Epidemiology module of practical skills for mbbs student

Table 4. Age-specific lung cancer death rates per 1 000 person-years at risk, by age group and amount smoked

	Average number of cigarettes smoked daily
	0		1-14		15-24		25+
Age group	Number	Rate	Number	rate	Number	rate	Number	rate
35-44	1	0,05	1	0,07	0	-	1	0,11
45-54	0	-	3	0,31	9	0,62	8	0,75
55-64	0	-	3	0,48	20	2,31	26	3,88
65-74	0	-	9	2,69	17	5,16	14	6,48
75+	2	1,11	6	2,68	8	7,27	8	16,33
Total all ages (Standardized)	3	0,07	22	0,57	54	1,39	57	2,27

The relation between lung cancer mortality and cessation of smoking is shown in Table 5.

Table 5. Lung cancer deaths, rates in smokers, ex-smokers and non-smokers

Smoking habit	Number of lung cancer deaths	Rates per 1 000 person-years
Current smoker	124	1,28
Ex-smoker for:
<5 years	5	0,67
5-9 years	7	0,49
10-19 years	3	0,18
20+ years	2	0.19
Non-smoker	3	0,07

Questions re Cohort study

1. 
   What proportion of male physicians answered the questionnaire? How might this affect comparability of smokers and non-smokers in the study?
   

2. 
   From table 4, compute and interpret values of relative and attributable risk for the “total” group.
   

3. 
   What do the data in table 5 imply for health education services?
   

Practical exercise

Clinical trials

The following are various scenarios illustrating clinical trials. Read the scenarios and answer the questions.

Scenario 1

A new anti-tuberculosis medication has been developed which has shown good in vitro activity and low toxicity in phase 2 (safety) trials. A randomised control clinical trial is now being planned.

Which of the following groups would you consider to be the most appropriate controls and why? What are the disadvantages of the groups you do not select?

		No	Yes
1.	Patients who receive the current standard 2-drug therapy in your clinic
	If no: disadvantages
2.	Patients who receive a placebo drug
	If no: disadvantages
3.	Patients who receive only one of the two drugs in current use
	If no: disadvantages
4.	Two groups, one on the current two drugs and one on placebo
	If no: disadvantages

Scenario 2

In a trial of Haemophilus influenzae type b conjugate vaccine involving 60 000 infants, children born on odd dates during a twelve month period received the vaccine in a four dose schedule plus the routine infant immunisation schedule vaccines. Those born on even dates served as controls and received only the routine vaccine schedule.

What factors would have entered into the calculation of numbers of infants in the trial?

Was the method of allocation to vaccine and control groups satisfactory?

What alternative method might have been adopted?

How would you go about identifying cases of invasive Haemophilus disease?

What sources of bias might enter into case ascertainment?

Practical exercise

Advantages and Disadvantages of Different Study Designs

Complete this table by placing one or more “+” or “ –“ marks or short comments in the boxes to provide a quick comparison of the advantages and disadvantages of these different study designs.

	Experi-mental study	Prospective cohort study	Retro-spective cohort study	Case-control study	Cross-sectional study	Descriptive (non-analytic) study
Can measure disease incidence
Can evaluate more than one outcome from the same risk factor
Can evaluate several different risk factors for the same disease
Follows the natural time sequence of exposure and disease
Useful for rare exposures or risk factors
Useful for rare diseases
Useful for diseases with long latency periods
Logistics? (eg. keeping track of study groups? Quality control in method?)
Number of subjects needed
Cost
Results available quickly
Uses existing data

Practical Exercise

Sampling a population

We propose to do a study of knowledge of tuberculosis in a group who are members of a scientific society. We have obtained a list of the members (see next page) from the register of members of the association and we will obtain a sample of the members to administer a questionnaire.

• 
  List the frequencies of the variables gender, language, region, TB and BACT.
  
• 
  Select the first 8 members in the list.
  
• 
  List the frequencies of the same variables in the table.
  
• 
  Select every 5^th member in the list.
  
• 
  List the frequencies of the same variables in the table.
  

Variable	All members	First 8	Every 5^th
Gender
Men
Women
Don’t know
Language
French
English
Spanish
Don’t know
TB
Yes
No
Don’t know
Bact
Yes
No
Don’t know

1. 
   Are the three groups the same?
   

2. 
   Which of the samples is closest to the whole group?
   

3. 
   Why?
   

4. 
   What conclusion would you draw about the geographic focus of the society from the first eight cases?
   

5. 
   Why is this conclusion misleading?
   

6. 
   What will be the effect of those cases where the value is not known?
   

List of members of a scientific society

NUMBER	GENDER	LANGUAGE	REGION	TB	BACT
1	MAN	SPANISH	AML	NO	NO
2	MAN	ENGLISH	AMN	YES	NO
3	MAN	ENGLISH	MIE	NO	NO
4	MAN	ENGLISH	MIE	NO	NO
5	MAN	ENGLISH		NO	NO
6		FRENCH	AFR	YES	YES
7	MAN	ENGLISH	ASI	YES	YES
8	MAN	ENGLISH	ASI	NO	YES
9		ENGLISH	AFR	NO	NO
10	MAN	ENGLISH	AFR	NO	NO
11	MAN	ENGLISH	AMN	NO	NO
12	MAN	FRENCH	AFR	YES	YES
13	MAN	ENGLISH	AFR	NO	NO
14		ENGLISH	MIE	NO	NO
15	MAN	ENGLISH		NO	NO
16	MAN	ENGLISH	AFR	NO	NO
17	WOMAN	SPANISH	AFR	YES	NO
18	WOMAN	SPANISH	AML	NO	NO
19	MAN	ENGLISH	MIE	YES	YES
20	MAN	SPANISH	AML		NO
21		SPANISH	AML	NO	NO
22	MAN		AFR	YES	NO
23	MAN		AMN	YES	YES
24	WOMAN	ENGLISH	AMN	YES	NO
25	MAN	ENGLISH	AFR	NO	NO
26	MAN	ENGLISH	AFR	NO
27	MAN	ENGLISH	AFR	NO	NO
28		ENGLISH	AMN	YES	NO
29	MAN	ENGLISH	ASI	NO	NO
30		ENGLISH	ASI	YES	YES
31		FRENCH	AFR	NO	NO
32	MAN	ENGLISH	AMN	YES	NO
33	WOMAN	FRENCH	AFR	NO	NO
34	MAN	ENGLISH	AFR	NO	NO
35	WOMAN	ENGLISH	EUR	NO	NO
36	MAN	ENGLISH	AMN	YES	YES
37	MAN	ENGLISH	EUR	NO	NO
38	MAN	ENGLISH	AFR	NO	NO
39	MAN	ENGLISH	ASI	NO	NO
40	MAN	ENGLISH	AFR	YES	NO
41	MAN	FRENCH	AFR	NO	NO
42	MAN	FRENCH	EUR	YES	YES