Skip navigation

Mindstretcher - IUD use and the risk of ectopic pregnancy

The trouble is that it isn't as simple as the title suggests. Bandolier had a real problem getting its mind around a detailed and worthy systematic review and meta-analysis of case-control studies on this topic from Brussels [1].

It starts off in the usual way - searching for papers which examined ectopic pregnancy and IUD use (though only in English, French and Chinese). Finding 16 studies follows, and the data extraction, quality scoring and pooling of odds ratios is all pretty standard stuff. It is the results and what they mean that stretched Bandolier's mind on this occasion.


There is a fair amount of statistical to-ing and fro-ing in the paper, but the main results are not contentious, and are shown in the figure in their most conservative form. For current IUD use, women with an ectopic pregnancy are six times more likely to have an IUD when compared with women in the second or third trimesters of a pregnancy. But women with an ectopic pregnancy are no more likely to have an IUD than women who are not pregnant.


The issue is one of framing, or what question is being asked. That means that it is the choice of controls that makes the result. In an ideal situation, controls would be chosen from the same population as the cases, but that is not always easy, as the case of ectopic pregnancy indicates.

Women who have an ectopic pregnancy may:

  1. Not want to get pregnant
  2. Want to get pregnant
  3. Not be worried either way

The trouble is that we don't know what particular mix we have in any one study.

If we want controls for group 1, then we choose women from the non-pregnant age-matched population. These women probably would have the same likelihood of having an IUD as our cases - which is just the result we find. What this tells us is that women with an IUD in situ do not have an increased risk of having an ectopic pregnancy.

If we want controls for group 2, then we choose women who are pregnant (on the grounds that most of them would want to get pregnant). But these women would be less likely than all women of the same age to be using contraception, and hence IUDs. So finding a higher rate of IUD use in women with ectopic pregnancy compared with pregnant controls is the expected result. What this tells us is that a pregnancy with an IUD in situ is more often an ectopic one than a pregnancy with no IUD in place.

The past IUD results tell us that former IUD use can mildly elevate the risk of ectopic pregnancy.


This is a useful paper, not only on the risks of ectopic pregnancy, but also on the limitations and difficulty with case-control studies. You may need to read it more than once.


  1. X Xiong, P Buekens, E Wollast. IUD use and the risk of ectopic pregnancy: a meta-analysis of case-control studies. Contraception 1995 52: 23-34.

previous or next story in this issue