Abstract
Contraceptive Practices of Women with Epilepsy: Findings of the Epilepsy Birth Control Registry
Herzog AG, Mandle HB, Cahill KE, Fowler KM, Hauser WA, Davis AR. Epilepsia 2016;57(4):630–637. doi:10.1111/epi.13320
OBJECTIVE: To report the contraceptive practices of women with epilepsy (WWE) in the community, predictors of highly effective contraception use, and reasons WWE provide for the selection of a particular method. METHODS: These cross-sectional data come from the Epilepsy Birth Control Registry (EBCR) web-based survey regarding the contraceptive practices of 1,144 WWE in the community, ages 18–47 years. We report demographic, epilepsy, and antiepileptic drug (AED) characteristics as well as contraceptive use. We determined the frequency of use of highly effective contraception use, that is, methods with failure rate <10%/year, and conducted binary logistic regression analysis to determine predictors of highly effective contraception use. We report frequencies of WWE who consult various health care providers regarding the selection of a method and the reasons cited for selection. RESULTS: Of the 796 WWE at risk of unintended pregnancy, 69.7% use what is generally considered to be highly effective contraception (hormonal, intrauterine device [IUD], tubal, vasectomy). Efficacy in WWE, especially for the 46.6% who use hormonal contraception, remains to be proven. Significant predictors of highly effective contraception use are insurance (insured 71.6% vs. noninsured 56.0%), race/ethnicity (Caucasian 71.3% vs. minority 51.0%), and age (38–47, 77.5%; 28–37, 71.8%; 18–27, 67.0%). Of the 87.2% who have a neurologist, only 25.4% consult them regarding selection of a method, although AED interaction is cited as the top reason for selection. SIGNIFICANCE: The EBCR web-based survey is the first large-scale study of the contraceptive practices of WWE in the community. The findings suggest a need for the development of evidence-based guidelines that address the efficacy and safety of contraceptive methods in this special population, and for greater discourse between neurologists and WWE regarding contraception.
Differential Impact of Contraceptive Methods on Seizures Varies by Antiepileptic Drug Category: Findings of the Epilepsy Birth Control Registry
Herzog AG, Mandle HB, Cahill KE, Fowler KM, Hauser WA. Epilepsy Behav 2016;60:112–117. doi:10.1016/j.yebeh.2016.04.020.
PURPOSE: The aim of this study was to determine whether categories of contraception differ in their impact on seizures in women with epilepsy and whether the impact varies by antiepileptic drug category. METHODS: Retrospective survey data came from 2712 contraceptive experiences reported by 1144 women with epilepsy. We compared risk ratios for reports of increase and decrease in seizure frequency on hormonal versus nonhormonal contraception, stratified by antiepileptic drug categories. RESULTS: More women with epilepsy reported a change in seizures on hormonal (28.2%) than on nonhormonal contraception (9.7%) (p < 0.0001). The risk ratio for seizure increase on hormonal (18.7%) versus nonhormonal contraception (4.2%) was 4.47 (p < 0.0001). The risk ratio for seizure decrease on hormonal (9.5%) versus nonhormonal contraception (5.5%) was 1.71, p < 0.0001. On hormonal contraception, the risk ratio for seizure increase was greater than for decrease (1.98, p < 0.0001). In comparison to combined pills, both hormonal patch and progestin-only pills had greater risk ratios for seizure increase. Depomedroxyprogesterone was the only hormonal method with a greater risk ratio for seizure decrease than combined pills. Seizure increase was greater for hormonal than nonhormonal contraception for each antiepileptic drug category (p < 0.001). On hormonal contraception, relative to the non-enzyme-inducing antiepileptic drug category which had the lowest rate, each of the other categories had significantly greater risks for seizure increase, especially the enzyme-inhibiting (valproate) category (risk ratio = 2.53, p = 0.0002). CONCLUSION: The findings provide community-based, epidemiological survey evidence that contraceptive methods may differ in their impact on seizures and that this impact may vary by antiepileptic drug category.
Commentary
Effective contraception is important, as most women spend the majority of their reproductive years trying to avoid pregnancy. Among women with epilepsy, the use of effective contraception is even more critical as it enables healthcare providers to choose an antiepileptic drug (AED) with a lower teratogenic risk, provide folic acid supplementation, and optimize seizure control before a patient gets pregnant. Studies find that similar to the general population, 50% of women with epilepsy have unplanned pregnancies (1).
