- Email: [email protected]
Bridging emergency contraceptive pill users to regular contraception: results from a randomized trial in Jamaica
Contraception 81 (2010) 133 – 139
Original research article
Bridging emergency contraceptive pill users to regular contraception: results from a randomized trial in Jamaica☆ Dawn S. Chin-Queea,⁎, Maxine Wedderburnb , Conrad Otternessa , Barbara Janowitza , Mario Chen-Mokc a
Health Services Research Division, Applied Research Department, Family Health International, Research Triangle Park, NC 27713, USA b HOPE Enterprises Ltd., 25 Burlington Avenue, Kingston 10, Jamaica, West Indies c Biostatistics Division, Family Health International, Research Triangle Park, NC 27713, USA Received 17 June 2009; revised 26 August 2009; accepted 30 August 2009
Abstract Background: Emergency contraception research has shifted from examining the public health effects of increasing access to emergency contraceptive pills (ECPs) to bridging ECP users to a regular contraceptive method as a way of decreasing unintended pregnancies. Study design: In a randomized controlled trial in Jamaica, we tested a discount coupon for oral contraceptive pills (OCPs) among pharmacybased ECP purchasers as an incentive to adopt (i.e., use for at least 2 months) this and other regular contraceptive methods. Women in the intervention and control arms were followed up at 3 and 6 months after ECP purchase to determine whether they adopted the OCP or any other contraceptive method. Condom use was recorded but was not considered a regular contraceptive due to its inconsistent use. Results: There was no significant difference in the proportion of women who adopted the OCP, injectable or intrauterine device in the control group or the intervention group (p=.39), and only 14.6% of the sample (mostly OCP adopters) used one of these three methods. Condom use was high (44.0%), demonstrating that ECP users were largely a condom-using group. Conclusions: The discount coupon intervention was not successful. Although a small proportion of ECP users did bridge, the coupon did not affect the decision to adopt a regular contraceptive method. The study highlighted the need for bridging strategies to consider women's reproductive and sexual behaviors, as well as their context. However, in countries like Jamaica where HIV/AIDS is of concern and condom use is appropriately high, bridging may not be an optimal strategy. © 2010 Elsevier Inc. All rights reserved. Keywords: Randomized controlled trial; Emergency contraception; Oral contraception; Contraceptive bridging; Jamaica
1. Introduction Emergency contraceptive pills (ECPs) have been shown to be an effective back-up method for birth control after unprotected sex. However, use of ECPs has not had a public health impact; a recent systematic review revealed that despite interventions to increase access to ECPs, no decrease in abortion or pregnancy rates has been documented [1], prompting a meeting of technical experts in July 2006 to discuss the next steps. Among the topics discussed in this meeting was the potential indirect effect of reducing unintended pregnancies by transitioning ECP users to other ☆ The work on which this article is based was funded by the William and Flora Hewlett Foundation. ⁎ Corresponding author. Tel.: +1 919 544 7040; fax: +1 919 544 7261. E-mail address: [email protected] (D.S. Chin-Quee).
0010-7824/$ – see front matter © 2010 Elsevier Inc. All rights reserved. doi:10.1016/j.contraception.2009.08.015
methods. Thus, efforts to decrease unintended pregnancies have now turned to promoting the use of more effective contraceptive methods among ECP users [2]. This approach, borrowing from Trussell et al. [3], has been dubbed “bridging,” as it encourages women to cross over or transition from ECPs to ongoing contraceptive use. The concept of bridging has emerged at a time when many developing countries allow over-the-counter provision of ECPs [4]. Therefore, ECPs are very likely obtained in pharmacies more often than in clinics, as even 80% of the International Planned Parenthood Federation's distribution of ECPs is made through commercial retail outlets (Hodgson M., personal communication, 2008). Anecdotal accounts of widespread and repeated use of ECPs have accompanied the increased over-the-counter availability of ECPs in pharmacies. With estimates of 74–85% effectiveness [5] of progestin-only formulations, reports of repeated use or use of
134
D.S. Chin-Quee et al. / Contraception 81 (2010) 133–139
ECPs as a main method of contraception are worrisome. Thus, bridging the ECP user to a more effective method of contraception, such as pills or injectables, could lead to a reduction in unwanted pregnancies. We conducted a study in Jamaica that promoted bridging in pharmacies where most women obtained the only dedicated ECP product at that time, Postinor-2® (hereafter referred to as Postinor; 0.75 mg of levonorgestrel in each tablet), in addition to 21 brands of regular oral contraceptive pills (OCPs). The goal was to encourage women to adopt a more effective ongoing contraceptive method with the incentive of a discount coupon for one cycle of OCPs. Only one cycle of OCPs was discounted, as the Jamaican distributor of Postinor would be willing to continue that level of discount beyond the life of the study, assuming that the intervention proved successful. The coupon intervention focused on OCPs because they are the second most prevalent method of contraception in Jamaica and are available over-the-counter in pharmacies where most ECP users obtain Postinor. The coupon excluded condoms, as their relatively low cost (and already high use as the most prevalent method) would provide little incentive for ECP users to switch to this method. Shortly after the Jamaican government removed the prescription-only requirement for Postinor in 2003, pharmacists went on record as opposing the government's decision and recounted instances of repeated and inappropriate use of the product. Thus, pharmacists and other stakeholders were interested in testing interventions to reduce the repeated use of ECPs and in documenting the frequency of repeat ECP purchases. A collaborative partnership with officials from the Pharmaceutical Society of Jamaica, the National Family Planning Board, Medimpex Jamaica, Ltd. (the distributor of Postinor and three OCP brands in the private sector), and with research staff from HOPE Enterprises, Ltd. (a marketing research firm in Jamaica) and Family Health International (FHI) was formed to look into this issue.
2. Materials and methods 2.1. Study design The study was designed as a multisite randomized controlled trial. Twenty-one pharmacies in Kingston, Jamaica (a subset of high-volume sellers of Postinor identified by Medimpex) participated in the study. Postinor purchasers were assigned to the control or intervention (coupon recipient) group within each pharmacy and administered a brief interview. The completed intercept interviews were subsequently examined in the offices of HOPE Enterprises to identify and assemble contact information for women who were eligible for follow-up. In both study groups, women who were not already using an ongoing method of contraception consistently were followed up 3 and 6 months later to assess contraceptive use.
