‧ 文獻評讀概念

機構名稱：高雄市立小港醫院(委託財團法人私立高雄醫學大學經營）

報告單位：藥劑科

報 告 者：張秀美 組長 日 期：101.03.08

主題：實證醫學

2

課程大綱

‧ 文獻評讀概念

– VIP-RAMbo

‧ 文獻數值解析

‧ 文獻評讀工具介紹

– CASP

– CEBM-Oxford (CATmaker) – PRISMA

– STROBE – CONSORT – Jadad scale

‧ 證據等級評比

– GRADE – SIGN

– CEBM-Oxford

3

實證醫學五大步驟

• 提出問題 (Question Formulation)

• 搜尋證據 (Evidence Search)

• 嚴格評讀 (Critical Appraisal)

• 臨床運用 (Evidence Application)

• 結果評估 (Outcome Evaluation)

• 提出問題 (Question Formulation)

• 搜尋證據 (Evidence Search)

• 嚴格評讀 (Critical Appraisal)

• 臨床運用 (Evidence Application)

• 結果評估 (Outcome Evaluation)

4



隨便找一篇文章



花很多時間，但找到品質不是很好的文章



先看結果，再看討論，對於研究背景和方

法卻不詳讀



看完文章對照顧病人並沒有幫助

文獻評讀常犯的錯誤

5

搜尋最佳文獻證據

To find and determine current best evidence

‧ 直接搜尋已評讀文獻

– ACP Journal Club, Cochrane Library

‧ 自行評讀文獻

– appraisal with VIP(RAMbo)

‧ 文獻評讀工具應用

– CASP、CONSORT、CATmaker、Jadad scale…

6

臨床問題研究設計類型

In vitro / Animal research Case Series / Case Reports

Case Control studies Cohort studies

Randomized Controlled Trial

Systematic Review

Meta- analysis

Introduction to Evidence-Based Practice 2010

7

不同類型臨床問題最佳研究設計

Cohort study > Case control study > Case series study 世代研究 > 病例對照研究 > 病例系列研究

Prognosis 預後

Economic analysis 經濟分析

Cost effectiveness 成本效益

Randomised control trial (RCT) 隨機對照試驗

Prevention 預防

Randomised control trial (RCT) 隨機對照試驗

Therapy 治療

Cohort study > Case control study > Case series study 世代研究 > 病例對照研究 > 病例系列研究

Etiology 病因

Prospective, blinded cross-sectional study comparing with gold standard

前瞻性、盲法、與黃金標準進行比較之斷面研究 Diagnostic test

診斷性檢驗或檢查

Study design (研究設計) Question type

(問題類型)

Introduction to Evidence-Based Practice 2010

8

Critical appraisal

EBM

Ask

Assess Apply Appraise Acquire

Practice Importance Validity

objective blinded

Measurement Ascertainment Representative

9

What is the VIP?

Validity (Reliability) 效度/信度 V

Can we believe it ?

(研究方法的探討)

Importance (Impact) 重要性 I

Does it matter?

(研究結果的分析)

Practice (Applicability) 臨床適用性 P

Does it apply to our patients?

(如何在臨床運用)

10

What is the RAMbo?

• Representative

– Random selection/ consecutive/ inception cohort – If comparative, were groups comparable?

(Random allocation/ adjustment)

• Ascertainment/follow-up

– response rate/follow-up/verification>80%

• • Measurement M

– b

o

Validity (Reliability)效度/信度

“Can we believe it ?”

Introduction to Evidence-Based Practice 2010 11

Key validity issues for studies of Therapy:

• Randomization

• Concealed allocation

• Baseline similarity

• Intention-to-treat

• Follow-up complete

• Blinding

12

Importance (Impact)重要性

“ Dose it matter? ”

• What were the results?

– How large was the treatment effect?

– How precise was the estimate of the treatment effect?

– What was the relative risk reduction?

– What was the absolute risk reduction?

– What were the confidence intervals?

Introduction to Evidence-Based Practice 2010

13

Practice (applicability)臨床適用性

Does it apply to our patient?

• Were the study patients similar to my population of interest?

• Were all clinically important outcomes considered?

• Are the likely treatment benefits worth the potential harm and costs?

