Case reports have raised suspicions that broad-spectrum antibiotics in general might lower the effectiveness of hormonal contraceptives. Still, studies find that various broad-spectrum antibiotics do not lower hormone levels and, with one early exception (13), they have found no evidence of ovulation. Pregnancy rates are similar among women taking COCs alone and women taking both COCs and antibiotics (9, 12, 20). The 2003 Expert Working Group left broad-spectrum antibiotics in MEC category 1 (use in any circumstances).

The MEC previously categorized use of the antibiotics rifampicin and griseofulvin both as category 3 (not usually recommended) for most hormonal contraceptives because these drugs were thought to reduce contraceptive effectiveness. There are reports of pregnancies in users of hormonal contraceptives taking griseofulvin, and griseofulvin affects liver enzymes in mice, suggesting a possible impact on hormone metabolism. There are no published clinical or pharmacokinetic studies on interaction between griseofulvin and contraceptive hormones, however. The Expert Working Group reclassified use of griseofulvin to category 1 for users of combined or progestin-only injectables and category 2 (generally use) for users of other hormonal methods.

The Expert Working Group kept cervical intraepithelial neoplasia (CIN), noninvasive lesions considered a precursor to cervical cancer, in MEC category 2 (generally use) for hormonal contraceptives except progestin-only pills, for which CIN is classed as category 1 (use in any circumstances). The category 2 rating, assigned in the first edition of the MEC, reflects “some concern that COCs enhance the progression of CIN to invasive disease, particularly with long-term use” (that is, greater than five years) (51). According to a 2002 meta-analysis of over 30 studies presented to the Expert Working Group (41), the risk of developing invasive cervical cancer or one of its more immediate precursors increases with duration of use of COCs or DMPA. The association is statistically significant after five years of use. Only limited evidence, however, addresses the question of whether CIN is more likely to progress in women who use hormonal contraceptives. The few studies comparing COC use among women with low-grade cervical lesions and use among women with high-grade lesions yield inconsistent findings (1, 11, 16, 27, 34, 46). One study followed up women with low-grade lesions and found that progression was significantly more common in COC users than nonusers (7).

The Expert Working Group continued to recommend that women who are breastfeeding should generally not use progestin-only contraceptives (category 3) until six weeks postpartum. The systematic review of the evidence found no adverse effects of these methods on breastfeeding patterns and, while the evidence is more limited, no adverse effects on infant growth, development, or health when women using progestin-only methods started breastfeeding before six weeks postpartum (32). Lacking data on the effects of progestin in breast milk on the infant’s brain and liver development, however, the Expert Working Group did not shorten the 6-week restriction. Still, as the Expert Working Group had noted in 2000 and reaffirmed in 2003, in many settings the morbidity and mortality risks of pregnancy are high, and progestin-only contraceptives may be one of the few types of methods widely available to breastfeeding women immediately postpartum (51).