OBJECTIVES: To evaluate use of a single-tablet (levonorgestrel 1.5 mg) emergency contraceptive administered to young females under simulated over-the-counter conditions. Secondary objectives were to assess repeat use, pregnancy, and adverse events.
METHODS: Females aged 11–17 years requesting emergency contraception at teen reproductive health clinics in five cities were eligible to participate. Participants read the study product label and determined whether and how to use the product without interacting with providers. Study product was dispensed to participants who appropriately selected to use it; participants were contacted 1, 4, and 8 weeks later to assess use, pregnancy, and adverse events. The incidences of outcomes were calculated and regression analysis was used to assess the effect of age and use status (ever used or no previous use) on primary outcomes.
RESULTS: Of the 345 females enrolled, 279 were younger than age 17 years. Among the 340 participants included in the selection analysis, 311 (91.5%) (97.5% confidence interval 87.5– 94.5%) participants appropriately selected to use or not use product. Among the 298 participants who used product, 274 (92.9%) (97.5% confidence interval 88.8–95.8%) correctly used it as labeled. Selection and correct use were not associated with age. Fifty-seven participants (18.8%) used additional emergency contraception over the study period and seven (2.3%) participants who used product became pregnant; there were no unusual adverse events.
CONCLUSION: Restricting young females' use of a single-tablet emergency contraceptive by prescription only is not warranted, because females younger than 17 years can use it in a manner consistent with over-the-counter access.
LEVEL OF EVIDENCE: II
Females know when use of emergency contraception is appropriate and can correctly use a single tablet of levonorgestrel 1.5 mg without provider assistance.
From the Women's Health Research Institute, Division of Research, Kaiser Permanente Northern California, Oakland, California; Teva Women's Health Research, Division of Teva Branded Pharmaceutical Products R&D Inc, North Wales, Pennsylvania; and the Bixby Center for Global Reproductive Health, Department of Obstetrics, Gynecology and Reproductive Sciences, University of California, San Francisco, San Francisco, California.
Supported by research grants from Teva Women's Health Research, Division of Teva Branded Pharmaceutical Products R&D, Inc.
Accepted for presentation at the Annual Clinical Meeting of the American College of Obstetricians and Gynecologists, May 5–9, 2012, San Diego, California.
Corresponding author: Tina R. Raine, MD, MPH, Women's Health Research Institute, Division of Research, Kaiser Permanente Northern California, 2101 Webster Street, 20th Floor, Oakland CA 94612; e-mail: email@example.com.
Financial Disclosure Teva Pharmaceuticals provided research grant support to the University of California San Francisco (UCSF), and this grant support was used by UCSF to support the salaries of the Investigators (Tina Raine and Cynthia Harper, as well as the research staff, including Abby Sokoloff and Beth Brown) for work on the study and research-related expenses. Tina Raine traveled to two meetings with the FDA in Maryland for this research project (in 2007 and 2010). Teva directly covered travel expenses (air and hotel) for the meetings. Nancy Ricciotti and Amy Hummel are both full-time salaried employees of Teva.
Despite the wide range of contraceptive methods currently available, the percentage of pregnancies that are unintended in the United States remains highest among females age 19 years and younger.1 Emergency contraception provides an additional effective option for preventing pregnancy. In 2009, Plan B One-Step, an emergency contraceptive containing levonorgestrel 1.5 mg in a single tablet, was granted approval by the U.S. Food and Drug Administration (FDA) to be marketed without a prescription for consumers aged 17 years and older. It remains restricted for use by prescription only for females aged 16 years and younger. The FDA indicated that additional data were needed demonstrating that females aged 17 years and younger understand the key concepts needed for safe and effective use. For a product that treats a condition to be sold over-the-counter, the FDA requires evaluation of consumers' ability to self-diagnose the condition and determine if treatment with the product is appropriate for them.
We conducted an over-the-counter simulation study to assess the ability of females age 11–17 years to appropriately select and correctly use single-tablet levonorgestrel 1.5 mg according to the package label. Study participants were not evaluated by a health care provider and each had to determine on her own by reading the label whether the product was appropriate for her to use and how to use it. It is indicated for prevention of pregnancy after unprotected sex or a known or suspected contraceptive failure. It should be taken within 3 days (72 hours) after unprotected sex; however, it is used off-label up to 5 days (120 hours). The findings of this study may be used to guide decision-making about over-the-counter use of emergency contraception by young females.