Introduction to Evidence-Based Practice 2010

14

研究結果估算- ^{常見專有名詞}

• Therapy/Prevention (治療/預防)

– Relative risk reduction (RRR, 相對風險性降低度) – Absolute risk reduction (ARR, 絕對風險性降低度) – Number needed to treat (NNT, 益一需治數)

• Harm/Etiology (傷害/病因)

– Relative risk (RR, 風險比)標準值 “ 1 ” – Risk difference (RD, 風險差) 標準值 “ 0 ”

– Absolute risk increase (ARI, 絕對風險性增加度) – Number needed to harm (NNH, 害一需治數)

• Prognosis (預後)

– Event rate (事件發生率) – Odds ratio (OR, 勝算比)

• Diagnosis (診斷)

– Sensitivity (敏感度) – Specificity (特異度)

– Positive predictive value (陽性預測值) – Negative predictive value (陰性預測值) – Likelihood ratio (相似比)

15

Confidence interval (CI)：信賴區間或信賴範圍

公式為：

• 指在某一信賴程度內，由樣本統計量所求出預期可以包括母 群體的範圍。即用以評估實驗的不確定性的指標。

• 信賴區間通常被假設為95%，這個數值的範圍代表我們對於 母群體的推估具有確信95%為真的信心。

• 舉例而言，若一個CI為95%的NNT為10±5，我們將會有95%

信心確信NNT的數值是在5和15之間。

16

17

18

• Control event rate (CER)：對照組的發生（病）率

即在研究中未施與介入因子（即實驗的療法），產生研究所關心之 結果的頻率。

• Experimental event rate (EER)：實驗組的發生（病）率

即在施與介入因子（實驗療法）後的病患，被觀察到發生研究所關 心之結果的比率。

• RRR (relative risk reduction)：相對風險差

在實驗組及對照組間所產生的不良結果比率所之降低百分比。

算法： RRR=|EER-CER|/CER

舉例：設CI＝95%，以維生素A療法治療嚴重的糖尿病患者，實驗 組及對照組分別各有30人，實驗組發生感染的人數為1人，對照組 發生感染者有9人──

EER=1/30=0.033 CER=9/30=0.3

RRR= |EER-CER|/CER=|0.033-0.3|/0.3=0.267/0.3=0.89

19

• ARR (absolute risk reduction)：絕對風險差（風險差異之絕對值）

實驗組和控制組產生不良結果的比率差異，其算法為：

ARR =|EER - CER|

以上案例而言，則ARR=|EER - CER| =|0.033-0.3| = 0.267

• Number needed to treat (NNT)：需要被治療的數量 絕對風險的減少率的倒數： 1/ARR

即預防產生一個不良結果（或為使一位病人達到實驗所求之有益 結果）所需治療的數量。

例如：在信賴區間為95%下，非胰島素依賴型糖尿病人中發生糖 尿病型維生素A缺乏症的ARR為25%，則其NNT=1/25%=4。

20

• ARI (absolute risk increase)：風險增加之絕對值 實驗組和控制組產生不良或有害結果的差異，

即： |EER-CER|（和ARR公式相同）

例如：在95%信賴區間下，某種治療產生低血糖副作用的情形

，其EER為57%，CER為23%，

ARI=|EER-CER|=|57%-23%|=34%。

(ARI 也被使用於評估『危險因子』對於疾病的影響。)

• NNH (number needed to harm)：需要被傷害的數量

當病患接受了這個實驗性的治療後，就可能會有病人受到副作 用的傷害，這個數字表示--對多少病患進行此實驗療法，與對照 組做比較後，會有一個病人產生不良的結果。

算法： 1/ARI

呈上例：在信賴區間95%下，ARI為34%，則其 NNH =1/ARI=1/34% = 3

21

EER CER

3/101=3.0%

12/101=11.9%

NNT ARR

1/8.9%=12 11.9%-3.0%=8.9%

Example

47/328=14.3% 22

25/294=8.5%

NNH ARI

1/5.8% =17 14.3%-8.5% =5.8%

EER CER

Example

23 http://www.cebm.net/

24

http://www.cebm.net/index.aspx?o=1160

http://www.cebm.net/

25

定量評價療效指標

• 相對效應評價

RR(relative risk) OR(odds ratio)

RRR(relative risk reduction)

