RTI uses cookies to offer you the best experience online. By clicking “accept” on this website, you opt in and you agree to the use of cookies. If you would like to know more about how RTI uses cookies and how to manage them please view our Privacy Policy here. You can “opt out” or change your mind by visiting: http://optout.aboutads.info/. Click “accept” to agree.
Application of a new method for linking anonymous survey data in a population of soldiers returning from Iraq
Wilson, A. L. G., Hoge, C. W., McGurk, D., Thomas, J. L., Clark, J. C., & Castro, C. A. (2010). Application of a new method for linking anonymous survey data in a population of soldiers returning from Iraq. Annals of Epidemiology, 20(12), 931-938. https://doi.org/10.1016/j.annepidem.2010.08.008
PURPOSE: This study provides a unique application of a procedure to link anonymous survey records as a means to assess mental health problems among soldiers after return from deployment to Iraq. The research purpose is to validate the use of a linkage method on anonymous survey data for longitudinal analysis.
METHODS: This research analyzes surveys completed by soldiers at three (N = 4,903), six (N = 3,110), and 12 (N = 2,544) months post Operation Iraqi Freedom (OIF) deployment. To protect anonymity while also allowing for longitudinal analysis, the survey includes four self-generated codes for matching the surveys from individual soldiers. The surveys were linked using Link Plus (Centers for Disease Control and Prevention, Atlanta, GA). Three matched groups (3-6 months, 6-12 months, and 3-12 months) were created, matching on the anonymous codes and demographic variables. An internal validation of the linkage results was performed and selection bias was assessed.
RESULTS: The matched records showed very high agreement in variables not included in the matching procedure and were similar to the unmatched records with regard to demographics and post-traumatic stress disorder (PTSD) prevalence.
CONCLUSIONS: The use of self-generated codes protects anonymity while the method utilized and validated in this paper allows for longitudinal data analysis. The results support an absence of any notable selection bias. Ann Epidemiol 2010;20:931-938. 2010 (C) Elsevier Inc. All rights reserved.
