Beate Bokhof1, Lewin Eisele2, Raimund Erbel3, Susanne Moebus2. 1. Institute for Medical Informatics, Biometry and Epidemiology, University Hospital Essen, University of Duisburg-Essen, Germany. Electronic address: beatebokhof@web.de. 2. Institute for Medical Informatics, Biometry and Epidemiology, University Hospital Essen, University of Duisburg-Essen, Germany. 3. Clinic of Cardiology, West German Heart Center, University Hospital Essen, University of Duisburg-Essen, Germany.
Abstract
OBJECTIVE: The validity of participants' self-reports via questionnaires or interviews in epidemiological studies remains questionable. We examined the agreement of tumors, reported via different survey instruments, with medical records. METHODS: Within the Heinz Nixdorf Recall Study, comprising 4814 subjects aged 45-75 years, tumors were assessed via different survey tools at baseline and 8-year-follow up (FU): personal interviews (CAPI), self-administered questionnaires (SA-Q), physical examinations, short questionnaire/non-responder questionnaire (S-/N-Q) and telephone interviews. Information on each self-reported tumor was coded via ICD-10, WHO-Version 2010, and evaluated against medical records. RESULTS: During FU, 95% of 1083 self-reported incident tumors in 623 individuals, at baseline, 65% of 473 prevalent tumors in 406 individuals could be evaluated. Agreement of the main assessment tools, CAPI and SA-Q, with medical records was 90.1% and 88.4% (FU) and 91.0% (baseline-CAPI). Best agreement was in tumors of prostate (baseline-CAPI: 97.8%; 5-year-FU-CAPI: 96.9%, SA-Q: 95.7%) and breast (baseline-CAPI: 93.2%; 5-year-FU-CAPI: 100.0%, SA-Q: 98.8%). DISCUSSION: Agreement of CAPI and SA-Q with medical records was good. To assess incident tumors, SA-Q emerged as favorable, as it is least expensive and easy to be applied. Especially for tumors of prostate and breast, cost-intensive and time-consuming validation with medical records may not be necessary.
OBJECTIVE: The validity of participants' self-reports via questionnaires or interviews in epidemiological studies remains questionable. We examined the agreement of tumors, reported via different survey instruments, with medical records. METHODS: Within the Heinz Nixdorf Recall Study, comprising 4814 subjects aged 45-75 years, tumors were assessed via different survey tools at baseline and 8-year-follow up (FU): personal interviews (CAPI), self-administered questionnaires (SA-Q), physical examinations, short questionnaire/non-responder questionnaire (S-/N-Q) and telephone interviews. Information on each self-reported tumor was coded via ICD-10, WHO-Version 2010, and evaluated against medical records. RESULTS: During FU, 95% of 1083 self-reported incident tumors in 623 individuals, at baseline, 65% of 473 prevalent tumors in 406 individuals could be evaluated. Agreement of the main assessment tools, CAPI and SA-Q, with medical records was 90.1% and 88.4% (FU) and 91.0% (baseline-CAPI). Best agreement was in tumors of prostate (baseline-CAPI: 97.8%; 5-year-FU-CAPI: 96.9%, SA-Q: 95.7%) and breast (baseline-CAPI: 93.2%; 5-year-FU-CAPI: 100.0%, SA-Q: 98.8%). DISCUSSION: Agreement of CAPI and SA-Q with medical records was good. To assess incident tumors, SA-Q emerged as favorable, as it is least expensive and easy to be applied. Especially for tumors of prostate and breast, cost-intensive and time-consuming validation with medical records may not be necessary.
Authors: Sarah H Nash; Gretchen Day; Vanessa Y Hiratsuka; Garrett L Zimpelman; Kathryn R Koller Journal: Int J Circumpolar Health Date: 2019-12 Impact factor: 1.228