• 絕對效應評價

ARR(absolute risk reduction) ARI(absolute risk increase)

• 95% confidence interval

• NNT; NNH

26

比較四種不同治療效果的大小

治療組 療效值 效果大小

A RR 0.4

B RRR 60%

C RD 0.06%

D NNT 1667

對照組P1= 0.1% 實驗組P2= 0.04%

27

用於低危險人群: RRR ARR NNT 治療1000人 預防中風人數

不治療時的發生率(CER) 2.0%

治療可將發生率降至(EER) 1.2%

40% 0.8% 125 8 用於高危險人群:

不治療時的發生率(CER) 20%

治療可將發生率降至(EER) 12%

40% 8% 12.5 80

抗高血壓藥在不同人群種預防中風的效果

• 相對效果在不同人群趨於一致,意味著不同病人的絕對效果一定不同

• RRR無法呈現實際風險降低程度，沒有考慮起始風險

• ARR可以更準確表示治療效果，但亦不容易體會兩組的差別

28

如何進行文獻評讀

--工具介紹--

29

文獻評讀工具介紹

ˇ

PRISMA

ˇ ˇ

ˇ

STROBE

ˇ

Prognosis

ˇ ˇ

Diagnostic test Observation

study

ˇ

Case control

ˇ

Cohort study

ˇ ˇ ˇ ˇ

RCT

Jadad Scale CONSORT

CEBM

ˇ

CASP

ˇ

SR/Meta analysis 評讀工具

CASP: Critical Appraisal Skills Programme CEBM: Centre for Evidence Based Medicine

PRISMA: Preferred reporting items for systematic reviews and meta-analysis STROBE: Strengthening the Reporting of Observational Studies in Epidemiology CONSORT: CONsolidated Standards of Reporting Trials

http://www.sph.nhs.uk/sph-files/casp-appraisal-tools 30

31

32

33

34

Example- appraise the Cohort study

NEJM 2009; 360(3): 225-235.

35

A/ Are the results of the study valid?

Screening Questions

ANSWER:

What is the risk for sudden cardiac death associated with the use of 2 classes (typical and atypical) of antipsychotic agents?

V

36

A/ Are the results of the study valid?

Screening Questions V

ANSWER: Harm / Etiology

37

Is it worth continuing?

YES

A/ Are the results of the study valid?

Screening Questions

38

A/ Are the results of the study valid?

Detailed Questions

39

the defined population

• The cohort was restricted to persons 30-74 yrs of age, enrolled in Tennessee Medicaid for at least 730 days and to have been eligible for full pharmacy benefits and made regular use of medical care.

• The cohort also included two controls for each user of

antipsychotic drugs, matched for age, sex, and first day of follow-up, who were randomly selected from qualifying nonusers of antipsychotic drugs on the first day of follow up.

• Used antipsychotic drugs (44218 used single typical antipsychotic drugs and 46089 used single atypical antipsychotic drugs.

Inclusion Criteria

40

• Patients at high risk for death from

noncardiac causes were excluded from the cohort.

• Follow-up time did not include time during hospitalization and the 30 days after

discharge from the hospital.

• Residence in nursing home, serious illness, or drug dependence.

Exclusion Criteria

the defined population

41

A/ Are the results of the study valid?

Detailed Questions

V

42

A/ Are the results of the study valid?

Detailed Questions

43

Did they use subjective or objective measurements?

• Antipsychotic drugs and other study

medications were identified from Medicaid pharmacy files.

• Computerized pharmacy records are an

excellent source of medication data because they are not subject to information bias

and have high concordance with medication

use as reported by patients.

44

Do the measures truly reflect what you want them to (have they been validated)?

YES

Were all the subjects classified into exposure groups using the same procedure?

Current use

person filled the prescription and the end of the days of supply.

Indeterminate use

up to 90 days after the last current use Former use

any subsequent person-time that was not classified as current or indeterminate use.

Nonuse

person-days with no prescribed use of antipsychotic drugs on those days or at any time in the past.

45

A/ Are the results of the study valid?

Detailed Questions

V

46

A/ Are the results of the study valid?

Detailed Questions

47

database linkage to computerized death certificates

Did they use subjective or objective measurements?

Do the measures truly reflect what you want them to (have they been validated)?

