Type of hormone therapy and risk of misclassification at mammography screening

Sisse H Njor, Jesper Hallas, Walter Schwartz, Elsebeth Lynge, Anette Tønnes Pedersen

9 Citationer (Scopus)

Abstract

Objective: Current users of hormone therapy (HT) are known to have a lower accuracy of mammography screening than do never users. We studied whether the risk of misclassification depends on type of hormone, administration, regimen, and dose of the therapy. Methods: We linked data from mammography screening registers with drug prescription registers from Fyn, Denmark to identify current and never HT users among screening participants. We compared false-positive risks and interval cancer proportions between current users of different HT preparations, taking women's age, breast density, screen number, and age of comparison mammogram into account. Results: Estrogen therapy users had a significantly higher false-positive risk when the administration was by injection instead of oral (relative risk [RR], 2.37; 95% CI, 1.37-4.09). Women using sequential estrogen plus progestogen therapy had a significantly higher false-positive risk (RR, 1.94; 95% CI, 1.16-3.26) and a nonsignificantly higher interval cancer proportion (RR, 4.29; 95% CI, 0.69-26.53) when the administration of both hormones was transdermal instead of oral. Using tibolone instead of comparable hormones gave a nonsignificantly lower false-positive risk and a nonsignificantly higher interval cancer proportion. Conclusions: Our data showed increased risks of misclassification at mammography screening among women using estrogen injections or transdermal, sequential estrogen plus progestogen. Tibolone seems to offer no advantage regarding accuracy of screening mammography.

OriginalsprogEngelsk
TidsskriftMenopause
Vol/bind18
Udgave nummer2
Sider (fra-til)171-177
Antal sider7
ISSN1072-3714
DOI
StatusUdgivet - 1 feb. 2011

Fingeraftryk

Dyk ned i forskningsemnerne om 'Type of hormone therapy and risk of misclassification at mammography screening'. Sammen danner de et unikt fingeraftryk.

Citationsformater