PRACTICE

nature publishing group

Emergency Contraception Update: A Canadian Perspective VWY Leung1, JA Soon1,2 and M Levine1,3

Barriers to hormonal emergency contraceptive (EC) access in Canada and the United States led professional and lay groups to lobby for levonorgestrel (LNG) (PLAN B, Barr Pharmaceuticals, Pomona, New York) to be made available over-the-counter. In December 2000, British Columbia, Canada, granted EC prescriptive authority to pharmacists, followed by Quebec in December 2001 and Saskatchewan in September 2003. In April 2005, Health Canada placed LNG on non-prescription, behind-the-pharmacy-counter status with no age restriction. After much controversy, in August 2006, the Food and Drug Administration (FDA) approved over-the-counter access to LNG by adults in the United States. Results of our experience in Canada and recent information regarding mechanisms of action, effectiveness, adverse effects, and the effect of increased availability on reproductive health outcomes are presented here to help inform clinical practice. MECHANISMS OF ACTION

Multiple mechanisms of action have been proposed for the traditional Yuzpe EC regimen (two doses, each containing 0.1 mg of ethinyl estradiol and 0.5 mg of LNG or 1.0 mg of norgestrel, taken 12 h apart) and LNG (two doses of 0.75 mg of LNG, 12 h apart, or 1.5 mg taken as a single dose), the most likely of which involves inhibiting or delaying ovulation.1 In a recent pilot study, Novikova et al.2 used endocrine data to assess ovulation day among 99 women seeking ECs. Their findings are consistent with the conclusion that LNG is unlikely to interfere with postovulatory events.2 In contrast, Mikolajczyk and Stanford3 performed follicular growth modeling studies and concluded that a preovulatory effect alone cannot account for the estimated effectiveness of LNG. Further studies of sufficient size that combine the methods used by these investigators are needed, and they should be able to resolve the question of the predominant mechanism(s) of action of these EC agents.

EFFECTIVENESS

The terms ‘‘efficacy’’ and ‘‘effectiveness’’ usually refer to how well interventions work in clinical trial and routine practice settings, respectively. Placebo-controlled studies would be ideal for evaluating the efficacy of EC regimens. However, it would not be ethical to assign women to placebo because ECs are considered effective in preventing unwanted pregnancy.4 Therefore, efficacy and effectiveness of ECs must be estimated indirectly from the pregnancy rate observed after treatment, relative to the expected rate in the absence of treatment.4,5 Expected pregnancy rates following unprotected intercourse have been obtained from studies that estimated the probability of pregnancy among women trying to conceive.5 Pooled data from early studies using an assumed day of ovulation suggested that the Yuzpe and LNG regimens prevent 74 and 79% of expected pregnancies, respectively.5,6 In contrast, Wilcox et al.7 addressed uncertainty in the day of ovulation by estimating the conception probability on every cycle day after the onset of menstruation. Using this approach, Trussell et al.4 estimated that the Yuzpe regimen is only 46.8–53.0% effective. However, the data from which these estimates were made involved a self-reported cycle day and an estimated day of ovulation, which may not be accurate. Since then, Mikolajczyk and Stanford8 have argued that estimating pregnancy risk on the basis of the onset of menses without accounting for average cycle length also results in overestimation of expected pregnancy rates, and they developed a method incorporating biologic variability in ovulation and length of the previous cycle.8 However, without information on the previous menstrual cycle length, the approach of Wilcox et al.7 may be the least biased for estimating pregnancy risk. Using the Wilcox et al.7 method, Trussell et al.4 calculated that the expected pregnancy rate was 5.4% among 675 women from one trial and was only 3.9% among B20,000 women seeking ECs from the same centers during the study period. In British Columbia, from December 2000 to December 2002, B13,000 women who obtained ECs from

1

Faculty of Pharmaceutical Sciences, University of British Columbia, Vancouver, British Columbia, Canada; 2Department of Pharmacy, School of Pharmacy, University of Washington, Seattle, Washington, USA; 3Department of Pharmacy, Children’s and Women’s Hospitals of British Columbia, Vancouver, British Columbia, Canada. Correspondence: JA Soon ([email protected]) Published online 28 November 2007. doi:10.1038/sj.clpt.6100443