YES

48

Has a reliable system been established for

detecting all the cases (for measuring disease occurrence)?

• The study end point was sudden cardiac death occurring in the community.

• Sudden cardiac death was defined as a sudden

pulseless condition that was fatal, that was

consistent with a ventricular tachyarrhythmia,

and that occurred in the absence of a known non-

cardiac condition as the proximate cause of the

death.

49

Were the subjects and/or the outcome assessor blinded to exposure (does this matter)?

Were the measurement methods similar in the different groups?

YES

Yes Outcome: Death

Blinding is less critical

50

A/ Are the results of the study valid?

Detailed Questions

V

51

A/ Are the results of the study valid?

Detailed Questions

52

• To assess residual confounding by factors

associated with the use of antipsychotic drugs, we performed a secondary analysis, using propensity scores (i.e., the predicted probability that a person would be a user of antipsychotic drugs) to identify a nonuser control group with a similar psychiatric-

illness profile.

• This matched cohort excluded users of antipsychotic drugs who had a diagnosis of schizophrenia or

related psychosis.

53

54

A/ Are the results of the study valid?

Detailed Questions V

V

55

A/ Are the results of the study valid?

Detailed Questions

Nonuser 624591 person-year Former user 189981 person-year Typical user 86735 person-year Atypical user 79589 person-year Indeterminate user 45381 person-year Current multiantipsy 15883 person-year

1042160

56

57

A/ Are the results of the study valid?

Detailed Questions

V

V

58

B/ What are the results?

59

• During the 1,042,160 person-years of cohort follow- up, there were 1870 sudden cardiac deaths, or 17.9 per 10,000 person-years.

• The unadjusted rate increased from 4.7 deaths per 10,000 for persons 30 to 34 years of age at

baseline to 47.6 per 10,000 for those 70 to 74 years of age.

• More than twice as high for men as for women (27.1 vs. 12.9 per 10,000).

Major Outcome- Sudden cardiac deaths

60

1.14 (0.93-1.39)

61

62

63

1.08(0.82-1.43)

64

Conclusion

• Current users of typical antipsychotic drugs

and of atypical antipsychotic drugs in the study cohort had a similar dose-related increased risk of sudden cardiac death.

• This finding suggests that with regard to this

adverse effect, the atypical antipsychotic

drugs are no safer than the older drugs.

65

B/ What are the results?

V

V

66

B/ What are the results?

• The most plausible explanation is that antipsychotic drugs increase the risk of serious ventricular arrhythmias, probably through blockade of potassium channels and prolongation of cardiac repolarization.

• Other mechanisms may be involved, including autonomic effects,

inhibition of other ion channels, and other acute cardiotoxic effects, such as the myocarditis associated with the use of clozapine

V

67

C/ Will the results help me locally?

V

68

C/ Will the results help me locally?

69

Not all antipsychotics display the same level of risk: the newer atypical antipsychotics olanzapine, risperidone, and quetiapine display a much lower level of risk than the older typical antipsychotics, especially those in the phenothiazine group.

Expert opinion

70

Our review suggests that overdose of atypical antipsychotic

medications (AAPM) is unlikely to cause significant cardiovascular toxicity.

71 American Journal of Emergency Medicine (2009) 27, 607–616

72

Atypical and typical antipsychotic drug use were both strongly associated with raised risks, as were SSRIs. Tricyclic

antidepressants were not associated with raised risks.

Case–control study

73

C/ Will the results help me locally?

V

74

Summary

Number : 93300 persons Inclusion criteria :

– 30-74 y (mean age 46 y ; 65% men)

– 46089 p’ts used single atypical antipsychotic drugs – 44218 p’ts used single typical antipsychotic drugs – Enrolled in Medicaid for ≧730 days

Exclusion criteria :

– P’t at high risk for death from noncardiac causes – 30 d after discharge from the hospital

– Residence in nursing home, serious illness, drug dependence

Patients (Question 3)

Cohort study Study design

NEJM 2009; 360: 225-235.

reference

What is the risk for sudden cardiac death associated with the use of 2 classes (typical and atypical) of

antipsychotic agents?

Question

75

Summary

• Each person-day of medicaid enrollment from Jounary 1,1990, through December 31, 2005.

• Means follow up 2.2-2.9 years.

