Cohort studiesDownloads

A researcher is studying whether a new knee implant is better than existing alternatives in terms of pain after knee replacement. She designs the study so that it includes all the surgeries performed at a certain hospital. Interestingly, she notices that patients who underwent surgeries on Mondays and Thursdays reported much better pain outcomes on a survey compared with those who underwent the same surgeries from the same surgeons on Tuesdays and Fridays. Upon performing further analysis, she discovers that one of the staff members who works on Mondays and Thursdays is aware of the study and tells all the patients about how wonderful the new implant is. Which of the following forms of bias does this most likely represent?

You have been entrusted with the task of finding the causes of low birth weight in infants born in the health jurisdiction for which you are responsible. In 2017, there were 1,500 live births and, upon further inspection of the birth certificates, 108 of these children had a low birth weight (i.e. lower than 2,500 g), while 237 had mothers who smoked continuously during pregnancy. Further calculations have shown that the risk of low birth weight in smokers was 14% and in non-smokers, it was 7%, while the relative risk of low birth weight linked to cigarette smoking during pregnancy was 2%. In other words, women who smoked during pregnancy were twice as likely as those who did not smoke to deliver a low-weight infant. Using this data, you are also asked to calculate how much of the excess risk for low birth weight, in percentage terms, can be attributed to smoking. What is the attributable risk percentage for smoking leading to low birth weight?