Multiple contraceptive methods are available; however, among women with epilepsy, the choice can be complicated. Hormonal contraception is the most common contraceptive method chosen by women, including women with epilepsy (2), but bidirectional interactions can impact the effectiveness of contraception (e.g., cytochrome P450 enzyme-inducing AEDs reduce the efficacy of hormonal contraception) or the effectiveness of some AEDs (e.g., estrogen-containing contraceptives reduce concentrations of lamotrigine and valproate) (3, 4). We need to better understand which contraceptive methods women with epilepsy are using and what factors influence that choice. In addition, improved understanding of the relationship between specific contraceptive methods, AEDs, and epilepsy will enable practitioners to advise women about effective methods that have minimal effect on either contraceptive efficacy or seizure control. Herzog and colleagues have aimed through the development of an online registry to address these questions.
The Epilepsy Birth Control Registry (EBCR) is a Web-based survey and educational site that collected data on the following: 1) contraceptive practices of women with epilepsy in the community, 2) the decision-making process, 3) the impact of various contraceptive practices on seizures, stratified by AED type, 4) frequencies and reasons for discontinuation of various contraceptive methods, 5) frequencies of unintended pregnancy on various contraceptive methods, stratified by AED type, and 6) rates of folic acid use and factors that determine its use. Subjects were directed to the registry from multiple sources including epilepsy organization web sites, social media, Internet searches, and study brochures posted in hospital and community clinics. All participants were women of reproductive age (between 18 and 47 years old) who self-reported a diagnosis of epilepsy. Eligible women completed a 40-question survey gathering demographics, seizure type (generalized convulsive, complex partial, simple partial), AED (no AED, enzyme-inducing AED, glucuronidated AED, non–enzyme-inducing AED, mixed categories), and reproductive history. Contraceptive methods were categorized into the following groups: 1) none, 2) withdrawal, 3) barrier (condom, diaphragm), 4) systemic hormonal (oral contraceptive pill—combined estrogen and progestin or progestin only, hormonal patch, vaginal ring, implanted progestin, depomedroxyprogesterone), 5) intrauterine device (IUD—progestin or copper), 6) tubal ligation, or 7) partner with vasectomy. All data was self-reported without collection of medical records for confirmation. Methods classified as highly effective (<10% per year failure rate with typical use) included systemic hormonal methods, IUD, tubal ligation, and vasectomy.
The process of choosing a contraceptive method was explored by asking which healthcare professionals were consulted regarding selection of contraceptive method and the top three reasons for contraceptive choice. The investigators generated a list of most frequently reported unstructured responses including 1) convenience, 2) AED interaction, 3) cost, 4) sexually transmitted infection prevention, 5) cycle regulation, 6) pelvic pain, 7) effectiveness, 8) side effects, and 9) other.
Investigators also addressed four specific questions regarding the relationship between contraceptive practices and seizures: 1) Is there a difference in reported seizure frequency between subjects on hormonal contraceptive methods and nonhormonal contraceptive methods? 2) Are there differences in seizure frequency among more specific categories of contraception and subcategories of hormonal contraception and IUDs? 3) Are the frequencies of reports of changes in seizure frequency on hormonal contraception and nonhormonal contraception affected by prescribed AED category? 4) What are the odds of seizure increase or decrease with the use of various combinations of contraceptive and AED categories?
Two published papers report the results from 1,144 women enrolled in the EBCR. In order to understand the findings and investigators’ conclusions, it is necessary to review the population studied as well as the statistical methods employed. As in any clinical observational study, generalizability and lack of bias are important in order to decide whether these reports should affect and influence how we treat women with epilepsy.
Most (87%) of the women enrolled lived within the United States. Demographic features are reported only for these women. Enrolled women were predominantly young, as more than 50% were under the age of 27, a finding reflective of young people being more likely to enroll in a Web-based registry. Over 90% were white, and over 90% were non-Hispanic. Over 80% had at least some college education, and almost 50% had at least an associate degree. Consistent with the age of the cohort, most individuals (>60%) had an annual household income of less than $60,000. More than 65% of registered subjects had commercial insurance. In sum, the enrolled group was predominantly white and non-Hispanic with at least some college education, and with commercial insurance. As stated previously, the method of ascertainment (i.e., Web-based) likely influenced who participated. Limited information is provided about minority women as well as for women with limited education and women with either no insurance or government-assisted insurance (Medicare or Medicaid). Surprisingly, only 7.7% of enrolled women were seizure free; most studies report that between 60 and 70 percent of persons with epilepsy have controlled seizures (5). Almost 60% of enrolled women reported having generalized convulsive seizures during the year before the survey. The lack of seizure control as well as the high percentage of women having generalized convulsive seizures also questions the generalizability of the findings and suggests that there may be bias, particularly when subjects are answering questions about the effects of a contraceptive method and the effects of changing a contraceptive method on seizure control.
