Iron Deficiency in Young Lebanese Children: Association With Elevated Blood Lead Levels
-
Upload
independent -
Category
Documents
-
view
1 -
download
0
Transcript of Iron Deficiency in Young Lebanese Children: Association With Elevated Blood Lead Levels
JOBNAME: No Job Name PAGE: 1 SESS: 34 OUTPUT: Tue Oct 9 18:30:08 2007/v2451/blackwell/journals/jbr_v5_i4/jbr_084
E V I D E N C E U T I L I Z A T I O N
Effectiveness of education in evidence-based healthcare: thecurrent state of outcome assessments and a framework forfuture evaluations
Mona Nabulsi xxxx, Janet Harris xxxx, Luz Letelier xxxx, Kathleen Ramos xxxx, KevorkHopayian xxxx, Claire Parkin xxxx, Franz Porzsolt xxxx, Piersante Sestini xxxx, Mary Slavinxxxx, William Summerskill xxxxXxxxxxxxxxxxxxxxxxxxxxxxxxsxxxxxxxxxxxxxxxxxxxx
AbstractBackground A discipline which critically looks at the evidence for practice should itself be critically examined.Credible evidence for the effectiveness of training in evidence-based healthcare (EBHC) is essential. We attempted tosummarise the current knowledge on evaluating the effectiveness of training in EBHC while identifying the gaps.
Methods A working group of EBHC teachers developed a conceptual framework of key areas of EBHC teachingand practice in need of evidence mapped to appropriate methods and outcomes. A literature search was conductedto review the current state of research in these key areas. Studies of training interventions that evaluated effectivenessby considering learner, patient or health system outcomes in terms of knowledge, skills, attitude, judgement,competence, decision-making, patient satisfaction, quality of life, clinical indicators or cost were included. There wasno language restriction.
Results Of 55 articles reviewed, 15 met the inclusion criteria: six systematic reviews, three randomised controlledtrials and six before–after studies. We found weak indications that undergraduate training in EBHC improvesknowledge but not skills, and that clinically integrated postgraduate teaching improves both knowledge and skills.Two randomised controlled trials reported no impact on attitudes or behaviour. One before–after study found apositive impact on decision-making, while another suggested change in learners’ behaviour and improved patientoutcome. We found no studies assessing the impact of EBHC training on patient satisfaction, health-related qualityof life, cost or population-level indicators of health.
Conclusions Literature evaluating the effectiveness of training in EBHC has focused on short-term acquisition ofknowledge and skills. Evaluation designs were methodologically weak, controlled trials appeared inadequatelypowered and systematic reviews could not provide conclusive evidence owing to weakness of study designs.
Key words: education, effectiveness, outcome assessment.
Background
Although the number of training programs in evidence-based healthcare (EBHC) is increasing rapidly, most of theseprograms have not demonstrated the effectiveness of inter-ventions to teach EBHC.1 The popularity of EBHC resultsfrom the belief that trainees become skilled at using currentbest research to provide the most effective and efficienthealthcare to their patients.2 However, there is little scientific
evidence that links training in evidence-based practice (EBP)to improved learner, patient or health system outcomes.3
In September 2003, delegates attending the SecondInternational Conference for Teachers and Developersof Evidence-Based Healthcare (‘Signposting the future ofEBHC’) suggested a list of issues in EBHC that are in need offurther evaluation.4 One outcome was the Sicily Statement,a consensus statement of what EBP means, a description ofthe skills required to practise in an EB manner and a curricu-lum that outlines the minimum requirements for train-ing health professionals in EBP.5 Another task set by theconference was the production of a framework for the
Correspondence: Dr Kevork Hopayian, Seahills Leiston Road, Alde-burgh, Suffolk IP15 5PL, UK. Email: [email protected]
doi:10.1111/j.1479-6988.2007.00084.x Int J Evid Based Healthc 2007; 5: 437–445
© 2007 The AuthorsJournal Compilation © Blackwell Publishing Asia Pty Ltd
12
3
4
5
6
7
8
9
10
11
12
13
14
1516
17
1819202122
23
242526272829
30
3132
33
34
35
36
37
38394041424344
45
4647
484950515253545556575859
11
JOBNAME: No Job Name PAGE: 2 SESS: 34 OUTPUT: Tue Oct 9 18:30:08 2007/v2451/blackwell/journals/jbr_v5_i4/jbr_084
comprehensive evaluation of training in EBHC. The task wasdelegated to a working group at the conference. A secondmeeting of the group was convened in October 2004 inUlm, Germany to carry this forward. Our goal was to sum-marise the current state of the art on evaluation of trainingEBHC in order to identify the gaps. This paper describes theposition arrived at by the group.
Methods
Working groupThe first stage of the study was a working group meetingwithout formal consensus methods to identify key areas ofEBHC training for evaluation of effectiveness. The groupspanned four continents (Europe, Asia and North and SouthAmerica), five disciplines (nursing, public health, occupa-tional health, physical therapy and medicine), and includedboth full-time academics and practising clinicians with aca-demic commitments. The results of this stage of enquiryform the methods of the subsequent stage and are pre-sented here.
We classified what we considered relevant outcomes intothree areas: learners, patients and health systems (Table 1).Outcomes were identified for each group. Learner outcomeswere further divided into three different domains: affective,cognitive and behavioural. The affective domain includesattitudes, beliefs and intentions. The cognitive domainincludes knowledge acquisition and skills development. Thebehavioural domain includes use of evidence in clinical prac-tice. Patient outcomes were further divided into patient sat-isfaction, health-related quality of life and improved patienthealth. System outcomes focused on cost-effectiveness.
