1 The Family Planning Association of Hong Kong, 9/F 130 Hennessy Road, Hong Kong SAR, 2 Department of Obstetrics and Gynaecology, The University of Hong Kong, Queen Mary Hospital, Pokfulam Road, Hong Kong SAR, China and 3 Department of Obstetrics and Gynaecology, University of Edinburgh, Department of Public Health Policy, University of London School of Hygiene and Tropical Medicine and NHS Lothian Family Planning and Well Woman Services, 18 Dean Terrace, Edinburgh EH4 1NL, UK
4 To whom correspondence should be addressed at: The Family Planning Association of Hong Kong, 9/F Southorn Centre, 130 Hennessy Road, Wanchai, Hong Kong SAR, China. Email: stlo{at}famplan.org.hk
![]() |
Abstract |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Key words: advanced provision/contraceptive behaviour/emergency contraception/levonorgestrel
![]() |
Introduction |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
![]() |
Materials and methods |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Recruitment and study procedure
All women agreeing to participate in the trial were educated about EC and given an information sheet to keep and completed a form recording demographic details.
Women randomized to the control group were instructed to attend the clinic for EC whenever they needed it. Women in the intervention group were given three courses of Norlevo® (two tablets of 0.75 mg levonorgestrel; HRA Pharma Laboratorie, France) to keep at home. They were instructed (verbally and in writing) to take one tablet of levonorgestrel as soon as possible within 72 h after a single episode of unprotected intercourse, and the second tablet 12 h later. If they used all three courses of EC, women in the intervention group were asked to attend the clinic for counselling about regular contraception. Three more courses of levonorgestrel were dispensed, if appropriate. If all six courses were used, no further supplies of study drug were dispensed but the follow-up continued until the end of the study.
All women were instructed to return to the clinic for a pregnancy test if menses was overdue by >1 week following EC use. If the pregnancy test was positive, women who chose to have the pregnancy terminated were given the option to remain in the study whereas those who chose to continue the pregnancy were withdrawn.
Outcomes measured
The primary objective was to study how women behave when given EC to keep at home. The two outcomes measured were the number of course(s) of EC used and the use of other methods of contraception (contraceptive behaviour). Contraceptive behaviour was assessed in terms of changes in method use (switching), consistency of contraceptive use and the frequency of unprotected intercourse.
The secondary objective was to assess whether women could self-administer EC correctly. The outcomes measured included the timing of taking the first pill, whether it was taken after a single episode of unprotected intercourse or contraceptive failure (or after more than one episode), and whether the second dose was taken 12 h later. Although the study was not powered to detect a significant difference in pregnancy rates between the two groups of women, the rates were recorded.
All women were asked to keep a diary of contraceptive use (including EC) and sexual intercourse. They were contacted every 3 months by telephone. The contraceptive method and consistency of use, the number of acts of intercourse and the number of episodes of EC use each month and any pregnancies were documented during the phone call.
Women with a supply of EC at home were instructed to complete a drug intake record whenever they used EC. The record sheet was either mailed to the principal investigator after completion (33 forms), or kept until the next follow-up contact when the details were given to the researcher over the telephone. The record documented the reason for using EC, the date of the last menstrual period, the timing and number of episodes of unprotected intercourse or contraceptive accidents in that cycle, the time when EC was taken and whether pregnancy had resulted or not.
The same drug intake record was collected for women in the control group if they attended the clinic for EC, or if EC was obtained from another provider.
Statistical analysis
The number of women requesting EC in the YHCC each month is 20%. Based on these data it was estimated that 20% of women would use EC at least once a year. A 10% increase in the use of EC resulting from advanced provision was regarded as clinically significant. The sample size necessary to achieve 90% power, with a significant difference at the 5% level, was estimated to be 389 women in each arm of the study. Assuming 20% drop-out, the target sample size for each arm was 490. Recruitment was continued until the end of February despite reaching the target sample size in mid February. The final number of subjects recruited was 515 in each arm.
SPSS (version 11.5 for Windows) was used for statistical analysis. Descriptive statistics were presented for demographic factors. The pattern of EC use and the significance of any difference were tested with 2-test for trend. Student's t-test was used to compare mean age, parity and previous number of abortions between the intervention and control groups. McNemar's test was used to test for changes in contraceptive behaviour over time in all women. P<0.05 was statistically significant.