The primary objective of the study was to determine whether provision of the discount coupon resulted in a higher adoption of a contraceptive method among ECP clients who were not already using a family planning method consistently compared to their counterparts who did not receive a discount coupon. Adoption was defined as continuous use of OCPs for at least 2 months. Although the coupon specifically discounted oral contraceptive products available in a pharmacy, it could also serve as a first step to encourage the adoption and long-term use of any ongoing contraceptive method obtained from the private or public sector. Therefore, we also counted two consecutive injectable contraceptive administrations (or intention to obtain the second injection) and retention of an intrauterine device (IUD) for at least 1 month as adoption of an ongoing or regular contraceptive method. Thus, for the purpose of this study, an ongoing contraceptive method (henceforth referred to as regular method) is a method employed for the prevention of pregnancy and used before the sexual act occurs and — in the case of these three methods — is independent of coitus. We excluded condoms in this definition because a pilot study we conducted prior to trial initiation documented that women who self-identified as habitual users of condoms did not use this method consistently over the last 3 months. The secondary objective was to document the frequency of ECP purchase by study participants. 2.2. The discount coupon intervention A business-size card, made from card stock for durability, served as the coupon. At the time of the study, the prices of the 21 pill brands in pharmacies ranged from US$1.39 to US $12.55 for one cycle (exchange rate US$1=J$65). Prices between US$5 and US$8 for a cycle of OCPs were considered reasonable and affordable for the typical pharmacy client and were cheaper than the price of one pack of Postinor at over US$10 at that time (lower-income women tended to patronize public health clinics where ECPs, while not as conveniently or readily accessed, are available at subsidized or no cost). The coupon discount ranged from 3% to 27% off all pill brands. However, for Medimpex's three oral contraceptive products, the discount was set at 13–20%, established during negotiations with the first author as part of a collaborative effort to build on the sustainability of the coupon intervention by a local stakeholder. The coupon's expiration date gave participants between 2 1/2 and 4 months (depending on the time of recruitment) to redeem the coupon. On the reverse side of the coupon, we included a message that urged women to use condoms, as OCPs do not provide protection against sexually transmitted diseases. 2.3. Sample size We estimated that we would need to intercept a minimum of 800 clients to obtain at least 80% power to detect an increase of at least 10% in the adoption of a regular
D.S. Chin-Quee et al. / Contraception 81 (2010) 133–139
contraceptive method in the intervention group compared to the control group, based on a one-sided test with a 5% significance level. This sample size estimation assumed that regular contraceptive use in the control group would be 20%, as estimated based on the pilot study. Also, based on results from this pilot study, the sample size was adjusted to account for refusal (5%) and loss to follow-up (20%). 2.4. Participant recruitment and randomization Pharmacists directed potential study participants to waiting interviewers, explaining that a study of Postinor was being conducted onsite and that, if they were interested, they could get additional information from the interviewer. Recruitment of ECP users took place between midDecember 2006 and mid-February 2007. Interviewers screened female Postinor clients in order to eliminate minors (under 16 years old) and women who were not purchasing Postinor for their own use. However, contact information for the Marge Roper Counseling Hotline run by the National Family Planning Board was provided to all ECP purchasers on another business card. After informed consent had been obtained from eligible women, interviewers randomly assigned participants to the intervention group or the control group using a 1:1 allocation ratio by opening sequentially numbered opaque sealed envelopes containing group assignment. An FHI statistician developed random allocation sequences using a computer random number generator and permuted blocks of sizes 6, 12 and 18. Separate allocation sequences were generated for each pharmacy and placed in randomization envelopes. If assigned to the intervention group, the Postinor client received a discount coupon, in addition to a Marge Roper card. Women in the control group received the Marge Roper card only. Brief interviews about their ECP purchase on that day and contraceptive history were then administered. All participants were asked to provide their telephone number in order for them to be contacted for possible follow-up interviews by HOPE Enterprises researchers. Given the nature of the intervention, participants and data collectors were not blinded to treatment group. Data managers and analysts at FHI and HOPE Enterprises were also not blinded. If participants reported that they had been using the OCP continuously for at least 3 months or that they returned to the clinic late for their injectable contraceptive but intended to resume the use of these methods after taking Postinor, they were considered consistent users of a regular contraceptive method. As specified in the study protocol, these women were not followed up at 3 and 6 months, and their contact information was destroyed. 2.5. Participant follow-up Interviewers contacted all nonregular contraceptive method users by telephone 3 and 6 months after their intercept interview. Computer-assisted telephone interviewing (CATI) was used to administer the interview instrument.
135
Participants were asked about current contraceptive use, additional ECP purchases and coupon redemption, distinguishing between those who reported receiving the coupon directly from the interviewer at a study pharmacy and those who may have received the coupon from someone else. At the 6-month interview, women who were determined to have adopted a regular contraceptive method were asked about any changes in contraceptive use since their 3-month interview, as well as the average length of continuous use of a regular contraceptive method. 2.6. Data analysis We estimated the proportion of women who adopted a regular contraceptive method and the proportion of women who used ECPs over the 6-month period for the intervention and control groups. Differences were tested using chi-square tests. As planned, tests were conducted for one-sided alternatives in favor of the intervention group, with a 5% significance level. For the main outcome of adoption of a regular contraceptive method, clients who were lost to follow-up at the 3-month or 6-month interview were assumed to have not initiated any method. This assumption provided a conservative estimate of method adoption. We used a similar assumption for all percent estimates and included women lost to follow-up in the reference group, thus providing conservative estimates. All analyses were performed using SAS (version 9.1). The reporting of this study complies with CONSORT guidelines for reporting randomized trials [6]. 2.7. Protection of human subjects The research protocol was approved by the ethical review committee of FHI and supported by the Pharmaceutical Society of Jamaica and the Jamaica National Family Planning Board. Participant names in the intercept interview were replaced by numerical identifiers in CATI files and in data sets created for data management and analysis. The numerical identifiers were maintained in a master list and secured in a locked file cabinet or drawer at HOPE Enterprises. A Data Monitoring Committee was not used for the trial, and no interim looks at the data were planned or performed.
