Beruflich Dokumente
Kultur Dokumente
Rujukan
Douglas Badenoch and Carl Heneghan,
Evidence based tool kit, BMJ Books,2002
Materi pelatihan EBM Workshop, Centre
for clinical epidemiology & EBM,
RSCM/FKUI, 2013.
Apa EBM?
Integrasi bukti penelitian
terbaik dengan
keahlian klinis dan
nilai2 (patient value)
Bukti penelitian
Keahlian klinis
EBM
Nilai2 pasien
2.
3.
4.
5.
6.
PICO
Evaluasi
Efektifitas
Efisiensi
searching
Kritisi jurnal
Integrasi
hasil dari jurnal,
kahlian dan
nilai pasien
I=
C=Compariso Outcome:
Intervention
n : versus,
Effect, result
or indicator or compare
index test or
exposur:
Does, would
In healthy
children
Exposed in
utero to
cocaine
Compare to
children not
exposed
Have
increased
incidence of
learning
disabilities at
6 years
I=
C=Compariso Outcome:
Interventionor n : versus,
Effect, result
indicator or
compare
index test or
exposur:
Does, would
In an
otherwise
healthy 7
year old boy
with sore
throat
I=
C=Compariso Outcome:
Interventionor n : versus,
Effect, result
indicator or
compare
index test or
exposur:
Does, would
In Child with
frequent
febrile
sizures
Would
Compare to
anticonvulsan no treatment
t therapy
Result in
seizure
reduction?
I=
C=Compariso Outcome:
Interventionor n : versus,
Effect, result
indicator or
compare
index test or
exposur:
Does, would
In patient
Does the
Better than To predict
with sepsis level of
C reactive in hospital
procalcitoni protein
mortality?
n
Alamat internet
Iuran tahunan
Google scholar
http://scholar.google.com.br/
gratis
Medline-pubmed
www.ncbi.nml.nih.gov/PUBMED
Terbesar di dunia
gratis
emedicine
www.emedicine.com
gratis
medscape
www.medscape.com
gratis
Medical matrix
www.medmatrix.org/index.asp
gratis
Clinical practice
guideline
www.guideline.gov
gratis
www.cma.caa/cpgs
gratis
Alamat internet
Bandolier: www.jr2.ox.ac.uk/Bandolier
TRIP: www.tripdatabase.com
Best evidence:
Secondary
jurnals
The cochrane
libary
Iuran tahunan
Langkah searching
1.
2.
3.
Judul tulisan
Nama Penulis
Nama/no/edisi jurnal
Abstract
Latar belakang
Tujuan
Method
Hasil
Diskusi
Kesimpulan
Kritisi Jurnal
Therapi
1. Is the study valid? 2. Are the result important?
1.1 Was there clearly define research question? PICO
1.2 Were the group randomised? (RCT= Randomized control
trial)
1.3 Were all patient accounted all? Drop out rate <80%,
length of study:enough outcome to be manifest, intention
to treat (all patient randomized to be analysed even if it
is not complete the treatment they actually got)
1.4 Were the research participant blinded? Double blind
randomization ( clinican and patient do not know)
1.5 Equal treatment: no co intervention to one group
1.6 Did randomization produce comparable group at the start
of the trial? Uji homogenitas.
Kritisi Jurnal
Terapi
2. Are the result important:
2.1 p value: < 0.01= there is a less than 1 in 100
probability of the result occuring by chance, p
value < 0.05= less then 1 in 20 probability by
chance.
2.2 quantifying the risk and benefit and harm: RRR
(relative risk reduction), ARR (Absolut Risk
Reduction), NNT (Number Needed to treat), CI
(Confidence Interval), RR (Relative Risk)
Kritisi Jurnal
terapi
2. 2. quantifying the risk and
benefit and harm:
CER=control event rate=a/
(a+c)
EER=Experiment event
rate=b/(b+d)
RRR=Relative risk
reduction= (CEREER)/CER
ARR=Absolut Rate
Reduction=CER-EER
NNT=Number needed to
treat=1/ARR
control experi
ment
event
No
event
Kritisi Jurnal
terapi
ARR are more
clinically relevant
measure to use than
RRR, because RRR
1
factor out the baseline
risk, so that small
difference in risk can
seem significant when 2
compare to small
baseline risk
CE EE AR RR
R
R
R
R
0.36
0.34
0.02
0.056
0.036
0.034
0.002
0.056
Kritisi Jurnal
Terapi
NNT= Number Needed
to Treat=the most
useful measure of
benefit= the absolut
number of patients
who need to be
treated to prevent one
bad outcome=1/ARR
Mortality in patient
surviving MCI for at
least 3 days with left
ventricular ejection
fraction <40%
RRR=
Relative
Risk
Reductio
n
ARR=
Absolut
e Risk
Reducti
on
NTT=Nu
mber
Needed
to treat
Palcebo
control
event
rate
=CER
Catropil
experime
nt control
rate=
EER
(CEEREER)/CE
R
CEREER
1/ARR
275/1116
=0.2464
228/115=
0.2045
(0.24640.2045)/0
.2464=0.
17
0.24640.2054=0
.0419
1/0.0419
=24
Kritisi Jurnal
terapi
CI= are used to represent sampling error.
95% CI= there is a 95% chance that the population
is true value lies between the two limit.
95% CI on the ARR=
/1.96x{(CERx(1-CER)/#control patients } +
{(EERx(1-EER)/# experiment patients } If CI
crosses the line of no difference (the point at
which benefit becomes a harm), then we
conclude that the result are not statistically
significant.
Kritisi Jurnal
terapi
RR=Relative Risk Ratio=EER/CER
RR<1=less risk of the event in the
experimental group
Kritisi Jurnal
Diagnosis
1. Is the study valid? 2. Are the result important?
1.1 Was there is clearly defined question? PICO
1.2 Was the presence or absence of the target
disorder confirmed with a validated test (gold
or reference standard)
1.3 was the test evaluated or an apropiate
spectrum of patient
1.4 Was the reference standard applied to all
patients
1.5 Is it clear the test was carried out?
1.6 Is the test result reproduceable?
Kritisi Jurnal
diagnosis
2. Are the result importance
2.1 sensitivity=
a/(a+c)
2.2 spesicity=
d/(b+d)
2.3 Positive predictive value=a/(a+b)
2.4 Negative predictive value=d/(c+d)
2.5 prevalence=pre test probability=(a+c)/
(a+b+c+d)
2.6 Likelihood ratio for a positive test
result= sens/(1-spec)
2.7 likelihood ratio for a negative test
result=(1-sens)/spec
2.8 pre test odds= prevalence/(1prevalence)
2.9 post test odds=pre test oddsx
likelihood pos
2.10 post test probability=post test odss/
(post test odds+1)
disease
present
Disease
absent
Diag a
test 26
+
b
69
a+b
95
Diag c
test - 46
d
249
c+d
295
total
b+d
318
A+b+c+d
a+c
72
390
Kritisi jurnal
prognosis
2. Are the result of the study importance?
2.1 How likely are the outcomes overtime?
1-5 years survival
Median time ( e.g. the length of follow up by
which 50% of patient have died)
Event curve, survival curve
2.2 How percise is this prognostic estimate?
95% CI