Question

A. You conduct a study on family history of diabetes and pulmonary hypertension. Since you are concerned about the presence of a third variable - personal history of diabetes, you run a stratified analysis. From a previous analysis, you know that the crude odds ratio is 3.4. What are the stratum-specific ORs & 95% CIs when stratified on personal history of diabetes? Be sure to jot down the stratum-specific ORs and the confidence intervals for each stratum.  

95% Cori dence Interval Tabe Odds Ratio for family history cf cbesitwexpcsure lyes no) For cohort Pulmonary Hypertension (outcome) E yes For cohort Pulmonary Hypertension (outcome) 1.869 619 5.641 1.16 276 1.408 Odds Ratio for family history cf cbesitwexpcsure lyes no) For cohort Pulmonary Hypertension (outcome) E yes For cohort Pulmonary Hypertension (outcome) 1.701 5443 1.374 2.655 Total Odds Ratio for family history cf cbesitwexpcsure lyes no) For cohort Pulmonary Hypertension (outcome) E yes For cohort Pulmonary Hypertension (outcome) 1.759 7.00B 3.511 1.654

B. Without hypothesizing and solely based on the results of your analysis in Question 15 above, what do you think personal history of diabetes could be:

a. an effect measure modifier

b. a confounding

c. a mediator

Answer 1

A. stratum-specific ORs & 95% CIs when stratified on personal history of diabetes

OBESE	valid cases	percentage[valid case/149x100]
Yes [Having Pulmonary Hypertension]	81/149	54.36%
No [Having Pulmonary Hypertension]	68/149	45.63%

B. Without hypothesizing and solely based on the results of your analysis in Question 15 above, what do you think personal history of diabetes could be:

A Confounding [as it has primary factor/ relationship with the potential factor]