Initiation of Oral Contraceptives—Start Now!
Article Outline
Abstract
Purpose
Conventional practice for initiating oral contraceptive (OC) pills involves waiting to start the pills with the next menstrual period. We investigated whether immediate initiation of OCs would lead to improved continuation rates and therefore decreased pregnancy rates in adolescents aged 12–17 years.
Methods
Study subjects were recruited from adolescent women presenting to 2 inner city clinics requesting OCs. A total of 539 adolescents between 12 and 17 years old were randomized to conventional initiation of the OC pill (Conventional Start [CS]) versus immediate, directly observed OC pill ingestion in the clinic (Quick Start [QS]). At 3 and 6 months the participants completed interviews that questioned them about their OC continuation and pregnancies.
Results
In all, 86% of our adolescents completed follow-up interviews at 3 months, and 77% at 6 months. There were 45 pregnancies during the study period. QS was associated with continuing OCs to a second pack (adjusted OR 1.8, 95% CI 1.1–3.3). There was no difference in OC continuation rates at 3 or 6 months. Only 26% of adolescents continued OCs at 6 months and we identified 45 pregnancies during follow-up.
Conclusion
We conclude that directly observed, immediate initiation of oral contraceptives (QS) with adolescents briefly improves continuation although overall continuation rates are discouraging low. Health care providers could use this simple strategy to start adolescents on OCs at the initial visit. The low 6-month OC continuation rates highlight the need to seek novel ways to provide adolescents with the necessary tools to be successful at contraception.
Keywords: Oral contraceptives, Adolescents, Quick Start, Contraceptive initiation
See Editorial p. 419
Oral contraceptive (OC) pills are the most popular hormonal method chosen by adolescents [1], and are extremely effective in preventing pregnancy. Conventional OC initiation typically requires waiting for the next period to begin. Many young women who receive OC prescriptions or actual packs of OCs never begin this method [2]. Moreover, many young women who intend to start conventionally may become pregnant while waiting to start. Young adolescents may also become confused about the starting instructions and/or may lose motivation to begin contraception [3].
Substantial evidence shows that oral contraceptives are not teratogenic or harmful to a developing fetus [4]; however, many providers ask women to wait for their menses before beginning OCs to avoid a scientifically unfounded risk for possible pregnancy. As many as 24% of young adolescents prescribed OCs and told to take their OCs using the Conventional Start (CS) method never take the first pills [5]. Quick Start (QS) is a novel method of OC initiation designed to overcome barriers to contraception [6]. Using QS, the woman who is an appropriate candidate for OCs takes the first pill directly observed in the office irrespective of her menstrual cycle day after a negative pregnancy test. Simultaneously, emergency contraception is also provided if needed. QS eliminates complicated starting instructions and offers immediate gratification to adolescents who are seeking effective contraception, as there is no wait time required. In addition, there is no increased intermenstrual bleeding in women who start OCs with the QS method as compared with the CS method [7].
We investigated the QS method in a randomized clinical trial. We hypothesized that the QS method would lead to improved continuation rates in this population as well as a decrease in pregnancies when compared with adolescents using CS of the OC. This analysis focuses on adolescents.
Methods
This randomized clinical trial was conducted between March 2003 and February 2005 in young women <25 years of age in urban clinics in New York City, Atlanta, Georgia, and Dallas, Texas. This analytic sample examines OC continuation rates only in those subjects <18 years of age recruited from two adolescent medicine clinics in New York City and Atlanta, Georgia. Because the Texas site enrolled few subjects aged <18 years of age, that site is not included in the present analysis. All clinical sites are all partially federally funded under Title X. The Institutional Review Boards in New York City and Georgia approved this study. This study was monitored by the coordinating center, Columbia University, as well as by an external Data Safety and Monitoring Board.
After receiving care from a primary care provider, the adolescents were referred to study staff for screening to determine eligibility for the study. Eligibility criteria included: (1) a current negative pregnancy test, (2) no use of an OC within 7 days, (3) no use of Depo-Provera (DMPA) within the past 6 months, (4) not seeking pregnancy within the next 6months, and (5) no lactational amenorrhea. After informed consent, participants were randomly assigned to QS or CS of the OC. QS consisted of directly observed ingestion of the first OC while still in the clinic, regardless of menstrual cycle day and with a negative pregnancy test. All patients randomized to receive CS consisted of taking the first OC at the onset of the next menses. All subjects received standardized pill taking instructions and at least 3 months supply of oral contraceptive pills at that initial visit. All clinics counseled their patients to use back-up methods for 1 week (until the presumed onset of OC effectiveness) and to always use condoms for protection against sexually transmitted infections.
