Latest National Estimates of Contraceptive Failure

Abstract & Commentary

By Leon Speroff, MD, Editor, Professor of Obstetrics and Gynecology, Oregon Health and Science University, Portland, is Editor for OB/GYN Clinical Alert.

Synopsis: The latest estimates of contraceptive failure from the 2002 National Survey of Family Growth fail to demonstrate an improvement.

Source: Kost K, et al. Estimates of contraceptive failure from the 2002 National Survey of Family Growth. Contraception. 2008;77:10-21.

Kost and colleagues provide updated contraceptive failure rates derived from the 2002 National Survey of Family Growth.1 The objective was to see if the effectiveness with which American women used contraception has improved. The results reflect the typical use of a method in a general population, not the efficacy when a method is used perfectly. There were 7643 women, ages 15-44, surveyed in this national sample. Sixty-two percent (38.1 million) of American women in reproductive age in 2002 were using some method of contraception. It is well-recognized that induced abortions are underreported by respondents; therefore the results were corrected for the number of abortions resulting from contraceptive failure. The estimated failure rates after one year of use were as follows:

    Typical Use Perfect Use
  Injectable methods 6.7% 0.3%
  Pill 8.7% 0.3%
  Male condom 17.4% 2.0%
  Withdrawal 18.4% 4.0%
  Rhythm methods 25.3% 5.0%

All methods


The current estimates of failure were not significantly different compared with the previous estimates from the 1995 national survey. Women over the age of 30 are less likely to experience failure than young women; teens are more than twice as likely to experience a failure than older women. Hispanic women and even more so, Black women, experience higher failure rates. Groups that were less likely to experience contraceptive failure were women who did not intend to have a subsequent birth and women who had no previous births. Married women experienced the lowest failure rates and cohabiting women the highest. The most important determinants of pill failure, therefore, were: age, intention toward a future birth, parity, and marital status. Interestingly, once these factors were accounted for, duration of use, race, ethnicity, and poverty status no longer affected the risk of pill failure. The same factors influence condom use, but when corrected for these factors, race, ethnicity, and poverty affected the risk of condom failure.


This is a subject of great interest because the rate of unintended pregnancies in the U.S. continues to be high. About one-half (3.1 million) of all pregnancies in the U.S. are unintended, and nearly half of those occur in women using a method of contraception. Here is a more striking statistic: one of every two American women aged 15-44 has experienced an unintended pregnancy.

An important piece of good news was the fact that women at each end of the economic spectrum, the poorest and the wealthiest, experienced a decrease in failure rates from 1995 to 2002, although the poorest women continued to have a higher failure rate than did the better-off women. Also, although the overall failure rate was not statistically significant comparing 1995 and 2002, there was about a 2.5% improvement; this missed mathematical significance but it may reflect a meaningful change in our population. This change is probably due to an increase in pill and injectable methods and a decrease in condom use during this period of time. It will be of great interest in the next survey to see if the newly marketed patch, vaginal ring, and implant favorably affect these statistics.

Women living in poverty who must rely on condoms or withdrawal (male-dependent methods) have about a two-fold increase in failure rates, but if they can use the pill, their failure rates are the same as better-off women. The message is clear: we need to make the more effective methods available for low-income women.

There is an alarming message in this report. Overall use of contraception among women at risk of unintended pregnancy decreased from 92.5% in 1995 to 89.3% in 2002. The use of contraception among low-income women at risk of pregnancy decreased from 92.1% in 1995 to 86.3% in 2002. I think this indicates an American problem; women are having more difficulty obtaining effective contraception.

What do women have to overcome to achieve good contraceptive efficacy, and if they are already using a method, to switch to a more effective one? Choices must be available for various methods. The technique of using a method must be compatible with an individual and her lifestyle. Some methods require partner cooperation. Once chosen and obtained, the individual must exert dedication to its use. The failure to substantially improve contraceptive failure rates from 1995 to 2002 indicates that we are not making enough progress with each of these variables.

It is not enough to say the obvious—that we need greater education—but we need to learn where and when education is most effective, where is money best spent, and how to maximize the choices available for all women. This isn't a task just for professional health care providers; it is a widespread social problem that requires policy and budgeting decisions. Trussell and Wynn, in an excellent editorial that accompanies the article, point out that technological achievements are not enough.2 The problems are more sociologic, such as cost and insurance coverage (and I would add the ridiculous practice of providing pills only one month at a time). These are reasons why other countries have lower percentages of women at risk for unintended pregnancies. In my view, this requires governmental action at both the federal level and the state level. So, whom are you voting for?


  1. Kost K, et al. Estimates of contraceptive failure from the 2002 National Survey of Family Growth. Contraception. 2008;77:10-21.
  2. Trussell J, Wynn LL. Reducing unintended pregnancy in the United States. Contraception. 2008;77:1-5.