CLINICAL PHARMACOLOGY & THERAPEUTICS | VOLUME 83 NUMBER 1 | JANUARY 2008

177

PRACTICE

pharmacists gave written consent for treatment that included demographic and reproductive health information, the dates of onset of the last menstrual cycle, and time of unprotected intercourse. Using anonymized versions of these consents, we estimated the overall probability of pregnancy among women who obtained the Yuzpe or LNG regimen from pharmacists to be 4.1%.9 As shown in Figure 1, these women tended to request ECs around cycle days 14–15, although a substantial proportion presented earlier or later after the onset of menses. Thus, because the early methods overestimated expected pregnancy rates, and for a number of other reasons, estimates of the effectiveness of ECs in the literature have been biased upward.6 On the basis of the trial data summarized by Stanford and Mikolajczyk6 and results of other studies,10,11 we calculated the pooled observed pregnancy rates for women who received the Yuzpe or LNG regimen to be 2.0 and 1.8%, respectively. With the rough estimates currently available, the use of ECs in routine practice may reduce the overall pregnancy rate from an expected 4–5% to B2%, an absolute risk reduction of 2–3% (number needed to treat (NNT): 33–50) and a relative reduction of 50–60%. Comparison of the Yuzpe and LNG regimens

Only two clinical trials have compared the effectiveness of the Yuzpe and LNG regimens.12,13 Ho and Kwan12 found that the pregnancy rates among B800 women randomized to receive the Yuzpe or LNG regimen were 2.7 and 2.4%, respectively. A WHO-sponsored multinational study involving nearly 2,000 women reported pregnancy rates of 2.9 and 1.0% (relative risk 0.36; 95% confidence interval: 0.17–0.73) in the Yuzpeand LNG-treated groups, respectively, when the four preexisting pregnancies were excluded.13 The latter study is the only direct evidence that LNG is more effective than the Yuzpe regimen, and it is frequently, and probably erroneously, cited to claim that LNG is B89% effective. To obtain a population-based comparison of effectiveness, we are currently linking the British Columbia health data (a province-wide linkable set of patient-specific health and pharmacy data) to

Figure 1 Probability of pregnancy among women requesting ECs from pharmacists in British Columbia. Bars represent the predicted probability of pregnancy on each day of the menstrual cycle. Open circles (Yuzpe cohort) and open triangles (LNG cohort) represent the women’s self-reports of the day in the menstrual cycle on which unprotected intercourse occurred.9 Reprinted with permission from the J. Obstet. Gynaecol. Can. 178

our database of anonymized consents for treatment to compare the pregnancy rates among women who received the Yuzpe or LNG regimen. This will provide the first estimate of effectiveness of these agents under conditions of routine care. Timing

The Yuzpe and LNG regimens were approved for use up to 72 h after unprotected intercourse, and studies tended to exclude women who presented beyond this period. On the basis of several underpowered studies, current Canadian guidelines recommend that ECs be offered up to 5 days after unprotected intercourse.14 Clearly, larger studies are needed to confirm an effect of ECs at 72–120 h after unprotected intercourse. ADVERSE EFFECTS