Telephone interviews took place at 3 and 6 months to determine OC continuation, interval pregnancy, sexual activity, and possible adverse reactions. Using records that included sonography from obstetric visits and or abortion clinics we were able to estimate the timing of onset of all study pregnancies. Further details of study methods are presented elsewhere [7].
Our main hypothesis was that QS of the OC in a young population would lead to improved continuation rates and a lower rate of pregnancy as compared with CS. We used univariate and multivariate logistic regression to estimate odds ratios (ORs) and confidence intervals (CIs). We used SPSS version 13.0 software (SPSS Inc., Chicago, IL) to analyze the data. The sample size for the main study was selected to have >90% power to detect a 10% difference in OC continuation between the intervention groups. In our analysis we had 80% power to identify a 12% difference (two-sided) in OC continuation, assuming a 30% continuation rate in the control group (i.e., the group using the CS method).
Results
We enrolled 539 adolescents, 12–17 years of age; we obtained 3- and 6-month follow-up information from 86% and 77%, respectively. Although loss to follow-up was slightly greater among our youngest adolescents, the overall loss to follow-up was equal in the QS and CS groups. The QS and CS subjects had similar baseline characteristics (Table 1). Subjects were mainly African-American and Latina and were economically poor, with 90% either receiving or eligible for Medicaid. The majority had never been pregnant, and only 22.4% had previous experience using the OC.
Table 1. Baseline characteristics of study participants in Conventional Start (CS) and Quick Start (QS) groups
| Characteristic | Total enrolled, n (%) | CS, n (%) | QS, n (%) |
|---|---|---|---|
| Total | 539 (100%) | 267(100%) | 272(100%) |
| Site | |||
| Emory, Atlanta | 210(39.0%) | 104(39.0%) | 106(39.0%) |
| Mount Sinai, New York | 329(61.0%) | 163(61.0%) | 166(61.0%) |
| Age (y) | 15.9 ± 1.0 | 16.0 ± 1.0 | 15.9 ± 1.0 |
| Race | |||
| African-American | 314(58.3%) | 155(58.1%) | 159(58.5%) |
| Latina | 181(33.6%) | 91(34.1%) | 90(33.1%) |
| Other | 44(8.2%) | 21(7.9%) | 23(8.5%) |
| Education | |||
| High school grad/GED | 16(3.0%) | 7(2.6%) | 9(3.3%) |
| Previous OC use (%) | 121(22.4%) | 59(22.1%) | 62(22.8%) |
| Previous pregnancies (%) | 85(16.1%) | 45(17.2%) | 40(15.0%) |
| Pregnancies unplanned (%) | 89.4% ± 30.0% | 88.9% ± 30.0 | 90.0% ± 30.4% |
| Feelings about pregnancy in next 6 months (n = 534) | |||
| Very unhappy | 235(44.0%) | 112(42.6%) | 123(45.4%) |
| Somewhat unhappy | 175(32.8%) | 79(30.0%) | 96(35.4%) |
| Plan to use pills >6 months (n = 530) | 494(93.2%) | 245(93.5%) | 249(92.9%) |
| Certain about wanting to use OCs | |||
| Very sure | 351(65.1%) | 173(64.8%) | 178(65.4%) |
| Not sure | 188(34.9%) | 94(35.2%) | 94(34.6%) |
| Partner knows of plan to use OCs 331 | (61.5%) | 170(63.7%) | 161(59.4%) |
Based on follow-up information at 3 months, 86% of subjects completed the first pack of pills and started a second pack of pills. Only 44% completed at least 3 months of the OC, and 26% were still continuing users of the OC at the final 6-month interview (Table 2). Subjects randomized to QS were more likely to continue to a second pack of pills (adjusted OR 1.8, 95% CI 1.1–3.3). At 3 and 6 months, however, the groups had similar continuation rates (OR 1.0, 95% CI .7–1.6 and 1.1, 95% CI .7–1.8). In a multivariate analysis incorporating variables previously found to influence continuation, planning to use the OC for at least 6 months, partner knowledge of OC use, and certainty about wanting to use the OC remained associated with continuation of the OC to the second pack and at 3 and 6 months (Table 3). Age, by year, was not associated with continuation rates (OR 1.1, 95% CI .8–1.5). Previous OC use was not associated with OC continuation rates after adjustment for age.