We applied the structured approach of EBHC to develop aconceptual framework to search and to evaluate the litera-ture for the second stage. We used the PIO format (popula-tion, intervention, outcome) to formulate focused questionsfor each domain. For learner outcomes, we asked ‘Doesteaching evidence-based practice to health professionalsproduce a change in knowledge, attitude, or skills?’ Forpatient outcomes, ‘Does teaching evidence-based practiceto health professionals lead to better patient outcomes interms of improved satisfaction, health related quality of life,
or improved health?’ For systems outcomes, ‘Does teachingevidence-based practice to health professionals producemore cost effective services?’
Search strategy and study selectionTwo authors (MN and JH) independently searched MEDLINE(1980–March 2006), EMBASE (1980–March 2006), CINAHL(1982–March 2006) and COCHRANE (1998–March 2006).Both thesaurus and free text terms were used to identifyevaluations of learner, patient and health system outcomes.We searched both as MeSH and free text terms:• Health professionals AND• ((Evidence based health care OR Evidence based medi-
cine) AND (Teaching OR Education OR Training)) AND• ((Knowledge OR Attitudes OR Skills) OR (Patient satisfac-
tion OR Decision making OR Quality of life) OR (Practicepatterns OR Judgement OR Competence OR Clinical indi-cators OR Cost effectiveness))Papers in any language were considered and reference
lists of selected articles were searched for additional rele-vant studies. No attempt was made to identify unpublishedmaterial.
We selected papers if they described a teaching or trainingintervention for EBHC, offered in healthcare settings andevaluated impact in terms of learner, patient or system out-comes. The study type was left deliberately broad in order toobtain an overview of the various approaches to evaluation,including randomised controlled trials (RCTs), systematicreviews, before–after studies and other designs that assessedthe effects of any method of training in EBP. Papers wereexcluded if they failed to meet the previously mentionedselection criteria, used a non-validated assessment tool or ifthey focused on training for guidelines. Guidelines wereexcluded from consideration because their focus is not EBPtraining, even though their production and disseminationare partly educational. Studies identified in the search thatwere already discussed in one of the systematic reviewsincluded in our overview were not evaluated separately.
Articles were categorised by study design into systematicreviews, RCTs, before–after studies (with or without acontrol), and other designs that assessed the effects of anymethod of training in EBHC as an intervention, on anyhealthcare learner. Data were extracted by (i) study type; (ii)
Table 1 Classification of possible outcomes of evidence-based healthcare training
Learner outcomes Patient outcomes System outcomes
Affective domains: • Increased patient satisfaction • Population health• Satisfaction with training • Health-related quality of life • Cost• Attitudes toward evidence-based practice • Improved patient health• Intentions to use an evidence-based practice approach
Cognitive domains:• Increased objective knowledge• Improved skills
Behavioural domains:• Transfer of knowledge and skills to the workplace
438 M Nabulsi et al.
© 2007 The AuthorsJournal Compilation © Blackwell Publishing Asia Pty Ltd
1234567
8
9
1011121314151617181920212223242526272829303132333435363738394041
42
43
44
45464748
495051
525354
555657
585960616263646566676869707172737475767778798081828384858687888990919293949596
JOBNAME: No Job Name PAGE: 3 SESS: 34 OUTPUT: Tue Oct 9 18:30:08 2007/v2451/blackwell/journals/jbr_v5_i4/jbr_084
learner population; (iii) content of teaching; (iv) mode ofdelivery; (v) duration of teaching; and (vi) outcome. Sam-pling, confounding, definition of outcomes and measure-ment tools were reviewed with the aim of assessing therigour of educational evaluation methods.
Results
Search resultsThe search identified 3016 articles of which 2961 wereexcluded based on the title and/or abstract (Fig. 1). Adetailed review was conducted for 55 articles. Fifteen studiesmet our selection criteria and were included in this overview:six systematic reviews,6–11 three RCTs12——14 and six before–after studies15–20 (Tables 2–4). The population of learnersvaried among studies and included medical students, resi-dents, practising physicians, nurses and physiotherapists.Training interventions were heterogeneous in terms of thebackground and skills of the teacher (resident, peer ortutor teaching); mode of delivery (lectures, seminars, andproblem-based learning); content of teaching (focusing onelements of question formulation, searching skills, criticalappraisal, and use of evidence for decision-making); dura-tion of teaching (number of seminars amd length of time);and provision of supporting materials or resources such ascomputer access to critically appraised topics or electronicdatabases. All six systematic reviews, one controlled trial14
and four before–after studies15,16,19,20 assessed knowledgeand skills as the main outcome. Change in behaviour,defined as the use of research evidence in clinical practice,was the primary outcome in one RCT,12 while change inattitude was the focus of one cluster RCT.13 Three uncon-trolled before–after studies assessed the impact of EBHCtraining on change in practice behaviour and decision-
making17,18,20 with one only reporting the impact onpatients’ outcomes.18 We found no studies assessing theeffects of training in EBHC on patient satisfaction, health-related quality of life, cost or population-level indicators ofhealth.
Learner outcomesAffective domain: attitudes, beliefs and intentionsAttitudes were variously defined as attitudes toward medicalliterature,7 attitudes toward participating in an elective,15
perceived importance for clinical practice,16 attitudes towardthe use of research information,12 perceived skills and confi-dence20 and attitudes toward facilitating integration of EBpractice into the workplace.18 We found one cluster-randomised trial that assessed learners’ attitudes as aprimary outcome of EBHC training.13 In this trial, musculosk-eletal physiotherapists of mixed experience were allocated toreceive either an EB package (critical appraisal, literaturesearching and training in low-back pain management withopinion leaders taking part in training sessions), or in-servicetraining on management of knee dysfunction. A self-administered questionnaire used to assess attitude towardEBP failed to detect any significant change. Participantsreported that they depended more upon opinion leadersand textbooks to inform treatment decisions, than onprimary resources or literature searching.13 However, thisstudy was inadequately powered with 16 subjects in theintervention group and 11 in the control group. In addition,groups were different at baseline, and had too few clustersto use multilevel analysis.