Follow up (Question 7)

• Using propensity scores to identify a nonuser control group with a similar psychiatric-illness profiles

• Excluded users of antipsychotic drugs who had a diagnosis of schizophrenia or related psychosis Adjust for

confounding factors

(Question 6)

End point was sudden cardiac death occurring in the community, database linkage to computerized death certificates

Outcome

measurement (Question 5)

Antipsychotic drugs and other study medications were identified from Medicaid pharmacy files.

Exposure measurement (Question 4)

76

Summary

•Current users of typical and atypical

antipsychotic drugs in the study cohort had a similar dose-related increased risk of sudden cardiac death.

•This finding suggests that the atypical

antipsychotic drugs are no safer than the older drugs.

Conclusion (Question 10) Results

(Question 8) (Question 9)

http://www.cebm.net/ 77

1.CATmaker 可以幫助執行重要 的臨床統計；同時儲存我們的 問題、搜尋策略及評讀結果。

2.CATmaker 也可以製作成檔案 夾， 存放於個人資料庫中。

3.進到CATmaker 首頁先確定所 要評讀的文獻為系統性文獻回 顧或是治療、診斷、預後以及 病因，再來進行評讀。

4.適合初學者的自學應用工具，

其中有一個critical appraisal guides，若有問題可以隨時進 入這個導讀工具，以利學習。

78

79

80

81

82

Representative 族群代表性

– Random selection/ consecutive/ inception cohort – If comparative, were groups comparable?

(Random allocation/ adjustment)

Ascertainment / follow-up 足夠的確認/追蹤 – Intention-to-treat (ITT) analysis

– response rate/follow-up/verification>80%

Measurement 結果量測適當性M

– blinded or objective measures

V I P

83

84

85

• A meeting was held in Ottawa, Canada, in June 2005 with 29 participants, including review authors, methodologists, clinicians, medical editors, and a consumer. The objective of the Ottawa meeting was to revise and expand the QUOROM checklist and flow diagram, as needed.

• An international survey of review authors, consumers, and groups commissioning or using systematic reviews and meta-analyses was completed, including the International Network of Agencies for Health Technology Assessment (INAHTA) and the Guidelines International Network (GIN).

• PRISMA Statement is to help authors report a wide array of systematic reviews to assess the benefits and harms of a health care intervention.

• PRISMA focuses on ways in which authors can ensure the transparent and complete reporting of systematic reviews and meta-analyses.

• The PRISMA Statement consists of a checklist and a flow diagram.

http://prisma-statement.org/

86

87

88

89

90

• STROBE stands for an international, collaborative initiative of epidemiologists, methodologists, statisticians, researchers and journal editors involved in the conduct and dissemination of observational studies, with the common aim of STrengthening the Reporting of OBservational studies in Epidemiology.

• Workshop in Bristol, United Kingdom, in September 2004 and established the STROBE.

• Observational research makes use of many different study designs, STROBE had to be clearly defined early on.

• It decided to focus on the 3 study designs that are used most

widely in analytical observational research: cohort, case – control, and cross-sectional studies.

• The STROBE Statement was developed to assist authors when writing up analytical observational studies, to support editors and reviewers when considering such articles for publication, and to help readers when critically appraising published articles.

http://www.strobe-statement.org/

91

92

http://www.consort-statement.org/ 93

• CONSORT, which stands for Consolidated Standards of Reporting Trials, encompasses various initiatives developed by the CONSORT Group to alleviate the problems arising

from inadequate reporting of randomized controlled trials (RCTs).

• It offers a standard way for authors to prepare reports of trial findings, facilitating their complete and transparent reporting, and aiding their critical appraisal and interpretation.

• The CONSORT Statement comprises a 25-item checklist and a flow diagram, along with some brief descriptive text.

• The checklist items focus on reporting how the trial was designed, analyzed, and interpreted; the flow diagram displays the progress of all participants through the trial.

