
ogra
phic
s, a
nd s
ocia
l-and
-beh
avio
ral i
nfor
mat
ion
•O
ptim
ize
the
usag
e of
hos
pita
l res
ourc
es,
trea
tmen
tpr
oces
s, a
nd o
utco
me
•A
ddre
ss in
divi
dual
pat
ient
con
ditio
ns f
or e
arly
dia
gnos
is,
and
desi
gn o
f per
sona
lized
evi
denc
e-ba
sed
trea
tmen
t
•R
educ
e in
fect
ions
and
sur
gery
err
ors
and
prov
ide
best
qual
ity t
reat
men
t pr
oced
ures
OB
JEC
TIV
ES
& P
AR
TN
ER
S:
Des
ign
Thi
nkin
g A
ppro
ach
Pro
ject
Ove
rvie
w a
nd
Des
crip
tio
n
Res
earc
h P
ersp
ecti
ve
OB
JEC
TIV
ES
& P
AR
TN
ER
S:
Des
ign
Thi
nkin
g A
ppro
ach
Pat
ient
Indi
vidu
aliz
ed t
reat
men
t ex
perie
nce
and
trea
tmen
t ou
tcom
e
Red
uctio
n of
hos
pita
l sta
y
Red
uced
ris
k fo
r in
fect
ion
Pro
vide
r
Red
uctio
n of
cos
t
Red
uctio
n of
tim
e
Bet
ter
sche
dule
pla
nnin
g
Hea
lth s
yste
m
Less
var
iabi
lity
Red
uctio
n of
tim
e
Impr
oved
util
izat
ion
of
med
ical
dat
a
Pay
er
Red
uctio
n of
cos
t
71
Pro
ject
Ove
rvie
w a
nd
Des
crip
tio
n
Co
ntr
ibu
tio
n t
o In
du
stry
an
d A
cad
emia
•C
are
team
s ca
n m
ake
mor
ein
form
edde
cisi
ons
and
achi
eve
pers
onal
ized
,tim
ely,
evid
ence
-bas
ed,
and
appr
opria
teca
re•
Bet
ter
outc
omes
are
achi
eved
ata
low
erco
st,
satis
fyin
gth
e ne
eds
ofal
lsta
keho
lder
s•
Hea
lthca
rere
sear
cher
sca
n be
bette
r eq
uipp
ed w
ithkn
owle
dge
and
tool
sto
ana
lyze
and
ben
efit
from
med
ical
data
•P
atie
nts
rece
ive
bette
r qu
ality
car
e•
Jour
nalp
aper
san
dco
nfer
ence
pres
enta
tion
for
broa
ddi
ssem
inat
ion
CO
NT
RIB
UT
ION
: How
is th
is d
iffer
ent
than
rel
ated
res
earc
h?A
pp
roac
hA
PP
RO
AC
H:
Exp
erim
enta
l App
roac
h
Ben
chm
ark
revi
ew
Dat
a co
llect
ion
and
extr
actio
n
Dat
a cl
eani
ng a
nd in
tegr
atio
n
Bui
ld p
redi
ctiv
e m
odel
s us
ing
mac
hine
lear
ning
and
der
ive
know
ledg
e fo
r pe
rson
aliz
ed tr
eatm
ent,
res
ourc
e ut
iliza
tion
and
trea
tmen
t pro
cedu
re o
ptim
izat
ion
for
best
out
com
e
Run
pilo
tpro
ject
sw
ithop
timiz
edpr
oced
ures
Ap
pro
ach
OV
ER
VIE
W O
F R
ES
ULT
S:
Exp
ecte
d M
ilest
ones
2018
2019
pppppppppppp
oac
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
Dec
2019
Feb
Mar
5/1/
2018
-7/
31/2
018
Deve
lop
pred
ictiv
e m
odel
s
4/20
/201
8 -7
/31/
2018
Run
pilo
t pro
gram
7/1/
2018
-8/
31/2
018
Anal
ysis
of th
e pi
lot p
rogr
am
9/1/
2018
-10
/31/
2018
Final
ize ro
ot ca
use(
s) li
st a
nd d
evel
op a
ctio
n pl
an to
max
imize
MR
syst
em u
tility
11/1
/201
8 -
3/31
/201
9Im
plem
ent a
ctio
n(s)
and
cond
uct s
ubse
quen
t ana
lyse
sAp
rd
May
o uJun
pJu
li Au
gs
Sep
i Oct
8
4/1/
2018
-6/
1/20
18Lit
erat
ure
revi
ew
5/1/
2018
-9/
1/20
18Co
nduc
t exp
erim
ents
8/1/
2018
-3/
1/20
19Di
ssem
inat
e fin
ding
s
Obj
ectiv
e 1
Obj
ectiv
e 2
Obj
ectiv
e 3
Jul
4/1/
2018
-5/
1/20
18EM
R da
ta e
xtra
ctio
n, cl
eani
ng, a
nd in
tegr
atio
n
5/1/
2018
-8/
15/2
018
Esta
blish
dise
ase
and
heal
th tr
ends
via
unsu
perv
ised
and
sem
i-sup
ervi
sed
lear
ning
8/1/
2018
-12
/1/2
018
Unco
ver p
atte
rns o
f car
e ch
arac
teris
tics v
ia su
perv
ised
lear
ning
e ii
11/1
/201
8 -2
/1/2
019
Desig
n ev
iden
ce-b
ased
exp
ert d
ecisi
on su
ppor
t sys
tem
12/1
/201
8 -3
/1/2
019
Desig
n an
d op
timize
per
sona
lized
trea
tmen
t pla
ns
2/1/
2019
-3/
30/2
019
Diss
emin
ate
know
ledg
e
Obj
ectiv
e 4
Ap
pro
ach
•In
dust
ry p
ract
ition
ers
can
mak
e m
ore
info
rmed
dec
isio
ns a
nd a
chie
ve c
are
that
is p
erso
naliz
ed, t
imel
y, e
vide
nce-
base
d,an
d ap
prop
riate
•R
educ
e w
aste
, ris
k, a
nd c
ost a
ssoc
iate
dw
ith p
roce
dure
s an
d op
erat
ions
•T
his
rese
arch
can
be
repl
icat
ed to
oth
erca
ses
and
site
s to
impr
ove
qual
ity a
ndpr
oces
s
OV
ER
VIE
W O
F R
ES
ULT
S:
Ben
efits
to In
dust
ry
72
Theme Champions:
73
74
PI:
Care Coordination Activities for Individuals with Spinal Cord Injury
Patient1 Tapan Mehta, Allyson Hall
Patient Experience
$50,000 NO 2
This is Phase 2 of a project designed to develop and pilot-test a care coordination program for people newly diagnosed with spinalcord injury (SCI). Phase 1 activities, which are currently underway, focused on developing the care coordination program. This phaseincludes (1) a review of relevant literature and (2) in-depth interviews with patients with SCI, their caregivers, physicians, other healthcare workers who specialize in SCI, and staff at the Lakeshore Foundation. Based on findings from these two activities, a pilotintervention will be developed. Phase 2 will implement the care coordination program developed in Phase 1 and assess the extent towhich it improves the quality of life of participants.
Several studies have documented the effectiveness of care coordination/transitions of care activities. Most of these studies focus onthe general population and do not address the specific and unique needs of individuals newly diagnosed with SCI. The proposedproject aims at addressing the needs of individuals with significant mobility limitations. In additions, the project specifically addresseshow a local disability focused community organization can partner with an academic medical center to improve the quality of life ofindividuals with SCI.
Phase 2 will be a mixed methods study with the following components:(1) Participant self-reported assessments of their quality of life using validated instruments from the PROMS inventory.
- Pre/Post comparisons among individuals receiving care coordination- Comparisons between individuals who did and did not receive care coordination
(2) Compare hospital use between intervention and control groups(3) In-depth interviews with both groups. These interviews will focus on care experiences following their SCI diagnosis.
Implementation of care coordination interventionQuality of life assessments and hospital data acquisitionAnalysis on hospital use and quality of lifeIn-depth interviewsFinal reportPeer review manuscript
Evidence of the effectiveness of a care-coordination programon improving the quality of life of SCI patients
Final report documenting study findingsSubmitted manuscript for publication
75
Car
e C
oo
rdin
atio
n A
ctiv
itie
s fo
r In
div
idu
als
wit
h S
pin
al C
ord
Inju
ry
Pro
ject
Lea
ders
: Tap
an M
ehta
, Ph.
D., A
llyso
n G
. Hal
l, Ph
.D.,
Rob
ert W
eech
-M
aldo
nado
, MBA
, Ph.
D.
Stud
ent
Nam
e(s)
: Ree
na Jo
seph
, MH
A, N
eera
j Pur
o, M
HA
, Gan
ishe
r D
avly
atov
, R
ache
l Wis
niew
ski,
Nat
han
Miy
asak
i
Enga
ged
IAB
mem
bers
: Lak
esho
re F
ound
atio
n, U
AB
Hea
lth S
yste
m
Enga
ged
CH
OT
Uni
vers
ity S
ites:
The
Uni
vers
ity o
f Ala
bam
a at
Bir
min
gham
Ob
ject
ive:
•In
divi
dual
s w
ith tr
aum
atic
spi
nal c
ord
inju
ries
(SC
Is)
face
uni
que
chal
leng
es in
rei
nteg
ratin
g in
to s
ocie
ty a
nd h
ave
a hi
gher
ris
k of
un
plan
ned
emer
genc
y de
part
men
t vis
its a
nd h
ospi
tal r
eadm
issi
ons
•C
are
coor
dina
tion
and
care
tran
sitio
n m
odel
s de
velo
ped
for
thes
e pa
tient
s ha
ve m
ixed
res
ults
•T
his
proj
ect w
ill d
evel
op a
car
e co
ordi
natio
n m
odel
for
indi
vid
uals
ne
wly
dis
char
ged
from
an
inpa
tient
set
ting.
Par
tner
s / R
elev
ance
:
•P
rovi
ding
tar
gete
d re
sear
ch th
at is
mea
ning
ful a
nd a
pplic
able
to
the
heal
th s
yste
m to
add
ress
the
need
s of
this
vul
nera
ble
popu
latio
n
•D
emon
stra
te fe
asib
ility
of a
car
e co
ordi
natio
n pr
ojec
t bet
wee
n he
alth
sy
stem
s an
d co
mm
unity
org
aniz
atio
ns fo
r pe
ople
with
SC
I
OB
JEC
TIV
ES
& P
AR
TN
ER
S:
Des
crip
tion
of P
roje
ct
Pro
ject
Ove
rvie
w a
nd
Des
crip
tio
nS
cope
of t
he p
robl
em•
~28
8,00
0 pe
ople
live
with
SC
I
•17
,700
new
cas
es e
ach
year
•~
30%
are
re-
hosp
italiz
ed o
ne o
r m
ore
times
dur
ing
the
year
follo
win
g in
jury
SC
I pat
ient
s ne
ed:
•F
requ
ent,
spec
ializ
ed, i
nter
disc
iplin
ary
heal
thca
re
•C
onne
cts
with
in th
e co
mm
unity
to
prov
ide
soci
al a
nd o
ther
sup
port
s
OB
JEC
TIV
ES
& P
AR
TN
ER
S: D
esig
n T
hink
ing
App
roac
h
Nat
iona
l Spi
nal C
ord
Inju
ry S
tatis
tical
Cen
ter,
Fac
ts a
nd F
igur
es a
t a G
lanc
e. B
irmin
gham
, AL:
Uni
vers
ity o
f Ala
bam
a at
B
irmin
gham
, 201
8.
Pro
ject
Ove
rvie
w a
nd
Des
crip
tio
nR
esea
rch
Per
spec
tive
•To
impr
ove
the
qual
ity o
f life
and
com
mun
ity in
tegr
atio
n fo
r pa
tient
s fo
llow
ing
disc
harg
e fr
om th
e re
habi
litat
ion
faci
lity
•R
educ
ing
read
mis
sion
s by
impr
ovin
g ca
re c
oord
inat
ion
Co
ntr
ibu
tio
n t
o In
du
stry
an
d A
cad
emia
•To
dem
onst
rate
the
feas
ibili
ty o
f a c
are
coor
dina
tion
proj
ect
focu
sed
on p
atie
nts
with
SC
Is
•T
he s
peci
fic n
eeds
of t
his
popu
latio
n ha
ve n
ot b
een
adeq
uate
ly r
esea
rche
d
•V
iabi
lity
of a
par
tner
ship
bet
wee
n lo
cal d
isab
ility
foc
used
co
mm
unity
org
aniz
atio
n an
d ac
adem
ic m
edic
al c
ente
r
OB
JEC
TIV
ES
& P
AR
TN
ER
S:
Des
ign
Thi
nkin
g A
ppro
ach
76
Ap
pro
ach
Ph
ase
I (C
urr
entl
y O
ng
oin
g)
Goa
l: D
evel
op In
terv
entio
n
•Li
tera
ture
Rev
iew
•Q
ualit
ativ
e in
terv
iew
s•
Pat
ient
s w
ith S
CI
•P
rovi
ders
•S
tatis
tical
ana
lyse
s to
det
erm
ine
pred
icto
rs o
f rea
dmis
sion
AP
PR
OA
CH
: E
xper
imen
tal A
ppro
ach
Ap
pro
ach
Pro
po
sed
Ph
ase
II
•P
ilot-
test
car
e co
ordi
natio
n pr
ogra
m f
or p
atie
nts
new
ly
diag
nose
d w
ith S
CI
–A
ssis
tanc
e w
ith o
btai
ning
nee
ded
heal
th c
are,
reh
ab, c
omm
unity
su
ppor
ts a
nd c
omm
unity
eng
agem
ent
–F
ocus
on
care
coo
rdin
atio
n w
ith th
e La
kesh
ore
Fou
ndat
ion
•M
ixed
met
hods
eva
luat
ion
with
the
follo
win
g co
mpo
nent
s:–
Pre
/pos
t com
paris
ons
amon
g in
divi
dual
s re
ceiv
ing
care
co
ordi
natio
n ve
rsus
thos
e no
t rec
eivi
ng c
are
coor
dina
tion
•C
ompa
re h
ospi
tal u
se b
etw
een
inte
rven
tion
and
cont
rol g
roup
s
•P
artic
ipan
t sel
f-re
port
ed a
sses
smen
ts o
f the
ir qu
ality
of l
ife u
sing
val
idat
ed
inst
rum
ents
(P
RO
MIS
)
–In
-dep
th in
terv
iew
s w
ith b
oth
grou
ps•
Car
e ex
perie
nces
follo
win
g th
eir
SC
I dia
gnos
is.
