Community-Based Studies May Be Biased
In cardiac study, non-participants had a higher mortality rate, were older and had more co-morbidities
THURSDAY, Aug. 9 (HealthDay News) -- Community-based studies, particularly those with a low participation rate, may be prone to participation bias, researchers report in the August issue of the Mayo Clinic Proceedings. In a myocardial infarction study, non-participants had a higher mortality rate, more co-morbidities, were older and were more likely to be non-white than participants.
Veronique L. Roger, M.D., of the Mayo Clinic in Rochester, Minn., and colleagues conducted a study of 2,277 patients with elevated cardiac troponin T levels, who were invited to participate in a study of myocardial infarction rates. Of these, 1,863 (82 percent) accepted while 414 refused to participate. The records of 375 of these patients were authorized for inclusion in the study.
During the course of the study, 558 patients had a first myocardial infarction, of whom 67 (12 percent) had refused to participate. The members of this sub-group were more likely to have more co-morbid conditions such as peripheral vascular disease, chronic pulmonary disease, heart failure and impaired creatinine clearance. They were also likely to be non-white and older.
"Non-participants experienced increased mortality rates largely because of older age and excess co-morbidities. These differences should be kept in mind when study results are being interpreted, particularly in settings with low participation rates," the authors conclude.