QuarterWatch™ (2 Years of Data Ending June 2017): Safety Signals for Pharmaceutical Contraceptives
The latest issue of ISMP’s QuarterWatch™ (see box at end of article) compares the markedly different safety profiles of the five leading pharmaceutical approaches to contraception:
- Combination oral contraceptive pills, which combine ethinyl estradiol with a chemical variant of progesterone, and sometimes a third component
- Emergency contraception pills, intended for single use after failure of another contraceptive method or unprotected sex, which contain a substantial dose of synthetic progesterone
- Long-acting etonogestrel implants, which provide reversible, long-acting contraception using a thin plastic rod containing etonogestrel (a form of progesterone) implanted in the upper arm
- Levonorgestrel intrauterine devices (IUDs), which provide long-acting contraception with a device emitting a synthetic form of progesterone
- Copper-releasing IUDs, which provide emergency and long-term contraception through emission of copper
These methods of contraception are used by an estimated 15 million women ages 15-44 years, making them one of the most widely used drug interventions in medicine.
To conduct the analysis, we used real-world data from 43,342 adverse event reports submitted to the US Food and Drug Administration (FDA) over 2 years ending in June 2017, including 14,759 serious and 28,583 non-serious events. The number of events reported for each of the five contraceptive methods appears in Table 1. Our analysis did not distinguish between different brand products in the same group, and excluded patches, rings, and injectable forms of contraception. For combination oral contraceptive pills, we also excluded progesterone-only oral products and combinations without ethinyl estradiol.
General Key Findings
Overall safety record. The overall safety record for these five pharmacological contraceptive methods was very good. Despite a very large population of women using these products, we identified 72 reported deaths over 2 years, 316 reports of life-threatening events, 1,491 events that resulted in hospitalization, and 217 events requiring intervention to prevent harm. This strong overall safety record is of interest given the known serious adverse events associated with these products, particularly blood clots with combination oral products, complications such as uterine perforation with IUDs, and ectopic pregnancies with IUDs and implant products.
Unintended pregnancies. Systematic surveys report a probability of contraception failure over 12 months of 7.2% for combination oral products, 6% for all IUDs, and 1.4% for implants. In our study, unintended pregnancies were reported for all five contraceptive methods, but the proportion of reports was highest for combination oral products, mostly due to missed doses, and lowest for levonorgestrel IUDs (Table 1). However, reports of unintended pregnancies associated with implants were higher than expected, due in part to being left in place beyond the recommended efficacy period. Real-world efficacy of these products may be weaker than women expect due to both the properties of the contraceptive itself as well as the actions or omissions of the patient or healthcare professional.
Psychiatric symptoms. While some studies have shown that hormonal contraceptives are associated with an increase in depression and other psychiatric side effects, others have reported no association or a positive overall effect on mental health. Nevertheless, in our large data set, psychiatric symptoms were 4 to 7 times more likely to be reported with contraceptives providing sustained doses of hormones compared to the non-hormonal copper IUD (Table 1). The most common symptoms included depression, anxiety, mood swings, loss of libido, and irritability. The strongest signals were seen for oral combination pills and levonorgestrel IUDs. However, we also found that psychiatric symptoms are reported substantially less frequently than many other adverse effects reported for contraceptive products.
Key Findings for Specific Methods
Levonorgestrel IUDs (MIRENA, SKYLA, LILETTA, KYLEENA). Levonorgestrel IUDs had the weakest safety profile of the five methods of contraception by several measures: It accounted for the largest number of serious injury reports, the largest number of direct reports to FDA, and the most cases of reported psychiatric symptoms, notably depression (Table 1). The large number of reports has persisted over many calendar quarters, which may be related to the serious warnings in the product labeling, the most prominent of which notes up to half of the pregnancies that occur may be ectopic. Another reason for the weakest safety profile is that the product category combines all the risks associated with insertion, dislocation, and expulsion of the device, along with the risks associated with hormonal contraceptives. However, levonorgestrel IUDs had the fewest reports associated with unintended pregnancies.
Copper IUDs (PARAGARD T 380A). This method, along with the levonorgestrel IUDs, shared a large number of device-related complaints, including expulsion and dislocation. There were also a large number of complaints that the device broke, became embedded in the uterus, or caused vaginal hemorrhage or muscle spasms necessitating removal. They do not appear to cause the psychiatric symptoms observed with the sustained hormonal methods (Table 1). Health professionals need to be aware of the need for expertise in safely inserting and removing the copper IUD, and women need to be aware of the risk of device expulsion.
Emergency contraception pills (nonprescription PLAN B ONE-STEP [levonorgestrel], prescription ELLA [ulipristal]). Emergency contraception pills had the best safety profile of the five groups, given the one-time rather than ongoing exposure to the products. The safety profile for this method was dominated by a large number of nonserious reports directly from consumers about abnormal menses, and a few reports of other adverse effects such as nausea and vomiting.
Etonogestrel implants (NEXPLANON). The safety profile for etonogestrel implants raised three safety concerns. First, this method differed from the others in the large number of complaints regarding implantation, including reports that it was difficult to use, that it broke, or concerns about product quality. Problems were even reported with the newer Nexplanon product, which replaced IMPLANON, despite a redesigned tool. This suggests the need for improved education and training of medical providers and the need to reevaluate the design of the insertion tool. Second, hundreds of reports indicated that the device had migrated, was dislocated, or caused a complication at removal. The third safety signal was apparent use beyond the 3-year recommended treatment period. Except for patient device identification cards provided in the implant packages, we were not able to identify any system to ensure that women were notified after 3 years of the need to remove or replace the implant. Studies are needed to more accurately assess the extent of the problem and identify cost-effective solutions.
Oral combination pills. The safety data for oral combination products highlight two prominent risks: unintended pregnancy through missed doses and serious injuries from pulmonary embolism and deep vein thrombosis. The adverse event data confirm these well-documented risks but do not identify new safety issues. By a wide margin, oral combination pills are the most commonly used method of pharmaceutical contraception while accounting for the smallest number of adverse event reports (Table 1).
The full QuarterWatch™ report with references can be found here.
What is QuarterWatch™ ?
QuarterWatch™ is an independent ISMP surveillance program that monitors adverse drug events reported to the FDA Adverse Event Reporting System (FAERS) by manufacturers, health professionals, and consumers. The goal is to identify signals that may represent important drug safety issues. The sheer number of case reports in these data have scientific weight. However, because of variation in reporting rates, they reveal little about how frequently events occur in the broader patient population, and do not prove that the suspect drug caused the event described—only that an observer suspected a relationship. Thus, identified safety issues often require further investigation to determine their frequency and establish a causal relationship to the suspect drug.