Internet-based control recruitment for a case-control study of major risk factors for stroke in Korea: Lessons from the experience

Background: This study aimed to estimate the population-attributable risks (PARs) of 9 major risk factors for stroke in Korea through a case-control study and to test the feasibility and validity of internet-based control recruitment. Methods: From April 2008 to September 2009 controls were enrolled via internet after providing consent for participation through a web-based survey. The cases included patients who were admitted to the participating centers due to acute stroke or transient ischemic attack within 7 days of onset during the study period. Each control was age- and sex-matched with 2 cases. Adjusted odd ratios, age-standardized prevalence, and PARs were estimated for the 9 major risk factors using the prevalence of risk factors in the control group and the age and sex characteristics from Korea's national census data. Results: In total, 1041 controls were matched to 2082 stroke cases. Because of a shortage of elderly controls in the internet-based recruitment, 248 controls were recruited off-line. The PARs were 23.44%, 10.95%, 51.32%, and 6.35% for hypertension, diabetes, smoking, and stroke history, respectively. Hypercholesterolemia, atrial fibrillation, obesity, coronary heart disease, and a family history of stroke were not associated with stroke. Comparison with education and religion of the control group with that mentioned in the national census data showed a notable difference. Conclusions: The study results imply that internet-based control recruitment for a case- control study requires careful selection of risk factors with high self-awareness and effective strategies to facilitate the recruitment of elderly participants.