AP
PR
OA
CH
: E
xper
imen
tal A
ppro
ach
Ap
pro
ach
Fin
aliz
e ca
re-c
oord
inat
ion
inte
rven
tion
Obt
ain
IRB
app
rova
l
Enr
oll s
tudy
par
ticip
ants
•In
terv
entio
n an
d co
ntro
l
Sta
tistic
al a
naly
sis
•C
ompa
re h
ospi
taliz
atio
ns a
nd q
ualit
y of
life
Con
duct
in-p
erso
n in
terv
iew
with
stu
dy p
artic
ipan
ts•
The
mat
ic a
naly
sis
Fin
al r
epor
ts a
nd p
ublic
atio
ns
OV
ER
VIE
W O
F R
ES
ULT
S:
Exp
ecte
d M
ilest
ones
Ap
pro
ach
Del
iver
able
s•
Fin
al r
epor
t doc
umen
ting
stud
y fin
ding
s
•P
eer-
revi
ewed
man
uscr
ipt
for
publ
icat
ion
in a
n ac
adem
ic
jour
nal
Imp
ort
ance
to
ind
ust
ry
•E
vide
nce
of th
e ef
fect
iven
ess
of a
car
e-co
ordi
natio
n pr
ogra
m o
n im
prov
ing
the
qual
ity o
f life
of
patie
nts
with
S
CIs
•F
easi
bilit
y of
fost
erin
g co
mm
unity
-bas
ed p
artn
ersh
ips
OV
ER
VIE
W O
F R
ES
ULT
S:
Ben
efits
to In
dust
ry
77
78
PI:
Generating Tailored Recommendations Automatically with Explanations via an Interactive Dialog-based System
Patient2 Prasenjit Mitra
Patient Experience
$50,000 NO 1
While certain medical diagnostic and treatment tasks can be automated, others require the decision-making by a medical practitioner.In such cases, a well-designed automated system can be of great value to the practitioner. The current state-of-the-art involvesretrieval systems that largely perform basic information retrieval. In this project, we seek to go further. Based on the level ofinteraction desired by the practitioner, our tool will take in a textual input ranging from a few keywords to all the notes available byinterviewing the patient during intake and existing past notes. Using this textual input, the system will request additional information orprovide recommendations for further investigation or treatment. The practitioner will be provided a clear explanation as to how therecommendations were generated quoting existing guidelines or best-practices in the field. Furthermore, using the latest developmentin chatbot technology, the system will enable the practitioner to refine the information and/or the treatment plan. Based on theseinteractions with the practitioner, the system will learn and adapt to make future interactions with a particular practitioner, similarpractitioners and even all practitioners better.
Currently, there exists no system that extracts relevant information with high accuracy, learns from interactions, provides explanationsabout how the suggestion was generated, and adapts using conversations to retrieve different suggestions. To generate therecommendations, our system will use (but require substantial retooling of) state-of-the-art technology on adaptive summarization thatwe have successfully used in generating Wikipedia articles. Generating explanations of why information was chosen succinctly will bedone using state-of-the-art deep learning techniques. To the best of our knowledge, explaining recommendations using deep-learninghas not been addressed in a medical recommendation system scenario. Adaptive recommendations based on dialogs using deeplearning has not been addressed either. Finally, an integrated system that provides summarized information, a tailoredrecommendation, explanations of recommendations, adapts recommendations based on dialog with high accuracy and usersatisfaction has not been designed or demonstrated. Our system will not be the final answer in one year but it will be one step closer.
1. Retool our abstractive summarizer to generate summarized recommendations based on textual input. Evaluate output usingdomain experts and adapt the information selection and summarization algorithms based on their inputs.2. Generate an explanation of the recommendations. Provide a link to the other alternatives and allow a domain expert to indicatewhether the other alternatives should have been selected instead. Perform user study to rate the quality of the explanation and theaccuracy of the explanations. Find where the explanations were lacking and try to regenerate explanations that are lacking.3. Re-purpose chatbot technology and couple it with our system to interact with medical practitioners. Evaluate if the interactionbetween the practitioner and chatbot helps refine the recommendation & increases the satisfaction of the end-users via user study.4. Evaluate the entire integrated system for improvement in productivity and the quality of care provided by the medical practitioner.Detect the shortcomings of the system for the next round of research to improve interactive medical recommendation systems.
Month 1-3: Generation of summaries, guidelines, and recommendations automatically.Month 4: Evaluation of the generated summaries and adaptation of the algorithms. Generation of explanations for generated content.Month 5: Generation of summaries, guidelines, and recommendations with higher accuracy.Month 6: Evaluation of the explanations and identification of what needs to be improved.Month 7: Improve the explanation generation module.Month 8: Evaluate the explanation generation module and perform experiments to validate its efficacy.Month 9-11: Repurpose chatbot technology to create an interactive module. Use interactions with the end-user to refinerecommendations or explanations.Month 12: Evaluate the interactive dialog-based system. Evaluate overall system. Identify successes and failures for future research.
1. The basic technology that will be developed in this projectwill influence products being developed by the medicalinformatics industry in this area.2. The changes required to make the information output highquality and highly relevant may be useful in making the generaltechnology better. Then, this technology will be useful ingeneral to any user who wants to build an interactiverecommendation system that explains what it is doing.
1. A (very basic) prototypical software that generatesrecommendations and allows end-users to refine them viadialogue based on deep-learning. All code will be releasedopen-source.2. Systematic evaluation of the methods used to generateabstractive summaries, recommendations, and explanations.3. Papers describing the core technology and the improvementsmade to the systems utilized in this project.
While certain medical diagnostic and treatment tasks can be automated, others require the decision-making by a medical practitioner.In such cases, a well-designed automated system can be of great value to the practitioner. The current state-of-the-art involvesretrieval systems that largely perform basic information retrieval. In this project, we seek to go further. Based on the level ofinteraction desired by the practitioner, our tool will take in a textual input ranging from a few keywords to all the notes available byinterviewing the patient during intake and existing past notes. Using this textual input, the system will request additional information orprovide recommendations for further investigation or treatment. The practitioner will be provided a clear explanation as to how therecommendations were generated quoting existing guidelines or best-practices in the field. Furthermore, using the latest developmentin chatbot technology, the system will enable the practitioner to refine the information and/or the treatment plan. Based on theseinteractions with the practitioner, the system will learn and adapt to make future interactions with a particular practitioner, similarpractitioners and even all practitioners better.
Currently, there exists no system that extracts relevant information with high accuracy, learns from interactions, provides explanationsabout how the suggestion was generated, and adapts using conversations to retrieve different suggestions. To generate therecommendations, our system will use (but require substantial retooling of) state-of-the-art technology on adaptive summarization thatwe have successfully used in generating Wikipedia articles. Generating explanations of why information was chosen succinctly will bedone using state-of-the-art deep learning techniques. To the best of our knowledge, explaining recommendations using deep-learninghas not been addressed in a medical recommendation system scenario. Adaptive recommendations based on dialogs using deeplearning has not been addressed either. Finally, an integrated system that provides summarized information, a tailoredrecommendation, explanations of recommendations, adapts recommendations based on dialog with high accuracy and usersatisfaction has not been designed or demonstrated. Our system will not be the final answer in one year but it will be one step closer.
1. Retool our abstractive summarizer to generate summarized recommendations based on textual input. Evaluate output usingdomain experts and adapt the information selection and summarization algorithms based on their inputs.2. Generate an explanation of the recommendations. Provide a link to the other alternatives and allow a domain expert to indicatewhether the other alternatives should have been selected instead. Perform user study to rate the quality of the explanation and theaccuracy of the explanations. Find where the explanations were lacking and try to regenerate explanations that are lacking.3. Re-purpose chatbot technology and couple it with our system to interact with medical practitioners. Evaluate if the interactionbetween the practitioner and chatbot helps refine the recommendation & increases the satisfaction of the end-users via user study.4. Evaluate the entire integrated system for improvement in productivity and the quality of care provided by the medical practitioner.Detect the shortcomings of the system for the next round of research to improve interactive medical recommendation systems.
Month 1-3: Generation of summaries, guidelines, and recommendations automatically.Month 4: Evaluation of the generated summaries and adaptation of the algorithms. Generation of explanations for generated content.Month 5: Generation of summaries, guidelines, and recommendations with higher accuracy.Month 6: Evaluation of the explanations and identification of what needs to be improved.Month 7: Improve the explanation generation module.Month 8: Evaluate the explanation generation module and perform experiments to validate its efficacy.Month 9-11: Repurpose chatbot technology to create an interactive module. Use interactions with the end-user to refinerecommendations or explanations.Month 12: Evaluate the interactive dialog-based system. Evaluate overall system. Identify successes and failures for future research.
1. The basic technology that will be developed in this projectwill influence products being developed by the medicalinformatics industry in this area.2. The changes required to make the information output highquality and highly relevant may be useful in making the generaltechnology better. Then, this technology will be useful ingeneral to any user who wants to build an interactiverecommendation system that explains what it is doing.
79
Gen
erat
ing
Tailo
red
Rec
omm
enda
tions
Aut
omat
ical
ly
with
Exp
lana
tions
via
an
Inte
ract
ive
Dia
log-
base
d Sy
stem
Pro
ject
Lea
ders
: P
rase
njit
Mit
ra, P
enn
Stat
e C
olle
ge o
f Inf
orm
atio
n Sc
ienc
e an
d Te
chno
logy
Co-
lead
ers:
Con
rad
Tuck
er, P
enn
Stat
e C
olle
ge o
f Eng
inee
ring
Stud
ent
Nam
e(s)
: TB
D
Enga
ged
IAB
mem
bers
: AT
&T,
Sie
men
s
Enga
ged
CH
OT
Uni
vers
ity S
ites:
Penn
Sta
te
Proj
ect O
verv
iew
and
Des
crip
tion
Obj
ectiv
e:D
evel
op a
pro
toty
pe a
utom
ated
de
ep le
arni
ng-d
riven
sys
tem
. The
sy
stem
use
s te
xtua
l inp
ut fr
om
patie
nt re
cord
s du
ring
inta
ke a
nd
past
not
es to
gen
erat
e re
com
men
datio
ns, b
ased
on
best
-pr
actic
es in
the
field
.
Part
ners
/ R
elev
ance
:D
evel
op th
e ba
sic
tech
nolo
gy to
bu
ild in
tera
ctiv
e re
com
men
datio
n sy
stem
s th
at d
rive
futu
re m
edic
al
info
rmat
ics
prod
ucts
for
heal
thca
re p
rovi
ders
and
pr
actit
ione
rs.
OB
JEC
TIV
ES
& P
AR
TNE
RS
: Des
crip
tion
of P
roje
ct
Text
ual I
nput
from
pa
tient
reco
rds,
test
re
sults
, and
not
es
Adap
tabl
e de
ep
lear
ning
sys
tem
Rec
omm
enda
tions
of
fere
d to
aid
in
deci
sion
-mak
ing
Proj
ect O
verv
iew
and
Des
crip
tion
With
less
tim
e al
loca
ted
per
patie
nt, h
ealth
care
pro
vide
rs
mus
t qui
ckly
cap
ture
and
as
sess
pat
ient
dat
a, m
ake
deci
sion
s, a
nd c
omm
unic
ate
impo
rtant
info
rmat
ion
to
patie
nts.
Pra
ctiti
oner
s ne
ed to
incr
ease
ef
ficie
ncie
s to
incr
ease
the
time
spen
t on
valu
e -ad
ded
task
s th
at d
irect
ly re
late
to
patie
nt c
are.