Cognitive domain: assessment of knowledge and skillsFive systematic reviews,6–9,11 one controlled trial14 and fourbefore–after studies15,16,19,20 assessed knowledge of EBP asthe primary outcome of training in EBP. It is important tonote that the definitions of knowledge or skills varied amongindividual studies. Knowledge was defined variously asknowledge of information sources,7,12,20 information retrievalskills,7 skills in critical appraisal,6,7,14 knowledge about con-cepts in critical appraisal,12 knowledge of statistics,6,7,19,20
knowledge of epidemiological concepts,7 development ofproblem solving skills and development of clinicalskills.14,16,18 The level of knowledge was further defined inone study as ‘superficial’ versus ‘deep’.15 Superficial knowl-edge was assessed via ability to reproduce facts by correctlyanswering questions about basic principles of EBP, whiledeep knowledge was explored via clinical scenarios thatwere designed to assess ability to use evidence conceptsin proxy decision-making situations.14–16 Two reviews9,10
acknowledged the heterogeneity of outcome definitions butcontained no information on the respective definitions.
The measurement tools used in the different studies wereheterogeneous. Some studies used a published validatedinstrument such as the Fresno test19,21 or Berlin question-naire,15,22 or modified the validated Fresno test,20 and othersused a locally validated tool but lacked detail on the valida-tion process.12,16 Variations in outcome definitions and in
3016 potentially relevant articles identified and screened for retrieval
2961 excluded (no evaluation of learner, patient or system outcomes; focus on guidelines training)
55 selected for full text review
40 excluded (training intervention not described; non-validated assessment tool)
15 included (6 systematic reviews; 3 controlled trials; 6 before-after studies)
Figure 1 Flow diagram of studies.
Effectiveness of EBHC education 439
© 2007 The AuthorsJournal Compilation © Blackwell Publishing Asia Pty Ltd
12345
6
7
89
101112131415161718192021222324252627282930313233
34
35
3637383940
41
424344454647484950515253545556575859606162636465
66
6768697071727374757677787980818283848586878889909192
JOBNAME: No Job Name PAGE: 4 SESS: 34 OUTPUT: Tue Oct 9 18:30:08 2007/v2451/blackwell/journals/jbr_v5_i4/jbr_084
Tab
le2
Sum
mar
yof
stud
ies
asse
ssin
gef
fect
iven
ess
ofev
iden
ce-b
ased
heal
thca
rete
achi
ngan
dle
arni
ng:
Syst
emat
icre
view
findi
ngs
Aut
hor
(yea
r)St
udy
desi
gn(n
)Le
arne
rIn
terv
entio
nO
utco
me
Valid
ityC
oncl
usio
n
Nor
man
&Sh
anno
n(1
998)
6
CT
(4)
RCT
(3)
Med
ical
stud
ents
;re
side
nts
Crit
ical
app
rais
alsk
ills
Know
ledg
e;de
cisi
on-m
akin
gLi
mite
dse
arch
stra
tegy
;m
etho
dolo
gica
lsco
re:
50–8
3%;
smal
lsam
ple
size
s;kn
owle
dge
asde
fined
inva
rious
stud
ies
not
defin
ed
Und
ergr
adua
te:
imp
rove
dkn
owle
dge;
smal
leffe
ctsi
zes.
Post
grad
uate
:m
inim
alch
ange
inkn
owle
dge;
noch
ange
inde
cisi
on-m
akin
gTa
ylor
,et
al.
(200
0)7
CT
(10)
Med
ical
stud
ents
;re
side
nts
Crit
ical
app
rais
alsk
ills
Perc
eive
dco
mp
eten
cein
abili
tyto
criti
cally
app
rais
e;Ev
iden
ce-
seek
ing
beha
viou
r;Kn
owle
dge
bios
tatis
tics,
epid
emio
logy
and
sear
chin
g;at
titud
esto
war
dev
iden
cean
dus
eof
liter
atur
e
Med
ian
qua
lity
scor
eof
stud
ies
3/10
(wea
k);
inst
rum
ents
not
sens
itive
enou
ghto
mea
sure
true
valu
eof
CA
Sp
rogr
ams
Imp
rove
dkn
owle
dge;
inco
nclu
sive
resu
ltsfo
rot
her
outc
omes
Park
eset
al.
