K - 1 Terapi EBM

"Evidence-Based Medicine"
Therapy
Rozaimah Zain-Hamid
Community Research Program ('KBK')
FacuIty of Medicine
Universitas Sumatera Utara
Evidence Evidence- -Based Medicine` Based Medicine`
Critical Appraisal` Critical Appraisal`
Research Methodology Research Methodology
Zain-Hamid, R; CRP-`KBK`, Faculty of Medicine,
Universitas Sumatera Utara.
Health Services Health Services
"Lv|dence8ased Med|c|ne"
(apers |n Iourna| of Med|c|ne ] nea|th Sc|ences)
'VaIid' MethodoIogy of study
'Important' ResuIt of study
'AppIicabiIity' iscussion
'VIA'
Zain-Hamid, R; CRP-'KBK', FacuIty of Medicine,
@reatment ] @herapeut|c |ntervent|on
EBM : VaIid, important and appIicabIe
to particuIar patients
High rank of hierarchy of evidence
(Systematic reviews /meta-anaIyses;
Randomized ControIIed cIinicaI triaI / RCT)
@reatment ] @herapeut|c |ntervent|on
AnswerabIe CIinicaI Question (ACQ)
P : Patient, ProbIem, PopuIation
I : Intervention
C : Comparison
O : Outcome
Hypertensive patient - shouId I start
ACE inhibitors?
Patient : middIe aged man with
diastoIic 100 mm Hg
Intervention : ACE inhibitors
Comparison : iuretics
Outcome : prevent heart disease;
stroke; end-organ damage?
"L8M" for therapy
1. Reports of individuaI studies
2. Reports of systematic reviews ('RCT')
5. Reports of quaIitative study
3. Reports of CIinicaI ecision AnaIyses
(CA)
4. Reports of economic anaIyses
Integrative Iiterature
Review articIe :
* unsystematic
Systematic review :
* in gathering, evaIuating, presenting evidence
* no formaI statisticaI method
Meta-anaIysis :
* systematic review + formaI statisticaI anaIysis
Integrative Iiterature
ev|ew art|c|e
Systemat|c rev|ew
Metaana|ys|s
eports of |nd|v|dua| stud|es
High rank of hierarchy of evidence
(Randomized ControIIed cIinicaI triaI / RCT)
Key eIements of RCT
Randomization
ControI BIinding
Reports of individuaI studies
1. Are the resuIts of individuaI studies vaIid?
2. Are the vaIid resuIts of individuaI studies,
important?
3. Are the vaIid, important resuIts
of individuaI studies;
appIicabIe to our patient?
#Ia||d|ty' of |nd|v|dua| stud|es
1. Are the resuIts of individuaI studies,
vaIid?
Primary guides:
1. 1. Was the assignment of patients
to treatments randomized?
1. 2. Was follow-up of patients sufficiently long
and complete?
1. 3. Were patients analyzed in the groups
to which they were randomized?
- revent|ng b|as
- Lqua| chance
- 8a|ance of sub[ects character|st|c
@oast ] co|n
S|mp|e random|zat|on (random tab|e)
8|ock random|zat|on
Strat|f|ed random|zat|on
11 Was the ass|gnment of pat|ents
to treatments random|zed?
2a|nnam|d C'k8k' Iacu|ty of Med|c|ne
Un|vers|tas Sumatera Utara
1 2 Was fo||owup of pat|ents suff|c|ent|y |ong
and comp|ete?
ost to foIIow up no more than 20%
ournals like Evidence-Based Medicine and
ACP ournal Club
won't publish trials with > 80% follow-up.
1 3 Were pat|ents ana|yzed |n the groups
to wh|ch they were random|zed?
lnLenLlon Lo LreaL analysls"
All paLlenLs are analyzed ln Lhe groups
Lo whlch Lhey were lnlLlally asslgned
A sLraLegy for analyzlng daLa
ln whlch all parLlclpanLs are lncluded
ln Lhe group Lo whlch Lhey were asslgned wheLher or
noL Lhey compleLed
Lhe lnLervenLlon glven Lo Lhe group
1 Are the resu|ts of the study va||d?
econdary guldes
1 7 Was Lhe randomlzaLlon concealed?
1 3 1 3 Aslde from Lhe experlmenLal lnLervenLlon
were Lhe groups LreaLed equally?
1 4 1 4 Were paLlenLs healLh workers
and sLudy personnel bllnd Lo LreaLmenL?
1 6 1 6 Were Lhe groups slmllar aL Lhe sLarL
of Lhe Lrlal?
1 4 at|ents hea|th workers and study
personne| b||nd to treatment?
9revenL blas
lngle bllnd
uouble bllnd
1rlple bllnd
S|m||ar
]Iorm
]Co|or
]@aste
]Drug of adm|n|strat|on
#8||nd' |ntervent|on
SU8ILC@S
LIILC@
LIILC@
CINICA TRIA ESIGNS
@LA@MLN@ GU
CN@ GU
1AAL DLSIGN @W GUS
"Washout
er|od"
SU8ILC@S
2 CSSIL DLSIGN
@LA@MLN@
A
@LA@MLN@
8
LIILC@
LIILC@
@LA@MLN@
8
@LA@MLN@
A
LIILC@
LIILC@
#Important' of |nd|v|dua| stud|es
2 1 What |s magn|tude
of the treatment effect?
2 2 now prec|se |s the est|mate
2 Are the va||d resu|ts of |nd|v|dua| study
|mportant?
2 L|ements of |mportant
of the treatment effect
p va|ue I A AI NN@ NNn
2 1 Magn|tude of the treatment effect
2 2 now prec|se |s the treatment effect
Conf|dence Interva| (CI)
Study of stat|n to prevent stroke
(S years' fo||owup)
]Stroke occurred among S7 of pat|ents
random|zed to contro| group
(#Contro| Lvent ate CL')
]Stroke occurred among 43 of pat|ents
random|zed to exper|menta| group
(#Lxper|menta| Lvent ate LL')
1 2 1 L|ements of magn|tude
] (CL LL) ] CL
]CL (study for prevent|ng stroke) S7
2 1 L|ements of magn|tude
] (CL LL) ] CL
(S7 43) ] S7 2S
Stat|n therapy decreased the r|sk of stroke

