Parity and use-effectiveness with the contraceptive sponge

Contraceptive Failure in the United States: A Critical Review of the Literature

“…The randomized sponge/diaphragm trials show rates of 12.9 percent in the US (slightly higher than the figure of 12.5 percent reported in an earlier analysis of the same data by Edelman et al, 1984) and 10.9 percent in London (McIntyre and Higgins, 1986;Bounds and Guillebaud, 1984). However, some women in these trials were diaphragm users reassigned to the diaphragm and analyzed as if they were initiating use.…”

“…US data on 722 women from 13 clinics yielded a first-year failure rate of 17.0 percent (vs. 12.5 percent for the diaphragm) (Edelman et al, 1984).57 The British trial was smaller (126 women), and the failure rate was even higher at 24.5 percent (vs. 10.9 percent for the diaphragm) (Bounds and Guillebaud, 1984). Reanalysis of the US data (which included one additional sponge user) revealed, however, that the sponge failure rate among parous users (28.3 percent) is much higher than among nulliparous users (13.9 percent); no difference in failure rates by parity was found for diaphragm users (McIntyre and Higgins, 1986). These results show that the sponge is as effective as the diaphragm among nulliparous women but much less effective than the diaphragm among parous women, presumably because the sponge comes in only one size (too small for parous women).58 The manufacturer has disputed this interpretation, reasoning that pregnancy intention (prevent vs. delay) is a confounding variable whose effects are not controllable because no information was collected.…”

“…Because the randomized trials of the sponge and diaphragm yielded the same failure rates for both methods among nulliparous users (and the rate for the diaphragm did not differ by parity) (McIntyre and Higgins, 1986), we assigned the diaphragm rate to nulliparous users of the sponge. However, because the failure rate among parous sponge users was so high, we cannot summarize the sponge failure rate with only one number; to do so would be grossly misleading.…”

“…In the analysis, they were incorrectly counted as starting at month 0 (Edelman et al, 1984;McIntyre and Higgins, 1986); the failure rate for the diaphragm (relative to that for the sponge) was thereby biased downward, although not by very much, because only a small fraction of women were reassigned to the diaphragm (see the discussion in the section entitled Diaphragmn).…”

Clinical Research in Obstetrics and Gynecology: More Tips for Busy Clinicians

“…Investigators compose subgroups based on characteristics present at enrollment and explore different outcomes in the various subgroups. For example, researchers have performed subgroup analyses to see if parous women using the vaginal sponge (30) and obese women using the contraceptive patch (31) had different contraceptive efficacy than other women. Given the many participant demographic Clinical Research in Obstetrics and Gynecology • Supplement S57 characteristics available for study, the nearly limitless comparisons can create havoc.…”