(200
1)8
RCT
(1)
Hea
lthp
rofe
ssio
nals
Crit
ical
app
rais
alsk
ills
Know
ledg
e;p
roce
ssof
care
;p
atie
ntou
tcom
es
Onl
yon
est
udy
incl
uded
;sm
alls
amp
lesi
zeM
odes
tim
pro
vem
ent
inkn
owle
dge;
noev
iden
cefo
und
for
othe
rou
tcom
esC
oom
aras
amy
etal
.(2
003)
9RC
Ts(4
);C
ontr
olle
d,no
n-ra
ndom
ised
(6);
Befo
re–a
fter
(9)
Post
grad
uate
sEB
M;
criti
cala
pp
rais
alKn
owle
dge;
skill
s;at
titud
e;be
havi
our
Het
erog
eneo
usst
udy
feat
ures
and
met
hodo
logi
calq
ualit
y
Imp
rove
dkn
owle
dge;
noco
mm
ent
onef
fect
size
s;no
chan
gein
othe
rou
tcom
esBr
ettle
(200
3)10
RCTs
,C
ohor
t,q
ualit
ativ
e(2
4)H
ealth
care
pro
fess
ions
Info
rmat
ion
skill
sSk
ills;
Patie
ntca
reM
arke
dhe
tero
gene
ityof
stud
yde
sign
sLi
mite
dev
iden
ceon
imp
rove
dsk
ills
orch
ange
inp
atie
ntca
reC
oom
aras
amy
&Kh
an(2
004)
11
RCTs
(4)
Con
trol
led
non-
rand
omis
ed(7
)Be
fore
–aft
er(1
2)
Post
grad
uate
sSt
and-
alon
eEB
Mte
achi
ngvs
.C
linic
ally
inte
grat
edEB
Mte
achi
ng
Know
ledg
e;Sk
ills;
Att
itude
;Be
havi
our
Het
erog
eneo
usst
udy
feat
ures
and
met
hodo
logi
calq
ualit
y
Stan
d-al
one:
imp
rove
dkn
owle
dge
only
.C
linic
ally
inte
grat
ed:
imp
rove
dkn
owle
dge,
skill
s,at
titud
e&
beha
viou
r;no
com
men
ton
effe
ctsi
zes
Num
ber
inp
aren
thes
is,
num
ber
ofst
udie
sin
clud
edin
the
revi
ew.
CT,
non-
rand
omis
edco
ntro
lled
tria
l;EB
M,
evid
ence
-bas
edm
edic
ine;
RCT,
rand
omis
edco
ntro
lled
tria
l.
440 M Nabulsi et al.
© 2007 The AuthorsJournal Compilation © Blackwell Publishing Asia Pty Ltd
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38
JOBNAME: No Job Name PAGE: 5 SESS: 34 OUTPUT: Tue Oct 9 18:30:08 2007/v2451/blackwell/journals/jbr_v5_i4/jbr_084
Tab
le3
Sum
mar
yof
stud
ies
asse
ssin
gef
fect
iven
ess
ofev
iden
ce-b
ased
heal
thca
re(E
BHC
)te
achi
ngan
dle
arni
ng:
cont
rolle
dtr
ials
Aut
hor
(yea
r)St
udy
desi
gnLe
arne
rIn
terv
entio
nO
utco
me
Valid
itySt
udy
conc
lusi
on
Fors
etlu
ndet
al.
(200
3)12
RCT
Nor
weg
ian
phy
sici
ans
EBH
Cw
orks
hop
;ne
wsl
ette
r;in
form
atio
nse
rvic
e;el
ectr
onic
data
base
;el
ectr
onic
disc
ussi
onlis
t
Beha
viou
r:us
eof
EB-r
esea
rch;
attit
udes
;kn
owle
dge
ofin
form
atio
nso
urce
san
dco
ncep
ts;
self-
effic
acy;
deci
sion
toad
opt
Ade
qua
tera
ndom
isat
ion
and
blin
ding
;va
lidat
edin
stru
men
ts;
adju
sted
anal
yses
;<8
0%re
spon
se;
noIT
Tan
alys
es
No
chan
gein
beha
viou
r;in
crea
sein
know
ledg
e
Stev
enso
net
al.
(200
4)13
Clu
ster
RCT
Phys
ioth
erap
ists
EBp
rogr
am;
low
-bac
kp
ain
man
agem
ent;
opin
ion
lead
ers
taki
ngp
art
intr
aini
ngse
ssio
ns.
Con
trol
:in
-ser
vice
trai
ning
onm
anag
emen
tof
knee
dysf
unct
ion
Att
itude
tow
ard
EBP
Smal
lsam
ple
size
(E=
16;
C=
11);
blin
ding
?;gr
oup
sdi
ffere
ntat
base
line;
self-
adm
inis
tere
dva
lidat
edin
stru
men
t
No
chan
gein
attit
ude;
gett
ing
evid
ence
into
pra
ctic
evi
alit
erat
ure
sear
chin
gw
asa
low
prio
rity
and
not
ahi
ghly
rate
dfa
ctor
inch
oice
oftr
eatm
ent
Maj
or-
Kinc
ade
etal
.(2
001)
14
Qua
sira
ndom
ised
cont
rolle
dtr
ial
Paed
iatr
icho
use
staf
f;N
ICU
rota
tion
45–6
0m
inof
EBet
hics
,in
clud
ing
criti
cal
app
rais
al
Know
ledg
e,ap
pra
isal
skill
s,de
cisi
on-m
akin
gin
ethi
cali
ssue
s
Non
-val
idat
edq
uest
ionn
aire
;cl
inic
alsc
enar
ios;
app
rais
alof
pro
gnos
isst
udie
s
Imp
rove
dkn
owle
dge
ofsh
ort-
term
outc
omes
;no
imp
act
onlo
ng-t
erm
outc
omes
;no
imp
act
oncr
itica
lap
pra
isal
skill
sor
ethi
cal
deci
sion
-mak
ing
EB,
evid
ence
-bas
ed;
EBP,
evid
ence
-bas
edp
ract
ice;
ITT,
xxx;
NIC
U,
xxxx
;RC
T,ra
ndom
ised
cont
rolle
dtr
ial.