by 2S re|at|ve to those who rece|ve p|acebo
]LL (study for prevent|ng stroke) 43
2. 1. EIements of magnitude
The situation in which the experimentaI
treatment increase the risk of a good event
as the 'ReIative Benefit Increase (RBI)' or
the risk of bad event as
'ReIative Risk Increase(RRI)
aIso can use the same formuIa (RRR):
RRR = RBI = RRI = {(CER - EER) / CER}

RRR = RBI = RRI =(5.7% - 4.3%) / 5.7% = 25%
] ReIative Risk Reduction (RRR) :
is the percent reduction in risk in treated group
compared to the controI group
] The RRR is measure of how the treatment studied
has reduced the frequency of an adverse event
] AbsoIute Risk reduction (ARR):
is the difference in risk between the controI group
and the treatment group
The situation in which the experimentaI
treatment increase the risk of a good event
as the 'AbsoIute Benefit Increase (ABI)' or
the risk of bad event as
'AbsoIute Risk Increase(ARI)
aIso can use the same formuIa (ARR):
ARR = ABI = ARI = (CER - EER)

ARR = ABI = ARI =(5.7% - 4.3%) = 1.4%
lgnlflcance of 8elaLlve 8lsk 8educLlon
Negat|ve ( 38) treatment may do harm pat|ents
g|ven the new treatment m|ght be
38 more ||ke|y to d|e than the contro| pat|ents
of 0 no treatment effect or benef|t
os|t|ve (S0) pat|ents rece|v|ng the new
treatment m|ght have |ess than 1]2 r|sk of dy|ng
compared to not treated
@he greater the re|at|ve r|sk reduct|on the
more effect|ve the therapy (
eff|cacy of therapy)
2.1. EIement of magnitude
RRR = {(CER - EER) / CER}
CER : ControI Event Rate (without treatment/pIacebo)
EER : ExperimentaI Event Rate (with treatment)
]Number Needed to @reat (NN@)
* Number of pat|ents shou|d be treated
to avo|d 1 (one) bad outcome
* Number of pat|ents shou|d be treated
to have add|t|ona| good outcome
2 1 L|ements of magn|tude
]Number Needed to narm (NNn)
Number needed to harm 1 (one) pat|ent
from the therapy
]Number Needed to Treat (NNT) = 1 / ARR
* NNT = 1 / 1.4% = 72
we need to treat 72 peopIe with a statin

(rather than pIacebo) for 5 years
to prevent one additionaI person
from suffering a stroke
Therapy / treatment - case study
A randomized doubIe bIind controI cIinicaI triaI
on 3 month - 5 year oId chiIdren with miId to
moderate croup (Iaryngotracheobronchitis).
The experimentaI group : 2 mg (4 mI) nebuIized
budesonide.
The controI group : 4 mI nebuIized normaI saIine.
Event being prevented : hospitaI admission due
to upper-airway obstruction.
27
N S4
27
1
26
7
20
@he study protoco|