OB
JEC
TIV
ES
& P
AR
TNE
RS
: Des
ign
Thin
king
App
roac
h
Less
tim
e pe
r pa
tient
can
lead
to
med
ical
err
ors
Patie
nt E
xpec
tatio
nsVs
.R
ealit
y
Wha
t pro
blem
is th
is p
roje
ct s
eeki
ng to
add
ress
?Pr
ojec
t Ove
rvie
w a
nd D
escr
iptio
nR
esea
rch
Pers
pect
ive
OB
JEC
TIV
ES
& P
AR
TNE
RS
: Des
ign
Thin
king
App
roac
h
•Im
prov
ed e
ffici
ency
in u
tiliz
ing
avai
labl
e da
ta fr
om e
lect
roni
c m
edic
al re
cord
s to
ass
ist
prac
titio
ners
in m
akin
g re
com
men
datio
ns fo
r tre
atm
ent.
•Le
ss ti
me
will
be
spen
t by
prac
titio
ners
on
non-
valu
e ad
ded
func
tions
to a
llow
mor
e tim
e fo
r pa
tient
s, e
nhan
cing
thei
r ex
perie
nce.
•R
educ
e th
e ch
ance
of c
ritic
al
info
rmat
ion
bein
g om
itted
from
th
e de
cisi
on-m
akin
g pr
oces
s.
80
Proj
ect O
verv
iew
and
Des
crip
tion
CO
NTR
IBU
TIO
N: H
ow is
this
diff
eren
t tha
n re
late
d re
sear
ch?
Con
trib
utio
n to
Indu
stry
and
A
cade
mia
Cur
rent
sta
teon
ly p
erfo
rms
basi
c in
form
atio
n re
triev
al in
the
med
ical
spa
ce.
Ther
e is
cur
rent
ly n
o sy
stem
that
:•
extra
cts
rele
vant
info
rmat
ion
with
hig
h ac
cura
cy•
lear
ns fr
om in
tera
ctio
ns•
expl
ains
how
sug
gest
ions
are
ge
nera
ted
•ad
apts
usi
ng c
onve
rsat
ions
to
retri
eve
diffe
rent
sug
gest
ions
Abs
tract
ive
sum
mar
izer
, D
eep
lear
ning
and
C
hatb
ox T
echn
olog
y
Inte
ract
ive
and
adap
tive
syst
em
that
gen
erat
es h
igh
qual
ity, r
elev
ant
reco
mm
enda
tionsB
asic
info
rmat
ion
Ret
rieva
l
App
roac
hA
PP
RO
AC
H: E
xper
imen
tal A
ppro
ach
Ret
ool a
bstra
ctiv
e su
mm
ariz
er to
ge
nera
te s
umm
ariz
ed
reco
mm
enda
tions
bas
ed o
n te
xtua
l in
put a
nd e
valu
ate
usin
g do
mai
n ex
perts
Gen
erat
e re
com
men
datio
n ex
plan
atio
ns w
ith s
ourc
e re
ferra
ls a
nd
have
dom
ain
expe
rts re
view
for
appr
opria
tene
ss
Enha
nce
chat
box
tech
nolo
gy fo
r us
e w
ith m
edic
al p
ract
ition
ers
and
eval
uate
if th
e in
tera
ctio
n in
crea
ses
satis
fact
ion
of e
nd-
user
s vi
a a
user
stu
dy
Eva
luat
e en
tire
inte
grat
ed
syst
em fo
r inc
reas
ed
prod
uctiv
ity a
nd q
ualit
y of
ca
re b
y m
edic
al
prac
titio
ners
Iden
tify
shor
tcom
ings
and
us
e ev
alua
tion
to fe
ed in
to
new
roun
d of
rese
arch
App
roac
hO
VE
RV
IEW
OF
RE
SU
LTS
: Exp
ecte
d M
ilest
ones
Mon
th 1
-3: G
ener
atio
n of
su
mm
arie
s, g
uide
lines
, an
d re
com
men
datio
ns
auto
mat
ical
ly
Mon
th 4
: Eva
luat
ion
of
gene
rate
d su
mm
arie
s an
d ad
aptio
n of
the
algo
rithm
s an
d ge
nera
tion
of
expl
anat
ions
for t
he
gene
rate
d co
nten
t
Mon
th 5
: Gen
erat
ion
of
sum
mar
ies,
gui
delin
es,
and
reco
mm
enda
tions
w
ith h
ighe
r acc
urac
y
Mon
th 6
: Eva
luat
ion
of
expl
anat
ions
and
id
entif
icat
ion
of a
reas
for
impr
ovem
ent
Mon
th 7
: Im
prov
e th
e ex
plan
atio
n ge
nera
tion
mod
ule
Mon
th 8
: Eva
luat
e th
e ex
plan
atio
n ge
nera
tion
mod
ule
and
perfo
rm
expe
rimen
ts to
val
idat
e ef
ficac
y
Mon
th: 9
-11:
Enh
ance
ch
atbo
xte
chno
logy
to
crea
te in
tera
ctiv
e m
odul
e,
use
end-
user
inte
ract
ions
to
refin
e
Mon
th 1
2: E
valu
ate
com
pone
nts
and
over
all
tech
nolo
gy, u
se fa
ilure
s fo
r fut
ure
rese
arch
App
roac
h
•P
roto
type
sys
tem
that
ge
nera
tes
reco
mm
enda
tions
an
d al
low
s en
d-us
ers
to
refin
e vi
a di
alog
ue b
ased
on
deep
lear
ning
•S
yste
mat
ic e
valu
atio
n of
m
etho
ds u
sed
to g
ener
ate
abst
ract
ive
sum
mar
ies,
re
com
men
datio
ns, a
nd
expl
anat
ions
•P
aper
s de
scrib
ing
the
core
te
chno
logy
•P
roto
type
sof
twar
e th
at c
an
be in
tegr
ated
into
hea
lthca
re
syst
ems
OV
ER
VIE
W O
F R
ES
ULT
S: B
enef
its to
Indu
stry
Appppppppppppppppppppppppppppppppppppp
rrrrrrrrrrrrrrrroooooooooooooaaaaaaaaaaaaaaaaacccccccccccccccc
hhhhhhhhhhhhhhhh
ns on
Pat
ient
dat
a su
mm
ary
com
pile
d.
Wou
ld y
ou li
ke to
vie
w
reco
mm
enda
tions
now
? Ye
s/ N
o
Yes
show
reco
mm
enda
tions
.
Bas
ed o
n th
e pa
tient
s m
edic
al re
cord
, cl
ick
here
for r
ecom
men
datio
ns.
Del
iver
able
s:
81
82
PI:
Embedding Routine Informal, Family Caregiver Assessment of Delirium Superimposed on Dementia into Acute Care
Patient3 Andrea Sillner
Patient Experience
$100,000 YES 1
The purpose of this pilot study is to assess initial accuracy and feasibility of communication of observed symptoms of delirium in olderadults with complex multiple chronic conditions dementia by family caregivers utilizing app-based delivery of the Family ConfusionAssessment Method (FAM-CAM) in the acute care setting.
Current standards in diagnosing delirium rely on diagnostic criteria of the Diagnostic and Statistical Manual of Mental Disorders, 5thEdition (DSM-5) and the International Statistical Classification of Disease and Related Health Problems, 10th Revision (ICD-10);however, there are no specific diagnostic criteria for delirium in persons with preexisting dementia. Recommended assessment toolsfor delirium, such as the Confusion Assessment Method (CAM), take into account changes from normal, but often this is unknown toformal healthcare providers. For older adults with complex multiple chronic conditions and cognitive impairment, the person who maybe best able to assess baseline cognitive function is the family caregiver. A modification of CAM, FAM-CAM, allows family caregiversto report their observations of symptoms of delirium in a standardized method. The FAM-CAM shows potential to improve recognition,and therefore, management of delirium in the acute care setting.
Aim 1: to assess the initial agreement of communication of observed symptoms of delirium in older adults with complex multiplechronic conditions by family caregivers utilizing app-based delivery of FAM-CAM compared to trained observers.
Aim 2: to determine feasibility of embedding the app-based FAM-CAM within the electronic medical record.
To meet the aims of the proposed project, the following steps will be undertaken:1) development of application based FAM-CAM for use within the acute care setting,2) assessment of the accuracy and feasibility of using this tool within the acute care setting, and3) determination of how to sync FAM-CAM with the acute care electronic medical record so that the communication of symptomsnoted by informal caregiver at the bedside is routinely and accurately delivered to healthcare providers, so that action can take place.
Understanding how we can allow informal family caregivers atthe bedside to routinely communicate observed signs andsymptoms of common hospital adverse events to medical staffusing app-based technology and standardized screeninginstruments.
Dissemination of findings by publications and presentations tostakeholders at all levels of care, including, but not limited tohealthcare providers, patients, informal caregivers, and industrypartners.
83
Embe
ddin
g R
outin
e In
form
al,
Fam
ily C
areg
iver
Ass
essm
ent o
f D
eliri
um S
uper
impo
sed
on
Dem
entia
into
Acu
te C
are
Pro
ject
Lea
ders
: And
rea
Yevc
hak
Silln
er, P
enn
Stat
e C
olle
ge o
f Nur
sing
Stud
ent
Nam
e(s)
: TB
D
Enga
ged
IAB
mem
bers
: Her
shey
Med
ical
Cen
ter,
Hig
hmar
k
Enga
ged
CH
OT
Uni
vers
ity S
ites:
FAU
/PSU
Proj
ect O
verv
iew
and
Des
crip
tion
Obj
ectiv
e:Ex
plor
e fe
asib
ility
and
accu
racy
of u
sing
dat
a fro
m
fam
ily c
areg
iver
sto
est
ablis
h ba
selin
e co
gniti
ve fu
nctio
n in
th
e ac
ute
care
set
ting
for o
lder
ad
ults
with
com
plex
mul
tiple
ch
roni
c co
nditi
ons
and
dem
entia
to in
crea
se th
e lik
elih
ood
of re
cogn
izin
g sy
mpt
oms
of d
eliri
um.
Part
ners
/ R
elev
ance
:G
ive
heal
thca
re p
rovi
ders
m
ore
data
obs
erva
tions
usi
ng
a st
anda
rdiz
ed m
etho
d fro
m
thos
e w
ho b
est k
now
the
patie
nt to
impr
ove
man
agem
ent a
nd o
utco
mes
.
OB
JEC
TIV
ES
& P
AR
TNE
RS
: Des
crip
tion
of P
roje
ctD
elir
ium
is a
stat
e of
wor
se-th
an-u
sual
m
enta
l con
fusio
n, b
roug
ht o
n by
som
e ty
pe
of u
nusu
al st
ress
on
the
body
or
min
d.
Diff
icul
ty fo
cusi
ngM
emor
y Pr
oble
ms
Vivi
d H
allu
cina
tions
Dis
orie
ntat
ion
Lang
uage
Diff
icul
tyR
estle
ssne
ss
Proj
ect O
verv
iew
and
Des
crip
tion
Wha
t pro
blem
is
this
pro
ject
see
king
to
add
ress
?C
urre
nt s
tand
ards
in
diag
nosi
ng d
eliri
um
rely
on
crite
ria th
at d
o no
t tak
e pr
eexi
stin
g de
men
tia in
to
acco
unt,
mak
ing
it di
fficu
lt to
ass
ess
base
line
cogn
itive
fu
nctio
n an
d m
onito
r sy
mpt
oms.
OB
JEC
TIV
ES
& P
AR
TNE
RS
: Des
ign
Thin
king
App
roac
h Sig
n of
oth
er
illnes
s or
stre
ssC
onfu
sed
pers
on
at h
ighe
r ris
k fo
r fa
lls a
nd in
jurie
sIn
crea
ses
leng
th o
f ho
spita
l sta
ys a
nd
chan
ce o
f dea
th.
Loss
of
inde
pend
ence
Acc
eler
ates
co
gniti
ve d
eclin
e
Del
irium
is a
n im
port
ant p
robl
em to
pr
even
t, de
tect
, and
man
age,
esp
ecia
lly
with
dem
entia
pat
ient
s.
rds
in
irium
th
at d
o st
ing
ng it
ss
tiv
e on
itor
Sig
nilln
esC
onfu
at h
igfa
lls a
Incr
eho
spi
chan
cLo
ssin
dep
Acc
eco
gni
Proj
ect O
verv
iew
and
Des
crip
tion
Res
earc
h Pe
rspe
ctiv
eH
ealth
car
e pr
ovid
ers,
usi
ng
info
rmal
fam
ily
care
give
r dat
a, w
ill be
ab
le to
det
ect
OB
JEC
TIV
ES
& P
AR
TNE
RS
: Des
ign
Thin
king
App
roac
h
eeeeeeeeeeeeeeeeee
delir
ium
sym
ptom
s ea
rlier
to g
uide
car
e de
cisi
ons
if it
can
be e
asily
cap
ture
d in
a
patie
nt’s
ele
ctro
nic
heal
th re
cord
.
84
Proj
ect O
verv
iew
and
Des
crip
tion
Con
trib
utio
n to
Indu
stry
and
A
cade
mia
•A
ddin
g th
e co
nstru
ct o
f the
Fam
ily
Con
fusi
on A
sses
smen
t Met
hod
(FA
M-C
AM
) dat
a w
ill al
low
he
alth
care
pro
vide
rs b
ette
r ins
ight
in
to tr
eatin
g de
liriu
m p
atie
nts
with
pr
eexi
stin
g de
men
tia.
