JOURNAL

PRESENTATION

PRESENTED BY
ATHENA SAJI
ATUL VINCENT
CONTENTS

• Introduction

• Objective

• Methodology

• Result

• Discussion

• Conclusion
Title
Objective of Study

• To evaluate the risks imposed by tobacco smoking on the whole and

by various tobacco smoking methods individually in the development
of Bronchogenic carcinoma

(Time not specified)

Introduction
• Association of tobacco smoking and lung cancer is well documented in
the west but there is relative paucity of such data in India

• Smoking habits of Indian population is different from that of West

• Rising trend of Lung Cancer during 1997

• Increasing prevalence of Bidi Smoking in India

These nessecitated the need to study relationship between tobacco

smoking and lung cancer in this part of the country.
 Methodology
• STUDY DESIGN- Observational,Analytical, Case-Control Study

• STUDY PERIOD- About July 1997-December 1997

• STUDY POPULATION- Patients who attended the OPD of department of

tuberculosis and Chest diseases and other departments of Gandhi
Memorial and Associated hospitals attached to the KG’s Medical
college, Lucknow and thier healthy bystanders.
Methodology(Contd.)

• CASES- 52 newly diagnosed and histopathologically proven patients of

bronchogenic carcinoma attending the department of TB and Chest
diseases and other departments of Gandhi Memorial and Associated
hospitals attached to the college

• CONTROLS- 156 healthy individuals who were

bystanders(friends,relatives) of the lung cancer patients, matched for
age,sex,socioeconomic status
Methodology(Contd.)
Inclusion Criteria:

• Newly diagnosed and histopathologically proven patients of

bronchogenic carcinoma are included in cases

• Healthy bystanders are included in controls

• Age,sex, socioeconomic status were matched

Methodology(Contd.)

Exclusion criteria

• Exclusion criteria is not mentioned in the study

No data regarding exclusion of Individuals with -

• incomplete data

• pre existing lung diseases

• Individuals with h/o exposure to known lung carcinogens

Exposure status

• Non smokers is defined as:

Persons who had never consistently smoked following items a day for 1
year
• 1 cigarette,
• 1 bidi or
• 24 g tobacco in chillam or hookah
Sample size

• The sample size of this study is not calculated.

• Sample is selected by universal sampling method

• Consecutive 52 newly diagnosed histopathologically proven lung

cancer patients are taken as cases

• For each case 3 controls were selected

Data collection

• The researchers used a questionnaire consisting of smoking frequency

and the Modified Kuppuswamy socioeconomic status scale for data
collection.

• Every individual was personally interviewed and the questionnaire

form were filled by the same interviewer throughout the study
period.
Data Variables
• In our study all variables are categorised variables and presented as
proportion
• To analyse association between lung cancer and smoking they have
done odds ratio
• Statistical signifance was measured by chi2
RESULTS
 Table 1(Charecteristics of patients with bronchogenic carcinoma
and controls)

Charecteristics Cases Controls

Total number 52 156
Males 37(71.2%) 111(71.2%)
Females 15(28.8%) 45(28.8%)
Male-Female ratio 5:2 5:2
Mean(S.D) Age(years)
All 54.96(9.27) 54.89(9.00)
Males 55.24(9.23) 55.16(9.01)
Females 54.27(9.67) 54.22(9.03)
Table 1(Contd.)
Charecteristics Cases Controls
Age (Years) distribuition
31-40 2(3.9%) 5(3.2%)
41-50 16(30.8%) 50(32.1%)
51-60 23(44.2%) 62(39.7%)
61-70 9(17.3%) 29(18.6%)
71-80 2(3.9%) 10(6.4%)
Socio economic status
I 0 0
II 7(13.5%) 21(13.5%)
III 17(32.7%) 51(32.7%)
IV 28(53.8%) 84(53.8%)
V 0 0
Table 1(Contd.)
Charecteristics Cases Controls
Religion
Hindus 42(80.8%) 120(76.9%)
Muslims 10(19.2%) 36(23.1%)
Residence
Urban 26(50.0%) 58(37.2%)
Rural 26(50.0%) 98(62.8%)

• P value is not given, hence significance of the data cannot be

verified
Table 2 (Distribution of smoking habits in
lung cancer patients and control)
Smoking CASES CONTROLS
habit
Male Female Total Male Female Total
Non 4 9 13(25.0) 58 43 101(64.7)
smokers
Smokers 33 6 39(75.0) 53 2 55(35.3)
Bidi 20 6 26(66.7%) 38 2 40(72.7%)
Mixed 6 0 6(15.4%) 2 0 2(3.6%)
Cigarette 5 0 5(12.8%) 10 0 10(18.2%)
Hookah 2 0 2(5.1%) 3 0 3(5.5%)
Table 3 -Odds ratio of Lung Cancer with
smoking(both sexes)
Smoking Status Cases Controls Odds ratio(95% CI)
Non smokers 13(25.0%) 101(64.7%)
Smokers(overall) 39(75.0%) 55(35.3%) 5.5
(2.56-12.02)
p<0.001
Bidi smokers 26(50.0%) 40(25.6%) 5.05
(2.21-11.7)
p<0.001
Table 3(Contd.)