MATERIALS AND METHODS
Between October 22, 2008, and November 26, 2010, we conducted an over-the-counter simulation study in females requesting emergency contraception at teen reproductive health clinics in five metropolitan areas in the United States (San Francisco, Minneapolis, Atlanta, Pittsburgh, and Philadelphia). Study sites in Minneapolis, Atlanta, Pittsburgh, and Philadelphia started enrolling participants in 2010 and were restricted to enrolling participants aged 16 and younger because the study stopped enrolling 17-year-olds in 2009. When the study was initiated, the study product was an investigational drug and had an approval from the FDA Investigational New Drug program; in 2009, the FDA granted approval for Plan B One-Step (levonorgestrel tablet, 1.5 mg) to be marketed under dual-label packaging (by prescription and over-the-counter). The study protocol, informed consent, and data collection forms were approved by the central Institutional Review Board, Copernicus, and the University of California, San Francisco, or by the Institutional Review Board associated with the other study sites. The study was conducted according to the laws, regulations, and administrative provisions relating to the implementation of Good Clinical Practice as applicable by the United States 21 Code of Federal Regulations Part 11, 50, 54, 56, and 312, and in accordance with the Declaration of Helsinki and its updates. The FDA also reviewed the protocol, and a number of the agency's comments were incorporated into the study design and analysis plan.
Eligible participants were females age 11–17 years who presented to a study clinic requesting emergency contraception for their own use, had not participated in a previous levonorgestrel 1.5-mg emergency contraception study, were able to read and understand English, and agreed to complete study follow-up procedures. A key component of this study was that participants had to agree to receive and use the study emergency contraceptive after reading the information on the front and back panels of the product package without any assistance or guidance from study staff or health care providers regarding who should or should not use the study product, when to take it, and potential risks and side effects. Study staff approached potentially eligible females before they initiated their visit with a counselor or clinician. Study staff used a script to strictly control the amount of information that potential participants were provided about the study product. Females who met the inclusion criteria and were interested in participating in the study were consented and enrolled in the study. Parental consent was waived because minors can consent to receiving contraceptive services without parental consent in the states where the study was conducted. The consent form included general information about emergency contraception; however, detailed information about product indications, contraindications, and instructions for use were not included on the consent form. Participants were reimbursed $10 for enrolling in the study and $5 for completing each follow-up.
Enrolled participants completed a self-administered questionnaire that asked them to indicate in writing their reason for requesting emergency contraception. Participants were then given an empty sealed over-the-counter prototype package of levonorgestrel tablet 1.5 mg and were asked to read the information provided on the front and back panels of the package. Participants were not permitted to open the box and access the inner card or consumer information leaflet because they would not have access to this in an over-the-counter environment. Participants were then given another self-administered questionnaire that asked them to indicate if they wanted to use the study product. If participants indicated they did not want to use the product, then they were asked to record their reasons. Participants were then asked to complete a self-administered baseline questionnaire to obtain limited medical history including allergies to medicines, age at first menses, age at first sex, type of birth control ever used, pregnancy history, current pregnancy status (eg, results of pregnancy tests performed by the participant or by a health care provider and whether the participant was already pregnant), and whether she had ever received or used emergency contraception in the past.
Study staff reviewed the verbatim responses on the two selection questionnaires and the baseline questionnaire to determine if participants appropriately selected to use or not use the study product. Because participants were expected to determine if treatment with the product is appropriate for them without provider assistance or guidance, study staff were not permitted to question participants' recorded information on the questionnaires or ask any additional probing questions to clarify a participant's responses. Research staff determined if participants appropriately selected to use or not use the product according to indications on the package label (Box 1). Participants who appropriately selected to use the product were provided with one package containing the study product for a single use. Participants did not pay for the study drug because emergency contraception is generally provided at no cost in the study clinics. Participants who appropriately selected not to use or inappropriately selected to use the product were not dispensed study product and were referred to clinic staff for further assessment and treatment according to the clinic's standard procedures. Participants who opted out of using the product because they wanted to talk to a provider or for other reasons were also not dispensed study product and were referred to clinic staff. No information regarding use was communicated to participants by study staff.
