School of Public and Environmental Health, Hawassa University Two by two tables Obesity Hypertensive Normal Total
Yes a b a+b
No c d c+d
Total a+c a+d a+b+c+d
Example Water Diarrhoea No diarrhoea supply Total Unsafe 165 547 712 a b a+b Safe 91 550 641 c d c+d Total 256 1097 1353 a+c b+d a+b+c+d • Relative risk (RR): RCT and cohort RR= (a/(a+b)) / (c/(c+d)) • OR: case-control and cross-sectional OR=ad/bc • Both are effect measures (these measure strength of association) • We can also test hypothesis when we calculate 95% CI • As the prevalence of the disease becomes below 5%, OR approaches RR OR versus RR real-time example.xlsx (95% CI for OR) • Change the OR into the natural log form • Consider 1.96 as a reliability factor • Calculate the standard error for lnOR Use this formula to construct 95% CI for OR Back transformation using exp function Interpretation of values • OR/RR of 1 is called the null hypothesis value, i.e no association between the exposure and the outcome • OR/RR >1 implies increased risk of developing the disease if exposed • OR/RR<1 implies the protective effect of the exposure • 1-RR is the protective efficacy • If the 95% CI for OR/RR includes 1, p>0.05, i.e no association • As the value becomes far from one on both sides the strength of association increases; as it approaches 1, it decreases. Interpretation • OR=(165*550)/(547*91)=1.82 • 95% CI for OR= 1.36-2.44, and hence we can reject the null hypothesis as 1 is not included in this interval and we can interpret the value of OR 1.82 as there is 82% increased risk of developing diarrhoea if one consumes unsafe water. Interpret these value • OR (95% CI): 0.56 (0.22-0.89) • OR (95% CI): 1.45 (0.87-3.99) • OR (95% CI): 2.4 (1.6-7.8)