Research Article
von Euler-Chelpin M, Wu W, Vej
Abstract
Background: We found previously that women with a false-positive result in mammography screening used significantly more anxiolytic- and antidepressant drugs in the two years prior to their screen than did women who later had a negative screen. When the European Agency for Evaluation of Medicinal Products in 2003 recommended restricted use of hormone therapy suggested alternatives for treatment of menopausal symptoms included antidepressants and other drugs for neurological disorders. Our hypothesis was that antidepressants use to some extent had replaced hormone use, and that this change in particular had affected women at risk of later becoming false-positive in mammography screening. Methods and findings: We undertook a difference-indifferences study including women with a false-positive screen and matched controls of women with a negative screen from the Copenhagen screening program 1997-2006, and their prescribed drugs in 1995-2008. For each group of women defined by calendar period (1997-2003 or 2004-2006), screening result (falsepositive /negative), and timing (2 years before/2 years after the screening event) we calculated the proportion of drug users. Risk ratios (RRs) were used for comparison of two proportions, and ratios of risk ratios (RRRs) for comparison of four proportions. For all women, the use and doses of hormone therapy decreased from 1997-2003 to 2004-2006. While the proportion of women using antidepressant drugs remained fairly stable, the doses increased. No statistically significant difference was found in use of anxiolytic- and antidepressant drugs from before to after the screening event between women with false-positive and negative screens neither in 1997-2003 (RRR 1.04, 95% confidence interval (CI) 0.92-1.17) nor in 2004-2006 (RRR 0.97, 95% CI 0.78-1.21). However, before the screening event in 1997-2003 women with a later false-positive screen had a statistically significantly higher use of these drugs than women with a later negative screen (RR 1.13, 95% CI 1.05-1.21). In 2004-2006 this difference was not statistically significant (RR 1.07, 95% CI 0.95-1.21), but the difference in doses reached RR 1.27 (95% CI 0.93-1.63). Conclusion: We found no impact of a false-positive screen on use of anxiolytic- and antidepressant drugs either before or after the EMEA recommendations of restrictive use of hormone therapy.