•A
pp-b
ased
tech
nolo
gy, u
tiliz
ing
stan
dard
ized
scr
eeni
ng in
stru
men
ts,
will
enab
le m
ore
obse
rvat
ions
of
sym
ptom
s fro
m th
ose
who
bes
t kn
ow th
e pa
tient
’s b
asel
ine
cogn
itive
st
ate,
fam
ily c
areg
iver
s.
CO
NTR
IBU
TIO
N: H
ow is
this
diff
eren
t tha
n re
late
d re
sear
ch?
App
roac
hW
hat d
o yo
u pl
an o
n do
ing?
•A
IM 1
: Obt
ain
agre
emen
t of
com
mun
icat
ion
of
obse
rved
sym
ptom
s fro
m fa
mily
ca
regi
vers
com
pare
d to
trai
ned
obse
rver
s.•
AIM
2: D
eter
min
e fe
asib
ility
of
embe
ddin
g th
e ap
p-ba
sed
FAM
-CA
M
with
in th
e el
ectro
nic
med
ical
reco
rd.
AP
PR
OA
CH
: Exp
erim
enta
l App
roac
h
App
roac
hTi
mel
ine
and
Ove
rvie
w o
f Ex
pect
ed R
esul
ts•
Dev
elop
app
bas
ed o
n FA
M-
CA
M (m
onth
s 1-
3)
•A
sses
smen
t of a
ccur
acy
and
feas
ibili
ty (m
onth
s 4-
8)
•D
eter
min
atio
n of
how
to s
ync
FAM
-CA
M w
ith a
cute
car
e el
ectro
nic
med
ical
reco
rd fo
r ap
prop
riate
mon
itorin
g(m
onth
s 9-
12)
OV
ER
VIE
W O
F R
ES
ULT
S: E
xpec
ted
Mile
ston
espp of
n FA
M-
acy
and
) to s
ync
care
App
roac
h
Del
iver
able
s:•
Dis
sem
inat
ion
of
findi
ngs
by p
ublic
atio
ns
and
pres
enta
tions
to
stak
ehol
ders
•Fu
nctio
nal m
obile
app
th
at c
aptu
res
FAM
-CA
M
Dat
a
OV
ER
VIE
W O
F R
ES
ULT
S: B
enef
its to
Indu
stry
85
86
Research Theme #5:
ACCESS TO CARE
Theme Champion : Dr. Christopher Johnson, University of Louisville
Access 1
Telemedicine in Primary Care and in the Management of Chronic Conditions: Exploring Patient & Provider Perspectives Access 2 Ask Me 3®: A Home Health Intervention to Address Health Literacy Barriers, Increase Patient Engagement, and Improve Patient Experi-ence and Outcomes
87
88
PI:
Telemedicine in Primary Care & in the Management of Chronic Conditions: Exploring Patient & Provider Perspectives
Access1 Lee, Lerouge, Tucker, Borkowski, Kash, Johnson, Agarwal
Access to Care
$200,000 YES 2
Timely access to quality healthcare service is a real challenge—as outlined in the 2015 IOM report—and misalignment of resourcesand demands result in long delays for care. Telehealth can offer alternative and timely care to rural area patients who lack sufficienthealthcare options. Telehealth can also help to improve health conditions and to promote active patient engagement, which isparticularly important for chronic disease management. This project identifies drivers and barriers of patient engagement by populationgroups and chronic conditions and provides recommendations for implementing appropriate telehealth/telemedicine interventionsthrough multiple care settings given governmental policies, reimbursement payments, and delivery of care.
1) Technology readiness model for telemedicine in primary caresettings2) Conference presentations3) Piloted technology in Primary Care Assessment Survey4) Optimized point-of-access for study sites5) A low-cost personalized prototype remote patient monitoringdevice
The adoption of telemedicine and the level of patient engagement and services provided across healthcare facilities remain unevenand far from optimal. Little evidence, particularly in the form of understanding from the viewpoint and situation of providers, is availableto guide stakeholder organizations as they consider introducing telemedicine into primary care practice. This study examines issuesincluding point-of-access, administrative logistics, timely primary care, monitoring chronic disease and mental health, and providingequal and affordable care to the poor and rural areas. We also investigate and design a personalized remote patient monitoringsystem to connect patients and providers. By exploring successful application in multiple settings such as the rural and primary caresetting, this study will define the terms telehealth and telemedicine.
1) Literature review of existing primary care literature to identify (a) various iterations of the telemedicine service provision in primarycare contexts, (b) possible forces affecting adoption and innovation, and (c) indicators of technology readiness factors and pathwaysfor primary care practices to implement telemedicine services in primary care.2) Secondary data analysis of former CHOT landscape project to identify (a) various iterations of the telemedicine service provision inprimary care contexts and (b) possible forces affecting adoption and innovation.3) Create a low-level prototype showcasing the design of a personalized remote patient monitoring system. This system will integratetelehealth devices using a smartphone application to connect patients and providers.
1) Perform systematic literature review2) Identify gaps in care through gap analysis3) Data and system modeling including optimizing point-of-access4) Develop and administer survey instrument to be completed by providers and carry out pilot study5) Collect feedback from partner sites and conduct further analyses to characterize barriers to the use of telehealth services andassess their level of telemedicine readiness
1) Understanding the adoption and diffusion of telemedicine inprimary care can inform decision making regarding servicedesign, implementation, operations, and provider engagement.2) An assessment tool based on these forces can help assessindividual primary care organizational readiness fortelemedicine innovation to promote organizational success indelivering this mode of service.
89
Tel
emed
icin
ein
Pri
mar
y C
are
and
in
the
Man
agem
ent
of
Ch
ron
icC
on
dit
ion
s: E
xplo
rin
g P
atie
nt
&P
rovi
der
Per
spec
tive
s
Co-
lead
ers:
Eva
Lee
, Geo
rgia
Tec
h; C
ynth
ia L
eRou
ge, U
W,
Con
rad
Tuc
ker,
Pen
n St
ate.
Oth
er In
vest
igat
ors:
N B
orko
wsk
(U
AB
), B
Kas
h (T
AM
U),
C
John
son
(UL)
, A A
garw
al (
FAU
)
Enga
ged
IAB
mem
bers
: C
olla
bora
tive
Enga
ged
CH
OT
Uni
vers
ity S
ites:
Col
labo
rativ
e
Pro
ject
Ove
rvie
w a
nd
Des
crip
tio
n•
Exp
lore
and
def
ine
the
itera
tions
of t
elem
edic
ine
serv
ices
and
mod
els
in p
rimar
y ca
re a
nd fo
r th
e ca
re o
f chr
onic
dis
ease
•Id
entif
y dr
iver
s an
d ba
rrie
rs t
o pa
tient
eng
agem
ent i
n te
lem
edic
ine
serv
ice
lines
by
popu
latio
n gr
oups
and
chr
onic
con
ditio
ns t
ypic
ally
m
anag
ed in
prim
ary
care
set
tings
(e.
g., d
iabe
tes,
dep
ress
ion)
•Id
entif
y th
e pr
ovid
er d
river
s an
d ba
rrie
rs t
o in
tegr
atin
g te
lem
edic
ine
in p
rimar
y ca
re fr
om t
he p
ersp
ectiv
e of
pro
vide
rs r
esul
ting
in a
prov
ider
tech
nolo
gy r
eadi
ness
fact
or m
odel
•In
vest
igat
e an
d de
sign
a p
erso
naliz
ed r
emot
e pa
tient
mon
itorin
gsy
stem
to c
onne
ct p
atie
nts
and
prov
ider
s.
•P
rovi
de r
ecom
men
datio
ns fo
r im
plem
entin
g ap
prop
riate
tele
med
icin
e in
terv
entio
ns th
roug
h m
ultip
le c
are
setti
ngs
give
ngo
vern
men
tal p
olic
ies,
rei
mbu
rsem
ent p
aym
ents
and
del
iver
y of
careO
BJE
CT
IVE
S &
PA
RT
NE
RS
: D
escr
iptio
n of
Pro
ject
on
Pro
ject
Ove
rvie
w a
nd
Des
crip
tio
nO
BJE
CT
IVE
S &
PA
RT
NE
RS
: D
escr
iptio
n of
Pro
ject
Wh
at p
rob
lem
is t
his
pro
ject
see
kin
g t
o a
dd
ress
?
•W
ithou
t cha
nges
to h
ow p
rimar
y ca
re is
del
iver
ed, t
he g
row
th in
prim
ary
care
phy
sici
an s
uppl
y w
ill n
ot b
e ad
equa
te to
mee
t dem
and
in 2
020,
with
a p
roje
cted
sho
rtag
e of
20,
400
phys
icia
ns.
•N
early
150
mill
ion
Am
eric
ans
suffe
r at
leas
t one
chr
onic
con
ditio
n,an
dne
arly
30
mill
ion
are
livin
g w
ith fi
ve c
hron
ic c
ondi
tions
or
mor
e,w
hich
acc
ount
for
abou
t 12
perc
ent o
f the
U.S
. adu
lt po
pula
tion
and
mor
e th
an 4
0 pe
rcen
t of U
.S. h
ealth
spe
ndin
g.
•Te
lem
edic
ine,
in v
ario
us it
erat
ions
, pro
vide
em
ergi
ng m
odel
s of
deliv
erin
g pr
imar
y ca
re a
nd m
anag
ing
chro
nic
dise
ase.
•In
form
hea
lthca
re p
rovi
ders
abo
ut th
e ef
fect
iven
ess
of c
urre
ntte
lem
edic
ine
syst
ems
and
may
sug
gest
way
s to
leve
rage
tele
med
icin
e fo
r pr
imar
y ca
re a
nd c
hron
ic d
isea
se m
anag
emen
t.
Pro
ject
Ove
rvie
w a
nd
Des
crip
tio
nR
esea
rch
Per
spec
tive
Ph
ase
1 -
Lit
. Rev
iew
Lite
ratu
re r
evie
w to
iden
tify:
•V
ario
us it
erat
ions
of t
he te
lem
edic
ine
serv
ice
prov
isio
n in
pri
mar
yca
re c
onte
xts
•P
ossi
ble
forc
es a
ffect
ing
prov
ider
ado
ptio
n an
d in
nova
tion
and
indi
cato
rs o
f tec
hnol
ogy
read
ines
s fa
ctor
s an
d pa
thw
ays
for
prim
ary
care
pra
ctic
es to
impl
emen
t te
lem
edic
ine
serv
ices
in p
rimar
y ca
re
•P
atie
nt p
opul
atio
ns' e
ngag
emen
t in
succ
essf
ul u
se o
fte
lem
edic
ine/
tele
med
icin
e op
tions
.
OB
JEC
TIV
ES
& P
AR
TN
ER
S:
Des
ign
Thi
nkin
g A
ppro
ach
90
Pro
ject
Ove
rvie
w a
nd
Des
crip
tio
nR
esea
rch
Per
spec
tive
Ph
ase
2 –
Pro
vid
er R
ead
ines
s S
urv
ey•
Sec
onda
ry d
ata
anal
ysis
of f
orm
er C
HO
Tpr
ojec
t to
iden
tify
a) v
ario
us it
erat
ions
of t
hete
lem
edic
ine
serv
ice
prov
isio
n in
prim
ary
care
con
text
s an
d b)
pos
sibl
e fo
rces
affe
ctin
g ad
optio
n an
d in
nova
tion
•B
ased
on
resu
lts o
f lite
ratu
re r
evie
w a
nd
seco
ndar
y da
ta a
naly
sis,
dev
elop
apr
ovid
er s
urve
y in
stru
men
t aim
ed a
tm
easu
ring
the
barr
iers
affe
ctin
g ad
optio
n of
tele
med
icin
e in
prim
ary
care
set
tings
and
leve
ls o
f tec
hnol
ogy
read
ines
s .
OB
JEC
TIV
ES
& P
AR
TN
ER
S:
Des
ign
Thi
nkin
g A
ppro
ach
Pro
ject
Ove
rvie
w a
nd
Des
crip
tio
nR
esea
rch
Per
spec
tive
Ph
ase
3–
Pro
toty
pe
Bas
edon
Pha
se1
and
2re
sults
,th
ete
am w
illus
e a
desi
gnsc
ienc
eap
proa
chto
cre
ate
alo
w-le
vel
prot
otyp
esh
owca
sing
the
desi
gnof
ape
rson
aliz
edre
mot
epa
tient
mon
itorin
gsy
stem
. Thi
ssy
stem
will
inte
grat
ete
lehe
alth
devi
ces
usin
ga
smar
tpho
neap
plic
atio
nto
con
nect
patie
nts
and
prov
ider
s.
