1 design and analysis techniques for case- control studies instructor: 李奕慧...
TRANSCRIPT
Lecture Overview
1. Case-Control Study
2. Example: ”Risk factors associated with lung cancer in Hong Kong”
3. OR for multiple exposure levels
4. Confounding factors
5. Methods of Controlling (adjusting for) confounders
2
Epidemiologic Study Design
Analytical studies
Intervention studies Clinical trials
Observational studiesCohort studiesCase-control studies
3
Case-control study
Study Population
Cases
Controls
Exposed
Non-exposed
Exposed
Non-exposed
4
Selection of cases
Establish a strict diagnostic criteria for the disease: Examples:Type 1 diabetes in children: severe
symptoms, very high BG, marked glycosuria, and ketonuria.
Type 2 diabetes: few if any symptoms, Slightly elevated BG, diagnosis “complicated”.
5
Selection of cases
Population-based cases: include all subjects or a random sample of all subjects with the disease at a single point or during a given period of time in the defined population: Danish childhood diabetes register
Hospital-based cases:
All patients in a hospital department at a given time
6
Selection of ControlsPrinciples of Control Selection: Study base:
Controls can be used to characterise the distribution of exposure
Comparable-accuracy Equal reliability in the information obtained from cases
and controls no systematic misclassification
Overcome confounding Elimination of confounding through control selection
matching or stratified sampling
7
Selection of Controls
General population controls:registries, households, telephone samplingcostly and time consumingrecall biaseventually high non-response rate
Hospitalised controls:Patients at the same hospital as the casesEasy to identifyLess recall biasHigher response rate
8
Ascertainment of Disease and exposure status External sources:
Death certificates, disease registries, Hospital and physicians records etc.
Internal sources: Questionnaires and interviews, information
from a surrogate (spouses or mother of children), biological sampling( e.g. antibody)
9
Bias in Case-Control studies
Selection biasNon-responseDetection bias
cases and controls are identified not independently of the exposure
Observation biasRecall Bias: Cases are more likely to remember
exposure than controls
10
Strengths in Case-control Quick, inexpensive Well-suited to the evaluation of
diseases with long latency period Rare diseases Examine multiple etiologic factors for a
single disease
11
Limitations in Case-control
Case-control study Not rare exposure Incidence rates cannot be estimated
unless the study is population based Selection Bias and recall bias
12
Risk factors associated with lung cancer in Hong Kong
Lung Cancer 40 (2003) 131-14013
14
Chi-Square Tests
Value dfAsymp. Sig.
(2-sided) Exact Sig. (2-sided)
Exact Sig. (1-
sided)
Pearson Chi-Square
0.257a 1 .613
Risk Estimate
Value
95% Confidence Interval
Lower Upper
Odds Ratio for Marital (other / married)
.880 .535 1.446
Multiple Exposure Levels
B1High A1
DNot exposed C
CasesExposurelevel
B2Medium A2
B3Low A3
OR1
OR2
OR3
Reference
Controls OR
15
Multiple Exposure Levels
16Lung Cancer 40 (2003) 131/140Lung cancer.sav
A significant (P<0.05) increasing trend in the OR was found between nonsmokers, ex- and current smokers; and increasing amount of smokingamong the ever smokers.
17
smoking * case Crosstabulationcase
Totalcase controlsmoking nonsmoker Count 52 96 148
% within case 24.5% 45.3% 34.9%exsmoker Count 68 87 155
% within case 32.1% 41.0% 36.6%current smoker Count 92 29 121
% within case 43.4% 13.7% 28.5%Total Count 212 212 424
% within case 100.0% 100.0% 100.0%
Chi-Square Tests
Value dfAsymp. Sig.
(2-sided)Pearson Chi-Square 48.212a 2 .000
Likelihood Ratio 50.088 2 .000
Linear-by-Linear Association
42.734 1 .000
N of Valid Cases 424
a. 0 cells (.0%) have expected count less than 5. The minimum expected count is 60.50.
抽煙與罹患肺癌有關,Case中抽煙者佔較高的比例 (43% vs 13.7%)
18
Data > Select cases >
19
只選 smoking=2, 3的資料進行分析
Risk Estimate
Value
95% Confidence Interval
Lower Upper
Odds Ratio for smoking (exsmoker / nonsmoker)
1.443 .908 2.293
For cohort case = case 1.249 .942 1.656
For cohort case = control .865 .721 1.039
N of Valid Cases 303
Exsmoker 罹患肺癌是 nonsmoker 的 1.4倍 ,95%CI (0.9, 2.3)Exsmoker 與nonsmoker罹癌機率沒有顯著差異
20
Confounding factors (干擾因素)
Confounder:
Variable is associated with both the disease and the exposure variable.
21
Method for control for confounders Study design:
restriction/ matching/ randomization Statistical adjustment:
1. Standardization; e.g. age standardized (where age is a confounder)
2. Stratified by confounder (Mantel-Haenszel test)
3. Incorporate the confounder into a regression analysis as a covariate. (logistic regression approach)
22
Restriction
Example研究主旨:二手煙 (ETS, exposure)與罹患肺癌(disease)的關係confounder: 研究對象本身是否抽煙
為了避免干擾只分析 ETS 對 nonsmoker的影響
Stratified Analysis
23
將性別當作分層 (stratum)的因子
24
smoking * case * sex CrosstabulationCount
sexcase
Totalcase controlmale smoking ex- and current smoker 160 116 276
nonsmoker 52 96 148Total 212 212 424
female smoking ex- and current smoker 13 6 19
nonsmoker 106 113 219Total 119 119 238
Lung cancer2.sav
Sex-Specific OR for smoking
25
Risk Estimate
sex Value
95% Confidence Interval
Lower Uppermale Odds Ratio for smoking (ex- and
current smoker / nonsmoker)2.55 1.68 3.85
N of Valid Cases 424female Odds Ratio for smoking (ex- and
current smoker / nonsmoker)2.31 0.85 6.30
N of Valid Cases 238
Lung cancer2.sav
可以將男士的 OR與女士的 OR合併嗎?怎麼併?
Thank you!
26