Transcript 投影片 1 - 衛生福利部胸腔病院

實證醫學文獻
查證與嚴格評讀
高雄榮民總醫院
洪碧蓮藥師
藥劑部
課程大綱

提出問題 (Question Formulation)

搜尋證據 (Evidence Search)

嚴格評讀 (Critical Appraisal)

恰當運用 (Evidence Application)

結果評估 (Outcome Evaluation)
99/06/09
2
Clinical Question

無法使用Rifampicin的結核病患者，Rifabutin是
否可以取Rifampicin，作為6個月短程治療的替代
藥物？
99/06/09
3
結核病患者可以Rifabutin取代Rifampicin作為6個
月短程治療的替代藥物？
Keywords
P 結核病患者 Tuberculosis, pulmonary
I Rifabutin
Rifabutin
C Rifampicin Rifampicin
O 治療成功
99/06/09
Sputum culture, CXR,
Clinical symptoms
4
搜尋證據
Evidence Search
檢索策略
System
Summaries
Synopses
Syntheses
Studies
99/06/09
6
檢索策略
是
是否有符合臨床問題的
綜合分析(Meta-analysis)
嚴格評讀
找出結論
提供建議
是
嚴格評讀
找出結論
提供建議
99/06/09
否
隨機分配
研究(RCT)
否
非隨機分配
研究(RCT)
7
Search Database
PubMed
 Cochrane Library

99/06/09
8
Cochrane Library_
Advanced Search
P and I and C and O
99/06/09
9
PubMed檢索程序

個人化登入My NCBI

選取合適的MeSH當作檢索字

結合MeSH與所知的natural language

利用AND和OR

設定Alert

應用書目管理軟體EndNote
99/06/09
10
99/06/09
11
99/06/09
12
99/06/09
13
99/06/09
14
99/06/09
15
Queries 1
N
Queries
#4 #1 AND #2 AND #3
Result
163
#3 "rifampin"[MeSH Terms] OR "rifampin"[All Fields] OR
"rifampicin"[All Fields]
19439
#2 "rifabutin"[MeSH Terms] OR "rifabutin"[All Fields]
1264
#1 "tuberculosis"[MeSH Terms] OR "tuberculosis"[All Fields]
99/06/09
182794
16
P and I and C and O
20091008
99/06/09
C
O
P
I
17
18
篩選有研究品質的藥物資訊
C
O
163
SR
P
I
RCT
in process
從交集PICO的語法而得的 163篇中,分別與搜尋SR/RCT文獻及in process的語法作交集,
20091008
18
以搜尋得高品質研究方法的文獻篇數
99/06/09
99/06/09
19
Queries 2_Clinical Queries
- Search by Clinical Study Category
Search Queries
Result
#5
(#4) AND Therapy/Narrow[filter] 11
RCT
99/06/09
20
Queries 3_Clinical Queries
- Find Systematic Reviews
Search Queries
Result
#6
8
(#4) AND systematic[sb]
SR
99/06/09
21
Queries 4 _in
process
Search
Queries
Result
#7
(#4) AND (In process[sb] OR
publisher[sb] OR
pubmednotmedline[sb])
7
in process
99/06/09
(檢索日期：2010/06/08)
22
Queries 5_Combination
Search
Queries
Result
#8
#5 OR #6 OR #7
24
163篇
24篇
99/06/09
23
Studies Extraction
PubMed (24)
Cochrane(1)
References (25)
Meta-Analysis 2010 (1)
References 2010 (4)
Excluded from
Title & Abstract (3)
Meta-Analysis 2010 (1)
RCT (0)
99/06/09
24
Critical Appraisal
嚴格評讀

Are the results of the article valid? (V)

What are the result? (I)

Will the results help me in caring for patients? (P)
99/06/09
26
Grade of
Level of Therapy/Prevention,Aetiology/Harm
Recommendation Evidence
A
1a
SR (with homogeneity) of RCT
1b
Individual RCT (with narrow CI)
1c
All or none
2a
SR (with homogeneity) of cohort study
2b
Individual cohort study (included low
quality RCT; e.g., <80% follow-up)
2c
“Outcomes”research
3a
SR (with homogeneity) of case-control
3b
Individual case-control study
C
4
Case-series (and poor quality cohort
and case-control studies)
D
5
Expert opinion
B
99/06/09
27
Validity: 研究方法的討論

Randomized

Concealed assignment

Sufficiently long & complete follow-up (> 80%)