Table 2. Distribution of oral contraceptives (OCs) continuation to second pack, at 3 months and at 6 months
| Continuation | Total enrolled, n (%) |
|---|---|
| Total | 539(100%) |
| Second pack (n = 462) | |
| One pack or more | 397(85.9%) |
| Less then one pack | 65(14.1%) |
| 3 Months (n = 462) | |
| Yes | 203(43.9%) |
| No | 259(56.1%) |
| 6 Months (n = 414) | |
| Yes | 109(26.3%) |
| No | 305(73.7%) |
Table 3. Baseline characteristics associated with oral contraceptive pill (OC) continuation to second pack, at 3 months and at 6 monthsa
| Baseline characteristics | 2nd pack OR (95% CI) | 3 Month OR (95% CI) | 6 Month OR (95% CI) |
|---|---|---|---|
| Quick Starta | 1.8(1.1,3.3) | 1.0(0.7,1.6) | 1.1(0.7,1.8) |
| Plan to use pills more than 6 month | 3.2(1.3,7.4) | 3.1(1.2,8.1) | 4.6(1.0,20.1) |
| Partner knows of plan to use OCs | 2.1(1.2,3.7) | 1.8(1.2,2.7) | 1.1(0.7,1.9) |
aAs compared to participants in CS. All OR were adjusted for age, site, access to OC, and certainty about wanting to use the OC. |
We identified 45 pregnancies during follow-up (8.3% of subjects). All pregnancies occurred among subjects who discontinued the OC for 7 days or longer. A total of 17 pregnancies occurred among the 272 subjects in the QS group, whereas 28 pregnancies occurred among the 267 subjects in the CS group (6.5% vs. 10.5%, p = .08).
Discussion
We studied the QS intervention, designed to facilitate initiation of oral contraception, with the aim to assess whether this would improve continuation and decrease pregnancy. Strengths of this analysis are the large number of adolescent OC users enrolled and the high rate of follow-up. Our enrollment of young, economically poor, minority women attending inner-city family planning clinics may not be generalizable to other populations. However, given the high rates and serious consequences of teen pregnancy in these young women [8], our results may have most significance in just this population. QS of the OC had only a brief impact on continuation; however, the trend toward a deficit of pregnancies in the QS group (6.5% vs. 10.5%) suggests there may be a key benefit to even this brief effect. Although this analysis did not have sufficient statistical power to compare pregnancy rates with precision, the magnitude of the observed difference seems to have clinical significance. In the larger QS clinical trial, including women up to age 24, there was also a brief increase in continuation after QS, but a far smaller difference in pregnancies [9]. Thus the possible pregnancy benefit that we observed may be limited to adolescents. Rickert et al [10], in a study of the immediate initiation of DMPA among young women, also found little difference in method continuation but found a substantial reduction in pregnancy in the intervention group.
The immediate initiation of OC may also decrease clinic time, as the need for lengthy counseling on how to start OC is eliminated [11]. This additional time could be spent discussing the myriad of issues common to adolescents. In this study as well as others, the adolescents preferred the QS approach rather than waiting for their period to begin OC (CS) [12], [13]. In QS, the health care provider gives OC packs to the patient and the patient is directly observed taking the first pill after a negative pregnancy test. Consequently, we cannot extrapolate our findings to health care settings in which the adolescent receives a prescription and has to go the pharmacy.
