In the table shown below are the odds ratios and 99% confidence intervals for ea
ID: 3324069 • Letter: I
Question
In the table shown below are the odds ratios and 99% confidence intervals for each risk factor for ischaemic stroke in a large case-control study. Although all of the cases had been admitted to a hospital with a first acute stroke, controls, with no history of stroke, were drawn from the hospital or the community. After the study was done, the authors split the sample by type of control (hospital or community) and compared the cases to each group of controls separately for every risk factor investigated.
1) Explain why the investigators probably found it necessary to conduct separate analyses, depending on the different sources of controls. What was their chief concern?
Odds of Ischaemic Stroke Associated with Risk Factors e of Control Hospital Controls Community Controls (n=1,602) 2.86 (2.33-3.51) 2.09 (1.65-2.64) 2.26 (1.68-3.06) 1.41 (1.12-1.79) (n 1,379) 2.01 (1.64-2.46) 2.53 (1.98-3.24) 1.24 (0.96-1.60) 1.26 (0.98-1.62) Self-Re ertension Current Smoker Diabetes Mellitus Walst to Hip Ratio Middle Tertile vs. Lowest Tertile Walst to Hip Ratio Highest Tertile vs. Lowest Tertile 1.99 (1.56-2.54) 1.46 (1.14-1.87) 1.98 (1.46-2.68) 1.67 (1.25-2.22) 1.10 (0.86-1.40) 0.98 (0.76-1.26) 1.27 (0.98-1.26) 1.53 (1.21-1.93) hosocial Stress Depression Diet Risk Score: Middle Risk to Lowest Risk (Healthiest Diet) Diet Risk Score: Unhealthlest Diet to Healthiest Diet 1.32 (1.03-1.71) 1.38 (1.08-1.78) 0.63 (0.44-0.88) 0.75 (0.57-1.00) 0.78 (0.55-1.11) 0.89 (0.66-1.18) lar Physical Activi Alcohol Intake: 1-30 drinks per month Alcohol Intake: >30 drinks per month or binge drinker Cardiac Cases heart disease) 1.36 (1.00-1.86) 1.59 (1.13-2.23) 3.32 (2.21-5.01) 2.53 (1.74-3.69)Explanation / Answer
1. The investigators found it necessary to conduct separate analyses depending on the different sources of control as two different controls always throw light on the reasons that could have caused the result. The causes of a first acute stroke might only be attributed to the genetic history or other related ailments if only hospital is used a control, but if community is also used as a control then it will throw light on other causes of a first acute stroke as well, hence the investigators found it necessary to conduct separate analyses
2. After looking throught the odds ratio of most risk factors we find there are no major differences as the Confidence intervals nearly almost overlap. This is a cause of concern as people in both the controls have the same risk of having an acute stroke, regardless of the control. Hence everyone needs to take of their health and be aware of the dangers.