NEJM – Early vs. Late IUD Insertion

Published - Written by Lisa Rosenbaum

Here is a riddle: What do grocery shopping while hungry, discharge planning post-MI, and immediate insertion of an IUD following first trimester abortion all have in common?

Stumped? Perhaps I can explain.

In this week’s NEJM, the Post-Aspiration IUD Randomization (PAIR) study group report their trial results in the article titled, “Immediate Versus Delayed IUD Insertion after Uterine Aspiration.” This non-inferiority trial randomized 575 women undergoing uterine aspiration for induced or spontaneous abortion to either immediate or delayed (2 to 6 weeks later) IUD insertion.

Though it is known that immediate insertion of an IUD following abortion decreases the rate of repeat abortion, there has been lingering concern that immediate insertion may be associated with higher rates of IUD expulsion. To address this, the primary endpoint of the trial was the rate of IUD expulsion six months following insertion. At six months, the expulsion risk was 5% in the immediate insertion group, and 2.7% in the delayed-insertion group, which meet the non-inferiority standard. The rates of pelvic infection and uterine perforation did not differ significantly between the groups.

Given that the criteria for clinical safety were met, what else can we learn from this study? An important observation is simply the differences in the rates of IUD use between the two groups.  At six months, 92.3% of those in immediate insertion group were using an IUD versus only 76.6% in the delayed-insertion group.

So what do these results have to do with hungry grocery shopping and post-MI discharge planning?

The hot-cold empathy gap refers to the incongruity between decisions we make in a “hot,” state, (hungry, angry, or sexually-aroused), versus those we make while “cold.”[i] But the principle, which was described by the economist/psychologist Dr. George Lowenstein, extends beyond the obvious fact that our emotions impact our decisions. What most of us predictably overlook is that our interpretation of the world around us is indeed “state-dependent.” When we are sad, we often forget that there will again be joy. While hungry, it’s tough to imagine ever being full. In essence, we consistently fail to keep our own feelings in perspective. Therein lies the gap.

While avoiding the grocery store when you’re hungry is one oft-used trick to keep the hot-cold empathy gap at bay, health delivery offers ample opportunities to use  this cognitive quirk to promote better health.  There is no hotter moment than an acute MI. Patients are terrified, and in the immediate aftermath, they will tell you that they have decided to change their lives.  This is precisely why widespread quality measures now take advantage of this “teaching moment” to initiate medications like statins right then, rather than later in the outpatient setting. If you start a statin prior to discharge your patients are more likely to actually keep taking it.   And they are more likely to be alive one year later.  The change began while they were hot.

All the women entered the PAIR study because they wanted IUDs when they presented for abortion. So what happened to the women in the delayed-group who went home with scheduled follow up appointments for their IUDs? Hot versus cold, again.

Of the 6.4 million pregnancies that occur in just the US each year, 49% are unintended.[ii] Mathematical modeling suggests that more than 70,000 unintended pregnancies could be avoided each year if we switched from delayed to immediate IUD insertion. World-wide adaptation could prevent hundreds of thousands more.  Thanks to this study, we now know this strategy is safe. The beauty of good science is that it fires you up just enough to take advantage of cold, hard facts.


[i] Lowenstein, G. “Hot-Cold Empathy Gap and Medical Decision Making.” Health Psychology 2005. Vol 24. No 4(Suppl).S49-256

[ii] Finer LB, Henshaw, SK. Disparities in Rates of Unintended Pregnancy in the United States, 1994 and 2001 Perspectives on Sexual Reproductive Health, 2006:38:90–96.