The most notable minor adverse effects reported in comparative trials are nausea, vomiting, dizziness, fatigue, breast tenderness, and irregular bleeding or spotting, which have been observed more frequently with the Yuzpe regimen than with the LNG regimen.12,13 There is little evidence of serious adverse effects of ECs. Venous thrombosis is a possibility, isolated cases of which have been reported among women taking ECs.15 Vasilakis et al.16 observed no thromboembolic events in a cohort exposed to a total of 100,615 prescriptions for the Yuzpe regimen. Some researchers advocate using a progestin-only regimen instead of the estrogen-containing Yuzpe regimen in women with ‘‘significant contraindications to estrogen,’’ including a history of thrombosis, stroke or heart attack, migraine headache with neurological symptoms, or smokers over 35 years of age.14 The risk should otherwise be very low, given that ECs are short-course treatments.15 Few post-marketing data are available on the outcomes of pregnancies among women who became pregnant after receiving ECs. The latest Cochrane review noted that four women exposed to the LNG regimen and two exposed to the Yuzpe regimen delivered healthy infants.17 An earlier report documented the outcomes of 119 continuing pregnancies after EC treatment, and major abnormalities were reported in five of these cases.18 This incidence is not more than would be expected in the general population, and in each case, the mother had been exposed to tobacco or co-medication.18 Spontaneous reports of ectopic pregnancy in the literature include five cases after the Yuzpe regimen and more than 30 cases after the LNG regimen,19 but the incidence cannot be estimated from these reports. As a precaution, follow-up monitoring has been recommended, ‘‘particularly in women with a previous ectopic pregnancy, fallopian tube surgery, or pelvic inflammatory disease’’.20 The potential teratogencity of LNG has been evaluated in a small study by De Santis et al.21 No association was found between LNG failure and increased risk of major congenital malformations, pre- or peripartum complications, or an adverse pregnancy outcome. Moreover, studies of regular oral contraceptives have found no risk or an extremely low risk at the most, even at high doses.15 VOLUME 83 NUMBER 1 | JANUARY 2008 | www.nature.com/cpt

PRACTICE

ECs, risk-taking behavior, and sexually transmitted infections

IMPACT OF INCREASED AVAILABILITY

Critics of wider EC availability have expressed concern over its potential to increase sexual risk-taking behavior, especially among young women.22 Raine et al.22 observed that the availability of additional EC packages (a practice known as ‘‘advance provision’’) increased EC use but was not associated with increased rates of unprotected intercourse, pregnancy, chlamydial infection, or herpes simplex at 6 months. Among teenagers in their study, advance provision was associated with greater EC use, but no differences were observed for other outcomes.23 Thus, available data from these and other studies24 do not suggest an increase in risk-taking behavior or sexually transmitted infections with increased availability and use of ECs. During the 2 years after prescriptive authority was granted, pharmacists in British Columbia instructed women to consult a physician in 35.3 and 3.2% of cases for birth control and sexually transmitted infection screening, respectively.25 Now that ECs are available without a prescription in Canada and the United States, the impact of the potential reduction in screening for sexually transmitted infections warrants investigation.

The FDA’s decision to grant over-the-counter status to Plan B for women aged 18 years and above is anticipated to result in increased EC use. In British Columbia, granting of prescriptive authority to pharmacists resulted in a doubling of the number of EC prescriptions during the first 2 years (Figure 2).25 Also, direct pharmacist-provided ECs enabled women to access treatment within 24 h of unprotected intercourse in 56% of cases in British Columbia and in 54% of cases in Ontario.25,26 Since Plan B was approved for behind-the-counter sale in Canada in 2005, sales to pharmacies and hospitals have more than doubled (Figure 3). These data were provided by IMS Health Canada, which maintains comprehensive national prescription databases of purchases made by Canadian retail pharmacies and hospitals from wholesalers and manufactures. Anecdotal reports suggest a similar increase in Plan B sales in the United States.27 Easier access to ECs does not appear to increase repeat use among adults or adolescents.22,23,28 In British Columbia, there was no change in the repeat use of ECs after pharmacists were granted prescriptive authority, and more than 96% of women used ECs only once or twice per year.25 Also, women aged 10–17, 18–24, and 25–49 years had almost identical rates of repeat use of ECs (J.A. Soon and M. Levine, unpublished data). Although LNG use after every act of intercourse was not associated with any serious adverse effects among 295 women,29 safety and effectiveness data are lacking for this scale of repeat use. Several studies have demonstrated cost-effectiveness of ECs in Canada and the United States.4 Using our populationbased cohort, we modeled costs and savings associated with ECs and found that the provincial government could spend as much as CDN$2.4 million to increase women’s awareness and use of ECs, and treatment would still be cost saving.30 Therefore, when used appropriately, ECs may be cost saving to third-party payers.

Health policy change

Number of ECs

20,000

Physician

Pharmacist

15,000 10,000 5,000 0

1996

1997

1998

1999

2000

2001

2002

Year

Figure 2 Pattern of EC provision by type of provider and by year in British Columbia.

