You are on page 1of 16

Risk estimates

Dr: Ibrahim Kabbash


Prof of Public Health & Community Medicine
Faculty of Medicine - Tanta University
Risk estimates
• Used to quantify the risk of occurrence of
outcome among persons with certain
exposure
• Two type:
– Odds ratio
– Relative risk
• 95% confidence interval (95%CI) is used to
determine significance
Risk estimates:
• Odds ratio (OR)
– Calculated from case control studies

– Considered indirect risk estimate

• Relative risk (risk ratio) (RR):


– Calculated from cohort studies

– Considered direct risk estimate (more powerful)


Risk estimates
• A value <1 mean no risk or may be protective!!
• A value of one mean equal outcome among
exposed and non exposed
• A value >1 mean occurrence of risk
• 95% confidence interval indicates whether the
risk is significant or not (having evidence
whether present or not present in the general
population
Risk estimate

axd
Odds ratio = ---------- Disease
Expo-
cxb
sure
+ve -ve Total

+ve a b a+b
a/a+b
Relative risk= ----------
c/c+d -ve c d c+d

Total a+c b+d


Estimating risk

Case control studies Cohort studies


Odds ratio = 5.674 Relative risk = 0.285
Answer this question
• A groups of patients of 350 suffering form
obesity were compared with a control group
of 250 non obese. Among patients 75 were
found to have regular exercises, while among
control those practicing regular exercises were
found to be 150. Estimate the risk of obesity in
relation to practice of regular exercises.
Answer
• First identify the type of the study
• Second identify the suitable risk estimate to
be calculated
• Write the equation used for calculation
• Draw 2X2 table
• Calculate the risk
• Give your interpretation of the results.
Answer
• The study is a case control study
• The odds ratio is the risk estimate to be
calculated
Exercise Obese Non obese Total
Regular 75 150 225
No exercise 275 100 375
Total 350 250 600

• OR =75X100/150X275 = 0.182
Answer this question
• A group of 400 persons was followed up for 10
years. In this group 200 were smokers and the
other 200 were non smokers. By end of the
study, 50 smoker persons developed lung
cancer. The second group of 200 non smoker
persons, 15 only developed lung cancer. What
is the risk of developing lung cancer
associated with tobacco smoking?
Answer
• The study is a cohort study
• The risk ratio is the risk estimate to be calculated

Exercise Lung cancer Non cancer Total


Smokers 50 150 200
Non smokers 15 185 200
Total 65 335 400

• RR =incidence among exposed/incidence among non exposed


• = 50/200 / 15/200 = 0.25/0.075 = 3.333
95% CI

• 95% CI = 0.983-2.584

• 95% CI = 1.234-2.321

• 95% CI = 0.234-0.821
• The 95% CI for risk of lung cancer among
a group of population was 2.345-4.562.
which of the following is the risk estimate
of this group?
a) 1.359
b) 2.324
c) 3.459
d) 4.789
• There was a significant risk of breast
cancer in relation to prolonged used of
contraceptive pills (RR= 2.897). Which of
the following should be the 95%
confidence interval of risk of cancer?
a. 0.324-0.897
b. 0.982-3.456
c. 1.298-3.895
d. 2.999-3.897

You might also like