Effectiveness of EBHC education 441
© 2007 The AuthorsJournal Compilation © Blackwell Publishing Asia Pty Ltd
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
JOBNAME: No Job Name PAGE: 6 SESS: 34 OUTPUT: Tue Oct 9 18:30:08 2007/v2451/blackwell/journals/jbr_v5_i4/jbr_084
Tab
le4
Sum
mar
yof
stud
ies
asse
ssin
gef
fect
iven
ess
ofev
iden
ce-b
ased
heal
thca
rete
achi
ngan
dle
arni
ng:
befo
rean
daf
ter
stud
ies
Aut
hor
(yea
r)St
udy
desi
gnLe
arne
rIn
terv
entio
nO
utco
me
Valid
itySt
udy
conc
lusi
on
Akl
etal
.(2
004)
15C
ontr
olle
d(t
wo
grou
ps)
Post
grad
uate
sTw
oEB
lect
ures
;re
side
ntte
achi
ng:
resi
dent
sear
ched
,ap
pra
ised
and
fed
back
onq
uest
ions
ofw
ard
team
Know
ledg
e;at
titud
esre
gard
ing
acce
pta
bilit
yRa
ndom
ised
but
imba
lanc
edba
selin
ech
arac
teris
tics;
cont
rol
blin
ded;
valid
ated
inst
rum
ent
(ada
pte
dBe
rlin)
Imp
rove
dkn
owle
dge
(NS)
;p
ositi
veat
titud
esto
war
dEB
P
Web
ersc
hock
etal
.(2
005)
16
Unc
ontr
olle
d;Va
lidat
edin
stru
men
t
Med
ical
stud
ents
Four
EBM
sem
inar
s;p
eer
teac
hing
Self-
asse
ssed
know
ledg
e;sk
ills
defin
edas
pee
rev
alua
tion
ofcl
inic
alsc
enar
ios;
acce
pta
bilit
yde
fined
asp
erce
ived
imp
orta
nce
for
clin
ical
pra
ctic
e
Inst
rum
ent
valid
ated
byas
sess
ing
tuto
rs’
per
cep
tion
ofdi
fficu
ltyw
ithp
rop
ortio
nof
corr
ect
resp
onse
s
Imp
rove
dkn
owle
dge
and
skill
s:q
uest
ion
form
ulat
ion,
sear
chin
g,cr
itica
lap
pra
isal
and
app
licat
ion
tocl
inic
alsc
enar
io
Stra
uset
al.
(200
5)17
Befo
re–a
fter
;no
cont
rol
Att
endi
ngp
hysi
cian
san
dre
side
nts
Com
ple
xin
terv
entio
n:se
ven
1-h
sess
ions
onEB
Msk
ills;
elec
tron
icEB
reso
urce
son
hosp
ital
netw
ork
and
war
d
Dec
isio
n-m
akin
g:p
rovi
sion
ofEB
ther
apie
sto
pat
ient
sas
sess
edth
roug
hp
re-p
ost
revi
ewof
disc
harg
esu
mm
arie
sof
com
mon
adm
ittin
gdi
agno
ses
Blin
ding
ofou
tcom
eas
sess
ors;
obje
ctiv
eou
tcom
eas
sess
men
t;hi
ghly
rele
vant
clin
ical
diag
nose
s
Incr
ease
inhi
gh-q
ualit
yEB
inte
rven
tions
Jeffe
ryet
al.
(200
4)18
Befo
re–a
fter
;no
cont
rol
Mac
edon
ian
doct
ors
and
nurs
es
Com
ple
xin
terv
entio
n:EB
Med
ucat
ion
deliv
ered
inta
ndem
with
infr
astr
uctu
rala
ndor
gani
satio
nals
upp
ort
Hea
lthsy
stem
:p
erin
atal
mor
talit
yra
te.
Lear
ner:
satis
fact
ion
with
trai
ning
,cl
inic
alco
mp
eten
cean
dp
robl
emso
lvin
g
Obj
ectiv
eou
tcom
eas
sess
men
t:q
uest
ionn
aire
,ob
serv
atio
n,cl
inic
alau
dit
data
,us
eof
com
put
er-b
ased
pro
toco
ls
EBed
ucat
ion
deliv
ered
with
ina
pro
gram
targ
etin
gsp
ecifi
cse
rvic
eim
pro
vem
ents
and
chan
ging
man
agem
ent
skill
sre
duce
dp
erin
atal
mor
talit
yra
teby
21.5
%in
2ye
ars
Din
kevi
chet
al.
(200
6)19
Unc
ontr
olle
dPa
edia
tric
resi
dent
sFo
ur2-
hEB
Mse
ssio
nsKn
owle
dge
and
skill
sVa
lidat
edto
ol(F
resn
o)Im
pro
ved
know
ledg
e&
skill
s
McC
lusk
y&
Lova
rini
(200
5)20
Unc
ontr
olle
dO
ccup
atio
nal
ther
apis
tsTw
o-da
yEB
Pw
orks
hop
;m
ultif
acet
edte
achi
ngap
pro
ach;
CAT
’sge
nera
tion,
follo
w-u
pou
trea
chsu
pp
ort
Know
ledg
e;sk
ills;
attit
udes
;be
havi
our
Valid
ated
inst
rum
ent
(ada
pte
dFr
esno
);q
uest
ionn
aire
;ac
tivity
diar
yof
EBP
activ
ities
;lo
ssto
follo
wup
>80%
Imp
rove
dkn
owle
dge
and
skill
sm
aint
aine
dfo
r8
mon
ths;
noch
ange
inbe
havi
our
CAT
,cr
itica
llyap
pra
ised
top
ics;
EB,
evid
ence
-bas
ed;
EBM
,ev
iden
ce-b
ased
med
icin
e;EB
P,ev
iden
ce-b
ased
pra
ctic
e.