35,000

BC pharmacist

Quebec pharmacist

SK pharmacist

Canada OTC

Number plan B treatments sold

30,000

25,000

20,000 Total plan B treatments

15,000

10,000

5,000

– Jul-00 Jan-01 Jul-01 Jan-02 Jul-02 Jan-03 Jul-03 Jan-04 Jul-04 Jan-05 Jul-05 Jan-06 Jul-06 Jan-07 Date

Figure 3 Plan B treatments sold to drugstores and hospitals in Canada from July 2000 to June 2007. CLINICAL PHARMACOLOGY & THERAPEUTICS | VOLUME 83 NUMBER 1 | JANUARY 2008

179

PRACTICE

IMPACT ON PREGNANCY AND ABORTION

The objective in making ECs more widely available is to reduce unwanted pregnancies and subsequent induced abortions. However, a recent systematic review indicated that increased access to ECs has not yet been shown to reduce unwanted pregnancy or abortion rates.31 While the studies included tended to be underpowered, and some had methodological weaknesses, there may be other reasons for this negative finding, and it is essential that this be investigated if population-wide EC use is to be evidence based. SUMMARY AND CONCLUSIONS

ECs appear to act largely by preventing ovulation; however, postovulatory mechanisms have not been ruled out. The true effectiveness of hormonal EC regimens under conditions of routine use cannot be determined but has most likely been overestimated to date. Under conditions of routine practice, EC use within 72 h of unprotected intercourse may result in an overall absolute reduction of 2–3% in pregnancy rate (NNT: 33–50). Use beyond 72 h may be effective and carries little or no risk of serious adverse effects. Should a woman become pregnant after exposure to the Yuzpe or the LNG regimen, there is no evidence of increased risk to the fetus. The effects of the increased availability and use of EC in Canada and the United States on unwanted pregnancy and consequent abortion rates must be investigated. ACKNOWLEDGMENTS We express our appreciation to Brian Carter and Anne-Maxime Dagenais of IMS Health Canada for the emergency contraception Canadian Drug Store and Hospital Purchases Audit data. This work was supported by the Canadian Institutes of Health Research, British Columbia Medical Services Foundation and the British Columbia Ministry of Health Planning. CONFLICT OF INTEREST The authors declared no conflict of interest.

8.

9.

10.

11.

12.

13.

14. 15. 16.

17. 18. 19.

20. 21.

22.

23.

24. & 2008 American Society for Clinical Pharmacology and Therapeutics

1. 2.

3.

4. 5.

6.

7.

180

Trussell, J. & Jordan, B. Mechanism of action of emergency contraceptive pills. Contraception 74, 87–89 (2006). Novikova, N., Weisberg, E., Stanczyk, F.Z., Croxatto, H.B. & Fraser, I.S. Effectiveness of levonorgestrel emergency contraception given before or after ovulation—a pilot study. Contraception 75, 112–118 (2007). Mikolajczyk, R.T. & Stanford, J.B. Levonorgestrel emergency contraception: a joint analysis of effectiveness and mechanism of action. Fertil. Steril. 88, 565–571 (2007). Trussell, J. et al. Estimating the effectiveness of emergency contraceptive pills. Contraception 67, 259–265 (2003). Trussell, J., Rodriguez, G. & Ellertson, C. Updated estimates of the effectiveness of the Yuzpe regimen of emergency contraception. Contraception 59, 147–151 (1999). Stanford, J.B. & Mikolajczyk, R.T. Methodological review of the effectiveness of emergency contraception. Curr. Womens Health Rev. 1, 119–129 (2005). Wilcox, A.J., Dunson, D.B., Weinberg, C.R., Trussell, J. & Baird, D.D. Likelihood of conception with a single act of intercourse: providing benchmark rates for assessment of post-coital contraceptives. Contraception 63, 211–215 (2001).

25.

26.

27.

28. 29.

30.

31.