3. Results A total of 1210 eligible women were intercepted at 21 pharmacies. Of these, 191 women refused to participate, and 11 women were inadvertently interviewed twice (their first interview was used), yielding a sample of 1008 women. The intervention and control groups were evenly split at 507 and 501 participants, respectively, including 54 pill users and 16 injectable users who met our definition of regular contraceptive method users and were excluded from further analyses. The final sample for analysis stood at 476
136
D.S. Chin-Quee et al. / Contraception 81 (2010) 133–139
women in the intervention group and 462 women in the control group identified as participants who were not using a regular contraceptive method consistently. 3.1. Background and ECP use characteristics at intercept The demographic characteristics of the control and intervention groups were very similar (Table 1). Participants were, on average, 25 years of age, about 50% had no children and just under 80% were single. The majority reported the Table 1 Baseline demographics, frequency of ECP purchase in the last 12 months and reason for recent ECP purchase of all clients at intercept Intervention group (received coupon) (n=476)
Control group (did not receive coupon) (n=462)
Age in years (mean) 25.6 25.6 Minimum/maximum 16–45 16–45 Number of children (%) 0 49 47 1–2 40 44 3 or more 10 9 Education (years) Less than secondary 1 b1 Secondary/high school 41 46 Postsecondary/university 47 42 11 11 Othera Employment status Employed 68 64 Not employed 31 35 No response 1 1 Marital status Single 79 79 Married/in union 19 19 Separated/divorced/widowed 2 2 No response 1 b1 Number of times purchased ECP in the past 12 months (including the time of enrollment) First time 33 36 One time 17 15 2–3 times 40 40 4 or more times 9 9 Do not know/no response 1 b1 Reasons for purchasing Postinor today (multiple responses provided) I used no method at all 61 58 The condom slipped/broke/tore 34 35 (failed) Forgot to take or missed pill(s) 1 3 Did not return on time to b1 1 get/buy pills Did not return on time to 1 b1 get next injection 3 2 Otherb No reason given 1 2 a Other includes evening class, nursing school, school of cosmetology/ beauty school, community college/business school, skills training/heart academy, fashion design/school of fashion design, preuniversity/certificate and women's center. b Other responses include “no,” “did not start off using condoms,” “back-up plan (just in case cycle changes or sex happens),” “just wanted to be absolutely sure,” “withdrawal,” “partner took off condom during sex” and “was taken off pill by doctor.”
attainment of secondary school education or higher, and more than two thirds were gainfully employed. Most participants were purchasing ECPs for the second or third time in the last 12 months. Approximately one third mentioned problems with condoms as the reason for purchasing ECPs on that day. In addition, of the approximately 60% who purchased ECPs because they said they had used no method, one third of this group or 20% of all women also reported that they considered themselves to be condom users. If those ECP purchasers are added to those who said that their condom failed (34%), we would find that 54% of ECP buyers were in fact condom users (data not shown). Although we did not explicitly ask participants if they purchased ECPs in advance of need, seven women (less than 1%) reported that they bought Postinor on that day “in case sex happens.” 3.2. Demographic characteristics of the follow-up sample Women contacted at 6 months had the same demographic characteristics as those interviewed at the 3-month follow-up (data not shown). Moreover, women lost to follow-up shared nearly identical age, parity, marital status and employment status as the remaining sample, but significantly more women lost to follow-up attained postsecondary or university education than women in the remaining sample (52% vs. 43%, respectively; p=.03). 3.3. Primary outcome: adoption of regular contraceptive methods Based on our definition of adoption of a regular method, 14.6% of participants had adopted the OCP, injectable or IUD within the 6-month study period (Fig. 1). There was no statistically significant difference in adoption rates between the intervention group and the control group (p=.39). With regard to adoption of the OCP, only 13% of the intervention group and 11% of the control group had used the pill for 2 months or more since the pharmacy intercept. This difference was also not statistically significant (p=.11). Condoms continued to be the method most frequently used by our sample at approximately 44% in both treatment groups, only slightly lower than the 54% estimated to be using condoms. 3.4. Coupon redemption The adoption rates for OCPs in the intervention group and for coupon redemption were at variance: out of 475 coupons distributed to women in the intervention group (with one refusal to accept), only four or less than 1% were redeemed. When asked during follow-up interviews why they did not redeem the coupon, over 70% of women said they did not want to use the OCP at that time, while another 15% reported that they forgot about the coupon. The remaining reasons ranged from their desire not to use any contraceptive method to providing no explanation.
D.S. Chin-Quee et al. / Contraception 81 (2010) 133–139
137
4. Discussion
Fig. 1. Adoption of pills, injectable or IUD during the 6-month study period.
3.5. Frequency of ECP use While over half of women in the intervention and control groups did not repurchase ECPs at all during the 6-month period, about 25% in both groups did repurchase the product. Of those who bought Postinor again, most did so one more time. Smaller proportions repurchased ECPs two or more times in that 6-month period (Table 2). Reasons for repurchase over the 6-month period were very similar and in the same proportion as the reasons given for buying Postinor at intercept. That is, the majority of repurchasers bought Postinor because they used no contraceptive method at all, followed by reports of condom failure.
Table 2 Repurchase of ECPs during the 6-month study period Intervention group Control group (received coupon) (did not receive (n=476) coupon) (n=462) Repurchase of ECP over study period (%) Did not repurchase 55 Repurchased once 14 Repurchased twice 5 Repurchased three or more times 6 No response b1 Lost to follow-up 19 Total 100
58 13 5 4 – 20 100
The discount coupon was redeemed by very few ECP users and, therefore, was not a successful bridging intervention for this group of women. Nevertheless, about 15% of all participants did adopt a regular method, mostly OCPs. This finding suggests that some women will bridge to a regular contraceptive method. There are several explanations for the low redemption rate of coupons: (a) there has been no history of use of consumer coupons in Jamaica; (b) the amount discounted per cycle was insufficient to motivate women to purchase OCPs; (c) three cycles, not one cycle, should have been provided at a discounted price; and (d) OCPs should have been provided at no cost rather than at a discounted price to the ECP user. Regarding the first explanation, Jamaicans have been known to save and redeem coupons. The largest local newspaper (The Gleaner) has, in the past, run a hugely successful incentive offer using coupons. In fact, recent qualitative research on general consumer shopping habits conducted by HOPE Enterprises shows great interest in coupons among adult females. This interest was often highest for products (whether consumer goods, novelty items or otherwise) sold in pharmacies. However, it does appear that the coupon offer itself did not overcome the majority of ECP users' disinterest in using the pill. While a larger discount might have been more attractive to the potential buyer, we also had to take into consideration the willingness of the distributors to agree to keep the discount in place after the project had been completed. To be a sustainable intervention, the discount rate set by Medimpex (the distributor of Postinor) had to be honored for the study. 4.1. Why were ECP users not interested in adopting a regular contraceptive method? More than two thirds of the women in our intervention group reported that they did not want to use the pill, and no further information was obtained when asked why they did not redeem the discount coupon. The demographic and behavioral profiles of these pharmacy-based ECP users may have affected their decision not to adopt OCPs. A national survey of 2400 sexually active 15- to 49-yearolds in Jamaica [7] found that those in visiting relationships (i.e., couples not living together) were more likely to use condoms than to use other methods. Nonbarrier method use (including OCPs) was dominant among those in cohabiting relationships. Our sample described themselves as “single” and, therefore, they may be similar to women in visiting relationships. Indeed, we know that women's partnership status affects their decisions on whether they should use a contraceptive method, when they should use a contraceptive method, how often they should use a contraceptive method and which method to use [8].