After completion of study enrollment procedures, participants who were dispensed study product were permitted to receive other clinic services (ie, pregnancy test, sexually transmitted diseases testing, birth control) per routine clinic procedures. To ensure the integrity of the study and to avoid introducing bias, clinic staff were advised not to discuss emergency contraception with participants who received the study product. At the conclusion of the participant's clinic visit, study staff confirmed with clinic staff that emergency contraceptive counseling was not provided as part of the participant's clinic services.
Study staff assessed participants by telephone or return clinic visit at approximately 1, 4, and 8 weeks after the date each participant was dispensed the study product. Follow-up questionnaires included information on study product use including date and time of use and reason for use. Safety information was collected on side effects and any medical problems. Information was also collected on use of additional emergency contraception other than the dispensed study product, any pregnancy tests performed, and pregnancies that occurred. Participants with pregnancies were followed-up until the resolution of the pregnancy. At the conclusion of the 8-week follow-up period, we conducted medical chart reviews of study participants at the clinical sites for emergency contraceptive use (other than the study product) and pregnancy tests performed on the day of study enrollment and over the 8-week follow-up period.
The primary outcomes for this study were the proportion of enrolled participants who appropriately selected to use or not use the study product based on their comprehension of the product label and the proportion of participants who were dispensed product, and who then correctly used it according to product labeling. Participants were considered to have appropriately selected to use the product if they indicated in their verbatim written response that they were requesting emergency contraception for a reason for which the product is indicated (Box 1) and if they did not indicate on their baseline questionnaire that they had an allergy to levonorgestrel, a positive pregnancy test within the previous 30 days, or that they were already pregnant. Participants were considered to have appropriately selected not to use the product if they selected not to use the product because it was not indicated based on their comprehension of the product label. Participants who opted out of using the product because after reading the product label they wanted to talk to a provider or for other reason were excluded from the appropriate selection analysis.
Correct use was considered to have occurred if the participant used the product according to the product label (for unprotected sex, not already pregnant, and less than 72 hours after unprotected sex). Time elapsed between unprotected sex and product use was calculated as the difference between the reported calendar date and time of unprotected sex and the reported calendar date and time of use of the study product. Because the package label indicated the product should be used “no later than 3 days (72 hours)” after unprotected sex, we also calculated time elapsed between unprotected sex and product use in days as the difference between the reported calendar date of unprotected sex and the reported calendar date of use of the study product. To avoid participant recall bias, information on date and time of study product use was only collected from participants who completed the 1-week follow-up. Participants who reported use of the study product after the 1-week follow-up were included in the safety analysis but were excluded from the correct use analysis.
Secondary outcomes of correct product use calculated using the alternative criteria of using the tablet no later than 120 hours after unprotected sex and the incidence of repeat use of emergency contraception (use of additional emergency contraception other than the study product) within the 8-week study period were also assessed. Information on pregnancies that occurred was collected; however, data on traditional efficacy measures was not collected.
The sample size was calculated to estimate the proportion of participants who will use the drug incorrectly. In the Raymond et al2 actual use study, incorrect use was 28% by strict definition (of using the two doses 12 hours apart). We assumed the proportion of incorrect use of 10% for our study, because it is a one-dose rather than two-dose regimen and thus easier to use correctly. With 97.5% confidence limits, rather than the traditional 95% limits, because the Bonferroni adjustment was applied for multiple (confidence interval [CI]) estimations in the study (selection and correct use), calculations indicated that 181 participants who used levonorgestrel tablet 1.5 mg were needed. It was expected that there would be a small number of participants who would correctly select not to use the study product (ie, assess that they were pregnant or unprotected sex more than 72 hours before request for emergency contraception) or inappropriately select not to use the study product and, thus, would not be dispensed study product. To ensure diversity of the study populations and to allow for subgroup analyses (ie, analysis stratified by ever-used emergency contraception compared with no previous use) as well as study attrition, target enrollment was set to at least 350 participants.
