Contraception Update: A Focus on Safety and Efficacy

What we know

In 2010, the Centers for Disease Control first published the United States Medical Eligibility Criteria for Contraceptive Use (US CDC MEC) to serve as a comprehensive resource on the safe use of contraceptives for persons who have certain characteristics or medical conditions. After an interim update to the CDC MEC in 2016, more evidence on the safety of different methods has become available, and an update was needed.

Study results

Based on expert input and review of new evidence, the 2024 MEC report replaces prior reports with updates including: recommendations for persons with chronic kidney disease (nephrotic syndrome, receiving dialysis), those being treated for or at high risk for HIV infection, and revisions to recommendations for persons with certain characteristics or medical conditions (i.e., breastfeeding, postpartum, postabortion, obesity, surgery, history of venous thrombosis/thrombophilia, valvular heart disease, peripartum cardiomyopathy, systemic lupus erythematosus, liver diseases, sickle cell disease, and solid organ transplantation). The guideline also includes new methods (i.e., barrier, vaginal pH modulators, fertility awareness-based). Additionally, the update also emphasizes the need for equitable access to reproductive health care for all and acknowledges the United States’ history of reproductive coercion and forced sterilization of multiple marginalized groups. Thus, there is a stronger statement for the need for shared-decision making throughout the document. Because the >100 page document may be difficult to access during a busy clinic, clinicians can download the free phone application to quickly access recommendations based on a particular contraceptive method or medical condition, available at https://www.cdc.gov/contraception/hcp/contraceptive-guidance/app.html.

What this changes

These guidelines better recognize the importance of shared decision-making in an equitable, noncoercive manner that respects the goals, values and autonomy of the individual. This includes prioritizing patient preference as the main decision maker and respecting the patient’s decision to discontinue or not use contraception rather than a sole focus on contraception and pregnancy prevention. This aligns with the American College of Obstetricians and Gynecologists 2022 Committee Statement on Patient-Centered Contraceptive Counseling, which encourages practitioners to recognize how their own biases, implicit or explicit, impacts contraception counseling and to prioritize the patients’ values and lived experiences. ¹

What we know

There are few contraindications to the use of POPs, and a 2022 survey of over 4000 reproductive age women showed that the majority were in favor of making oral contraceptive pills (OCPs) available over-the-counter (OTC) without a prescription. ^2,3

Although OCPs are the most widely prescribed reversible contraceptive, POPs make up a small minority of these prescriptions. This may be due to perceived worry about a higher failure rate given the recommendation to take norethindrone POPs during the same daily 3-hour time window. A recent comprehensive literature review investigating the effectiveness and efficacy of POPs indicated that among studies at low or moderate risk of bias, typical POP use was associated with a failure rate of 1–2%, depending on the formulation. ³ With the approval of the first OTC progestin OCP in the United States in 2023 (Opill, norgestrel 0.075 mg), clinicians may have concerns about its effectiveness.

Study results

In this randomized crossover study, 46 women aged 18–35 with a BMI <32 kg/m² completed a baseline 28-day cycle with correct daily pill use, taking the norgestrel 0.075 mg pill within a daily 3-hour time window. This was followed by two intervention cycles: a pill taken 6 hours late mid-cycle or missed completely. The study evaluated patients’ ovarian follicle formation using transvaginal ultrasonography (TVUS), cervical mucus assessment, and lab testing (serum progesterone, estradiol, follicle-stimulating hormone, and luteinizing hormone). After study completion, a committee of experts, who were not affiliated with the study, reviewed the data and assigned theoretical scores for degree of contraceptive protection based on the cervical mucus favorability, serum estradiol and progesterone levels, and TVUS findings of ovarian activity. They found there was no significant change to theoretical contraceptive protection when delaying or missing one dose of mid-cycle norgestrel 0.075 mg POP. However, the authors acknowledge that this study did not investigate pregnancy as an outcome. This study also did not assess other forms of variation from perfect use of POP, such as delaying pill intake for different time intervals or missing more than one day of use. Importantly, only 28% of patients had a BMI >25 kg/m², which may further limit generalizability.