138
D.S. Chin-Quee et al. / Contraception 81 (2010) 133–139
ECP users may also misperceive the effectiveness of the method. Research in East and West Africa suggests that ECP users have inaccurate perceptions of the effectiveness of ECPs for pregnancy prevention, believing that they are more effective than condoms and possibly as effective as OCPs and other regular contraceptive methods [9]. This may help to explain why unprotected sex occurs among women, especially among those who may find it more convenient and therefore desirable to use a contraceptive method after sexual intercourse. 4.2. In general, is bridging the best strategy for ECP users? Jamaica's HIV/AIDS prevention efforts have made the male condom the most prevalent contraceptive method in the country, and it was certainly the most used method by ECP users in our sample. Moreover, the National Family Planning Board has supported the introduction of the condom for dual-method use in efforts aimed at integrating the prevention of HIV and other sexually transmitted infections (STIs) into its programs. Thus, the context in which this bridging intervention was implemented may not have been optimal, since a strong precedent and preference for condoms have likely been established among ECP users who are mostly single. Results from the abovereferenced national survey and our own study point to the fact that many single women in visiting relationships use barrier methods and may not be interested in adopting a hormonal or long-acting contraceptive method — a logical decision for women who may not have a partner or sex on a regular basis. Nevertheless, our pharmacy-based ECP users — over half of whom inconsistently and incorrectly use condoms — present a special problem. These women may perceive that they have solved this problem by using ECPs when they have failed to use condoms. Therefore, should efforts focus on bridging ECP users to regular contraceptive methods or on improving condom use? The ideal, of course, would be dual-method use, but consistent use of two methods may be difficult to achieve. Consequently, women may need to weigh the risks of pregnancy and STI/HIV in making choices about contraception, including condom and ECP use. 4.3. Are ECPs being overused in Jamaica? The very postcoital nature of ECPs — and its ready availability without prescription — may make it appealing to women and, at the same time, worrisome to pharmacists who may find themselves selling it to the same individuals over and over again. However, our findings on the frequency of purchase and use do not support the perception of overuse of ECPs in Jamaica, nor does it support the notion that ECPs are being purchased in advance of need (i.e., before underprotected or unprotected sex actually occurs). In our study, most women did not buy ECPs more than once in the 6month study period; among those who did, the majority reported using ECPs only two to three times in the last 18
months (i.e., in the last year plus the 6-month study period). Less than 1% of participants at intercept indicated that they purchased Postinor before they had sex, and it was not clear whether any advance purchase occurred during the 3-month and 6-month follow-up periods. 4.4. Study limitations The wording of some questionnaire items may have limited the interpretability of our findings. For example, our measure of frequency of ECP use at the intercept interview may have produced underestimates in our participants' reports due to the lengthy time frame. Asking women how many times they purchased ECPs in the last 3 months or how often they used them in an average month may have been better than asking about ECP use in the last 12 months as we did. Nevertheless, we documented that ECP use over the 6month study period was low. The ability to generalize these findings is also limited by the fact that, in some countries such as the United States, OCPs are not available over-the-counter. As such, it would not be possible to bridge ECP users to another method solely in pharmacy settings, adding yet another hurdle to women's adoption of an ongoing contraceptive method. The strengths of the study reside in its randomization of participants (which resulted in well-balanced treatment groups) and the unequivocal finding that although the discount coupon per se did not encourage this group of women to adopt an ongoing contraceptive method, a small number did so on their own. Furthermore, the findings suggest that ECP users are not a monolithic group and that special effort should be undertaken to identify those who are amenable to a regular contraceptive method and for whom bridging to a regular contraceptive method would be an appropriate approach. A better understanding of why some women bridge while others fail to adopt a regular method is needed to design more effective bridging interventions. Given cultural differences, as well as beliefs about the effectiveness of ECPs and the advantages and disadvantages of other methods, information that expressly compares the advantages of regular methods over ECPs should be a key component in the materials and (when available) counseling provided to these women.
Acknowledgments The authors would like to express appreciation to Drs. John Stanback, Kelly L'Engle and Elizabeth Raymond for their helpful comments on earlier drafts of the manuscript. The first author would also like to thank the Pharmaceutical Society of Jamaica, the National Family Planning Board and Medimpex for their collaboration, support and participation through all phases of study development and implementation.
D.S. Chin-Quee et al. / Contraception 81 (2010) 133–139
References [1] Raymond EG, Trussell J, Polis CB. Population effect of increased access to emergency contraceptive pills: a systematic review. Obstet Gynecol 2007;109:181–8. [2] Trussell J, Schwarz EB, Guthrie K, Raymond E. No such thing as an easy (or EC) fix. Contraception 2008;78:351–4. [3] Trussell J, Stewart F, Guest F, Hatcher RA. Emergency contraceptive pills: a simple proposal to reduce unintended pregnancies. Fam Plann Perspect 1992;24:269–73. [4] International Consortium for Emergency Contraception. EC status and availability: searchable databases. http://www.cecinfo.org/database/pill/ pillData.php. New York: Family Care International; 2008. Accessed on December 14.
139
[5] Trussell J, Guthrie KA. Talking straight about emergency contraception. J Fam Plann Reprod Health Care 2007;33:139–42. [6] Altman DG, Schulz KF, Moher D, Egger M, Davidoff F, Elbourne D, et al. The revised CONSORT statement for reporting randomized trials: explanation and elaboration. Ann Intern Med 2001;134:663–94. [7] Bourne D, Wedderburn M. Report on dual method use in Jamaica. Kingston (Jamaica): Hope Enterprises; 2007. [8] Frost JJ, Darroch JE. Factors associated with contraceptive choice and inconsistent method use, United States, 2004. Perspect Sex Reprod Health 2008;40:94–104. [9] Wamatu C. EC use in Kenya: a profile of EC users and non-users. Bridging from emergency contraceptive pills to ongoing contraception: consultation for the Hewlett Foundation. Research Triangle Park (NC): Family Health International (FHI); 2007.