442 M Nabulsi et al.
© 2007 The AuthorsJournal Compilation © Blackwell Publishing Asia Pty Ltd
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42
JOBNAME: No Job Name PAGE: 7 SESS: 34 OUTPUT: Tue Oct 9 18:30:08 2007/v2451/blackwell/journals/jbr_v5_i4/jbr_084
assessment tools contributed significantly to heterogeneityamong studies, which limited meaningful comparison.
Two systematic reviews6,7 and one uncontrolled before–after study16 assessing the effect of EBHC training on knowl-edge and skills in undergraduate medical students reportedimprovement in knowledge but not skills. For postgraduates,the same two reviews6,7 and one controlled trial14 suggestedminimal improvement in knowledge only. Two subsequentsystematic reviews9,11 concluded that clinically integratedEBHC teaching improved knowledge and skills of residents,compared with stand-alone teaching, which improvedknowledge only. Both these reviews included randomisedand non-randomised controlled trials, as well as before–aftercomparison studies, and were heterogeneous with respectto study design and methodological quality. In addition,effect sizes of individual studies were not specified withoutcomes reported as change in knowledge and skills. Find-ings of two other systematic reviews8,10 evaluating the effectsof EBHC training on health professionals in general wereconsistent in terms of modest knowledge improvement andlimited evidence on improved skills. A controlled before–after study15 involving postgraduates found a non-significantimprovement in knowledge using the validated Berlin ques-tionnaire, while a more recent uncontrolled before–afterstudy reported improved knowledge and skills of paediatricresidents when assessed with the Fresno test.19
Behavioural domainOne before–after study assessed change in the frequencyof question formulation, evidence retrieval and criticalappraisal.20 This study followed participants over time,finding that although learners sustained an increase inknowledge of terms and resources for EBP, there was littleincrease in frequency of searching or critical appraisal. Wefound one RCT12, and one before–after study17 that assessedthe impact of EBHC training on physicians’ behaviours, usingdifferent methods. While the RCT12 reported that a multifac-eted training intervention did not lead to greater use ofevidence, the before–after study17 reported change in use ofevidence in decisions about treatment following EBHC train-ing of a group of residents and attending physicians in acommunity hospital.
Patient outcomesOnly one study18 reported on patients’ health outcomesfollowing EB training of health professionals. Doctors andnurses received a complex intervention consisting of educa-tion in EBHC that was delivered in tandem with infrastruc-tural and organisational support, and aimed at reducingperinatal mortality rate (PMR) in Macedonia, which has oneof the highest PMR rates in Europe. In addition to the healthsystem outcome, learner outcomes such as satisfaction withtraining, clinical competence and problem solving wereassessed through questionnaires, observation, clinical auditdata and use of computer-based protocols. This complexintervention was successful in reducing PMR by 21.5% over2 years. Although it is difficult to quantify the contribution ofEB education to this success, the study is a good example of
the importance of contextual factors in enhancing theimpact of teaching and learning EBP.
System outcomesTraining in EBHC can affect outcomes at the health servicelevel with respect to population health and quality of care.This was demonstrated by Jeffery et al.,18 who reported thata multistage project that began with training in EBP, whencombined with support in the implementation of evidence,brought about system-wide changes in practice. No otherstudies were found, however, that assessed the contributionof training in EBP to improvements in healthcare systems.
Cost-effectivenessIt is not known whether or not the systematic application ofEBHC is cost-effective. We found no study that assessed theimpact of EBHC training on cost of healthcare.
Discussion
Strengths and limitations of the studyEnrolment in the original workshop and invitation to theUlm workshop were unrestricted, so that the decisions takenare those of freely participating and independent individuals.It could be argued that the self-selected nature of the groupraises the possibility that its priorities may not be represen-tative. However, the broad span of its membership lendscredibility to its deliberations.
The current state of assessmentMost of the literature evaluating the effectiveness of EBHCtraining has focused on short-term acquisition of knowledgeand skills not exceeding 9 months. Knowledge has beenassessed through both validated and unvalidated tools.Although there are validated instruments for evaluating EBP,most instruments have assessed self-reporting of learners’perceptions knowledge, skill or performance. A recent sys-tematic review has found that retrospective self-assessmentsof skills are not an accurate indicators of skills acquisition.23
There are few assessment tools that capture changes inattitudes, behaviors or skills, particularly in relation to theimpact on patient outcomes.24 Systematic reviews could notprovide conclusive evidence of significant improvement inknowledge or skills following training in EBHC because of theweakness of study designs and their heterogeneity. Sourcesof heterogeneity included: variations in learner populationsin the different studies with potential variation in learners’needs; combinations of multiple simultaneous interventions;differences in training interventions in terms of backgroundand skills of the teacher, mode of delivery, duration of theteaching and learning, and availability of supporting mate-rials and resources; different definitions of knowledge orskills in the various studies; and variations in assessment toolsand time frames. The degree of heterogeneity is not surpris-ing, given the wide range of teaching interventions, partici-pants and study designs, but now that this variation hasbeen documented, future reviews should focus on exploringhow these factors influence outcome in greater depth.
Effectiveness of EBHC education 443
© 2007 The AuthorsJournal Compilation © Blackwell Publishing Asia Pty Ltd
123456789
10111213141516171819202122232425262728293031323334353637383940414243444546474849505152535455565758
5960616263646566676869707172737475
76
77
78798081828384858687888990919293949596979899
100101102103104105106107108109110111112113114115
JOBNAME: No Job Name PAGE: 8 SESS: 34 OUTPUT: Tue Oct 9 18:30:08 2007/v2451/blackwell/journals/jbr_v5_i4/jbr_084
Evidence from RCTs was weak because of either poorquality or threats to internal validity, such as small samplesize or differences in baseline characteristics between experi-mental and control groups. Studies that assessed skills oftenused a proxy measurement such as clinical scenarios. Proxymeasures may not capture the barriers in everyday practicethat prevent implementation of EB skills, and better obser-vational measures or indicators of changes in practice suchas those used by Straus et al.17 need to be used.