At the request of the FDA, the original protocol incorporated a goal to enroll a minimum of 25 participants in each of the 14-year-old to 17-year-old groups with an additional minimum of 25 participants of any age from 11 to 13 years. However, based on study experience over the course of a 2-year period and information from published literature showing that very low numbers of females aged 11–13 years present to reproductive health clinics requesting emergency contraception, the FDA agreed to eliminate the requirement for a minimum of 25 participants in the 11-year to 13-year age group. The incidences of appropriate and inappropriate selection were calculated in subgroups by age. The Wilcoxon rank-sum test was used to compare age with the primary outcome because age is not normally distributed. Multiple logistic regression analysis was performed to determine the relationship between selection and correct use and use status (ever use and no previous use) when adjusting for age. The CIs were computed using exact method. Analyses were performed using Stata 11.0.
A total of 379 females were screened for the study. The majority of the 34 participants who did not meet the inclusion criteria were outside the age requirement, preferred to talk to a provider, or were unwilling to comply with study follow-up procedures (Fig. 1). Of the 345 participants who met the inclusion criteria for enrollment and agreed to be in the study, 279 (80.9%) were younger than 17 years old. Three participants aged 13 years were enrolled over a 2-year study period; 35 participants aged 14 years, 100 aged 15 years, 141 aged 16 years, and 66 aged 17 years were enrolled. No participants younger than age 13 years were screened or enrolled. The majority of the participants (92.2%) were from clinic sites in the San Francisco Bay Area; the study population was ethnically and racially diverse (Table 1). The mean age of coitarche was 14.5 years, with teens reporting having ever used condoms (59.3%), birth control pills (30.5%), no method (29.5%), and withdrawal (22.3%) in the past (Table 1). More than one third (39.4%) of the participants had ever used emergency contraception in the past; most reported getting it from a clinic, although 1 in 10 reported getting it from a pharmacy.
Of the 345 females enrolled in the study, five opted out of using the study drug after reading the product label. Three hundred eleven females (91.5%) (97.5% CI 87.5–94.5%) appropriately selected to use (n=308) or not use (n=3) the product (Table 2). Within individual age groups, appropriate selection was highest among 15-year-olds (92.9%); however, the differences by age were not significant (P=.78 Wilcoxon rank-sum). Among participants who had no previous use of emergency contraception (n=169), 87.8% (97.5% CI 81.7–92.4%) appropriately selected to use the study product. Regression analysis of selection revealed that females who had no previous use of emergency contraception were less likely to appropriately select to use the product than those who had ever used emergency contraception (odds ratio 0.20) (97.5% CI 0.06–0.73; P=.005). Age was not associated with selection.
The most frequently cited reasons for requesting emergency contraception were that participants had sex but a condom was not used, broke, or fell off (54%); the participant had sex but did not use birth control or was not sure (40%); and that the participant did not want to get pregnant (37%). The main reason for inappropriate selection was indicating that they thought they either were or suspected they were pregnant. Only one participant, a 16-year-old, indicated she wanted to use emergency contraception in place of regular birth control, which is contrary to package labeling. Three participants appropriately selected not to use the study product according to the package labeling because their unprotected sex had occurred more than 3 days or 72 hours previously.
Study product was dispensed to 308 participants who appropriately selected to use the product. Of these, 298 (96.7%) reported having used the study product by the 1-week follow-up. Among the participants who reported date and time of use of use (n=295), 274 (92.9%) correctly used it less than 72 hours after unprotected sex (97.5% CI 88.8–95.8%). Among participants who had no previous use of emergency contraception (n=169), 158 (93.5%) (97.5% CI 87.9–93.7%) correctly used it less than 72 hours after unprotected sex. Neither age nor previous use of emergency contraception was associated with correct use in the regression analysis. Calculating elapsed time using days, 284 (95.6%) (97.5% CI 92.2–97.9%) of those who reported date of use (n=297) used it no more than 3 days after unprotected sex. For the small proportion of participants who used the study product more than 72 hours after unprotected sex, the range was 73.5 to 118 hours between product use and the unprotected sex.
Approximately one in five (18.8%) participants who used the study product reported additional use of emergency contraception (product other than the study drug) within the 8-week follow-up period. The proportion of females reporting additional emergency contraceptive use (one or more times) did not differ by age (P=.78) Chart review data were available for 23 of the 29 participants who inappropriately selected to use the product and were not dispensed study product by the research staff; 20 of these 23 participants received emergency contraception from clinic staff on that day.