Original research article
Bridging emergency contraceptive pill users to regular contraception: results from a randomized trial in Jamaica☆ Dawn S. Chin-Queea,⁎, Maxine Wedderburnb , Conrad Otternessa , Barbara Janowitza , Mario Chen-Mokc a
Health Services Research Division, Applied Research Department, Family Health International, Research Triangle Park, NC 27713, USA b HOPE Enterprises Ltd., 25 Burlington Avenue, Kingston 10, Jamaica, West Indies c Biostatistics Division, Family Health International, Research Triangle Park, NC 27713, USA Received 17 June 2009; revised 26 August 2009; accepted 30 August 2009
Abstract Background: Emergency contraception research has shifted from examining the public health effects of increasing access to emergency contraceptive pills (ECPs) to bridging ECP users to a regular contraceptive method as a way of decreasing unintended pregnancies. Study design: In a randomized controlled trial in Jamaica, we tested a discount coupon for oral contraceptive pills (OCPs) among pharmacybased ECP purchasers as an incentive to adopt (i.e., use for at least 2 months) this and other regular contraceptive methods. Women in the intervention and control arms were followed up at 3 and 6 months after ECP purchase to determine whether they adopted the OCP or any other contraceptive method. Condom use was recorded but was not considered a regular contraceptive due to its inconsistent use. Results: There was no significant difference in the proportion of women who adopted the OCP, injectable or intrauterine device in the control group or the intervention group (p=.39), and only 14.6% of the sample (mostly OCP adopters) used one of these three methods. Condom use was high (44.0%), demonstrating that ECP users were largely a condom-using group. Conclusions: The discount coupon intervention was not successful. Although a small proportion of ECP users did bridge, the coupon did not affect the decision to adopt a regular contraceptive method. The study highlighted the need for bridging strategies to consider women's reproductive and sexual behaviors, as well as their context. However, in countries like Jamaica where HIV/AIDS is of concern and condom use is appropriately high, bridging may not be an optimal strategy. © 2010 Elsevier Inc. All rights reserved. Keywords: Randomized controlled trial; Emergency contraception; Oral contraception; Contraceptive bridging; Jamaica
1. Introduction Emergency contraceptive pills (ECPs) have been shown to be an effective back-up method for birth control after unprotected sex. However, use of ECPs has not had a public health impact; a recent systematic review revealed that despite interventions to increase access to ECPs, no decrease in abortion or pregnancy rates has been documented [1], prompting a meeting of technical experts in July 2006 to discuss the next steps. Among the topics discussed in this meeting was the potential indirect effect of reducing unintended pregnancies by transitioning ECP users to other ☆ The work on which this article is based was funded by the William and Flora Hewlett Foundation. ⁎ Corresponding author. Tel.: +1 919 544 7040; fax: +1 919 544 7261. E-mail address: [email protected] (D.S. Chin-Quee).
0010-7824/$ – see front matter © 2010 Elsevier Inc. All rights reserved. doi:10.1016/j.contraception.2009.08.015
methods. Thus, efforts to decrease unintended pregnancies have now turned to promoting the use of more effective contraceptive methods among ECP users [2]. This approach, borrowing from Trussell et al. [3], has been dubbed “bridging,” as it encourages women to cross over or transition from ECPs to ongoing contraceptive use. The concept of bridging has emerged at a time when many developing countries allow over-the-counter provision of ECPs [4]. Therefore, ECPs are very likely obtained in pharmacies more often than in clinics, as even 80% of the International Planned Parenthood Federation's distribution of ECPs is made through commercial retail outlets (Hodgson M., personal communication, 2008). Anecdotal accounts of widespread and repeated use of ECPs have accompanied the increased over-the-counter availability of ECPs in pharmacies. With estimates of 74–85% effectiveness [5] of progestin-only formulations, reports of repeated use or use of
134
D.S. Chin-Quee et al. / Contraception 81 (2010) 133–139
ECPs as a main method of contraception are worrisome. Thus, bridging the ECP user to a more effective method of contraception, such as pills or injectables, could lead to a reduction in unwanted pregnancies. We conducted a study in Jamaica that promoted bridging in pharmacies where most women obtained the only dedicated ECP product at that time, Postinor-2® (hereafter referred to as Postinor; 0.75 mg of levonorgestrel in each tablet), in addition to 21 brands of regular oral contraceptive pills (OCPs). The goal was to encourage women to adopt a more effective ongoing contraceptive method with the incentive of a discount coupon for one cycle of OCPs. Only one cycle of OCPs was discounted, as the Jamaican distributor of Postinor would be willing to continue that level of discount beyond the life of the study, assuming that the intervention proved successful. The coupon intervention focused on OCPs because they are the second most prevalent method of contraception in Jamaica and are available over-the-counter in pharmacies where most ECP users obtain Postinor. The coupon excluded condoms, as their relatively low cost (and already high use as the most prevalent method) would provide little incentive for ECP users to switch to this method. Shortly after the Jamaican government removed the prescription-only requirement for Postinor in 2003, pharmacists went on record as opposing the government's decision and recounted instances of repeated and inappropriate use of the product. Thus, pharmacists and other stakeholders were interested in testing interventions to reduce the repeated use of ECPs and in documenting the frequency of repeat ECP purchases. A collaborative partnership with officials from the Pharmaceutical Society of Jamaica, the National Family Planning Board, Medimpex Jamaica, Ltd. (the distributor of Postinor and three OCP brands in the private sector), and with research staff from HOPE Enterprises, Ltd. (a marketing research firm in Jamaica) and Family Health International (FHI) was formed to look into this issue.
2. Materials and methods 2.1. Study design The study was designed as a multisite randomized controlled trial. Twenty-one pharmacies in Kingston, Jamaica (a subset of high-volume sellers of Postinor identified by Medimpex) participated in the study. Postinor purchasers were assigned to the control or intervention (coupon recipient) group within each pharmacy and administered a brief interview. The completed intercept interviews were subsequently examined in the offices of HOPE Enterprises to identify and assemble contact information for women who were eligible for follow-up. In both study groups, women who were not already using an ongoing method of contraception consistently were followed up 3 and 6 months later to assess contraceptive use.
