Comparison of two procedures for tracing participants in an epidemiologic cohort study

Lorraine P. Hahn, J. Michael Sprafka, Gregory L. Burke, David R Jacobs Jr

Research output: Contribution to journalArticlepeer-review

7 Scopus citations


The effectiveness and accuracy of using a commercial firm for tracing survey participants were compared with the results of an established, University of Minnesota tracing protocol. Studying a cohort of 622 survey participants ages 25–74 who were living in a single Midwestern community, we were unable to locate 142 (23%) at their original address after four years. We traced them according to the Minnesota protocol, using information obtained from contact persons (friends or relatives) and other sources. Overall, current address information was found for 96% of the participants lost to follow-up, and 113 (81%) of 139 addresses were supplied through the contact person. An independent attempt to obtain similar information using only credit bureau tracing through Associated Credit Services provided addresses for 123 of the participants. Associated Credit Services provided correct, current addresses for 39 (29%) of the 136 surviving participants (six fatal cases were excluded) and supplied correct, current addresses for a majority (58%) of the study participants found by protocol to be living in the study area and eligible for a follow-up clinic examination. We support the use of commercial services as secondary sources of address information in tracing participants after the initial use of contact persons to optimize participant contact. We recommend the use of commercial services as primary sources when contact persons are unavailable.

Original languageEnglish (US)
Pages (from-to)157-160
Number of pages4
Issue number2
StatePublished - Mar 1990


  • Data collection
  • Epidemiologic methods
  • Epidemiology
  • Tracing


Dive into the research topics of 'Comparison of two procedures for tracing participants in an epidemiologic cohort study'. Together they form a unique fingerprint.

Cite this