In the sample 796 of 1,144 women were considered at risk for unplanned pregnancy (currently sexually active, not pregnant or seeking pregnancy and denied previous hysterectomy or infertility). Similar to other reports (2), in the EBCR, hormonal contraception is the most commonly reported method of contraception (47%) among these 796 women. Most women (70%) used contraceptive methods considered to be highly effective. Considering that these women predominantly have uncontrolled seizures, 30% using ineffective methods is problematic, as these women are at risk for unplanned pregnancy. Optimal care for women with epilepsy includes seizure control, particularly control of generalized convulsive seizures, prior to their getting pregnant. Not surprising, demographic predictors for effective contraceptive use included insurance (p = 0.003), race/ethnicity (p = 0.007), and age (p = 0.015), with more effective contraception being used by women with insurance, white women, and older women. Given that the studied cohort included predominantly white women with insurance, the use of effective contraception among a more representative cohort would likely be less. Overall, these data suggest that we as community need to do better.
The majority of women reported having a healthcare visit in the year prior to completing the survey. Of the women at risk for getting pregnant, almost half consulted with their gynecologist about contraceptive methods, whereas only 25% consulted with their neurologist/epileptologist. This finding contradicts what should be happening in clinical practice; as an American Academy of Neurology (AAN) measure (No. 6) instituted in 2014 outlines: all female patients of childbearing potential (aged 12 to 44 years) who have been diagnosed with epilepsy should be counseled or referred for counseling annually regarding the effects of epilepsy and its treatment on contraception or pregnancy (6). Even though the women were not consulting with their neurologists/epileptologists, the number one reason cited for their choice of contraceptive method was AED interaction. Given that more than half consulted with their gynecologist, this finding suggests that gynecologists may be more aware of the interactions between contraceptive methods and AEDs than prior older reports have suggested (7).
In the second cited paper, when asked whether the use of hormonal or nonhormonal contraceptive methods had either increased or decreased their seizure frequency or had caused no change in their seizure frequency, the majority of women stated “no change”: 72% of those using hormonal contraception and 90% of those using nonhormonal contraception. The investigators then performed multiple statistical comparisons with the groups who reported a change. A χ2 analysis done comparing proportion with change in seizure frequency between those using hormonal contraception and those using nonhormonal contraceptive methods revealed a significant difference (p < 0.0001). More women reported a change in seizure frequency (either an increase or a decrease) when using hormonal contraceptive methods compared with nonhormonal methods. The investigators frame an argument that hormonal contraceptive methods are more likely to cause an increase in seizure frequency by stating that the risk ratio for seizure increase to decrease was greater for those using hormonal contraception. Subanalysis revealed no change with barrier method and minimal change with withdrawal method. There is no clinically plausible explanation as to why either barrier methods or withdrawal would affect seizure control, and therefore this finding is not surprising. The bottom line is that women using hormonal contraception were more likely to report a change in seizure frequency. The multiple risk ratios presented in the second cited paper are confusing and do not add to our understanding of the effects of contraceptive methods and in particular hormonal contraception on seizure control. It must be emphasized that these data are subjective as the enrolled subjects were asked, “Do you think that this method of birth control changed how often you had seizures?” Additionally, causality cannot be assessed given the many potential confounding factors that cannot be controlled for in this analysis. We certainly know that some forms of hormonal contraception affect specific AED concentrations, mainly lamotrigine and, to a lesser extent, valproate. Standard seizure diaries and AED concentrations are not part of this study, and therefore we have no available quantitative or prospective data to better assess and understand the relationships between contraceptive methods and seizure control. In addition, as stated before, the majority of enrolled women had refractory seizures with a large percentage (almost 60%) having generalized convulsive seizure(s) in the year before enrollment; therefore, the findings may be biased as these women are likely looking for explanations for continued seizures. A randomized well-controlled study is needed to address these limitations and answer the question as to whether hormonal contraceptive methods affect seizure control.
The efficacy of contraceptive methods and the interactions between different contraceptive methods, seizures, and AEDs among women with epilepsy has received limited study. The EBCR adds to our understanding as it highlights that among the women with epilepsy studied, almost one-third were using ineffective contraceptive methods. Furthermore, only a minority of women consulted their neurologist/epileptologist when choosing a contraceptive method. Given that this is a predominantly well-educated and insured cohort, one would anticipate that the use of effective contraceptive methods and consultation with their treating neurologist/epileptologist would be even lower in the general population of women with epilepsy. This registry and these findings emphasize the need for neurologists/epileptologists to educate women with epilepsy about effective contraception as advised by the AAN.