The primary objective of the study was to determine whether provision of the discount coupon resulted in a higher adoption of a contraceptive method among ECP clients who were not already using a family planning method consistently compared to their counterparts who did not receive a discount coupon. Adoption was defined as continuous use of OCPs for at least 2 months. Although the coupon specifically discounted oral contraceptive products available in a pharmacy, it could also serve as a first step to encourage the adoption and long-term use of any ongoing contraceptive method obtained from the private or public sector. Therefore, we also counted two consecutive injectable contraceptive administrations (or intention to obtain the second injection) and retention of an intrauterine device (IUD) for at least 1 month as adoption of an ongoing or regular contraceptive method. Thus, for the purpose of this study, an ongoing contraceptive method (henceforth referred to as regular method) is a method employed for the prevention of pregnancy and used before the sexual act occurs and — in the case of these three methods — is independent of coitus. We excluded condoms in this definition because a pilot study we conducted prior to trial initiation documented that women who self-identified as habitual users of condoms did not use this method consistently over the last 3 months. The secondary objective was to document the frequency of ECP purchase by study participants. 2.2. The discount coupon intervention A business-size card, made from card stock for durability, served as the coupon. At the time of the study, the prices of the 21 pill brands in pharmacies ranged from US$1.39 to US $12.55 for one cycle (exchange rate US$1=J$65). Prices between US$5 and US$8 for a cycle of OCPs were considered reasonable and affordable for the typical pharmacy client and were cheaper than the price of one pack of Postinor at over US$10 at that time (lower-income women tended to patronize public health clinics where ECPs, while not as conveniently or readily accessed, are available at subsidized or no cost). The coupon discount ranged from 3% to 27% off all pill brands. However, for Medimpex's three oral contraceptive products, the discount was set at 13–20%, established during negotiations with the first author as part of a collaborative effort to build on the sustainability of the coupon intervention by a local stakeholder. The coupon's expiration date gave participants between 2 1/2 and 4 months (depending on the time of recruitment) to redeem the coupon. On the reverse side of the coupon, we included a message that urged women to use condoms, as OCPs do not provide protection against sexually transmitted diseases. 2.3. Sample size We estimated that we would need to intercept a minimum of 800 clients to obtain at least 80% power to detect an increase of at least 10% in the adoption of a regular
D.S. Chin-Quee et al. / Contraception 81 (2010) 133–139
contraceptive method in the intervention group compared to the control group, based on a one-sided test with a 5% significance level. This sample size estimation assumed that regular contraceptive use in the control group would be 20%, as estimated based on the pilot study. Also, based on results from this pilot study, the sample size was adjusted to account for refusal (5%) and loss to follow-up (20%). 2.4. Participant recruitment and randomization Pharmacists directed potential study participants to waiting interviewers, explaining that a study of Postinor was being conducted onsite and that, if they were interested, they could get additional information from the interviewer. Recruitment of ECP users took place between midDecember 2006 and mid-February 2007. Interviewers screened female Postinor clients in order to eliminate minors (under 16 years old) and women who were not purchasing Postinor for their own use. However, contact information for the Marge Roper Counseling Hotline run by the National Family Planning Board was provided to all ECP purchasers on another business card. After informed consent had been obtained from eligible women, interviewers randomly assigned participants to the intervention group or the control group using a 1:1 allocation ratio by opening sequentially numbered opaque sealed envelopes containing group assignment. An FHI statistician developed random allocation sequences using a computer random number generator and permuted blocks of sizes 6, 12 and 18. Separate allocation sequences were generated for each pharmacy and placed in randomization envelopes. If assigned to the intervention group, the Postinor client received a discount coupon, in addition to a Marge Roper card. Women in the control group received the Marge Roper card only. Brief interviews about their ECP purchase on that day and contraceptive history were then administered. All participants were asked to provide their telephone number in order for them to be contacted for possible follow-up interviews by HOPE Enterprises researchers. Given the nature of the intervention, participants and data collectors were not blinded to treatment group. Data managers and analysts at FHI and HOPE Enterprises were also not blinded. If participants reported that they had been using the OCP continuously for at least 3 months or that they returned to the clinic late for their injectable contraceptive but intended to resume the use of these methods after taking Postinor, they were considered consistent users of a regular contraceptive method. As specified in the study protocol, these women were not followed up at 3 and 6 months, and their contact information was destroyed. 2.5. Participant follow-up Interviewers contacted all nonregular contraceptive method users by telephone 3 and 6 months after their intercept interview. Computer-assisted telephone interviewing (CATI) was used to administer the interview instrument.
135
Participants were asked about current contraceptive use, additional ECP purchases and coupon redemption, distinguishing between those who reported receiving the coupon directly from the interviewer at a study pharmacy and those who may have received the coupon from someone else. At the 6-month interview, women who were determined to have adopted a regular contraceptive method were asked about any changes in contraceptive use since their 3-month interview, as well as the average length of continuous use of a regular contraceptive method. 2.6. Data analysis We estimated the proportion of women who adopted a regular contraceptive method and the proportion of women who used ECPs over the 6-month period for the intervention and control groups. Differences were tested using chi-square tests. As planned, tests were conducted for one-sided alternatives in favor of the intervention group, with a 5% significance level. For the main outcome of adoption of a regular contraceptive method, clients who were lost to follow-up at the 3-month or 6-month interview were assumed to have not initiated any method. This assumption provided a conservative estimate of method adoption. We used a similar assumption for all percent estimates and included women lost to follow-up in the reference group, thus providing conservative estimates. All analyses were performed using SAS (version 9.1). The reporting of this study complies with CONSORT guidelines for reporting randomized trials [6]. 2.7. Protection of human subjects The research protocol was approved by the ethical review committee of FHI and supported by the Pharmaceutical Society of Jamaica and the Jamaica National Family Planning Board. Participant names in the intercept interview were replaced by numerical identifiers in CATI files and in data sets created for data management and analysis. The numerical identifiers were maintained in a master list and secured in a locked file cabinet or drawer at HOPE Enterprises. A Data Monitoring Committee was not used for the trial, and no interim looks at the data were planned or performed.
