> For every 100 people who use the pull out method perfectly, 4 will get pregnant. Are those 4-in-100 distributed randomly? Did the pregnancies result despite

> For every 100 people who use the pull out method perfectly, 4 will get pregnant. Are those 4-in-100 distributed randomly? Did the pregnancies result despite 100% adherence to the method, or was there occasional failure to adhere? There's a big difference between a method being ineffective from the perspective of a health care provider, and a method being ineffective in absolute terms. A provider has to care about what broad cross sections of people will actually do, rather than what they say they will do. If you're an individual person who knows they can adhere to the method perfectly, the fact people on average cannot or will not adhere perfectly has no particular relevance to you. I have no horse in this race, but the same difference in the meaning of "efficacy" arises in all sorts of aspects of health care, like advice on diet and exercise, or the prescription of specific exercises for physical therapy.