OB
JEC
TIV
ES
& P
AR
TN
ER
S:
Des
ign
Thi
nkin
g A
ppro
ach
Pro
ject
Ove
rvie
w a
nd
Des
crip
tio
nC
on
trib
uti
on
s to
Ind
ust
ry a
nd
Aca
dem
ia
•T
his
proj
ect
will
hel
p in
dust
ry m
embe
rs u
nder
stan
d th
e st
atus
and
oppo
rtun
ities
of l
ever
agin
g te
lem
edic
ine
to s
uppo
rtpr
imar
y ca
re a
nd m
anag
emen
t of
chr
onic
dis
ease
fro
m b
oth
the
patie
nt a
nd p
rovi
der
pers
pect
ive.
•In
dust
ry m
embe
rs w
ill g
ain
a be
tter
unde
rsta
ndin
g of
bot
h th
epa
tient
and
pro
vide
r co
ncer
ns a
nd in
tere
sts
in u
sing
tele
med
icin
e in
the
cont
ext o
f prim
ary
care
and
the
man
agem
ent
of c
hron
ic d
isea
se.
•It
will
con
cept
ualiz
e a
viab
le r
emot
e pa
tient
mon
itorin
gsy
stem
.
CO
NT
RIB
UT
ION
: How
is th
is d
iffer
ent
than
rel
ated
res
earc
h?A
pp
roac
hW
hat
do
yo
u p
lan
on
do
ing
?
•P
erfo
rm s
yste
mat
ic li
tera
ture
rev
iew
•D
evel
op a
nd a
dmin
iste
r a
tele
med
icin
e pr
imar
y ca
repr
ovid
er s
urve
y ai
med
at m
easu
ring
the
barr
iers
affe
ctin
gad
optio
n of
tele
med
icin
e in
prim
ary
care
set
tings
and
leve
lsof
tech
nolo
gy r
eadi
ness
•Id
entif
y ga
ps in
car
e ev
iden
ced
by th
e lit
erat
ure
revi
ew a
ndsu
rvey
thr
ough
gap
ana
lysi
s
•D
evel
op a
sys
tem
pro
toty
pe t
o ad
dres
s on
e or
mor
e ga
ps in
care
•C
onfe
renc
e su
bmis
sion
AP
PR
OA
CH
: E
xper
imen
tal A
ppro
ach
91
Ap
pro
ach
OV
ER
VIE
W O
F R
ES
ULT
S:
Exp
ecte
d M
ilest
ones
Exp
ecte
d M
ilest
on
esE
xpec
ted
Mile
sto
nes
1)Te
chno
logy
rea
dine
ssm
odel
for
tele
med
icin
e in
pr
imar
y ca
re s
ettin
gs
2)C
onfe
renc
e pr
esen
tatio
ns
3)P
ilote
d te
chno
logy
inP
rimar
y C
are
Ass
essm
ent
Sur
vey
4)O
ptim
ized
poi
nt-o
f-ac
cess
for
stud
y si
tes
5)A
low
-cos
t pe
rson
aliz
edpr
otot
ype
rem
ote
patie
nt
mon
itorin
g de
vice
ach
Ap
pro
ach
OV
ER
VIE
W O
F R
ES
ULT
S:
Ben
efits
to In
dust
ry
•U
nder
stan
ding
the
forc
es a
ffect
ing
adop
tion
and
diffu
sion
to
tele
med
icin
e in
prim
ary
care
set
tings
can
info
rm d
ecis
ion
mak
ing
rega
rdin
g se
rvic
ede
sign
, im
plem
enta
tion,
ope
ratio
ns,
and
prov
ider
enga
gem
ent.
•A
n as
sess
men
t to
ol b
ased
on
unde
rsta
ndin
g of
thes
e fo
rces
can
hel
p as
sess
indi
vidu
al p
rimar
yca
re o
rgan
izat
iona
l rea
dine
ss f
or te
lem
edic
ine
inno
vatio
n to
pro
mot
e or
gani
zatio
nal s
ucce
ss in
deliv
erin
g th
is m
ode
of s
ervi
ce.
92
PI:
An Intervention to Address Health Literacy Barriers, Increase Patient Engagement, & Improve Patient Experience & Outcomes
Value-based reimbursement in health care has resulted in an increasing focus on patient engagement as a mechanism to improvepost-acute care outcomes, particularly in reducing readmissions. However, health system strategies aimed at increasing patientengagement should account for health literacy and generational differences. Strategies that may work with a high literacy populationmay not be as effective among a population with low literacy.
Access2 Weech-Maldonado, Borkowski, and Lord
Access to Care
$40,000 YES 1
There has been limited research on the effectiveness of health literacy interventions in improving patient engagement and healthoutcomes, particularly in the home health context. This is a two-phase project. During the current first phase, we are conducting aliterature review to identify best practices/strategies in addressing health literacy barriers in a home health environment with theultimate goals of improving patient engagement and reducing hospital readmissions. We are proposing a second phase, which willconsist of a pilot intervention Ask Me 3® in a home health setting. Ask Me 3® is an educational program that encourages patients andfamilies to ask three specific questions of their providers to better understand their health conditions and what they need to do to stayhealthy: (1) what is my main problem?; (2) what do I need to do?; and (3) why is it important for me to do this?
A final report outlining the findings of the pilot project.
A pre-post experimental design: select two comparable health agency sites in the Birmingham metro area. In the experimental site,nurses will receive training on Ask Me 3® and patient data on health literacy. A second site will serve as a control group, with nursesproviding usual care. We will compare the intervention and control sites in terms of pre-post patient outcomes, such as patientengagement, patient experience, and readmissions.
Obtain IRB approval and finalize intervention Months 1-3Collect baseline data Months 4-5Implement intervention Months 6-7Collect post-intervention data Months 8-9Data analysis Months 10-11Final report Month 12
This pilot project will provide the foundation for futureinterventions of health system strategies to address barriersrelated to health literacy, increase patient engagement, andimprove patient outcomes.
93
Ask
Me
3®: A
Hom
e H
ealth
In
terv
entio
n to
Add
ress
Hea
lth
Lite
racy
Bar
riers
, Inc
reas
e Pa
tient
En
gage
men
t, an
d Im
prov
e Pa
tient
Ex
perie
nce
and
Out
com
es
Enga
ged
IAB
mem
bers
: A
laca
re
Enga
ged
CH
OT
Uni
vers
ity S
ites:
UA
B
Pro
ject
Lea
ders
: R
ober
t Wee
ch-M
aldo
nado
, Nan
cy B
orko
wsk
i; Ju
stin
Lor
d
Co-
lead
ers:
Sa
mik
a W
illia
ms
Stud
ent
Nam
e(s)
: R
eena
Jose
ph; N
eera
j Pur
o
Proj
ect O
verv
iew
and
Des
crip
tion
Obj
ectiv
e:•
Pilo
t tes
t a ta
rget
ed in
terv
entio
n (A
sk M
e 3®
) in
a ho
me
heal
th e
nviro
nmen
t to
redu
ce h
ealth
lite
racy
bar
riers
to
care
and
incr
ease
pat
ient
eng
agem
ent.
Part
ners
/ R
elev
ance
:•
Add
ress
ing
heal
th li
tera
cy b
arrie
rs to
car
e ca
n im
prov
e pa
tient
eng
agem
ent,
whi
ch c
an in
turn
impr
ove
heal
th
care
out
com
es, s
uch
as lo
wer
read
mis
sion
s an
d im
prov
ed p
atie
nt e
xper
ienc
e.
OB
JEC
TIV
ES
& P
AR
TNE
RS
: Des
crip
tion
of P
roje
ct
Proj
ect O
verv
iew
and
Des
crip
tion
Prob
lem
: •
Valu
e-ba
sed
reim
burs
emen
t: in
crea
sing
focu
s on
pa
tient
eng
agem
ent a
s a
mec
hani
sm to
impr
ove
post
-acu
te c
are
outc
omes
, par
ticul
arly
in re
duci
ng
read
mis
sion
s•
Stra
tegi
es th
at m
ay w
ork
with
a h
igh
liter
acy
popu
latio
n m
ay n
ot b
e as
effe
ctiv
e am
ong
a po
pula
tion
with
low
lite
racy
.Sc
ope:
Unn
eces
sary
hos
pita
l rea
dmis
sion
s co
st
arou
nd $
18 b
illio
n pe
r yea
r
OB
JEC
TIV
ES
& P
AR
TNE
RS
: Des
ign
Thin
king
App
roac
hPr
ojec
t Ove
rvie
w a
nd D
escr
iptio
n
Res
earc
h Pe
rspe
ctiv
e•
Prov
ider
: Ala
care
(hom
e an
d co
mm
unity
ba
sed
care
pro
vide
r)•
Patie
nt:
Pos
t-acu
te c
are
OB
JEC
TIV
ES
& P
AR
TNE
RS
: Des
ign
Thin
king
App
roac
h
94
Proj
ect O
verv
iew
and
Des
crip
tion
Con
trib
utio
n to
Indu
stry
and
Aca
dem
ia
CO
NTR
IBU
TIO
N: H
ow is
this
diff
eren
t tha
n re
late
d re
sear
ch?
•Li
mite
d re
sear
ch o
n th
e ef
fect
iven
ess
of h
ealth
lit
erac
y in
terv
entio
ns in
impr
ovin
g pa
tient
en
gage
men
t and
hea
lth o
utco
mes
, par
ticul
arly
in
the
hom
e he
alth
con
text
.
•Th
is p
ilot p
roje
ct w
ill p
rovi
de th
e fo
unda
tion
for
futu
re la
rge-
scal
e in
terv
entio
ns o
f hea
lth s
yste
m
stra
tegi
es to
add
ress
hea
lth li
tera
cy b
arrie
rs to
ca
re.
AP
PR
OA
CH
: Exp
erim
enta
l App
roac
h
•C
urre
nt fi
rst p
hase
(201
7-18
)•
Lite
ratu
re re
view
to id
entif
y be
st p
ract
ices
/stra
tegi
es in
ad
dres
sing
hea
lth li
tera
cy b
arrie
rs in
a h
ome
heal
th
envi
ronm
ent
•P
ropo
sed
seco
nd p
hase
(201
8-19
)•
Pilo
t tes
t int
erve
ntio
n in
a h
ome
heal
th s
ettin
g•
Ask
Me
3®: e
duca
tiona
l pro
gram
that
enc
oura
ges
patie
nts
and
fam
ilies
to a
sk th
ree
spec
ific
ques
tions
of t
heir
prov
ider
s to
be
tter u
nder
stan
d th
eir h
ealth
con
ditio
ns a
nd w
hat t
hey
need
to
do to
sta
y he
alth
y: 1
. Wha
t is
my
mai
n pr
oble
m?
(Dia
gnos
is);
2.
Wha
t do
I nee
d to
do?
(Tre
atm
ent);
3. W
hy is
it im
porta
nt fo
r me
to d
o th
is?
(Con
text
)•
Exp
erim
enta
l des
ign
•S
elec
t tw
o co
mpa
rabl
e he
alth
age
ncy
site
s in
the
Birm
ingh
am
met
ro a
rea.
In th
e ex
perim
enta
l site
, nur
ses
will
rece
ive
train
ing
on A
sk M
e 3
and
patie
nt d
ata
on h
ealth
lite
racy
. A s
econ
d si
te
will
ser
ve a
s a
cont
rol g
roup
, with
nur
ses
prov
idin
g us
ual c
are.
•
Com
pare
the
inte
rven
tion
and
cont
rol s
ites
in te
rms
of p
re-p
ost
patie
nt o
utco
mes
, suc
h as
pat
ient
eng
agem
ent,
patie
nt
expe
rienc
e, a
nd re
adm
issi
ons.