• Overall smokers experiensed 5.5 fold increased risk of lung cancer

than non smokers( X2 - 23.29, p <0.001)

• Bidi smokers also had 5 fold greater risk of lung cancer than non
smokers( X2 - 17.68, p<0.001)
Table 4 (Odds ratio of Lung cancer with
number of Bidis smoked per day)
Average no. of bidi Cases Controls Odds ratio(95% CI)
smoked per day
Non Smokers 13(25.0%) 101(64.7%)
Bidi Smokers
0-10 5(19.2) 21(52.5) 1.85
(0.94-3.63)
p>0.05
11-20 10(38.5) 11(27.5) 7.06
(5.41-11.24)
p<0.001
>20 11(42.3) 8(20.0) 10.68
(6.82-15.36)
p<0.001
Table 4 (Contd.)

• Average no.of bidis per day is calculated by summing up smoking

indices and dividing the whole by the duration of smoking in days

• N = average no. of sticks smoked per day during lifetime

• n = average no. of sticks smoked per day during duration dx

• D = Total duration of smoking in days

Table 5 (Odd ratio of Lung Cancer wth
duration of Bidi smoking)
Years of smoking Cases Controls Odds Ratio(95% CI)
Non-smoker 13 (25.0) 101(64.7)
Bidi Smokers
1-39 15(57.7) 34(85.0) 3.43
(1.62-.6.84)
p<0.01
>40 11(42.2) 6(15.0) 14.24
(8.34-24.31)
P<0.001

Increasing odds ratio seen in those who had smoked bidis equal to or more
Table 6 (Distribution of histopathological
types of Lung Cancer)
Histological Type Number
Squamous cell 24(46.2)
carcinoma
Adenocarcinoma 11(21.2)
Small cell 6(11.5)
carcinoma
Large cell 2(3.9)
carcinoma
Undifferentiated 8(15.3)
Mixed( small cell 1(1.9)
&squamous)
Total 52
Table 6(Contd.)

• 31 cases were in Kreyberg Group I (squamous cell & small cell

carcinoma)

• 13 cases were in Kreyberg Group II (Adenocarcinoma & Large cell

carcinoma)
Table 7 (Overall smoking status according to
histological type of lung cancer)
Histological Type Smokers Non Smokers
Squamous cell carcinoma 20(51.3) 4(30.8)
Adenocarcinoma 6(15.4) 5(38.8)
Small cell carcinoma 6(15.4) 0
Large cell carcinoma 1(2.6) 1(7.7)
Undifferentiated 5(12.7) 3(23.0)
Mixed( small cell 1(2.6) 0
&squamous)
Total 39 13
Table 8 (Status of bidi smoking according to
histological type of Lung Cancer)
Histological Type Number of bidi smokers
Squamous cell carcinoma 13(50.0)
Adenocarcinoma 5(19.2)
Small cell carcinoma 2(7.7)
Large cell carcinoma 0
Undifferentiated 5(19.2)
Mixed( small cell &squamous) 1(3.9)
Total 26
Table 8(Contd.)

• Squamous cell type was seen commonly in smokers

• Adenocarcinoma was seen commonly in non smokers

Table 9 (Distribution of non smokers and different
types of smokers in Kreyberg’s Groups I and II )
Smoking status Kreyberg’s Group I Kreyberg’s Group II Group I : Group II
Non smokers 4 6 0.7 : 1
Overall smoker 27 7 4:1
Bidi smoker 19 2 9.5:1
DISCUSSION
• Case control study design provides odds ratio rather than relative risk
when the outcome is rare

• Odds ratio is a close approximation of relative risk when the outcome is

rare.
CASES CONTROLS
EXPOSED a b
NON c d
EXPOSED
• Odds ratio • Relative risk
= =
= =
=
If the no. of cases are rare

= =
Discussion(Contd.)

• Our study has shown that risk of developing lung cancer in overall
tobacco smokers was 5.5 times high compared to non smokers

• Numerous studies in different part of world( ref- 1,2,3,4,5 ) has shown

3.3-12 times higher risk of lung cancer in smokers than in non smokers

• Indian studies (ref-6,7) had shown that Indian smokers has 2.45-16
times higher risk of developing lung cancer than non smokers
Discussion(Contd.)

• A dose response relationship was seen both with the number of bid
smoked per day and duration of bidi smoking

• This was also observed by other Indian studies (ref-6,7)

Discussion(Contd.)

Hills criteria is used to define Causality

• It includes 9 criteria

1. Strength of association - Higher Odds Ratio more strength

2. Dose response relationship

3. Biological Plausibility -rational and theoretical basis of finding

4. Temporality - Cause must precede effect

Discussion(Contd.)