There is a paucity of research evaluating the link betweentraining in EBHC and longer-term effects. As acknowledgedby Forsetlund et al.,12 evaluation tools may capture part ofthe effect (short-term improvement in knowledge and skills),but not the long-term sustainability of knowledge, skills andattitudes or their translation into changed behaviour ofhealth professionals and improved outcomes for patients. Inorder to answer whether EBHC training makes a difference,complex interventions are needed, which may representcombinations of multiple simultaneous interventions such asthose found in the Forsetlund study. The challenges of estab-lishing a causal link between exposure to an intervention,changes in behaviour and improved outcomes have beenextensively discussed elsewhere in evaluation research,25 andmethods for establishing causation must be tailored to par-ticular behaviours and outcomes.26 More studies need to beconducted that attempt to establish a relationship betweenteaching, adoption of EBP and changes in the process ofcare. The two studies by Straus et al.17 and Jeffery et al.18 areimportant because they attempted to establish such a link,nested the intervention in the practitioner context, andattempted to describe other contextual factors that mayhave influenced the impact of training in EBHC.
Future evaluationsIn evaluating the impact of training in EBHC, there is a needto shift the focus from educational interventions to evalua-tion of the teaching and learning pathway. Assessment ofthe ‘big picture’ of increased knowledge and skills, durabilityof attained skills, change in practice, and patient outcomeshould be a priority for interested researchers. RCTs arethe ideal study designs; however, they may be difficult toconduct, and it may be unethical to deprive the controlgroup from training in EBHC, especially if the outcome to beassessed is long-term retention of acquired skills. Given thewidespread popularity of EBHC concepts, it cannot beguaranteed that the control group remains unexposed totraining in EBHC for the duration of the study. Therefore,multi-site studies, perhaps with cluster randomisation, maybe needed to overcome the problems of contamination,inadequate power and generalisability. In addition, complexintervention designs such as the use of mixed qualitative andquantitative methods, or selection of multiple measurementtools may be helpful in capturing the various factors influ-enced by training in EBHC. Countries that have attempted toimprove their quality of care through re-organisations oftheir health service, similar to Jeffrey et al.’s study,18 are‘natural experiments’ where observational research could beconducted.
The primary goal of healthcare education is to producehealth professionals who deliver high-quality healthcare.However, there has been remarkably little investment intothe conceptualisation and study of the association betweenthe process of medical education and quality of care.27
Longer-term behaviour change, which includes consistentconsideration of evidence in clinical practice, is much moredifficult to attribute to a teaching intervention. Behaviourchange among health professionals does not automaticallyfollow from increased knowledge and skills, as it is a productof culminating experiences, rather than a single educationalintervention. Although patient outcomes have not beenextensively examined, they are the most relevant productof training in EBHC and therefore a priority for evaluationresearch. Hence, it is essential to identify appropriate andmeasurable patient health outcomes when examining theeffects of EBHC training. Implementation of EBP is very likelyto be affected by the extent of institutional support or resis-tance faced by the health professional. Barriers to practicewill certainly limit the observed effect on behaviour changeand consequently on patient health outcomes. It is impera-tive therefore, that when evaluating the impact of training inEBHC on health systems, the methods include description ofthe context in which the intervention takes place. Consider-ation needs to be given to contextual factors that supporteither the buttressing or the deterioration of an effect overtime. Future research evaluating training in EBHC should alsofocus on providing evidence for the cost-effectiveness of theinterventions. Information produced from this sort of studywould help to convince managers, department heads andpolicymakers to support the transfer of EB knowledge andskills to the workplace.
Conclusions
Most of the literature evaluating the effectiveness of trainingin EBHC has focused on short-term acquisition of knowledgeand skills. There is paucity of research evaluating the linkbetween training in EBHC and knowledge transfer. Providingcredible evidence on the effectiveness of training in EBHCrequires concerted efforts on the part of many educators andresearchers. We hope that this overview is useful to thoseinterested in evaluation research, and that some years fromnow we will be able to provide more conclusive answersabout the effectiveness of training programs for EBHC.
References1. Hatala R, Guyatt G. Evaluating the teaching of evidence-based
medicine. JAMA 2002; 288: 1110–2.2. Sackett DL, Straus SE, Richardson WS, Rosenberg W, Haynes RB.
Evidence-Based Medicine: How to Practice and Teach EBM.London: Churchill Livingstone, 2000.
3. Straus SE, McAlister FA. Evidence-based medicine: a commen-tary on common criticisms. CMAJ 2000; 163: 837–41.
4. Second International Conference of Evidence-Based Health CareTeachers and Developers. Reports from Working Groups, 2003.Accessed 11 Jun 2006. Available from: http://www.ebhc.org/2003/13-9.htm
444 M Nabulsi et al.
© 2007 The AuthorsJournal Compilation © Blackwell Publishing Asia Pty Ltd
123456789
10111213141516171819202122232425262728293031323334353637383940414243444546474849505152535455565758
5960616263646566676869707172737475767778798081828384858687888990
91
92
93949596979899
100101102
103
104
105106107108109110111112113114115
JOBNAME: No Job Name PAGE: 9 SESS: 34 OUTPUT: Tue Oct 9 18:30:08 2007/v2451/blackwell/journals/jbr_v5_i4/jbr_084
5. Dawes MG, Summerskill W, Glasziou P et al. Sicily statement onevidence-based practice. BMC Med Educ 2005; 5: 1. Accessed30 October 2006. Available from: http://www.biomedcentral.com/1472-6920/5/1
6. Norman GR, Shannon SI. Effectiveness of instruction in criticalappraisal (evidence-based medicine) skills: a critical appraisal.CMA J 1998; 158: 177–81.
