IUDs have actually been around for quite a long time. Their effectiveness has never really been questioned - they were always found to be highly effective contraceptives. Originally, it was thought that IUDs work almost exclusively by a local inflammatory response in a woman's endometrium. In other words, we used to believe that IUDs work almost exclusively post-fertilization.
About 20 years ago, data began to emerge that questioned that theory. It was found that IUDs also had effects on the local environment of the uterus that could result in some of its contraceptive efficacy. An increase in the thickness of the cervical mucous may result in inhibition of sperm motility. Likewise, the copper released form certain IUDs (not Mirena) may have a toxic effect on sperm. There is also some research that shows a possible effect of the IUD on ovulation. Remember, Mirena uses the same progestin analogue as Plan B (LNG). If these mechanisms are totally responsible for the contraceptive effect of Mirena, then we could confidently state that Mirena does not act after fertilization and would not have an effect on a developing human embryo.
Unfortunately, the evidence does not support this. The LNG that was added to the Mirena only has a local effect within the endometrium, and its blood levels never rise enough to have an effect on ovulation. In fact, this LNG was added primarily to decrease the amount of bleeding sometimes seen amongst copper IUD users. Furthermore, this study shows deleterious effects of this LNG dose on the endometrium, resulting in the conclusion that these changes may have "a pivotal role in the contraceptive effect of the LNG-IUS". This is clear evidence that the changes in the uterine lining not only can cause an embryo not to implant, but may also be a primary mechanism for the contraceptive effect of Mirena.
This is supported in many other places. Here is an article about various forms of IUDs. Its conclusions are clear:
In women with the LNG-IUS, the endometrium is abnormally thin and contains areas of superficial fragile vessels (Guttinger and Critchley, 2007). These features suggest that the uterus would be hostile to implantation.
In conclusion, IUDs may exert their contraceptive action at different levels. Potentially, they interfere with sperm function and transport within the uterus and tubes. It is difficult to determine whether fertilization of the oocyte is impaired by these compromised sperm. There is sufficient evidence to suggest that IUDs can prevent and disrupt implantation. The extent to which this interference contributes to its contraceptive action is unknown. The data are scanty and the political consequences of resolving this issue interfere with comprehensive research.As usual, this data come from peer-reviewed sources that are otherwise very hostile to the pro-life view. There is very little evidence that Mirena acts exclusively before fertilization and significant evidence that it effects the endometrium to an extent that implantation impairment may be an important mechanism to explain its efficacy. If one believes in the inherent value of human beings from the moment they become human beings, the use of Mirena is very much ethically problematic.