What this changes

By investigating deliberate nonadherence to the norgestrel 0.075 mg POP, this study suggests that the margin for error with this particular POP may not be as narrow as the current 3-hour recommendation. With the now widespread availability of norgestrel 0.075 mg OTC, this study bolsters clinicians’ confidence in the efficacy of this POP for patients seeking access to contraception. Given several important limitations of the study, further research needs to be done before changes can be made to the “3-hour window” recommendation.

What we know

Migraine headaches affect a large proportion of the general population, with women more likely to be affected. The CDC MEC 2024 guidelines note that combined hormonal contraception (CHC) use in the setting of migraine headache with aura poses an “unacceptable health risk,” citing concern for increased risk of ischemic stroke. In contrast, migraine without aura does not have similar restrictions. Stroke risk is correlated to estrogen dose, with contraceptives containing ethinyl estradiol (EE) >50 μg being associated with the greatest risk. ⁴ However, there is less data as to whether lower doses (EE ≤20 μg) are associated with a reduced risk.

Study results

The aim of this study was to understand how estrogen dose and a diagnosis of migraine impact ischemic stroke risk with CHC use. This case-control study looked at all women between ages 18 and 55 who have used CHCs at a large tertiary care center over the course of 10 years. Using this population, patients who had suffered from an ischemic stroke were identified using ICD codes, which were then validated by a team of physicians through chart and imaging review (case group). The control group was a random sample of 20,000 CHC users; the groups were matched based on multiple patient characteristics, including medical and family histories and stroke risk factors. Of the 203,853 CHC users identified, 127 had a confirmed stroke while on CHC (0.06%; CI: 0.05–0.07%). A higher number of patients in the case cohort had migraines compared to the control (26.8% versus 17.3%, p = 0.011). Those taking doses of EE ≥30 μg had increased odds of stroke compared to those using <30 μg (OR 1.52; CI: 1.02–2.26; p = 0.040). Interestingly, there was no increased risk of ischemic stroke in patients with migraines with aura compared to those without migraines, whereas migraines without aura showed increased risk (OR, 2.35; CI: 1.32–4.2; p = 0.004). Important limitations of the study include the retrospective design, as well as the overall small number of strokes, which may limit conclusions that can be drawn from subgroup analyses with respect to migraine subtype. The authors encourage shared decision-making when using CHCs in those with migraine, with or without aura.

What this changes

Historically, those who suffered from migraine with aura were counseled against CHC due to the increased risk of ischemic stroke. When contraceptives are used for the purposes of pregnancy prevention alone, CDC recommendations are to offer first the variety of progestin-only options available. At times, patients may prefer a CHC because of medical conditions (with or without need for pregnancy prevention), such as for treatment of acne, menorrhagia, mood change, etc. In these situations, shared decision-making using a patient-centered approach is recommended, including careful counseling of potential stroke risks for all migraineurs. If a CHC is chosen, formulations containing <30 μg EE would be preferred, though lower-dose formulations may change effectiveness and tolerability profiles.

What we know

Topiramate is prescribed in reproductive-age women for a variety of indications; concomitant use with OCPs may impact hormone levels via the CYP3A4 pathway. Strong CYP3A4 inducers (i.e., phenytoin, carbamazepine) have been shown to decrease OCP efficacy, which can be concerning in patients taking these medications as they can increase the risk of birth defects. Previous studies on topiramate, a moderate CYP3A4 inducer with known teratogenicity, have shown a dose-dependent interaction with 1 mg norethindrone and 35 mcg EE OCP, which has led to differing recommendations regarding the concomitant use of these medications. ^5,6

Study results

This retrospective cohort study, with an active comparator design, assessed the rates of unintended pregnancy when topiramate was taken with OCPs. A national private health insurance database was utilized, with data spanning the course of 13 years. The study population consisted of females aged 12–48 who suffered from migraines or chronic headaches, taking a daily low dose of topiramate (≤200 mg). The comparator group consisted of patients who were on the following medications in addition to an OCP (with at least 14 days of concomitant use): propranolol, metoprolol, amitriptyline, venlafaxine, and verapamil. There were 63,649 episodes of concomitant use of oral contraceptives and topiramate, of which 95% of patients were on ≤200 mg. The comparator group included 59,012 episodes of oral OCP with other maintenance therapies. There was no difference in contraception failure rate between the two groups, with both groups having unintended pregnancy rates of 1.3 per 100 person-years (95% CI: 1.1–1.6).