App
roac
h•
Mon
ths
1-3:
Obt
ain
IRB
app
rova
l and
fina
lize
inte
rven
tion
•M
onth
s 4-
5: C
olle
ct b
asel
ine
data
•
Mon
ths
6-7:
Impl
emen
t int
erve
ntio
n •
Mon
ths
8-9:
Col
lect
pos
t-int
erve
ntio
n da
ta
•M
onth
s 10
-11:
Dat
a an
alys
is•
Mon
th 1
2: F
inal
repo
rt
OV
ER
VIE
W O
F R
ES
ULT
S: E
xpec
ted
Mile
ston
esA
ppro
ach
•D
eliv
erab
les
–Tr
aini
ng o
f Ala
care
staf
f–
Mat
eria
ls a
nd re
sour
ces
–Fi
nal r
epor
t•
Info
rm fu
ture
stra
tegi
es th
at c
an a
ddre
ss
heal
th li
tera
cy b
arrie
rs w
ith th
e go
als
of
impr
ovin
g pa
tient
eng
agem
ent a
nd
outc
omes
OV
ER
VIE
W O
F R
ES
ULT
S: B
enef
its to
Indu
stry
95
96
RESEARCH PROJECT UPDATES AND OVERVIEWS
97
Research Project Updates Overview
Title % Completed $ Remaining
Patient Engagement and Hospital Readmissions: The Role of Health Literacy
25% $45,000
Impact of Direct-to-Consumer Telemedicine on Downstream Health Care Utilization and Costs
15% $50,000
Telehealth and Remote Patient Monitoring Systems to Improve Access and Promote Active Patient Engagement in Rural Communities
35% $100,000
Development of a Middleware Framework for Medical Device Integration for Telemedicine
0% $50,000
Using Care Coordination to Address Cost, Quality, and Access to Care Across Systems and Populations
80% $250,000
Effects of Care Coordination on the Improvement of Quality of Care
40% $65,000
Effects of Care Coordination on Care Transitions 25% $100,000
A Mobile Based Care Coordination System for Critical Care 25% $50,000
Integration of Population Health Data and Digital Assistants to Reduce Readmission Risks 10% $50,000
Gamification and its impact on the Population Health Management of Chronic Conditions 10% $50,000
Improving Employee and Patient Health through Population Data Mining 10% $50,000
Data-Driven Predictive Analytics to Improve Diagnosis, Treatment, Care Coordination, and Resource Utilization 60% $120,000
Machine Learning for Evidence –Based Practice, Resource Allocation, and Risk Prediction 60% $80,000
98
PROJECT TITLE:
PROJECT ID: PI:
RESEARCH THEME:
PROJECT START DATE:
BUDGET:
I/UCRC EXECUTIVE SUMMARY | PROJECT UPDATE
DESCRIPTION:
PROJECT OBJECTIVES: SCOPE:
HOW THIS IS DIFFERENT THAN RELATED RESEARCH:
MILESTONES TARGETED START DATE TARGETED END DATE
PERCENT COMPLETED OVERALL:
ISSUES AND RISKS:
NEXT STEPS:
BENEFITS TO INDUSTRY: EXPECTED DELIVERABLES:
MULTI-UNIVERSITY PROJECT:
UPDATED: OCTOBER 1, 2017
CHOT CONFIDENTIAL
Patient Engagement and Hospital Readmissions: The Role of Health Literacy
01-05171.UAB-FAU Robert Weech-Maldonado, Nancy Borkowski, & Justin Lord
Access to Care
6/15/2017
$45,000
Value-based reimbursement in healthcare has resulted in an increasing focus on patient engagement as a mechanism to improve post-acute care outcomes, particularly in reducing readmissions. However, interventions to address patient engagement should account for health literacy and generational differences. Strategies that may work with a high literacy population may not be as effective among a population with low literacy. Similarly, strategies used with millennials may not be as effective among baby boomers.
(1) Conduct a systematic literature review of the relationships among health literacy, generational differences, patient engagement, and hospital readmissions, (2) identify best practices to address barriers, and (3) propose a pilot project from findings.
Databases included PubMed, Scopus, and ABI Inform. Search terms include patient engagement, health literacy, and health outcomes/readmissions.
This research is attempting to examine the relationship between patient engagement and hospital readmissions as mediated by health literacy. Patient engagement is comprised of four separate but distinct elements. Personalizations (an element of engagement) has been linked to health literacy. It is from this aspect that the barriers to health literacy will be addressed using the social ecological model. This paper will also highlight the best practices/strategies to addressing health literacy barriers which will in turn improve patient engagement.
Systematic Literature Review 8/1/2017 10/30/2017
Synthesize Findings 11/1/2017 1/31/2018
Propose a Pilot Project 2/1/2018 5/31/2018
Draft of Report 6/1/2018 8/31/2018
25%
There is limited research that examines health literacy as a mediator between patient engagement and health outcomes. Our study is focused on reducing readmissions; however, the limited existing literature exploring patient engagement and health literacy in areas of different diseases and contexts. This may pose a possible limitation but provide a unique opportunity to contribute new findings to this field.
Complete the literature review to identify the relationship between patient engagement and readmissions as mediated by health literacy. Will start on the draft of the report.
This project will provide the foundation for future interventions of health system strategies to address barriers related to health literacy generational differences, increase patient engagement, and improve patient outcomes.
A final report outlining the findings of the literature review and the proposed pilot project.
UAB & FAU
99
PROJECT TITLE:
PROJECT ID: PI:
RESEARCH THEME:
PROJECT START DATE:
BUDGET:
I/UCRC EXECUTIVE SUMMARY | PROJECT UPDATE
DESCRIPTION:
PROJECT OBJECTIVES: SCOPE:
HOW THIS IS DIFFERENT THAN RELATED RESEARCH:
MILESTONES TARGETED START DATE TARGETED END DATE
PERCENT COMPLETED OVERALL:
ISSUES AND RISKS:
NEXT STEPS:
BENEFITS TO INDUSTRY: EXPECTED DELIVERABLES:
MULTI-UNIVERSITY PROJECT:
UPDATED: OCTOBER 1, 2017
CHOT CONFIDENTIAL
Impact of Direct-to-Consumer Telemedicine on Downstream Health Care Utilization and Costs
02-05171.TAM Cynthia LeRouge
Access to Care
12/15/2017
$50,000
Direct-to-consumer telemedicine (DTCT) provides patient-initiated, on-demand access to care for common non-emergent conditions using real-time interactive technologies. DTCT holds the promise of increased accessibility, improved service fit, and cost containment/savings. However, more research is needed to assess the impact of DTCT on downstream use of health care services and associated costs. We will use a retrospective observational study design to describe, analyze, and report information on service utilization and spending patterns among DTCT users and a control group of non-users across different patient populations and organizational settings.
Compare patterns of service utilization and cost between DTCT users and non-users among self-insured and Medicaid patients. Examine the extent to which DTCT visits replace in-person visits among Medicaid enrollees for leading DTCT conditions.
Health organizations offering DTCT services to self-insured and Medicaid patient populations. Data capture across multiple provider organizations with service areas across the United States (WA, IL, WI, OK).
The project represents a new and timely step in advancing DTCT research among different health organizations, patient populations, and types of non-emergent conditions. This study will be the first to explore downstream utilization and cost using datasets that combine longitudinal claims files with DTCT encounter data from a telemedicine vendor company, allowing for a novel examination of relationships between socio-demographic/encounter variables and downstream utilization and cost following DTCT use. We will also offer first insights into how DTCT effects underserved populations that may benefit most from the accessibility and low cost of DTCT.
Gather and Clean Data 12/15/2017 6/30/2018
Descriptive Analysis 2/1/2018 6/30/2018
Final Modeling 3/1/2018 8/31/2018
Reporting Deliverables 9/1/2018 12/14/2018
15%
Negotiating data use agreements with participating organizations. Potential challenges and delays with secure data capture/transfer. Potential challenges and delays with cleaning and management of claims data (i.e. missing data, file formatting).
Complete data use agreements with participating health organizations Securely transfer data to University of Washington servers Complete data cleaning and management needs Complete data descriptive analysis and statistical modeling Complete expected reporting deliverables
Provide insights regarding DTCT impact on service utilization and cost across different target markets and organizational settings. Share strategies to evaluate return on investment for DTCT.
PPT slide deck of key findings for CHOT members. Submit abstract(s) to conference proceedings.
TAM
100
PROJECT TITLE:
PROJECT ID: PI:
RESEARCH THEME:
PROJECT START DATE:
BUDGET:
I/UCRC EXECUTIVE SUMMARY | PROJECT UPDATE
DESCRIPTION:
PROJECT OBJECTIVES: SCOPE:
HOW THIS IS DIFFERENT THAN RELATED RESEARCH:
MILESTONES TARGETED START DATE TARGETED END DATE
PERCENT COMPLETED OVERALL:
ISSUES AND RISKS:
NEXT STEPS:
BENEFITS TO INDUSTRY: EXPECTED DELIVERABLES:
MULTI-UNIVERSITY PROJECT:
UPDATED: OCTOBER 1, 2017
CHOT CONFIDENTIAL
Telehealth & Remote Patient Monitoring Systems Improve Access & Promote Patient Engagement in Rural Communities
03-05171.UAB-GIT Shannon Houser, Darrell Burke, & Eva Lee
Access to Care
6/1/2017
$100,000
Timely access to quality healthcare service is a real challenge--as outlined in the 2015 IOM report--and misalignment of resources and demands results in a long delay for care. Telehealth can offer alternative and timely care to rural area patients who lack sufficient healthcare options. Telehealth can also help to improve health conditions and to promote active patient engagement, which is particularly important for chronic disease management. This project identifies drivers and barriers of patient engagement by population groups and chronic conditions and provides recommendations for implementing appropriate telehealth/telemedicine interventions.
(1) Examine patient populations' engagement in successful use of telehealth/telemedicine options and (2) explore the success of telehealth/telemedicine interventions in chronic conditions and design a remote patient monitoring system of practical usage .
(1) Databases: PubMed, Scopus, CINAHL, Embase, and ABI/Inform, (2) Search terms including telehealth, telemedicine, m-health, e-health, obesity, COPD, diabetes, and (3) A prototype remote-patient monitoring system to connect patients and providers.
The adoption of telemedicine and level of patient engagement and services provided across healthcare facilities remain uneven and far from optimal. There has been little research examining various patient populations engagement in the successful use of telehealth/telemedicine options. By exploring successful application in rural care setting, this study will define the terms telehealth and telemedicine. We will also investigate and design a personalized remote patient monitoring system to connect patients and providers.
Systematic Literature Review 8/1/2017 10/30/2017
Gap Analysis 11/1/2017 1/31/2018
Data and System Modeling 7/1/2017 5/31/2017
Draft of Report 6/1/2018 8/31/2018
35%
Availability of studies related to COPD, privacy, and issuance policy remain critical issues.
Complete the literature review to identify barriers to patient engagement and look at the successful interventions of telehealth/telemedicine in chronic conditions. Complete the design of the remote patient monitoring prototypical system.
This project will help industry members understand the benefits of implementation of telehealth/telemedicine practices to improve patients' health. It will establish a viable, low-cost remote patient monitoring system.
1) Systematic literature review 2) Prototype remote patient monitoring device
UAB & GIT
101
PROJECT TITLE:
PROJECT ID: PI:
RESEARCH THEME:
PROJECT START DATE:
BUDGET:
I/UCRC EXECUTIVE SUMMARY | PROJECT UPDATE
DESCRIPTION:
PROJECT OBJECTIVES: SCOPE:
HOW THIS IS DIFFERENT THAN RELATED RESEARCH:
MILESTONES TARGETED START DATE TARGETED END DATE
PERCENT COMPLETED OVERALL:
ISSUES AND RISKS:
NEXT STEPS:
BENEFITS TO INDUSTRY: EXPECTED DELIVERABLES:
MULTI-UNIVERSITY PROJECT:
UPDATED: OCTOBER 1, 2017
CHOT CONFIDENTIAL
Development of a Middleware Framework for Medical Device Integration for Telemedicine
04-05171.FAU Ankur Agarwal
Access to Care $50,000
An integration of mobile, wireless, and sensor technologies has the potential to greatly advance the ability to enable automated data collection for monitoring patient health status in real time, even from a remote environment. This could dramatically increase the ability to rapidly respond to a critical healthcare need. However, the networking capability of currently available health status monitoring devices is limited in functionality and primarily relies on proprietary communication protocols offered by a multitude of different vendors. Furthermore, current systems are missing critical elements of a truly robust system.
-Design a working prototype based on IEEE11073 protocol -Develop a hardware/software used for interfacing biosensors -Evaluate and Expand existing capabilities of the IEEE 11073 protocol to enable remote patient monitoring
The networking capability of currently available health status monitoring devices is limited in functionality and primarily relies on proprietary communication protocols offered by a multitude of different vendors and current systems are missing critical elements of a truly robust system. The development of a middle-ware layer framework in this project will be able to use the recorded data to continuously mine it in real-time to detect data inconsistencies due to network issues. Then, the intelligent system engine (knowledge base) could automatically detect potential health-related issues in patients and alert the caregivers.
Understand IEEE 11073 Protocol
Middleware Layer Completion
Develop a "MOT"
Testing Various Devices
0%
This project provides better remote patient monitoring while keeping costs down. The current project further provides unique technology to reduce the cost for patient data collection thereby helping the economics of home-healthcare providers, nursing homes.
A prototype which is based on IEEE 11073 protocol for various device integration. The system will be interface with various medical devices which are compatible with WiFi and Bluetooth for physiological data collection. The software will based on LAMP environment and hosted on amazon cloud in a secured manner.
FAU
102
PROJECT TITLE:
PROJECT ID: PI:
RESEARCH THEME:
PROJECT START DATE:
BUDGET:
I/UCRC EXECUTIVE SUMMARY | PROJECT UPDATE
DESCRIPTION:
PROJECT OBJECTIVES: SCOPE:
HOW THIS IS DIFFERENT THAN RELATED RESEARCH:
MILESTONES TARGETED START DATE TARGETED END DATE
PERCENT COMPLETED OVERALL:
ISSUES AND RISKS:
NEXT STEPS:
BENEFITS TO INDUSTRY: EXPECTED DELIVERABLES:
MULTI-UNIVERSITY PROJECT:
UPDATED: OCTOBER 1, 2017
CHOT CONFIDENTIAL
Data-Driven Predictive Analytics to Improve Diagnosis and Treatment -- Innovation to Cancer Treatment
05-05171.GIT Eva Lee, Cao Yu, Alistair Temple, Rui Yao, & Jame Chu
Analytics & Innovative Technologies
4/1/2017
$120,000
Personalized data-driven predictive analytics is one of the most significant topics in recent years. With the newly improvement of imaging technology, personalized high-quality data collection has become quick and reliable. Machine learning and statistics based high-level representation of these data induce accurate diagnosis. With the help of advanced optimization planning algorithm and equipment frontier, a very precise individualized treatment can be delivered. Care coordination and resource utilization remain an uphill-battle up until now. Data-driven novel analytics combined with innovative planning will definitely reform this field.