Double blinded

Groups similar & treated equally except for the therapy
99/06/09
28
Important: 結論的分析

The magnitude of the treatment effect

NNT: number needed to treat


NNT= 1/ARR = 1 / l CER –EER l, ARR: absolute risk reduction
NNH: number needed to harm

NNH= 1/ARI = 1 / l EER –CER l, ARI: absolute risk increase

RRR: relative risk reduction = (CER-EER)/CER

RRI: relative risk increase = (EER-CER)/CER
99/06/09
29
Calculation
Treatment
Event
Exposed
Positive
a
Negative
b
Not exposed
c
d
 Experimental Event Risk (EER) = a/(a+b)
Control Event Risk (CER) = c/(c+d)
Relative Risk (RR) = EER/CER = (a/(a+b))/ (c/(c+d))
 Experimental Event Odd (EEO) = a/b
Control Event Odd (CEO) = c/d
Relative Odds = Odds Ratio = EEO/CEO = (a/b)/ (c/d)
Odds: a ratio of events to non-events (勝算比)
99/06/09
30
網址: http://www.cebm.net
99/06/09
31
SYSTEMATIC REVIEW
Are the results of the review valid ?
Screening Questions (Yes/No)
What question (PICO) did the systematic
review address?
Is it unlikely that important, relevant
studies were missed?
Were the criteria used to select articles
for inclusion appropriate?
Were the included studies sufficiently
valid for the type of question asked?
Were the results similar from study to
study? (Cochrane Q ?, Q/df ?, I2 ?)
How are the results presented?
(Forest plot?)
Critical Appraisal Skills Programme
網址: http://www.phru.nhs.uk/pages/phd/CASP.htm
99/06/09
33
Critical Appraisal Tools
99/06/09
34
Critical Appraisal Skills
Programme (CASP)
Systemic Review
A. Is the study valid (效度)?
1. Did the study ask a clearly-focused
question?

Consider if the question is “focused” in
terms of :
 the
population studied
 the intervention given or exposure
 the outcomes considered
99/06/09
36
P
I
C
O
99/06/09
37
2. Did the review include the right type
of study?
 Consider:
 Address
the review’s question
 Have an appropriate study design
Is it worth continuing?
99/06/09
38
3.Did the reviewers try to identify all
relevant studies?
(沒有遺漏重要的文獻)?

Consider:
 which
bibliographic databases were used
 if there was follow-up from reference lists
 if there was personal contact with experts
 if the reviewers searched for unpublished studies
 if the reviewers searched for non-English-language
studies
99/06/09
39
搜尋資料來源
最後檢索時間
99/06/09
40
4. Did the reviewers assess the quality
of the included studies?

Consider:
 if
a clear, pre-determined strategy was
used to determine which studies were
included. Look for:
- a scoring system
- more than one assessor
99/06/09
41
5. If the results of the studies have been
combined, was it reasonable to do so?

Consider whether:
99/06/09
 the
 the
results of each study are clearly displayed
results were similar from study to study
(look for tests of heterogeneity)
 the reasons for any variations in results are
discussed
42
6.How are the results presented and
what is the main result?

Consider:
 how
the results are expressed (e.g. odds
ratio, relative risk, etc.)
 how large this size of result is and how
meaningful it is
 How you would sum up the bottom-line
result of the review in one sentence
99/06/09
43
7. How precise are these results?

Consider:
 if
a confidence interval were reported.
Would your decision about whether or
not to use this intervention be the same
at the upper confidence limit as at the
lower confidence limit?
 if a p-value is reported where
confidence interval are unavailable
99/06/09
44
8.Can the results be applied to the local
population?

Consider whether:
 the
population sample covered by the
review could be different from your
population in ways that would produce
different results
 your local setting differs much from
that of the review
 you can provide the same intervention in
you setting
99/06/09
45
9. Were all important outcomes
considered?

Consider outcomes from the point of
view of the:
Individual
 policy makers and professionals
 Family/carers
 Wider community

99/06/09
46
10. Should policy or practice change as a
result of the evidence contained in
this review?

Consider:
 whether
any benefit reported outweighs
any harm and/or cost. If this
information is not reported can it be
filled in from elsewhere?
99/06/09
47

Consider outcomes from the point of view
of the:
 individual
 policy
maker and professionals
 family/carers
 wider community

Consider whether:
 any
benefit reported outweighs any harm
and/or cost. If this information is not reported
can it be filled from elsewhere?
 policy or practice should change as a result of
the evidence contained in this trial
99/06/09
48