What this changes

Based on the results of this study, it appears that the extent to which topiramate alters OCP efficacy via CYP3A4 induction is dose dependent, primarily if the dose is >200 mg daily. At lower doses (≤200 mg daily), there appears to be no reduction in contraception effectiveness. This study provides reassurance to both patients and clinicians that when topiramate is taken concomitantly at lower doses, OCPs can help prevent unintended pregnancies. However, for many individuals, the teratogenicity of topiramate may warrant a secondary method of pregnancy prevention.

What we know

While IUDs are the most effective form of emergency contraception (EC), oral EC is the most widely utilized method of preventing unintended pregnancy. Currently there are two forms of oral EC, levonorgestrel and ulipristal acetate, with levonorgestrel being available OTC in the U.S. Though ulipristal acetate has a significantly lower failure rate than levonorgestrel EC, both forms of oral EC are only effective when administered prior to ovulation. Prostaglandins, particularly PGE2, play a key role in the ovulatory cycle including follicular maturation and rupture, ovulation, fallopian tube motility, and implantation. ⁷

Therefore, it is proposed that a cyclo-oxygenase (COX) inhibitor could interfere with these reproductive processes and act synergistically with oral EC to further improve outcomes.

Study results

In this randomized placebo-controlled trial, the study population consisted of women, aged 18 and older, who requested levonorgestrel EC within 72 hours of a single act of unprotected sexual intercourse in their current menstrual cycle. Patients had to have a history of regular menses, agree to further abstinence, and follow up over the subsequent 6 weeks. Exclusion criteria included current pregnancy, breastfeeding, recent abortion or postpartum without resumption of regular menses, IUD in place, or history of sterilization. Medication and medical history exclusion criteria also included current use of hormonal contraceptives (including recent oral EC), NSAIDs in the previous week, anticoagulants, cyclosporine, tacrolimus, corticosteroids, lithium, SSRI, quinolones, asthma, urticaria, ischemic heart disease, heart failure, hypertension, cerebrovascular disease, peptic ulcer disease, or gastrointestinal bleeding. Any history of allergic reactions to piroxicam, aspirin, or other NSAIDs was also excluded. Participants (n = 860, mean age 31; mean BMI 21.3 kg/m²) were randomly assigned to receive levonorgestrel plus piroxicam 40 mg or levonorgestrel plus placebo under medical supervision. Patients were to follow up 1–2 weeks after their next expected menstrual period and tested for pregnancy at that time, if indicated.

The primary outcome was the number of pregnancies prevented, and the secondary outcomes were pregnancy rate after EC, date and change in bleeding pattern of subsequent menstruation, occurrence of non-menstrual bleeding, and adverse events. At follow-up, 1 (0.2%) in the EC+piroxicam and 9 (1.7%) in the EC+placebo group became pregnant following treatment, with an odds ratio of 0.20 [95% CI: 0.02–0.91]; p = 0.036). The percentage of pregnancies prevented was significantly higher following piroxicam–levonorgestrel co-treatment (the addition of piroxicam prevented 94.7% of expected pregnancies vs 63.4% prevented with EC only). There was no difference in adverse events noted between the two groups. The most frequently reported adverse events were fatigue or weakness, nausea, lower abdominal pain, dizziness, and headache.

What this changes

In this trial, co-administration of an NSAID, piroxicam 40 mg, with levonorgestrel EC resulted in a significant reduction in unintended pregnancies with no associated increase in side effects. Piroxicam was chosen due to its long half-life, which allowed for a one-time administration, as well as its safety profile. However, piroxicam is only available via prescription, requiring a patient to seek medical care and thus presenting another barrier to treatment. It is unclear if the results of this study can be applied to other COX inhibitors that are available OTC, as the majority of these options have shorter half-lives requiring dosing several times a day for a number of days, which may impact adherence and efficacy. Additionally, the study population was primarily made up of women with BMI <26 kg/m², and the efficacy of levonorgestrel EC is reduced in patients with BMI ≥26 kg/m². Further research should include alternative COX inhibitors and a more diverse patient population.