J1

1. A study sought to determine the association between the use of appetite suppressant products containing phenylpropanolamine [PPA] and hemorrhagic stroke. This study was designed as a case-control study... with the cases being people who had experienced stroke, and the controls being people who had not experienced stroke. The case subjects were identied during hospitalization for their stroke and their history of use of PPA was determined by an in-person interview. The control subjects were recruited by random-digit phone dialing between 9:l]am and 5:l]pm. Subjects answering the phone were told about the study and. after consenting. were asked the same questions about PPA use as the case subjects. {a} Consider the selection of cases and mntrols for this study and comment on whether selection bias may be present. {b} Comment on whether misclassication bias {i.e. erroneous classication of disease status or exposure status] may be present in this study. (c) Comment on whether recall bias may inuence results from this study. {d} Interviewer bias may occur when interviewers collecting the data are not blinded to the disease status of the participant. How might interviewer bias play a role in this study? (e) "Bias by survival\" occurs when, e.g.. subjects die Erom disease {and hence are not avail- able for selection into the study} at unequal rat in the exposure groups. Might this bias be present in this study? {PPA can no longer be marketed in the US.)