5.Specificity- Effect has only one cause

6.Consistency- Multiple study showing same finding

7.Coherence- Causal association compatible with generally known facts

8.Experiment- Experimental evidence to make casual inference

9.Analogy- Are Similar agents known to cause similar disease in similar

circumstances
Discussion(Contd.)

• In our study Kreyberg group I was more common among overall

smokers and Group I and II was equal among non smokers

• Several other studies in differnent contries (ref-8,9,10,11)shows

significant relationship between Kreyberg’s Group I and smoking.
Limitations

• Limited number of cigarette smokers compared with bidi smokers,

Hence odds ratio of cigarette smoking could not be assessed
 Critical Analysis-Strength

• Case-Control Design: Suitable for studying rare outcomes like lung

cancer.

• Matching: Important confounding factors such as age, sex, and

socioeconomic status was matched.

• Clear Diagnosis: Cases were histopathologically confirmed (gold

standard method)
Strength(Contd.)
• Questionaire forms were filled by same Interviewer throughout the
study period

• Dose-Response Analysis: Provide valuable insights into the association

• Consideration of Histological Types: Identify potential differences in

smoking habits among various lung cancer types.
 Limitations
• Potential recall bias due to participants' ability to accurately
remember smoking habits.

• Selection bias: As controls were chosen from healthy bystanders of

lung cancer patients.

• Selection bias as consecutive 52 newly diagnosed patients were taken

as cases

• Urban and Rural demarcation was not clearly defined

Limitations(Contd.)
• Cannot generalise the study to entire Indian population as the study was
specific to a particular region

• Not all confounding factors are inclued in selection of cases and controls
- passive smoking,comorbidities, exposure to enviornmental pollutants,

• Temporal relationship: The study cannot determine if smoking caused

lung cancer or if lung cancer influenced smoking behaviour.

• Time frame is not specifically mentioned

Critical Appraisal Skills Programme(CASP)
SECTION A
1. Did the study address a clearly focused issue? - Yes
2. Did the authors use an appropriate method to answer their question? -
Yes
3. Where the cases recruited in an acceptable way? - Can’t tell
Comment- Selection bias in selecting cases, Population not defined
correctly, Time frame not specifically mentioned, Sample size calculation
not mentioned.
4.Where the controls selected in an acceptable way? - Can’t tell
Comment- Couldn’t eliminate all confounding factors
CASP(Contd.)
5. Was the exposure accurately measured to minimise bias?-Yes
Comment- Recall bias maybe present
6.(A) Aside from the experimental Intervention , were the groups
treated equally?
List- Enviornmental factors,Occupational factors, Comorbidities are not
not taken into consideration while selecting study population.
6.(B) Have the authors taken account of the potential confounding
factors in the design and/or in their analysis? - Can’t tell
CASP(Contd.)
SECTION B
7. How large was the treatment effect? - A strong association between
tobacco smoking and lung cancer was found in the study.
8. How precise was the estimate of the treatment effect ?
Size of CI is 95% and p value was significant in most cases(<0.05)
9. Do you believe the result ? - Yes
SECTION C
10. Can the result be applied to the local population? - Yes
11.Do the results of the study fit with other avaliable evidence? - Yes
 Reference
1. Wynder EL,Graham EA. Tobacco smoking as a possible aetiologic factor
in bronchogenic carcinoma. JAMA 1950;143: 336-38

2.Doll R, Hill AB . A study of the aetiology of carcinoma of the lung. Br Med

J 1952; 2:127-86

3.Wynder EL, Mabuchi K,Beattle EJ. Epidemiology of lung cancer.

JAMa:1970; 213:2221-28

4. Gao YT, Bld WJ. Lung cancer and smoking in Shanghal.Int J Epidemiol
1991;20:26-31
 Reference(Contd.)
5. Lubin JH,LI JY, XuanXZ. Risk of lung cancer amomg cigarette and pipe
smoker s in Southern China. Int J Cancer 1992;51:390-95

6. Notani P,Sanghvl LD. A retrospective study of lung cancer in Bombay.

Br J Cancer 1974:29:477-82

7. JussawalaDJ, Jain DK. Lung cancer in greater Bombay: Correlation with

religion and smoking habits. Br J Cancer 1979:40:437-48
Reference(Contd.)

8. Wakalal-k,Ohno YGenka-k,Olgmine K,Kawamura T, Tamakoshi A, Aoki

R,Kojima M,Lin YAoki k,Fukuma S.Smoking habits ,local brand ciggarattes
and lung cancer risk in Okinawajapan. J Epidemiol 1997:7:99-105

9. Pemg DW, Pemg RP,Kud BI, Chiang SC.The variation of cell type
distribution in lung cancer: a study of 10,910 cases at a medical center in
Taiwan between 1970 and 1993.Japan J Clin Oncol 1996: 26:229-33
Reference(Contd.)

10.Cigarette smoking and Health. America Thoracic Society. Am J Respir

Crit Care Med1996:153:861-5

11. Wynder EL, Muscat JE. The changing epidemiology of smoking

andlung cancerhistology . Environ Health Perspect.1995:103(suppl 8):
143-8
THANK YOU