Mikolajczyk, R.T. & Stanford, J.B. A new method for estimating the effectiveness of emergency contraception that accounts for variation in timing of ovulation and previous cycle length. Fertil. Steril. 83, 1764–1770 (2005). Levine, M. & Soon, J.A. Risk of pregnancy among women seeking emergency contraceptives from pharmacists in British Columbia. J. Obstet. Gynaecol. Can. 28, 879–883 (2006). Hamoda, H. et al. A randomized trial of mifepristone (10 mg) and levonorgestrel for emergency contraception. Obstet. Gynecol. 104, 1307–1313 (2004). Ngai, S.W. et al. A randomized trial to compare 24 h versus 12 h double dose regimen of levonorgestrel for emergency contraception. Hum. Reprod. 20, 307–311 (2005). Ho, P.C. & Kwan, M.S. A prospective randomized comparison of levonorgestrel with the Yuzpe regimen in post-coital contraception. Hum. Reprod. 8, 389–392 (1993). 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 352, 428–433 (1998). Black, A. et al. Canadian contraception consensus. J. Obstet. Gynaecol. Can. 26, 143–156 (2004). Norris Turner, A. & Ellertson, C. How safe is emergency contraception? Drug Saf. 25, 695–706 (2002). Vasilakis, C., Jick, S.S. & Jick, H. The risk of venous thromboembolism in users of postcoital contraceptive pills. Contraception 59, 79–83 (1999). Cheng, L. et al. Interventions for emergency contraception. Cochrane Database Syst. Rev. 3, CD001324 (2004). Cardy, G.C. Outcome of pregnancies after failed hormonal postcoital contraception—an interim report. Br. J. Fam. Plann. 21, 112 (1995). Matsushita, H., Takayanagi, T. & Ikarashi, H. Ectopic pregnancy following emergency contraception with ethinyloestradiol–levonorgestrel: a case report. Eur. J. Contracept. Reprod. Health Care 12, 184–186 (2007). Vinson, D.R. Emergency contraception and risk of ectopic pregnancy: is there need for extra vigilance? Ann. Emerg. Med. 42, 306–307 (2003). De Santis, M., Cavaliere, A.F., Straface, G., Carducci, B. & Caruso, A. Failure of the emergency contraceptive levonorgestrel and the risk of adverse effects in pregnancy and on fetal development: an observational cohort study. Fertil. Steril. 84, 296–299 (2005). Raine, T.R. et al. Direct access to emergency contraception through pharmacies and effect on unintended pregnancy and STIs: a randomized controlled trial. JAMA 293, 54–62 (2005). Harper, C.C., Cheong, M., Rocca, C.H., Darney, P.D. & Raine, T.R. The effect of increased access to emergency contraception among young adolescents. Obstet. Gynecol. 106, 483–491 (2005). Polis, C.B. et al. Advance provision of emergency contraception for pregnancy prevention [full rev.]. Cochrane Database Syst. Rev. 2, CD005497 (2007). Soon, J.A., Levine, M., Osmond, B.L., Ensom, M.H. & Fielding, D.W. Effects of making emergency contraception available without a physician’s prescription: a population-based study. CMAJ 172, 878–883 (2005). Dunn, S. et al. Pharmacy provision of emergency contraception: the Ontario emergency contraception pilot project. J. Obstet. Gynaecol. Can. 25, 923–930 (2003). Reuters. Morning-after pill sales jump as access eases /http:// edition.cnn.com/2007/HEALTH/08/24/contraceptive.reut/ index.htmlS. Accessed 24 August (2007). Glasier, A. & Baird, D. The effects of self-administering emergency contraception. N. Engl. J. Med. 339, 1–4 (1998). Task Force on Post-ovulatory Methods for Fertility Regulation. Efficacy and side effects of immediate postcoital levonorgestrel used repeatedly for contraception. Contraception 61, 303–308 (2000). Soon, J.A. et al. Modelling costs and outcomes of expanded availability of emergency contraceptive use in British Columbia. Can. J. Clin. Pharmacol 14, E326–E338 (2007). Raymond, E.G., Trussell, J. & Polis, C.B. Population effect of increased access to emergency contraceptive pills: a systematic review. Obstet. Gynecol. 109, 181–188 (2007).

VOLUME 83 NUMBER 1 | JANUARY 2008 | www.nature.com/cpt