3. Results A total of 1210 eligible women were intercepted at 21 pharmacies. Of these, 191 women refused to participate, and 11 women were inadvertently interviewed twice (their first interview was used), yielding a sample of 1008 women. The intervention and control groups were evenly split at 507 and 501 participants, respectively, including 54 pill users and 16 injectable users who met our definition of regular contraceptive method users and were excluded from further analyses. The final sample for analysis stood at 476
136
D.S. Chin-Quee et al. / Contraception 81 (2010) 133–139
women in the intervention group and 462 women in the control group identified as participants who were not using a regular contraceptive method consistently. 3.1. Background and ECP use characteristics at intercept The demographic characteristics of the control and intervention groups were very similar (Table 1). Participants were, on average, 25 years of age, about 50% had no children and just under 80% were single. The majority reported the Table 1 Baseline demographics, frequency of ECP purchase in the last 12 months and reason for recent ECP purchase of all clients at intercept Intervention group (received coupon) (n=476)
Control group (did not receive coupon) (n=462)
Age in years (mean) 25.6 25.6 Minimum/maximum 16–45 16–45 Number of children (%) 0 49 47 1–2 40 44 3 or more 10 9 Education (years) Less than secondary 1 b1 Secondary/high school 41 46 Postsecondary/university 47 42 11 11 Othera Employment status Employed 68 64 Not employed 31 35 No response 1 1 Marital status Single 79 79 Married/in union 19 19 Separated/divorced/widowed 2 2 No response 1 b1 Number of times purchased ECP in the past 12 months (including the time of enrollment) First time 33 36 One time 17 15 2–3 times 40 40 4 or more times 9 9 Do not know/no response 1 b1 Reasons for purchasing Postinor today (multiple responses provided) I used no method at all 61 58 The condom slipped/broke/tore 34 35 (failed) Forgot to take or missed pill(s) 1 3 Did not return on time to b1 1 get/buy pills Did not return on time to 1 b1 get next injection 3 2 Otherb No reason given 1 2 a Other includes evening class, nursing school, school of cosmetology/ beauty school, community college/business school, skills training/heart academy, fashion design/school of fashion design, preuniversity/certificate and women's center. b Other responses include “no,” “did not start off using condoms,” “back-up plan (just in case cycle changes or sex happens),” “just wanted to be absolutely sure,” “withdrawal,” “partner took off condom during sex” and “was taken off pill by doctor.”
attainment of secondary school education or higher, and more than two thirds were gainfully employed. Most participants were purchasing ECPs for the second or third time in the last 12 months. Approximately one third mentioned problems with condoms as the reason for purchasing ECPs on that day. In addition, of the approximately 60% who purchased ECPs because they said they had used no method, one third of this group or 20% of all women also reported that they considered themselves to be condom users. If those ECP purchasers are added to those who said that their condom failed (34%), we would find that 54% of ECP buyers were in fact condom users (data not shown). Although we did not explicitly ask participants if they purchased ECPs in advance of need, seven women (less than 1%) reported that they bought Postinor on that day “in case sex happens.” 3.2. Demographic characteristics of the follow-up sample Women contacted at 6 months had the same demographic characteristics as those interviewed at the 3-month follow-up (data not shown). Moreover, women lost to follow-up shared nearly identical age, parity, marital status and employment status as the remaining sample, but significantly more women lost to follow-up attained postsecondary or university education than women in the remaining sample (52% vs. 43%, respectively; p=.03). 3.3. Primary outcome: adoption of regular contraceptive methods Based on our definition of adoption of a regular method, 14.6% of participants had adopted the OCP, injectable or IUD within the 6-month study period (Fig. 1). There was no statistically significant difference in adoption rates between the intervention group and the control group (p=.39). With regard to adoption of the OCP, only 13% of the intervention group and 11% of the control group had used the pill for 2 months or more since the pharmacy intercept. This difference was also not statistically significant (p=.11). Condoms continued to be the method most frequently used by our sample at approximately 44% in both treatment groups, only slightly lower than the 54% estimated to be using condoms. 3.4. Coupon redemption The adoption rates for OCPs in the intervention group and for coupon redemption were at variance: out of 475 coupons distributed to women in the intervention group (with one refusal to accept), only four or less than 1% were redeemed. When asked during follow-up interviews why they did not redeem the coupon, over 70% of women said they did not want to use the OCP at that time, while another 15% reported that they forgot about the coupon. The remaining reasons ranged from their desire not to use any contraceptive method to providing no explanation.
D.S. Chin-Quee et al. / Contraception 81 (2010) 133–139
137
4. Discussion
Fig. 1. Adoption of pills, injectable or IUD during the 6-month study period.
3.5. Frequency of ECP use While over half of women in the intervention and control groups did not repurchase ECPs at all during the 6-month period, about 25% in both groups did repurchase the product. Of those who bought Postinor again, most did so one more time. Smaller proportions repurchased ECPs two or more times in that 6-month period (Table 2). Reasons for repurchase over the 6-month period were very similar and in the same proportion as the reasons given for buying Postinor at intercept. That is, the majority of repurchasers bought Postinor because they used no contraceptive method at all, followed by reports of condom failure.
Table 2 Repurchase of ECPs during the 6-month study period Intervention group Control group (received coupon) (did not receive (n=476) coupon) (n=462) Repurchase of ECP over study period (%) Did not repurchase 55 Repurchased once 14 Repurchased twice 5 Repurchased three or more times 6 No response b1 Lost to follow-up 19 Total 100
58 13 5 4 – 20 100
The discount coupon was redeemed by very few ECP users and, therefore, was not a successful bridging intervention for this group of women. Nevertheless, about 15% of all participants did adopt a regular method, mostly OCPs. This finding suggests that some women will bridge to a regular contraceptive method. There are several explanations for the low redemption rate of coupons: (a) there has been no history of use of consumer coupons in Jamaica; (b) the amount discounted per cycle was insufficient to motivate women to purchase OCPs; (c) three cycles, not one cycle, should have been provided at a discounted price; and (d) OCPs should have been provided at no cost rather than at a discounted price to the ECP user. Regarding the first explanation, Jamaicans have been known to save and redeem coupons. The largest local newspaper (The Gleaner) has, in the past, run a hugely successful incentive offer using coupons. In fact, recent qualitative research on general consumer shopping habits conducted by HOPE Enterprises shows great interest in coupons among adult females. This interest was often highest for products (whether consumer goods, novelty items or otherwise) sold in pharmacies. However, it does appear that the coupon offer itself did not overcome the majority of ECP users' disinterest in using the pill. While a larger discount might have been more attractive to the potential buyer, we also had to take into consideration the willingness of the distributors to agree to keep the discount in place after the project had been completed. To be a sustainable intervention, the discount rate set by Medimpex (the distributor of Postinor) had to be honored for the study. 4.1. Why were ECP users not interested in adopting a regular contraceptive method? More than two thirds of the women in our intervention group reported that they did not want to use the pill, and no further information was obtained when asked why they did not redeem the discount coupon. The demographic and behavioral profiles of these pharmacy-based ECP users may have affected their decision not to adopt OCPs. A national survey of 2400 sexually active 15- to 49-yearolds in Jamaica [7] found that those in visiting relationships (i.e., couples not living together) were more likely to use condoms than to use other methods. Nonbarrier method use (including OCPs) was dominant among those in cohabiting relationships. Our sample described themselves as “single” and, therefore, they may be similar to women in visiting relationships. Indeed, we know that women's partnership status affects their decisions on whether they should use a contraceptive method, when they should use a contraceptive method, how often they should use a contraceptive method and which method to use [8].