Utilizing the PET imaging and data-driven imaging reconstruction,we obtain individualized cancer biology and tumor cell distribution. Capture the nature of the tumor cells by data-driven deep learning further boost diagnosis accuracy and plan quality.
PET enables characterization of tumor cells. The robustness of data-driven approach leads us to find the nature in disease source, which results in an enhancement in treatment planning. Initial focus will be on prostate, head-and-neck and lung cancer.
This project analyzes heterogeneous types of data including imaging, clinical and biological patient data across multiple clinic sites and platforms using large-scale data and predictive analytics. The study involves precision medicine, utilizing individual patient-specific data to advance innovative disease treatment with lab diagnostics and imaging advances. It will contribute to the development of state-of-the-art system data analytics and real-time decision technologies with broad applicability. This study is the first biological image-guided cancer treatment. We have currently completed cancer cases of the prostate, lung, and head-and-neck.
PET Guided Treatment Planning Models 4/1/2017 8/31/2017
Optimization for Individualized Plan Design 9/1/2017 10/31/2017
Machine Learning to Characterize Plan Result 7/1/2017 2/28/2018
Plan Performance Comparison 11/1/2017 3/31/2018
60%
We have thus far obtained outstanding treatment results for prostate, head-and-neck and lung cancer patients, compare to current approaches. One of the remaining challenges is to efficiently reconstruct the imaging using advanced data-driven algorithms. A better dose delivery control measurement is needed to characterize the dose delivered to PET pockets. A clear interpretation of the data-driven machine learning result of large scale patients' data will be useful for generalization to multiple clinical sites.
More patient cases will be tested and further analyzed to explore tumor cells response to radiation with respect to normal cells. We will investigate real-time generation of tumor constraints. Machine learning will be used to analyze which models work best for what tumor and biological characteristics. Highly-integrated simulation and optimization system will be used to optimize care coordination and resource utilization.
Improve quality of care, improve tumor control, improve treatment outcome, improve care coordination, improve resource usage, and advance data analytics for innovative cancer treatment.
Improved treatment plans and a new data-driven model combining cancer biology with cells distribution for advanced personalized treatment. The system results in improved plan quality, improved treatment outcome, and reduce hospital resources (due to shorter treatment time).
GIT
103
PROJECT TITLE:
PROJECT ID: PI:
RESEARCH THEME:
PROJECT START DATE:
BUDGET:
I/UCRC EXECUTIVE SUMMARY | PROJECT UPDATE
DESCRIPTION:
PROJECT OBJECTIVES: SCOPE:
HOW THIS IS DIFFERENT THAN RELATED RESEARCH:
MILESTONES TARGETED START DATE TARGETED END DATE
PERCENT COMPLETED OVERALL:
ISSUES AND RISKS:
NEXT STEPS:
BENEFITS TO INDUSTRY: EXPECTED DELIVERABLES:
MULTI-UNIVERSITY PROJECT:
UPDATED: OCTOBER 1, 2017
CHOT CONFIDENTIAL
Machine Learning for Evidence-Based Practice, Risk Prediction, and Optimal Care Coordination and Outcome
06-05171.GIT Eva Lee, Joe Malecki, Chris Kwan, Ellie Cheng, & Cody Wang
Analytics & Innovative Technologies
4/1/2017
$80,000
Fueled by rapid digital media advances, healthcare systems are investing more in advanced sensors and robotics, communication technologies, and sophisticated data centers. This facilitates information and knowledge visibility and delivery standardization and performance efficiency through big data analytics. In this study, 20 years of millions of EMR records are used to perform machine learning and data mining to identify evidence and characteristics of best practice, uncover risk factors of different patient groups, develop effective clinical practice guidelines and disease management strategies, and optimize the service delivery to meet demand.
Apply machine learning, optimization and decision models to EMR to uncover evidence and treatment knowledge. We aim to determine timing, decision and treatment options within clinical patient care process that positively influence the outcome.
We will initially focus on chronic kidney disease (CKD) and prostate cancer. The work will expand to other chronic diseases including cardiovascular patients, diabetes, hypertension, arthritis and sepsis treatment control.
This project involves multi-units, multi-disease and stakeholders. It is the first study of this kind that includes a massive amount of data across heterogeneous hospital and provider sites. The data captures a diverse population across the United States with varying demographics, clinical practices, and outcome measures.
Data Extract from EMR and Deidentification 4/1/2017 8/1/2017
Text Mining and Clinical Concept Mapping 6/1/2017 10/1/2017
Clinical Process Maps and Simulation Models 7/1/2017 12/1/2017
Clustering and Outcome Prediction Model 10/1/2017 3/31/2018
60%
This project is the first study of this kind that combine machine learning with simulation models and process maps to identify major clinical bottle necks and optimize treatment timing, process and decision making using a holistic, evidence-based approach. It includes a massive amount of data across heterogeneous provider sites, and initially we tackle prostate cancer and all stages of CKD patients.
1) Use machine learning to uncover treatment evidence and critical clinical features to predict outcome. 2) Complete disease progression model to understand and optimize the current treatment process. 3) Build simulation models to identify bottlenecks and optimize resources allocation and other clinical processes for best outcome. 4) Design optimized evidence-based treatment plans and dissemination of this knowledge and best practice transfer to multiple sites.
1) Improve quality and efficiency of care across patient population. 2) Identify best practice, offer evidence-based care. 3) Optimize timing and individualized treatment 4) Reduce unnecessary resources/procedures
1) Identification of effective treatment plans and best practice characteristics. 2) Optimization of evidence-based treatment plans. 3) Detailed plans and methods for best practice transfer across hospital sites.
GIT
104
PROJECT TITLE:
PROJECT ID: PI:
RESEARCH THEME:
PROJECT START DATE:
BUDGET:
I/UCRC EXECUTIVE SUMMARY | PROJECT UPDATE
DESCRIPTION:
PROJECT OBJECTIVES: SCOPE:
HOW THIS IS DIFFERENT THAN RELATED RESEARCH:
MILESTONES TARGETED START DATE TARGETED END DATE
PERCENT COMPLETED OVERALL:
ISSUES AND RISKS:
NEXT STEPS:
BENEFITS TO INDUSTRY: EXPECTED DELIVERABLES:
MULTI-UNIVERSITY PROJECT:
UPDATED: OCTOBER 1, 2017
CHOT CONFIDENTIAL
Using Care Coordination to Address Cost, Quality, and Access to Care Across Systems and Populations
07-05171.TAM Bita Kash & Farzan Sasangohar
Care Coordination
6/1/2017
$250,000
Consumers access healthcare in a multitude of settings, ranging from acute care to home and community based services (HCBS). With a variety of access points, care coordination programs are essential to facilitate care from one system to another, which emphasizes the isolated nature of the US healthcare system. In an effort to bring about broader systemic changes, this project aims to develop a care coordination program that focuses on the utilization of a bio-psycho-social model and leveraging community resources to facilitate care coordination outcomes.
Identify best practices for improving care coordination across systems and populations. Develop a toolkit for providers to evaluate practice and identify gaps in care coordination
Use of PubMed (MEDLINE) database Use of CINAHL database Use of Embase database
Related research focuses on a piece meal approach to improving care coordination, often focusing on single visits and procedures, rather than the whole continuum of care. This project takes a more holistic approach by defining the continuum of care and developing models of collaborative best practices in care coordination that take into account the full system of care.
Systematic Literature Review 9/1/2017 12/20/2017
Team Care: Measurement Tools Needed 9/1/2017 1/31/2018
Evaluation of Selected CC Programs 9/1/2017 4/15/2018
Creation and Dissemination of Toolkit 3/1/2018 5/31/2018
80%
We do not foresee any risks with this project as it is mostly dependent on time and effort from PIs and graduate research assistants. A potential survey on care team effectiveness might produce response rate issues.
Write up final reportComplete list of tools and metrics, evaluation of care coordination programs, create and disseminate toolkit
This project will assist industry members in identifying best practices for improving care coordination across systems and populations.
1) Care coordination model for people with disabilities, theirfamilies, providers, and communities; 2) template for integratinginstitutional care and community care coordination; 3) structuredtoolkit for providers to access their practice and identify gaps incoordination; and 4) manuscript and conference presentations
TAM
105
PROJECT TITLE:
PROJECT ID: PI:
RESEARCH THEME:
PROJECT START DATE:
BUDGET:
I/UCRC EXECUTIVE SUMMARY | PROJECT UPDATE
DESCRIPTION:
PROJECT OBJECTIVES: SCOPE:
HOW THIS IS DIFFERENT THAN RELATED RESEARCH:
MILESTONES TARGETED START DATE TARGETED END DATE
PERCENT COMPLETED OVERALL:
ISSUES AND RISKS:
NEXT STEPS:
BENEFITS TO INDUSTRY: EXPECTED DELIVERABLES:
MULTI-UNIVERSITY PROJECT:
UPDATED: OCTOBER 1, 2017
CHOT CONFIDENTIAL
Effects of Care Coordination on the Improvement of Quality of Care
08-05171.UAB Nathan Carroll and Midge Ray
Care Coordination
6/1/2017
$65,000
Transitions in patient care include home/community to acute care to post-acute care back to home/community. During these transitions, gaps in care may occur, which can negatively impact quality as well as increase healthcare costs. Two examples are hospital-acquired infections (HAI) and reconciliation for discharged and/or transferred patients. Both remain high nationwide, meaning providers have an opportunity to improve the quality of care they offer during care transitions. This project summarizes best practices for HAIs and identifies peer-reviewed evaluations of programs for medication management and adherence to reduce readmission rates.
Demonstrate practices to reduce certain hospital-acquired conditions and prevent readmissions from medication adherence. Identify best practices to increase quality of care while reducing avoidable costs.
Various databases have been used to conduct a systematic literature review including CINAHL, PubMed, Scopus, Cochrane, and Embase
The literature contains numerous case studies of HAIs and medication non-adherence, but is limited in nature. HAIs have been studied, but literature is scarce on hospital-acquired clostridium difficile (C. diff) infection. The literature on programs made to improve medication adherence and reconciliation for discharged patients has yet to be considered a cohesive body. A systematic review of published evaluations of readmission reduction programs will allow researchers to identify best practices common to the most effective programs and will identify elements important to the programs' success and intervention characteristics that are less effective.
Develop a Search Strategy 6/15/2017 7/15/2017
Synthesize Findings 7/16/2017 11/15/2017
First Draft for Industry Feedback 12/1/2017 2/1/2017
Final Draft of Report 2/15/2017 4/1/2017
40%
Synthesis of findings may be challenging as numerous variables are involved in all the studies.
(1) Identifying predictors of hospital-acquired C. diff infection (2) Identifying interventions for reducing medication non-adherence related hospital readmissions.
This project will assist industry members in better understanding how to reduce certain hospital-acquired conditions, and to prevent readmissions relating to medication adherence. The "best practices" identified will help industry members to increase quality of care while reducing avoidable costs.
Final reports outlining the findings of literature reviews and recommendations.
UAB
106
PROJECT TITLE:
PROJECT ID: PI:
RESEARCH THEME:
PROJECT START DATE:
BUDGET:
I/UCRC EXECUTIVE SUMMARY | PROJECT UPDATE
DESCRIPTION:
PROJECT OBJECTIVES: SCOPE:
HOW THIS IS DIFFERENT THAN RELATED RESEARCH:
MILESTONES TARGETED START DATE TARGETED END DATE
PERCENT COMPLETED OVERALL:
ISSUES AND RISKS:
NEXT STEPS:
BENEFITS TO INDUSTRY: EXPECTED DELIVERABLES:
MULTI-UNIVERSITY PROJECT:
UPDATED: OCTOBER 1, 2017
CHOT CONFIDENTIAL
Effects of Care Coordination on Care Transitions
09-05171.FAU-UAB Ravi Behara, Tapan Mehta, & Robert Weech-Maldonado
Care Coordination
6/15/2017
$100,000
Care coordination programs and care transition programs often center on the intersection of acute care and chronic care (such as the transition from nursing home to hospital), but in reality people access healthcare in a multitude of settings. This project aims to develop a care coordination program that focuses on measuring care coordination program impact, utilize the bio-psycho-social and spiritual model, and create a model that continually improves provider collaboration specifically for older adults and individuals with disabilities. This approach can play a critical role especially when dealing with under-served and underrepresented populations.
Develop metrics to measure care coordination program impact Utilize the bio-psycho-social and spiritual model Build a community-based care coordination model for older adults and people with disabilities that is continually improving
Databases include PubMed, Scopus, CINAHL, and Google Scholar. Search terms include care coordination, care transition, and spinal cord injuries.
