Skip navigation

Contraceptive effectiveness

Bandolier has long been mystified by something called the Pearl index (the number of pregnancies per 100 woman years) a measure of how well a method prevents pregnancy. Is it the same as pregnancies in 100 women in one year, or are there other interpretations? Fortunately others also find this measure unhelpful, and perhaps wrong [1]. Bandolier found some new definitions for use in contraception studies helpful and clear.

Efficacy and effectiveness

Trussell and colleagues [1] define efficacy as being the inherent protection of the method in perfect-use cycles only, borrowing the term from epidemiology. Effectiveness is to do with typical use for all cycles - and therefore includes issues like compliance and has more 'real-world' utility, if, of course, effectiveness can be measured.

So condoms or pills may have high efficacy, but because people use them improperly, or forget to take them, the effectiveness of the method falls, and more pregnancies occur than would be expected. Oral contraceptives have high efficacy - but there are suggestions that effectiveness is much lower than the 99.9% efficacy often quoted.


There are many variables affecting conception and pregnancy rates - the capacity to conceive, frequency and timing of intercourse and the degree of compliance with a contraceptive method. Not all of the underlying nuances are captured by these broad terms, but one thing is likely to be universally true, and that is that the effectiveness of a contraceptive method will be lower than its efficacy.

Impact of effectiveness

Bandolier turned the question of effectiveness around, to know what level of effectiveness would provide what degree of protection against pregnancy. This was done for a couple where the woman had 13 cycles a year with four fertile days per cycle (52 fertile days a year), where intercourse happened twice a week, and in which only 1 in 5 conceptions led to a pregnancy.

By applying different rates of contraceptive effectiveness it is then possible to calculate (Table) how long (on average) it might be until a pregnancy occurs, or what are the odds of a pregnancy in any one year. Thus using a contraceptive with 75% effectiveness the odds are even (1:1) that a woman will be pregnant in any one year. Odds of 10:1 against being pregnant don't happen until effectiveness rates reach 97%, and 100:1 against until effectiveness is 99.7%.

Table: Bandolier's calculations of the impact of contraceptive effectiveness on the chance of a woman becoming pregnant in any one year


  How long does it take to become pregnant?  
Days a year able to conceive Contraceptive effectiveness (%) Days Months Years Odds of pregnancy in any one year
15 1 122 4 0.3 1:3
15 50 243 8 0.7 1:2
15 75 487 16 1.3 1:1
15 90 1217 39 3.3 3:1
15 95 2433 78 6.5 7:1
15 96 3042 98 8.2 9:1
15 97 4056 131 10.9 11:1
15 98 6083 196 16.4 16:1
15 99 12167 392 32.7 33:1
15 99.5 24333 785 65.4 65:1
15 99.7 40556 1308 109.0 110:1
15 99.8 60833 1962 163.5 165:1
15 99.9 121667 3925 327.1 335:1
Assumption 1: 13 cycles/year          
Assumption 2: 4 fertile days/cycle          
Assumption 3: 4x13 = 52 fertile days/year = odds of 1 in 7          
Assumption 4: sex on average twice a week = odds of 1 in 3.5          
Assumption 5: 1 in 5 conceptions lead to pregnancy          
Chance of conception for any one day without contraception = 1/7 x 1/3.5 = 1 in 24          
Chance of conception in any one year without contraception = 365/24 = 15 chances total          

Neck on the block

Yes, it is quite likely that some of the assumptions in these calculations are educated guesses. In the absence of any published figures that Bandolier could find they are still worth doing. The broad outcome of needing a very high level of effectiveness will stand.

The chance of conception will be high over a time because of the repeated number of individual opportunities, even though the chances for any particular opportunity are small. A bit like taking part in a raffle when you have bought all the tickets - most tickets won't win, but you still get the prize.

The important thing is that women and men understand the risks. As the Table shows, unprotected intercourse carries a risk of about 1 in 24 of conception. The risk of pregnancy may be smaller because we have assumed that only 4 in 5 conceptions will not lead to pregnancy.

Womens' choices

If a woman is to choose a contraceptive method, she should have information about its likely effectiveness - and therefore her risk of pregnancy. That might (perhaps should) be the prime factor. Secondary factors are the risks of common but reversible adverse effects or harms (altered cycling or weight gain) as well as those of rare but irreversible harm (heart disease, cancer). Providing that information in ways that can be understood by women of different ages and educational backgrounds seems to be a challenge which has yet to be robustly grasped.


  1. M Steiner, R Dominik, J Trussell, I Hertz-Picciotto. Measuring contraceptive effectiveness: a conceptual framework. Obsterics Gynecology 1996 88: 24S-30S.

previous or next story in this issue