138
D.S. Chin-Quee et al. / Contraception 81 (2010) 133–139
ECP users may also misperceive the effectiveness of the method. Research in East and West Africa suggests that ECP users have inaccurate perceptions of the effectiveness of ECPs for pregnancy prevention, believing that they are more effective than condoms and possibly as effective as OCPs and other regular contraceptive methods [9]. This may help to explain why unprotected sex occurs among women, especially among those who may find it more convenient and therefore desirable to use a contraceptive method after sexual intercourse. 4.2. In general, is bridging the best strategy for ECP users? Jamaica's HIV/AIDS prevention efforts have made the male condom the most prevalent contraceptive method in the country, and it was certainly the most used method by ECP users in our sample. Moreover, the National Family Planning Board has supported the introduction of the condom for dual-method use in efforts aimed at integrating the prevention of HIV and other sexually transmitted infections (STIs) into its programs. Thus, the context in which this bridging intervention was implemented may not have been optimal, since a strong precedent and preference for condoms have likely been established among ECP users who are mostly single. Results from the abovereferenced national survey and our own study point to the fact that many single women in visiting relationships use barrier methods and may not be interested in adopting a hormonal or long-acting contraceptive method — a logical decision for women who may not have a partner or sex on a regular basis. Nevertheless, our pharmacy-based ECP users — over half of whom inconsistently and incorrectly use condoms — present a special problem. These women may perceive that they have solved this problem by using ECPs when they have failed to use condoms. Therefore, should efforts focus on bridging ECP users to regular contraceptive methods or on improving condom use? The ideal, of course, would be dual-method use, but consistent use of two methods may be difficult to achieve. Consequently, women may need to weigh the risks of pregnancy and STI/HIV in making choices about contraception, including condom and ECP use. 4.3. Are ECPs being overused in Jamaica? The very postcoital nature of ECPs — and its ready availability without prescription — may make it appealing to women and, at the same time, worrisome to pharmacists who may find themselves selling it to the same individuals over and over again. However, our findings on the frequency of purchase and use do not support the perception of overuse of ECPs in Jamaica, nor does it support the notion that ECPs are being purchased in advance of need (i.e., before underprotected or unprotected sex actually occurs). In our study, most women did not buy ECPs more than once in the 6month study period; among those who did, the majority reported using ECPs only two to three times in the last 18
months (i.e., in the last year plus the 6-month study period). Less than 1% of participants at intercept indicated that they purchased Postinor before they had sex, and it was not clear whether any advance purchase occurred during the 3-month and 6-month follow-up periods. 4.4. Study limitations The wording of some questionnaire items may have limited the interpretability of our findings. For example, our measure of frequency of ECP use at the intercept interview may have produced underestimates in our participants' reports due to the lengthy time frame. Asking women how many times they purchased ECPs in the last 3 months or how often they used them in an average month may have been better than asking about ECP use in the last 12 months as we did. Nevertheless, we documented that ECP use over the 6month study period was low. The ability to generalize these findings is also limited by the fact that, in some countries such as the United States, OCPs are not available over-the-counter. As such, it would not be possible to bridge ECP users to another method solely in pharmacy settings, adding yet another hurdle to women's adoption of an ongoing contraceptive method. The strengths of the study reside in its randomization of participants (which resulted in well-balanced treatment groups) and the unequivocal finding that although the discount coupon per se did not encourage this group of women to adopt an ongoing contraceptive method, a small number did so on their own. Furthermore, the findings suggest that ECP users are not a monolithic group and that special effort should be undertaken to identify those who are amenable to a regular contraceptive method and for whom bridging to a regular contraceptive method would be an appropriate approach. A better understanding of why some women bridge while others fail to adopt a regular method is needed to design more effective bridging interventions. Given cultural differences, as well as beliefs about the effectiveness of ECPs and the advantages and disadvantages of other methods, information that expressly compares the advantages of regular methods over ECPs should be a key component in the materials and (when available) counseling provided to these women.
Acknowledgments The authors would like to express appreciation to Drs. John Stanback, Kelly L'Engle and Elizabeth Raymond for their helpful comments on earlier drafts of the manuscript. The first author would also like to thank the Pharmaceutical Society of Jamaica, the National Family Planning Board and Medimpex for their collaboration, support and participation through all phases of study development and implementation.
D.S. Chin-Quee et al. / Contraception 81 (2010) 133–139
References [1] Raymond EG, Trussell J, Polis CB. Population effect of increased access to emergency contraceptive pills: a systematic review. Obstet Gynecol 2007;109:181–8. [2] Trussell J, Schwarz EB, Guthrie K, Raymond E. No such thing as an easy (or EC) fix. Contraception 2008;78:351–4. [3] Trussell J, Stewart F, Guest F, Hatcher RA. Emergency contraceptive pills: a simple proposal to reduce unintended pregnancies. Fam Plann Perspect 1992;24:269–73. [4] International Consortium for Emergency Contraception. EC status and availability: searchable databases. http://www.cecinfo.org/database/pill/ pillData.php. New York: Family Care International; 2008. Accessed on December 14.
139
[5] Trussell J, Guthrie KA. Talking straight about emergency contraception. J Fam Plann Reprod Health Care 2007;33:139–42. [6] Altman DG, Schulz KF, Moher D, Egger M, Davidoff F, Elbourne D, et al. The revised CONSORT statement for reporting randomized trials: explanation and elaboration. Ann Intern Med 2001;134:663–94. [7] Bourne D, Wedderburn M. Report on dual method use in Jamaica. Kingston (Jamaica): Hope Enterprises; 2007. [8] Frost JJ, Darroch JE. Factors associated with contraceptive choice and inconsistent method use, United States, 2004. Perspect Sex Reprod Health 2008;40:94–104. [9] Wamatu C. EC use in Kenya: a profile of EC users and non-users. Bridging from emergency contraceptive pills to ongoing contraception: consultation for the Hewlett Foundation. Research Triangle Park (NC): Family Health International (FHI); 2007.