Emergency contraceptive use was not associated with any unusual pattern of adverse events in this study population and was consistent with the current FDA-approved labeling for levonorgestrel tablet 1.5 mg. Of the 299 participants included in the safety analysis, 45 participants (15%) reported at least one adverse event. In all, a total of 68 adverse events were reported, with the most frequent being headache (3%), nausea (3%), menstrual irregularity (2%), pelvic pain (3%), vaginal bleeding (3%), and vaginal spotting (1%). No unexpected serious adverse events were reported. Seven (2.3%) pregnancies were reported among the 303 who were known to have used the study drug. All pregnancies occurred in participants who correctly used the study product no more than 72 hours after unprotected sex (range 25–45 hours).
Currently, Plan B One-Step (levonorgestrel 1.5 mg), a single-tablet regimen, is approved for purchase without a prescription (behind-the-counter) for consumers ages 17 years and older; it is restricted for use by prescription only for females younger than 17. By approving to market it with a “dual label,” the FDA maintains the precedent that was established in 2006 when the FDA first granted approval for the two-tablet regimen for levonorgestrel 0.75 mg under a dual label, restricting access without a prescription to women 18 years and older. The FDA loosened the restriction to allow nonprescription use for 17-year-olds in 2009 after a federal court order.3 This over-the-counter simulation study, in which the majority of the participants were younger than 17 years of age, demonstrates that young females seeking emergency contraception can appropriately select and correctly use a single-tablet emergency contraceptive, levonorgestrel 1.5 mg, without provider assistance.
Our study population differed from that of the simulated over-the-counter study of the two-tablet regimen for levonorgestrel 0.75 mg conducted by Raymond et al2 in 2001 in which the median age of participants was 21 years and only 29 (5%) participants were younger than 17 years of age. In addition, our study incorporated a study design that allowed the investigators to assess whether study participants' reasons for wanting emergency contraception were appropriate before the product was dispensed. This design enabled us to standardize the protocol across sites, minimize the risk of inadvertently giving participants information on emergency contraception, and insure that those who needed assistance from clinic staff were appropriately referred. We confirmed that only one participant received information on emergency contraception from clinic staff; she is included in this intent-to-treat analysis. Our results were probably influenced by this design, which may have underestimated appropriate selection because research staff were instructed to not probe participants with vague or unclear written responses to minimize potential of inadvertently providing information to participants. A high proportion of participants who were deemed inappropriate for selection by study criteria and were not dispensed the study product by study staff and were instead referred to a clinic provider did receive emergency contraception from a clinic provider that day. This suggests that some participants may not have been able to adequately articulate their reasons for wanting to use emergency contraception in a written questionnaire but use was actually appropriate.
Correct use was also very high at 92.9%, and all participants who used the product more than 72 hours after unprotected sex used it within 120 hours. It is possible that some participants were aware that some providers recommend off-label use to 120 hours. Repeat use of emergency contraception in the follow-up period was limited and consistent with findings of previous studies of females and women aged 15–19 years.4 No unusual adverse events were observed among participants who used the product. The safety data in this study are consistent with previous research in adolescents aged 13–16 years demonstrating similar pharmacokinetics and side effect profiles as adult populations.5,6 The efficacy of levonorgestrel tablet 1.5 mg in preventing pregnancy was not assessed in this study because the efficacy of emergency contraception is well-established; however, the low pregnancy rate (2.3%) is comparable to that of previous studies of levonorgestrel emergency contraception.7–10
There are limitations in interpreting the findings of this study. The study population was racially diverse and recruited primarily from urban and suburban public family planning clinics in the San Francisco Bay Area. Although one would not expect reading or cognitive abilities to be higher in this population than in private clinic settings, results may not be generalizable to females in different settings. Of note, 12- to 17-year-olds recruited from a wide range of settings including shopping malls and family planning clinics demonstrated good comprehension of the prototype over-the-counter package label in the label-comprehension study that was used to inform the labeling of the study product.11 Although there was no difference in correct use in females who had no previous use of emergency contraception compared with those who had ever used it, females who had no previous use were less likely to appropriately select to use the study product, suggesting that individuals do learn about the method when they use it, as we might expect to see. However, appropriate selection among naïve users was high.
