Mirena coil

The short version: Mirena is a local progestin device with real, well-documented benefits for heavy bleeding and endometrial protection, but the breast cancer signal is real and deserves weight.

The evidence here is mostly clinical and mostly solid, but it is old. The newest record is from 2016 [10]. Nothing here was published in the last decade, so you are not getting current post-marketing data. The Finnish registry study is the one that matters most for long-term safety [5]. It followed real women over years and found a clear pattern: fewer endometrial, ovarian, pancreatic, and lung cancers than expected, but more breast cancer. About 1,542 breast cancers occurred where about 1,292 were expected. That is a 19% relative increase, and the confidence interval does not cross zero. No manufacturer funded that registry; it is national health data, which is the strongest kind of evidence in this whole file.

The bleeding and endometrial data are consistent across trials. In heavy menstrual bleeding, Mirena cut blood loss from about 420 mL to 42 mL at 12 months, while surgical resection went from 404 to 7 mL [3]. It beat oral contraceptive pills on treatment failure, 11% versus 32% [9]. For endometrial hyperplasia, regression rates ran 81% to 100% across studies, better than oral progestins in complex and atypical cases [10, 15]. In tamoxifen-treated breast cancer patients, it cut polyp formation sharply, from about 15% to under 2% [2]. These are hard clinical outcomes, not surrogates.

The harms are understated in this file. The 2001 trial lists pelvic adverse events in 13 of 20 Mirena users, plus bleeding disorders, pain, breast tenderness, headache, acne, and mood changes [3]. The adolescent series shows real failures: one expulsion, one no improvement, one family requesting removal [13]. Perforation and expulsion rates appear in the 2012 review, but the numbers are sparse [16]. The breast cancer finding is the one that should give any woman pause, and it is buried in a review that concludes no increased risk [16]. That review is contradicted by the registry data it cites.

The burden of proof question here is not whether Mirena works. It works for bleeding and endometrial protection. The question is whether the breast cancer signal is real, and the only long-term registry data in this file says it is. No trial in this file was designed to answer that question, and none followed women long enough. The 2012 review's claim of no increased risk is an assertion, not a finding, and it sits directly against the registry numbers.

My call: Mirena is effective for heavy bleeding and endometrial protection, but the breast cancer risk is a genuine concern that the evidence here does not resolve. Confidence: moderate.