Traditional research in care coordination is related to coordination between acute and/or chronic care providers. Yet, delivery of care to older adults and people with disabilities extends beyond traditional institutional-based clinical care to include home-based care, as well as services which are social, financial, legal, and spiritual in nature. This project broadens the perspective of care coordination by developing a system that also includes the non-traditional participants, such as social service and public health agencies, religious organizations, and end-of life services, in addition to traditional participants such as hospitals and skilled nursing homes.
Systematic Literature Review 8/1/2017 10/30/2017
Gap Analysis 11/1/2017 1/31/2018
Data and System Modeling 2/1/2018 5/31/2018
Draft of Report 6/1/2018 8/31/2018
25%
Availability of specific studies related to spinal cord injuries may be limited.
Complete the literature review to identify models of care coordination and care transition for patients transitioning from acute care to post-acute care setting.
This project will help industry members understand the benefits and the effects of care coordination on care transitions outside of traditional healthcare settings. If successful, partnerships between healthcare systems and community partners can be way to improve the quality of life of people with disabilities.
(1) Care coordination model, (2) template for institutional care, (3)structured toolkit for providers to assess their practice, (4) functionally-working cloud-based solution of a learning care coordination system, and (5) manuscripts and conference calls
FAU & UAB
107
PROJECT TITLE:
PROJECT ID: PI:
RESEARCH THEME:
PROJECT START DATE:
BUDGET:
I/UCRC EXECUTIVE SUMMARY | PROJECT UPDATE
DESCRIPTION:
PROJECT OBJECTIVES: SCOPE:
HOW THIS IS DIFFERENT THAN RELATED RESEARCH:
MILESTONES TARGETED START DATE TARGETED END DATE
PERCENT COMPLETED OVERALL:
ISSUES AND RISKS:
NEXT STEPS:
BENEFITS TO INDUSTRY: EXPECTED DELIVERABLES:
MULTI-UNIVERSITY PROJECT:
UPDATED: OCTOBER 1, 2017
CHOT CONFIDENTIAL
A Mobile Based Care Coordination System for Critical Care
10-05171.FAU Ankur Agarwal and Ravi Behara
Care Coordination
8/30/2017
$50,000
Currently, the healthcare industry is going through rapid transformations including readmission penalties, payment bundling, and patient compliance related medical coding. Such changes have given rise to Accountable Care Organizations (ACO) and Managed Care Organizations (MCO). These organizations are directly incentivized to reduce the cost of healthcare as well as improve quality in order to stay profitable. This project aims to develop a mobile-based care coordination system for critical care patients. The created system will provide a secured, asynchronous messaging system, which will ensure an instant communication with the entire care team for a patient.
Design a HIPAA compliant messaging platform to ensure a timely delivery of message to a patient's care team Facilitate tight communication, collaboration, and coordination among care team members.
The project will develop a mobile application for the medical practitioner (physicians, nurses) and a web-based system for backend data.
While communication or the lack of it is the main reason for missed diagnosis, hospital admission, readmission, and duplication of care, it has not yet been successfully addressed in any EHR system. Though several new mobile healthcare messaging applications have been implemented, they are basically HIPAA compliant text messaging among doctors (i.e. WhatsApp) and effectively create more silos. This project proposes to build a mobile EHR agnostic application connecting the patient with their outpatient and inpatient doctors, staff, and others related to care for intelligent communication, which has potential to improve healthcare and provide opportunities.
Compare and Analyze System 8/30/2017 10/30/2017
Develop Prototype System 10/15/2017 4/30/2018
Internal System Testing 5/1/2018 5/30/2018
Prepare Final Product 6/1/2018 7/30/2018
25%
System adaptability is always a challenge for this type of project. There needs to be a mind set for medical professionals to be open to adopting the new technology. Transitions to a new technology pose the challenge in terms of training and efficiency of usage and adaptability. Further, in an ideal scenario, the system should be able to communicate with an EHR system; such interface is available for modern solution via CDA interface. It becomes a training issue for the medical office to learn to make it communicate with their existing EHR systems.
We have completed the system analysis phase of the project by comparing various current systems in the medical space. The gaps among the current solutions have been identified. We have finalized a system design and the technology mapping. The next step in the project is to develop the system prototype and then take it for an internal system testing.
Care coordination is a major issue which has been clearly identified by healthcare companies. The proposed system provides a method and presents an approach to coordinate care among various care givers specifically for sick patients such as critical care where several medical professionals are engaged.
The deliverable for this project is a report comparing various care coordination systems and a working prototype for phase-2 deliverable. The prototype will be a cloud-based mobile application developed on a LAMP environment.
FAU
108
PROJECT TITLE:
PROJECT ID: PI:
RESEARCH THEME:
PROJECT START DATE:
BUDGET:
I/UCRC EXECUTIVE SUMMARY | PROJECT UPDATE
DESCRIPTION:
PROJECT OBJECTIVES: SCOPE:
HOW THIS IS DIFFERENT THAN RELATED RESEARCH:
MILESTONES TARGETED START DATE TARGETED END DATE
PERCENT COMPLETED OVERALL:
ISSUES AND RISKS:
NEXT STEPS:
BENEFITS TO INDUSTRY: EXPECTED DELIVERABLES:
MULTI-UNIVERSITY PROJECT:
UPDATED: OCTOBER 1, 2017
CHOT CONFIDENTIAL
Integration of Population Health Data and Digital Assistants to Reduce Readmission Risks
11-05171.PSU-GIT Conrad Tucker, Eva Lee, Alva Ferdinand, & Susan Feldman
Population Health
9/1/2017
$50,000
Frequently, the factors that influence medical readmissions exist outside the borders of a healthcare setting and include patient-level decisions and societal interactions. The objective of this project is to leverage the size and availability of population health data to model and predict readmission risk factors. Data will be acquired on a large scale by mining publicly-available websites. The collected data will then be used to segment, model, and identify patients at risk of medical readmissions. For patient segments at a high risk of readmission, digital assistants (e.g. IBM Watson) will provide interactive feedback in an attempt to mitigate the risks.
Create a model that predicts readmission risk factors from population health data. Evaluate the impact of digital assistants in reducing medical readmission through pilot study.
Acquire large scale publicly available health data my mining. Segment, model, and identify patients at risk. Intervene via digital assistants. Conduct an experiment to evaluate impact. Report results.
Typically, medical readmission research focuses on investigating clinical-level factors (such as age and medical condition) that have the potential of increasing medical readmission. Yet when patients leave the hospital, a wide range of factors may influence their risk profiles, such as their support system and social norms. This project includes population health data that provides a more holistic understanding of what happens to patients once they are discharged from the hospital and utilizes a digital assistant that can provide real time decision support to patients who have been predicted to be at a higher risk of readmission.
Identify Data Features that Predict Risk 9/1/2017 10/15/2017
Evaluation of Value of Digital Assistants 10/1/2017 11/20/2017
Integration into Healthcare System Pilot 1/1/2018 3/1/2018
Dissemination of Work/Publications 3/1/2018 3/31/2018
10%
Risk of not acquiring enough meaningful data from publicly available social media sites due to privacy settings. Recruiting pilot subjects with the criteria needed to evaluate sufficiently.
Create Model of the features in social media data that predict medical readmission Create Digital Assistant (e.g., IBM Watson) that is tailored to providing interventions for patients at risk of medical readmissions.
This project benefits industry in several ways: 1) Discovery of the value of publicly available social media data in modeling patient-specific outcomes 2) Quantify the impact of digital assistants in serving as ubiquitous decision support systems
1) Model of features in social media data that predict medical readmission 2) Digital assistant (e.g., IBM Watson) that is tailored to providing interventions for patients at risk of medical readmissions
PSU & GIT
109
PROJECT TITLE:
PROJECT ID: PI:
RESEARCH THEME:
PROJECT START DATE:
BUDGET:
I/UCRC EXECUTIVE SUMMARY | PROJECT UPDATE
DESCRIPTION:
PROJECT OBJECTIVES: SCOPE:
HOW THIS IS DIFFERENT THAN RELATED RESEARCH:
MILESTONES TARGETED START DATE TARGETED END DATE
PERCENT COMPLETED OVERALL:
ISSUES AND RISKS:
NEXT STEPS:
BENEFITS TO INDUSTRY: EXPECTED DELIVERABLES:
MULTI-UNIVERSITY PROJECT:
UPDATED: OCTOBER 1, 2017
CHOT CONFIDENTIAL
Gamification and its Impact on the Population Health Management of Chronic Conditions
12-05171.PSU-GIT Nilam Ram, Eva Lee, Sherry Lin, & Susan Feldman
Population Health
9/1/2017
$50,000
Despite advances is medical technologies and public awareness programs, the rates of chronic conditions such as diabetes and asthma continue to rise. For example, in the United States, more than 29 million individuals have been diagnosed with diabetes, with a new diagnosis occurring every 23 seconds. This data indicates the need to evaluate the clinical effectiveness and economic impact of different approaches to managing diabetes and asthma. Data analysis will measure the impact of gamification methods in changing the behavior of patients, toward better healthcare outcomes.
The objective of this project is to evaluate the efficacy of chronic condition treatment programs.
Evaluate the clinical effectiveness and economic impact of approaches to managing diabetes and asthma; exploring secondary data analysis of program operations and biometric data on participants; investigating impact that gamification methods.
Existing research has focused on predicting factors that influence chronic diseases, but a knowledge gap exists between research on chronic disease management and translating the recommended practices for disease management. This project aims to identify specific practices that best translate into practice, as well as explore the influence of gamification in chronic disease management to determine whether successful implementations in other settings (e.g. education and rehabilitation) can be adapted for management of chronic disease.
Review Existing Lit on Existing Mgt. Programs 9/1/2017 9/30/2017
Study Factors that Facilitate Adoption of tech 9/30/2017 11/30/2018
Quantify Impact of Gamification (experiment) 1/1/2018 3/1/2018
Dissemination of Research/Publication 3/1/2018 3/31/2018
10%
Ineffectiveness or loss of interest in gamification. An increased addiction to mobile games. Reduced participation due to design of game (i.e. adverse to 3-D play)
Create and demonstrate downloadable apps that uses gamification to manage diabetes and asthma in conjunction with an industry partner for possible commercialization.
This project benefits industry in several ways: 1) Understanding the sets of basic research that may contribute to better management of chronic conditions 2) Exploring the role of gamification in changing behavior, towards positive health outcomes
1) Knowledge about the factors that impact chronic diseases such as diabetes and asthma 2) Gamification model that demonstrates short-term and long-term behavior changes in patients with chronic health conditions.
PSU & GIT
110
PROJECT TITLE:
PROJECT ID: PI:
RESEARCH THEME:
PROJECT START DATE:
BUDGET:
I/UCRC EXECUTIVE SUMMARY | PROJECT UPDATE
DESCRIPTION:
PROJECT OBJECTIVES: SCOPE:
HOW THIS IS DIFFERENT THAN RELATED RESEARCH:
MILESTONES TARGETED START DATE TARGETED END DATE
PERCENT COMPLETED OVERALL:
ISSUES AND RISKS:
NEXT STEPS:
BENEFITS TO INDUSTRY: EXPECTED DELIVERABLES:
MULTI-UNIVERSITY PROJECT:
UPDATED: OCTOBER 1, 2017
CHOT CONFIDENTIAL
Improving Employee and Patient Health through Population Data Mining
13-05171.PSU Vasant Honavar
Population Health
9/1/2017
$50,000
The majority of patients' time is spent away from the healthcare facility, where there exists little to no ability of healthcare decision makers to monitor patients' health improvement or outcomes. A recent study by the CDC reported that 54% (for women) and 42% (for men) of the 33+ million injuries occurring between 2004-2007 were inside/outside of home. The emergence of ubiquitous sensing systems such as mobile phones and wearable sensors has enabled the rapid acquisition of health-related data at population scale. This indicates the need to explore the ability to mine ambulatory data in order to improve employee and patient health outcomes.
The objective of this project is to explore methods to effectively manage the health of a population, who typically spend a majority of their time outside the confines of a healthcare facility.
Explore the current barrier that exist to capturing health-related data at home/work; conduct interviews at medical facility to gain understanding of needs; design a mobile app that has the ability to capture patient-specific data to be aggregated for population.
Existing research related to population health is limited by data acquisition tools (e.g. mobile app) currently available. Rather than utilize existing data acquisition tools, this project will design and create a data acquisition tool that is based on patient and employee feedback. Such feedback will guide tool development to ensure it is highly customizable, user friendly, patient access friendly, and valued by the healthcare decision makers and patients.
Understand In-Situ / Health Data Challenges 9/1/2017 10/15/2017
Quantify Attributes Needed for Mobile App 10/15/2017 12/15/2017
Design/Dev. iOS and Android App to Deploy 1/1/2018 3/1/2018
Disseminate Research 3/1/2018 3/30/2018
10%
Lack of access to devices. Fear of loss of privacy.
Work with industry partner to optimize and commercialize iOS and Android applications.
1) Healthcare app that can be deployed to capture patient-specific data 2) Data mining tool that can be used to extract valuable information from patient-centered data
1) Knowledge of the needs of patients and employees 2) Functional iOS and Android app 3) Data mining model that informs provision of patient-centered health care
PSU
111
Thank you! 2018 SPRING INDUSTRY ADVISORY BOARD MEETING SPONSORS