7. Taylor R, Reeves B, Ewings P, Binns S, Keast J, Mears R. Asystematic review of the effectiveness of critical appraisal skillstraining for clinicians. Med Educ 2000; 34: 120–5.
8. Parkes J, Hyde C, Deeks J, Milne R. Teaching critical appraisalskills in health care settings. Cochrane Database Syst Rev 2001;3: CD001270.
9. Coomarasamy A, Taylor R, Khan KS. A systematic review ofpostgraduate teaching in evidence-based medicine and criticalappraisal. Med Teach 2003; 25: 77–81.
10. Brettle A. Information skills training: a systematic review of theliterature. Health Info Libr J 2003; 20: 3–9.
11. Coomarasamy A, Khan KS. What is the evidence that post-graduate teaching in evidence based medicine changes any-thing? A systematic review. BMJ 2004; 329: 1017–22.
12. Forsetlund L, Bradley P, Forsen L, Nordheim L, Jamtvedt G,Bjørndal A. Randomised controlled trial of a theoreticallygrounded tailored intervention to diffuse evidence-based publichealth practice. BMC Med Educ 2003; 3: 2.
13. Stevenson K, Dip G, Lewis M, Hay E. Do physiotherapists’attitudes towards evidence-based practice change as a result ofan evidence-based educational programme? J Eval Clin Pract2004; 10: 207–17.
14. Major-Kincade TL, Tyson JE, Kennedy KA. Training pediatrichouse staff in evidence-based ethics: an exploratory controlledtrial. J Perinatol 2001; 21: 161–6.
15. Akl EA, Izuchukwu IS, El-Dika S, Fritsche L, Kunz R, SchünemannHJ. Integrating an evidence-based medicine rotation into aninternal medicine residency program. Acad Med 2004; 79: 897–904.
16. Weberschock TB, Ginn TC, Reinhold J et al. Change in knowl-edge and skills of year 3 undergraduates in evidence-basedseminars. Med Educ 2005; 39: 665–71.
17. Straus SE, Ball C, Balcombe N, Sheldon J, McAlister FA. Teachingevidence-based medicine skills can change practice in a com-munity Hospital. J Gen Intern Med 2005; 20: 340–3.
18. Jeffery HE, Kocova M, Tozija F et al. The impact of evidence-based education on a perinatal capacity-building initiative inMacedonia. Med Educ 2004; 38: 435–47.
19. Dinkevich E, Markinson A, Ahsan S, Lawrence B. Effect of a briefintervention on evidence-based medicine skills of pediatric resi-dents. BMC Med Educ 2006; 6: 1.
20. McCluskey A, Lovarini M. Providing education on evidence-based practice improved knowledge but did not change behav-iour: a before and after study. BMC Med Educ 2005; 5: 40.
21. Ramos KD, Schafer S, Tracz SM. Validation of the Fresno test ofcompetence in evidence based medicine. BMJ 2003; 326: 319–21.
22. Fritsche L, Greenhalgh T, Falck-Ytter Y, Neumayer HH, Kunz R.Do short courses in evidence based medicine improve knowl-edge and skills? Validation of Berlin questionnaire and beforeand after study of courses in evidence based medicine. BMJ2002; 325: 1338–41.
23. Davis DA, Mazmanian PE, Fordis M, Van Harrison R, Thorpe KE,Perrier L. Accuracy of physician self-assessment compared withobserved measures of competence: a systematic review. JAMA2006; 296: 1094–102.
24. Shaneyfelt T, Baum KD, Bell D et al. Instruments for evaluatingeducation in evidence-based practice: a systematic review.JAMA 2006; 296: 1116–27.
25. Pawson R, Tilley N. Realistic Evaluation. London: Sage, 1998.26. Grol R, Grimshaw J. From best evidence to best practice: effec-
tive implementation of change in patients’ care. Lancet 2003;362: 1225–30.
27. Chen FM, Bauchner H, Burstin H. A call for outcomes researchin medical education. Acad Med 2004; 79: 955–60.
Effectiveness of EBHC education 445
© 2007 The AuthorsJournal Compilation © Blackwell Publishing Asia Pty Ltd
123456789
101112131415161718192021222324252627282930313233343536
373839404142434445464748495051525354555657585960616263646566676869707172
JOBNAME: No Job Name PAGE: 10 SESS: 34 OUTPUT: Tue Oct 9 18:30:08 2007/v2451/blackwell/journals/jbr_v5_i4/jbr_084
AUTHOR QUERY FORM
Dear Author
During the preparation of your manuscript, the questions listed below have arisen. Please answer all the queries(marking any other corrections on the proof enclosed) and return this form with your proofs.
QueryReferences
Query Remarks
q1 Au: Please supply the authors’ qualifications and affiliation addresses.
q2 Au: (Table 2) Please spell out: CAS.
q3 Au: (Table 3) Please clarify what the question mark ‘?’ (as in ‘blinding ?’)indicates.
q4 Au: (Table 3) Please define ITT and NICU.
q5 Au: (Table 4) Please spell out: NS.
SNP Best-set Typesetter Ltd.Journal Code: JBR Proofreader: MonyArticle No: 084 Delivery date: 9 October 2007Page Extent: 9 Copyeditor: Wendy