Artikel
Agreement between self-reported use and physician reports of hormone replacement therapy
Suche in Medline nach
Autoren
Veröffentlicht: | 14. September 2004 |
---|
Gliederung
Text
A large population-based case-control study with the goal to investigate the association between lifetime hormone replacement therapy (HRT) use and breast cancer risk is presently being conducted in two regions in Germany. In order to assess the accuracy with which women report current use of HRT and specific aspects of use, a validation study was performed among early study participants comparing self-reported use of HRT with data from physician reports. Cases were women aged 50-74 years with primary breast cancer. Controls were frequency matched by year of birth. A random sample of 224 cases and 225 controls were selected to provide equal strata by study region and age group (dichotomized at 60 years). Aided by photographic depictions of HRT medications prescribed over the past 30 years, participants gave detailed reports about their HRT use to trained interviewers. Data obtained from the interviews were compared with individual prescription data provided by gynaecologists via telephone interview. The results showed good to excellent agreement between study participants and physicians on ever/never use of HRT (kappa=0.78(0.71-0.84)), never/ex/current status (proportion of agreement=83%), duration of use (intraclass correlation coefficient ri= 0.82(0.77-0.85)), age at first intake (ri=0.88(0.85-0.91)), as well as for age at last intake (ri=0.98). Agreement for ever/never use was not differential by case-control status, study region or age. Agreement was slightly better for younger women. This study suggests that personal interviews in case-control studies represent a good tool for assessing lifetime HRT use with reasonably high accuracy.