Contraceptive failure in the United States: revised estimates from the 1982 National Survey of Family Growth.

It is estimated that only about half of all abortions occurring during the period covered by the 1982 National Survey of Family Growth were actually reported in the survey; thus, contraceptive failure rates calculated from these data are almost certainly inaccurate. An attempt to correct for the underreporting of abortion indicates that actual 12-month use-failure rates are more than one-third higher than those calculated without taking abortion underreporting into account, with rates ranging from six percent for the pill to 14-16 percent for the condom, diaphragm and rhythm and to 26 percent for spermicides. Patterns of contraceptive failure are similar to those found in earlier studies.