Knowledge of medical history information among proxy respondents for deceased study subjects |
| |
Authors: | A Tepper LB Connally P Haltmeier E Smith MH Sweeney |
| |
Affiliation: | Division of Surveillance, Hazard Evaluations, and Field Studies, National Institute for Occupational Safety and Health, Cincinnati, OH 45226. |
| |
Abstract: | ![]() Proxy respondents were interviewed for 96 decedents in an occupational cohort. A second respondent was interviewed for 59 decedents. Medical records were reviewed to validate questionnaire information. The percentage of respondents who answered "don't know" (non-response) to questions about medical condition ranged from 5% (cancer and heart disease) to 17% (ulcers). Non-response rates were lowest among spouses, intermediate among children, parents, and siblings, and highest among other relatives and friends. Among 41-55 pairs, depending on the condition, agreement between paired respondents was excellent (kappa > 0.75) for ulcers, cancer, diabetes, and lung disease. A higher percentage of medical records was obtained for decedents with spouse respondents and for decedents with more recent dates of death. Sixty percent or more of the medical records were obtained for patients with cancer (n = 30), heart disease (n = 26), stroke (n = 9), and liver disease (n = 10). The positive predictive value of the proxy respondent information for these conditions was 93, 81, 78, and 60%, respectively. |
| |
Keywords: | |
本文献已被 PubMed 等数据库收录! |
|