HospitaIized
HospitaIized
Non-HospitaIized
Non-HospitaIized
budesonide
normaI saIine
No Yes
Budesonide (E) 26 1 27
NaCI (C) 20 7 27
Upper-airway obstruction
X
2
df =1 p = 0.04
Important
No Yes
Budesonide (E) 26 1 27
NaCI (C) 20 7 27
CER = 7 / 27 = 0.26 ;
Important
EER = 1 / 27 = 0.04
RRR = (CER - EER) / CER
RRR = (0.26 - 0.04) / 0.26 = 85%
ARR = (CER - EER) = (0.26 - 0.04) = 0.22
NNT = 1 / ARR = 1 / 0.22 = 5
Budesonide vs normaI saIine;
CER EER ARR NNT
In the actuaI triaI 26% 4% 22% 5
In the hypotheticaI
triviaI case
0.00026 0.00004 0.00022 5000
Important
Conf|dence |nterva|
2 2 L|ements for dec|d|ng prec|s|on
( CI prec|ss|on of the treatment effect)
'AppIicabiIity' of individuaI studies
3 1 Is our pat|ent so d|fferent from those Is our pat|ent so d|fferent from those
|n the study that |ts resu|ts cannot app|y? |n the study that |ts resu|ts cannot app|y?
3 2 Is the treatment feas|b|e |n our sett|ng? Is the treatment feas|b|e |n our sett|ng?
3 Are the va||d |mportant resu|ts of th|s |nd|v|dua|
stud|es app||cab|e to our pat|ent?
3 3 What are our pat|ent's potent|a| benef|ts What are our pat|ent's potent|a| benef|ts
and harms from the therapy? and harms from the therapy?
3 4 What are our pat|ent's va|ues and expectat|ons What are our pat|ent's va|ues and expectat|ons
for both the outcome we are try|ng for both the outcome we are try|ng
to prevent and the treatment to prevent and the treatment
we are offer|ng? we are offer|ng?
ConcIusions ConcIusions
AppIication of good therapy must be AppIication of good therapy must be
supported by EBM supported by EBM
AbiIity to appraise the resuIts of many AbiIity to appraise the resuIts of many
kind of studies, reviews, anaIyses etc kind of studies, reviews, anaIyses etc
Arigatoo gozaimasu
Therapy Worksheet
THERAPY WORKSHEET
Are the results of this single preventive
or therapeutic trial valid?
Citation:
Was the assignment of patients
to treatments randomized?
Was the randomization Iist conceaIed?
Was foIIow-up of patients sufficientIy
Iong and compIete?
Were aII patients anaIyzed in the groups
to which they were randomized?
THERAPY WORKSHEET
Are the results of this single preventive
or therapeutic trial valid?
Citation:
Were patients, cIinicians,
and study personneI kept "bIind"
to treatment?
Were the groups treated equaIIy,
apart from the experimentaI treatment?
Were the groups simiIar at the start
of the triaI apart from
the experimentaI therapy?
THERAPY WORKSHEET
Are the vaIid resuIts of this randomized triaI
important?
What is the magnitude
How precise is the estimate
SAMPE CACUATIONS
Occurrence of diabetic
neuropathy at 5 years
among insuIin-dependent
diabetics in the CCT triaI
ReIative risk
reduction
(RRR)
AbsoIute risk
reduction
(ARR)
Number
needed
to treat
(NNT)
UsuaI
insuIin
Regimen
controI
event rate
(CER)
9.6 %
Intensive
insuIin
regimen
experimentaI
event
rate
(EER)
2.8 %
CER ~ EER
CER
9.6% ~ 2.8%
9.6%
CER ~ EER
9.6% ~ 2.8%
1/ARR
1/6.8%
=15 patients
71 %
6.8 %
95% CI a 4.4% to 9.2% 11 - 23
a 95% confidence intervaI (CI) on an NNT
=1/(Iimits on the CI of its ARR)
YOUR CACUATIONS
ReIative risk
reduction
(RRR)
AbsoIute risk
reduction
(ARR)
Number
needed
to treat
(NNT)
EER
CER ~ EER
CER
CER ~ EER
1/ARR
95% CI a
CER
THERAPY WORKSHEET
Can you appIy this vaIid, important evidence
about therapy in caring for your patient?
o these resuIts appIy to our patient?
Is our patient so different
from those in the study that
its resuIts cannot appIy?
Is the treatment feasibIe in our setting?
What are our patient's potentiaI benefits
and harms from the therapy?
Risk of the outcome in our patient,
reIative to patients in the triaI.
Expressed as a decimaI:______
NNT/ f =______/______=______
(NNT for patients Iike ours)
Our patient's expected event rate
if they received
the controI treatment (PEER) =______
1/(PEER RRR) = 1/________=______
(NNT for patients Iike ours)
094/: f
094/:
1/(PEER RRR)
THERAPY WORKSHEET
Are our patient's vaIues and preferences
satisfied by the regimen and its consequences?
o we and our patient have
a cIear assessment
of their vaIues and preferences?
Are they met by this regimen and
its consequences?
AdditionaI notes:

K - 1 Terapi EBM

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

K - 1 Terapi EBM

Uploaded by

Copyright:

Available Formats

"Evidence-Based Medicine"

Stat|n therapy decreased the r|sk of stroke

aIso can use the same formuIa (RRR):

RRR = RBI = RRI = {(CER - EER) / CER}

aIso can use the same formuIa (ARR):

ARR = ABI = ARI = (CER - EER)

we need to treat 72 peopIe with a statin

@he study protoco|

You might also like