內科學誌 94

2010：21：408-418

內科學誌95

2010：21：408-418

內科學誌96

2010：21：408-418

97

• 對研究者而言，可以促進隨機對照試驗報告的品質

• 對讀者、期刊編輯及同儕審查者而言，可以當做一 種有效的隨機對照試驗文章評讀工具

• 叢集隨機試驗(cluster randomized trials) 及不劣性試 驗(non-inferiority trials)需進一步作延伸應用

• 超過50% 主流醫學期刊認同CONSORT Statement，

許多期刊亦經由要求作者遵循CONSORT Statement

而改善了研究報導的品質

98

Jadad Scale

Try to answer the following questions:

Controlled Clinical Trials 17:1-12 (1996)

99

Jadad評估表分三大面向，以0分為最低分，

總分為5分，三大面向評估原則如下：

一、 Was the study described as randomized？(研究是否被描述為隨機?)

□ 基本分1 分：研究樣本是隨機分派，但未詳述如何產生隨機的方式。

□ 加 1 分(即得2 分)：有說明隨機分配的產生方式，如以電腦輔助產生的隨 機試驗序列。

□ 減 1 分(即得0 分)：有說明隨機分配的產生方式，但分配方式不恰當。

二、 Was the study described as double-blind？(研究是否被描述為雙盲試驗?)

□ 基本分 1 分：僅提及實驗是採雙盲實驗，但未交待如何進行雙盲實驗。

□ 加 1 分(即得2 分)：具體描述如何進行雙盲實驗的方法。

□ 減 1 分(即得0 分)：描述如何進行雙盲，但方法不當。

三、 Was there a description of withdrawals and drop-outs？(文獻是否有描述病 患追蹤的退出、踢除原因?)

□ 基本分 1 分：清楚說明失去追蹤（退出及失聯）的原因。

□ 減 1 分(即得0 分)：沒有說明失去追蹤的原因。

100

證據等級評比

101

What is GRADE?

• GRADE :Grading of Recommendations Assessment, Development, and Evaluation

• GRADE offers a transparent and structured process for

developing and presenting summaries of evidence, including its quality for systematic reviews and recommendations in health care.

• Although the GRADE system makes judgments about quality of evidence and strength of recommendations in a systematic and transparent manner, it does not eliminate the inevitable need for judgments.

• GRADE is most useful for three groups: authors of systematic reviews, groups conducting HTAs, and guideline developers.

GRADEprofiler About Quality of Evidence

102

The GRADE system

Any estimate of effect is very uncertain.

Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.

Further research is likely to have an

important impact on our confidence in the estimate of effect and may change the estimate.

Further research is very unlikely to

change our confidence in the estimate of effect.

DEFINITION

Case report

Very Low

♁^

Observational studies without special strengths

Low

♁♁^

Limitations or special

strengths of both randomized trials and observational

studies

Moderate

♁♁♁^

Randomized trials without important limitations

High

♁♁♁♁

Example GRADE

classifies the quality of evidence in one of four grades:

GRADEprofiler About Quality of Evidence Grading of Recommendations Assessment, Development and Evaluation

103

GRADE-Quality assessment criteria

GRADEprofiler About Quality of Evidence

104

Non-analytic studies, eg case reports, case series 3

D

Expert opinion 4

D

Case control or cohort studies with a high risk of confounding or bias and a significant risk that the relationship is not causal 2 -

Well conducted case control or cohort studies with a low risk of confounding or bias and a moderate probability that the relationship is causal

2+

C

High quality systematic reviews of case control or cohort

studies High quality case control or cohort studies with a very low risk of confounding or bias and a high probability that the relationship is causal

2++

B

Meta-analyses, systematic reviews, or RCTs with a high risk of bias

1 -

Well conducted meta-analyses, systematic reviews, or RCTs with a low risk of bias

1+

A

High quality meta-analyses, systematic reviews of RCTs, or RCTs with a very low risk of bias

1++

A

Therapy/Prevention,Aetiology/Harm Level of

Evidence Grade of

Recommendation

The grade of recommendation relates to the strength of the evidence on which the recommendation is based. It does not reflect the clinical importance of the recommendation.

http://www.sign.ac.uk/pdf/sign50.pdf

105

SIGN GRADES OF RECOMMENDATION

A At least one meta-analysis, systematic review, or RCT rated as 1++, and directly applicable to the target population; or A body of evidence

consisting principally of studies rated as 1+, directly applicable to the target population, and demonstrating overall consistency of results

B A body of evidence including studies rated as 2++, directly applicable to the target population, and demonstrating overall consistency of results; or Extrapolated evidence from studies rated as 1++ or 1+

C A body of evidence including studies rated as 2+, directly applicable to the target population and demonstrating overall consistency of results; or

Extrapolated evidence from studies rated as 2++

D Evidence level 3 or 4; or Extrapolated evidence from studies rated as 2+

http://www.sign.ac.uk/pdf/sign50.pdf

106

• How strong is the evidence that a treatment

works (or is harmful, or that a diagnostic test is accurate, etc.)?