It was anticipated that many more sites would have participated in this study; however, recruitment of clinical sites for this study posed several challenges. To protect privacy, only sites whose Institutional Review Board and state law permitted minors to provide consent for confidential reproductive health care including receipt of emergency contraception were considered for participation. In addition, because of the requirement for clinical sites to have a robust volume of sexually active young teens and research capacity, conducting the study in clinics in metropolitan centers with some university affiliation was more feasible, also limiting the study population somewhat.
It is possible that conducting the study in clinics with participants who had self-selected to go to a clinic to get emergency contraception may have biased the results. Females who do not seek such care may be different and results may not be generalizable to them. Although not ideal, family planning clinics are the best place to find significant numbers of potential users of emergency contraception. It would not have been logistically feasible or ethical to advertise or shop around in malls for females who plan to or might have unprotected sex in the future, as is typically performed for products for medical conditions that are more chronic such as H2 blockers for those with heartburn. It should also be noted that the ability to appropriately select is not the same as a decision to purchase the product. In the clinic study sites, patients generally do not have to pay for emergency contraception, and thus the study did not require them to pay for it. Teenagers may select to use the product, but the retail price may discourage them from purchasing the product.
Enrollment was dependent solely on the number of females who spontaneously presented to the respective teen clinic requesting emergency contraception. No 11- or 12-year-olds were screened for or enrolled in the study. Considering that approximately 3% of teens initiate sexual activity before age 13 years, and that only a fraction (fewer than 3%) of visits for any reason to the study clinics during the study period were by teens 13 years and younger, the ages of those who ultimately participated in this study represent the population at risk who might need emergency contraception.12 Despite these limitations, the results of this study provide a high level of support for the ability of females younger than age 17 years to appropriately select and correctly use a single tablet of levonorgestrel 1.5 mg in a manner consistent with over-the-counter access.
1. Finer LB, Zolna MR. Unintended pregnancy in the United States: incidence and disparities, 2006. Contraception 2011;84:478–85.
2. Raymond EG, Chen PL, Dalebout SM. “Actual use” study of emergency contraceptive pills provided in a simulated over-the-counter manner. Obstet Gynecol 2003;102:17–23.
4. Harper CC, Cheong M, Rocca CH, Darney PD, Raine TR. The effect of increased access to emergency contraception among young adolescents. Obstet Gynecol 2005;106:483–91.
5. Sambol NC, Harper CC, Kim L, Liu CY, Darney P, Raine TR. Pharmacokinetics of single-dose levonorgestrel in adolescents. Contraception 2006;74:104–9.
6. Harper CC, Rocca CH, Darney PD, von Hertzen H, Raine TR. Tolerability of levonorgestrel emergency contraception in adolescents. Am J Obstet Gynecol 2004;191:1158–63.
7. Arowojolu AO, Okewole IA, Adekunle AO. Comparative evaluation of the effectiveness and safety of two regimens of levonorgestrel for emergency contraception in Nigerians. Contraception 2002;66:269–73.
8. von Hertzen H, Piaggio G, Ding J, Chen J, Song S, Bartfai G, et al.. Low dose mifepristone and two regimens of levonorgestrel for emergency contraception: a WHO multicentre randomised trial. Lancet 2002;360:1803–10.
9. Task Force on Postovulatory Methods of Fertility Regulation. Randomised controlled trial of levonorgestrel versus the Yuzpe regimen of combined oral contraceptives for emergency contraception. Lancet 1998;352:428–33.
10. Ho PC, Kwan MS. A prospective randomized comparison of levonorgestrel with the Yuzpe regimen in post-coital contraception. Hum Reprod 1993;8:389–92.
11. Raymond EG, L'Engle KL, Tolley EE, Ricciotti N, Arnold MV, Park S. Comprehension of a prototype emergency contraception package label by female adolescents. Contraception 2009;79:199–205.
© 2012 The American College of Obstetricians and Gynecologists
12. Centers for Disease Control and Prevention. Youth Risk Behavior Surveillance—United States, 2009 Surveillance Summaries. MMWR Morb Mortal Wkly Rep 2010;59:98.