You Probably Don’t Understand Contraceptive Pregnancy Rates

Originally published in abridged form on the Our Bodies Ourselves blog

Failure Is Not What You Think.

“The male condom has a 2% failure rate.”

What does that mean?

There are a lot of ways to go wrong here. Let’s start simple. What does failure mean? One straightforward interpretation might have failure count as every time the condom breaks or slips.

That’s not what it means.

Another way to look at failure might be to count failure as every time a sperm fertilizes an egg. But that’s not what failure means here either.

A contraceptive failure really means that a pregnancy occurs. This confusion is also why academics use the term pregnancy rate. And it’s why this article will use the term pregnancy rate from here on. (Note that rates are sometimes framed as something like, “98% effective,” confusing matters even further; we’ll avoid this, too.)

Now for a further caveat. Not all pregnancies count; we don’t add pregnancies resulting in spontaneous abortions. For clarification, spontaneous abortions are when the fertilized zygote or preembrio either never attaches to the uterine wall, or it quickly detaches. These spontaneous abortions are quite common — around half of all pregnancies. In these cases, the woman won’t even have realized she was pregnant. If you were looking to avoid pregnancy, this is kind of like nature giving you a mulligan.

So which pregnancies do count towards the pregnancy rate? Pregnancies that result in abortion or miscarriage count. Of course, pregnancies resulting in birth also count.

We can hone our statement a bit more now:

"Among those using the condom, the probability of pregnancy is 2%."

How Long Are We Talking?

Another interpretation might be to say that pregnancy occurs in 2% of all condom uses.

Wrong again.

And that’s a good thing. If that were correct, then all contraceptive pregnancy rates would be high. Really high!

In fact, we’re talking about the probability of a pregnancy over the course of a year. The most reliable data on pregnancy rates come from using the National Survey on Family Growth (NSFG). This large-scale survey asked over 7,000 women aged 15–44 which contraceptive they were using and whether they experienced a pregnancy. The women responding referred to periods when they just starting to use a particular method, though it wasn’t necessarily the first time they ever used that method. The pregnancy rates from this survey were then converted into one-year periods and accounted for complicating factors such as underreported abortions.

We have more information now. This lets us clarify our pregnancy rate interpretation further:

"Among those using the condom over the course of a year, the probability of pregnancy is 2%."

What’s Typical?

Contraceptive pregnancy rates are also separated out into two categories: typical and perfect use. And there’s a huge difference.

Typical use is what we see in real life. Forgot to use the condom? Didn’t put it on right? Used it correctly and consistently? It all counts. Typical use is the average among everyone.

Perfect use is more of how it sounds. Only those that use the method correctly and consistently count here. Sometimes, however, it’s hard for researchers to get good data on perfect-use rates. Consequently, these researchers do their best to give good estimates based on the information they have.

Again, the difference between typical and perfect-use pregnancy rates is huge. And the condom’s 2% pregnancy rate that keeps getting referenced actually refers to its perfect-use rate. The typical (real world) pregnancy rate for condoms is 18%. That’s an incredible 9x higher!

We see this rate discrepancy not only with condoms but with all contraceptive methods that allow for user error. For example, early withdraw has a typical-use pregnancy rate over 5x its perfect-use rate. Well, maybe that one’s not too surprising. But you might be surprised with the U.S.’s most popular method, the pill. Its perfect-use pregnancy rate is only 0.3% while its typical-use rate is 9%. That’s a 30-fold difference!

For a bit of contrast, look at the pregnancy rates for for implant devices. Nexplanon has a perfect-use pregnancy rate of 0.05% — as good as contraceptives get. Its perfect-use rate? Also 0.05%. The same goes for the intrauterine contraceptive Mirena. Its perfect and typical-use rates are both 0.2%. These methods shrink typical-use pregnancy rates by either minimizing or eliminating the possibility for user error.

We’re at a point now where we have an even better idea of what our initial statement means:

"Among perfect condom users (using it every time, following directions), the percentage of those becoming pregnant after one year is 2%. Among typical condom users (includes everyone), the percentage of pregnancies after one year is 18%."

But What About You?

The thing about the NSFG survey results is that they refer to a large sample of people meant to tell us about the population at large. While we’d be incorrect to extrapolate that information to the individual level — like you — it’s hard for contraceptive pregnancy rates to be useful otherwise. And so in practice, we seem to go ahead and extrapolate away. First, we look at the typical-use rate. Then, we try to be honest with ourselves and figure whether we should move ourselves closer to the perfect-use rate or further away.

We might also consider factors like how frequently we have sex. Are we a bit older (particularly with women)? Are we younger? It’s tough to know where we fall for sure, but published pregnancy rates give us a good base rate.

Rolling the Dice Over Time

Perhaps there’s a bit more, however. After all, we don’t stop having sex after one year. And we don’t become indifferent to pregnancies after one year either. So maybe we should look at rates beyond one year?

Unfortunately, we don’t have good data on long-term pregnancy rates. But we can get an idea of the effect longer time periods have.

Let’s say you’re an average condom user, and your odds of experiencing a pregnancy in a year are 18%. But you aspire to be a better contraceptive user (good for you!). And so over the next five years you follow through. As an improving condom user, you lower your pregnancy risk by 4% each year until you hit the perfect-use pregnancy rate of 2% (ie. 18% rate the first year, then 14%, then 10%, then 6%, then 2%).

What are your chances of experiencing at least one pregnancy over this five-year span? Guesses?

We can actually calculate this out because we’re making assumptions about the rates here (even putting it in a positive light). The risk of you encountering at least one pregnancy over five years with a condom — while improving usage — is just over 40%. Ready for that baby?

Below is a figure from a paper I wrote in graduate school (way before the recently popular New York Times article) which highlights the difference of short-versus long-term pregnancy rates.

Dealing with the Numbers

This 40% risk shouldn’t be surprising. It explains why half of women aged 15–44 in the U.S. have experienced an unplanned pregnancy. That number also represents a lot of surprised new fathers!

Note that 90% of women at risk for an unplanned pregnancy really do use contraception. And while the majority of contraceptive users aren’t perfect, even perfect users still account for 5% of unplanned pregnancies.

So what can you do to get your own numbers down? It’s easy to just say that you should become a perfect user. That’ll certainly make a big difference, but it’s not a realistic option for everyone.

Using more effective methods is an approach. As discussed, a long-acting reversible contraceptive like an implant device or intrauterine contraceptive has a great pregnancy-rate track record. Further, over 75% of users keep using those methods after a year. That’s a better continuing-use rate than condoms or the pill. Unfortunately, long-acting methods are a heavily underutilized approach in the U.S. (they shouldn’t be). And men still have no such reversible long-acting option (this is awful, too).

Let’s say both you and your partner aren’t perfect users, and long-acting methods aren’t an option. Another route is to use at least two methods simultaneously, an approach taken by 15% of couples. By using two methods simultaneously, both methods have to fail at the same time in order for pregnancy to occur.

The most common example would be using the pill and condom together. But you could do other combinations. Pill and withdrawal, condom and rhythm method, diaphragm and pill — these all work. Any combination will be better than using one of the methods alone. (Note that the lack of contraceptive options for men limits the combinations.)

Successfully controlled parenthood has a lot to do with having options that work well for you and your partner while providing acceptable pregnancy risks. This is yet another reason why we need more contraceptive options — especially for men! With a better understanding of how pregnancy rates work, we can make better decisions for ourselves. And we can make more intelligent demands for new contraceptives.

Originally published at www.malecontraceptive.org on November 13, 2014.