• Should I use the treatment (or diagnostic test, etc.)?

The levels of evidence and grades of recommendation were developed in order to help people answer the

following questions in a systematic way:

107

108

NEW

109

3000多種常用治療的效果

實證文獻臨床應用

110

老年癡呆者的福音

• Pfizer和Eisai公司推出老年癡呆新藥

• 乙醯膽鹼脂酶抑制劑: 多奈蒎齊

• 效果: 可以延遲入院2年(廣告詞)

• 價格: 每年1200美元 (約10,000元人民幣)

• 臨床療效: 可以改善認知能力score1.4-3.9個單位

癡呆病人認知能力約為70個單位

正常人的認知能力score應為0-1個單位

實證文獻臨床應用

111

治療骨質疏鬆的重大突破

• 骨質疏鬆: 骨密度低於均值1-2.5個標準差

• 藥物: Alendronate

• 價格: 每年一萬美元

• 臨床療效: RCT顯示可降低髖骨骨折56%

68歲婦女，骨密度低於均值2.5個標準差者給予Alendronate 用藥4年；結果顯示—

Alendronate組骨折率0.2%，對照組為0.5%，即降低56%

的骨折發生率。

81位婦女連續服藥4年共花費近320萬美元，絕對有效治療 值NNT為333，即4年花費1300萬美元才能預防一例骨折。

10年後兩組骨折率沒有區別

實證文獻臨床應用

112

以病人為中心的實證選擇

‧ 不治療的危險是什麼? 有多大?

‧ 治療能改變什麼結果?

‧ 改變這個結果的機會有多大?

‧ 該治療有什麼副作用? 發生機率有多大?

‧ 病人覺得治療費用是否划算?

‧ 病人是否有更需要花錢的地方?

113

EBM…..

思考與行為模式

的改變

114

謝謝鈴聽 敬請指教

116

The Pyramid of Evidence

Studies: Medline / PubMed Synthesize: Cochrane Library

Synopsis: ACP Journal Club Summaries:

Clinical Evidence

System

117

Intention-to-treat (ITT)

A method of analysis for randomized trials in which all patients randomly assigned to one of the treatments are analyzed together, regardless of whether or not

they completed or received that treatment. This is

complex area, and there are many definitions of what

constitutes ITT.

118

119

120

121

診斷工具的特性

• Sensitivity (敏感度)

• Specificity (特異性)

• Positive predictive value (陽性預測值)

• Negative predictive value (陰性預測值)

• Likelihood ratio (相似比)

– Positive likelihood ratio (陽性相似比)

– Negative likelihood ratio (陰性相似比)

診斷文獻的評讀-李智雄醫師 高雄醫學大學 122

123 診斷文獻的評讀-李智雄醫師 高雄醫學大學

124

Test Characteristics

高度排除疾病

<1/10 LR(-)

中度排除疾病 1/5~1/10

LR(-)

低度排除疾病 1/2~1/5

LR(-)

不具臨床意義 1/2~2

低度排入疾病 2~5

LR(+)

中度排入疾病 5~10

LR(+)

高度排入疾病

>10 LR(+)

臨床意義 Likelihood ratio

"檢驗結果為陽性“

"驗有病且真正有病的人的百 分比"和"驗有病但真正沒病的 人的百分比" 的比例

"檢驗結果為陰性", "驗沒病 但真正有病的人的百分比"和

"驗沒病且真正沒病的人的百 分比"的比例

125

Test Characteristics

在做過該項檢查之後，醫師認為該病患確實罹患該疾病或確實 不罹患該疾病的機率

診斷文獻的評讀-李智雄醫師 高雄醫學大學

126

The Likelihood Nomogram

診斷試驗後的 疾病可能性

61.54%

診斷試驗前的 疾病可能性

(prevalence)

25%