A key result here is the dismal OC continuation rate. Only 26% of participants continued use at 6 months, even though they received their care in clinics specifically targeted to adolescents. Other published rates of continuation for adolescent OC users are 52% at 12 months with 30% pregnant [14] 33% at 12–15 months with 38% pregnant [5]. Across these different adolescent populations, OC continuation rates are universally poor. Factors cited in a previous study for inconsistent contraception use include seeking pregnancy, a history of inconsistent contraceptive use, and less partner discussion about pregnancy prevention [15]. In another study, OC continuation rates were higher in adolescents who made the appointment themselves and came in specifically for birth control [16]. In our study, we also did not limit our participants to those who came in specifically requesting OCs with an appointment and more than 50% of our adolescents did not have scheduled appointments and were seen as “walk-in” visits. Furthermore, only ∼25% of our study population came to the visits specifically requesting OCs. In contrast to adolescents who select OCs, adolescents who choose long-term contraception such as DMPA or implants have significantly lower pregnancy rates than OC users [14]. Further research that examines the acceptability of long-term methods in adolescents is needed. Despite these discouraging and consistent findings, OCs are most often chosen by adolescents as their hormonal contraceptive method [1]. As providers and researchers, we need to provide teens with the methods they will be most successful with, which are often the longer-term methods. These findings suggest that in counseling we need to have a dialogue with adolescent patients about strategies that might help them to continue with their chosen method, as well as informing them about the benefits of long acting reversible contraceptive methods.
Although QS had only a brief impact on OC continuation rates, it does allow streamlining of care with users able to initiate OCs on the day they are seen in the clinic. Waiting for a woman's next menses before starting OCs delays protection and can lead to confusion with instructions. Women in this study and others preferred immediate initiation of their method [1], [6], [9].
Other factors important in successful continuation with OCs were plan at baseline to use OCs for 6 months, certainty about wanting to use OCs and partner knowledge. In the larger study the subgroup of women who were ambivalent about using the OC had both a continuation and a pregnancy benefit of QS. It was disappointing to have such low continuation rates in specialized teen clinics that focus on reducing access barriers for this population. We really do not know how to use the results from this study to improve counseling for teens; clearly, additional evaluation of counseling interventions is needed. The QS approach is a simple and acceptable way to achieve brief improvement in OC continuation; however, the overall poor continuation rates seen here and elsewhere point to the importance of also having long-acting contraceptive methods available for teens.
Acknowledgments
This work was supported by National Institutes of Health grant R01 HD 42413.
References
- . Advances in hormonal contraception. Adolesc Med. 2006;17:653–671
- . Oral contraceptive use after an initial visit to a family planning clinic. Fam Plann Perspect. 1991;23:150–154
- Immediate recall of oral contraceptive instructions: Implications for providers. Am J Obstet Gynecol. 1999;180:1399–1406
- . Oral contraception and congenital malformations in offspring: A review and meta-analysis of prospective studies. Obstet Gynecol. 1990;76:552–557
- The use of levonorgestrel implants (Norplant) for contraception in adolescent mothers. N Engl J Med. 1994;331:1201–1206
- Quick Start: A novel oral contraceptive initiation method. Contraception. 2002;79:322–329
- Bleeding patterns after immediate vs conventional oral contraceptive initiation: A randomized controlled trial. Fertil Steril. 2003;79:322–329
- Sexually transmitted diseases, unintended pregnancies, and adolescent health promotion. Adolesc Med. 1999;10:87–108
- Initiation of oral contraceptives using a quick start compared with a conventional start. Obstet Gynecol. 2007;109:1270–1276
- Depo now: Preventing unintended pregnancies among adolescent young adults. J Adolesc Health. 2007;40:22–28
- . Immediate recall of oral contraceptive instructions: Implications for providers. Am J Obstet Gynecol. 1999;180:1399–1406
- . A prospective study of immediate initiation of depo medroxyprogesterone acetate contraceptive injection. Contraception. 2005;71:99–103
- . Acceptability and satisfaction using Quick Start with the contraceptive ring versus an oral contraceptive. Contraception. 2006;73:488–492
- Preventing repeat teen pregnancy: Postpartum Depot medroxyprogesterone acetate, oral contraceptive pills, or the patch?. J Pediatric Adolesc Gynecol. 2007;20:61–65
- Predictors of inconsistent contraceptive use among adolescent girls: Findings from a prospective study. J Adolesc Health. 2006;39:43–49
- . Identifying adolescents at risk for noncompliance with contraceptive therapy. J Pediatr. 1980;96:742–745
PII: S1054-139X(08)00265-6
doi:10.1016/j.jadohealth.2008.06.008
© 2008 Society for Adolescent Medicine. Published by Elsevier Inc. All rights reserved.
Refers to article:
- Start Now or Start Something Else?