![]() |
Results |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Study subjects
A total of 1030 women were recruited during the study period, 750 from the BCC and 280 from the YHCC. Of 515 women randomized to the control group, 487 subjects (94.6%) completed 12 months follow-up, 15 defaulted and 13 withdrew. In the intervention group, 499 subjects (96.9%) completed 12 months follow-up, eight defaulted and eight withdrew. The reasons for withdrawal are shown in Figure 1. The baseline demographic characteristics and reproductive histories did not differ between the two groups of women. Of the women recruited from the BCC, 85.9% were married compared with none of those recruited from the YHCC. The number of episodes of EC used during the 6 months before the study started was also similar for both groups of women (P=0.377) (Table I).
|
|
The pattern of EC use during the year of follow-up is summarized in Table II. Twice as many women (n=149, 29.9%) in the intervention group had used at least one course of EC compared with women in the control group (n=63, 12.9%). Women with a supply of EC at home used almost three times more EC than women who had to attend the clinic. The difference in the pattern of use was statistically significant (P<0.001).
|
|
|
The reasons for using EC are listed in Table V. Women who had a supply of EC at home were more likely to use it following an act of unprotected intercourse (rather than following a contraceptive accident) than women who had to attend the clinic to obtain EC (P<0.001).
|
Data for the time of intercourse and time the first pill was taken were available in 327 records (95 from the control group and 232 from the intervention group). With only one exception, all the women used EC within 72 h of intercourse. Women who had a supply at home used EC significantly sooner after intercourse than women who had to attend a clinic to obtain it. The mean±SD time interval from intercourse to first dose taken was 28.5±19.8 h for the control group and 13.9±14.4 h for the intervention group (P<0.001).
All subjects in the control group took the second dose correctly, i.e. 12 h after the first dose. In the intervention group, 38 subjects out of 229 (16.6%) did not take it correctly.
There were 321 records (84.3%) with the number of episodes of intercourse prior to EC use documented. On 284 occasions (88.5%) EC was used after a single act of unprotected intercourse or contraceptive failure. There were 37 episodes (15 from the group who had to obtain EC from a clinic versus 22 from the group who had EC at home, P=0.110) when EC was taken after more than one episode of unprotected intercourse or contraceptive accident. One of these women conceived.
Secondary outcome: pregnancy rate
The pregnancy rate was similar in both groups. There were nine unplanned intrauterine pregnancies in nine women from the control group and none of them had used EC in the cycle in which conception occurred. They did not take EC because four of them thought that the risk of becoming pregnant was low that time, four believed that the contraceptives they used were successful and one could not obtain EC. In the intervention group, there were eight unplanned intrauterine pregnancies in seven subjects. Five women, including the one who became pregnant twice, did not take EC despite having a supply at home. The main reasons for not taking EC on the six occasions were again perceived low risk of getting pregnant (four women) and two believed that the contraceptives were successful. Out of 373 cycles in which EC was used, there were two pregnancies, a rate of 0.54%. The 95% confidence interval for the odds ratio for pregnancy was 0.282.05.
![]() |
Discussion |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Easier access to EC does not appear, among this population of women, to lead to repeated use. Only a few women appeared to rely on EC for contraception. Only 13 women used more than three courses in 1 year. Admittedly, some women might have used more had they been given more than three courses at the start of the study, nevertheless only 25 (3%) used all three courses. The two young women who used it more than six times obtained the additional courses from elsewhere, suggesting that it was not advanced provision per se which encouraged repeated use.
Women who had a supply of EC at home used it on 71% of occasions following unprotected intercourse. In contrast, only just over half the women who had to attend the clinic for supplies had had unprotected intercourse, the rest using it after an accident with their contraceptive method. It is possible that some women who had EC at home had unprotected sex, intending to use EC as a planned post-coital contraceptive. There is no evidence, however, from the data collected at follow-up, that women in the intervention group used condoms less consistently than women in the control group, and other studies have not shown any increase in risk-taking behaviour. If easier access to EC encouraged less consistent contraceptive use, one might also have expected more pregnancies in the intervention group. It is possible that women who had to see a health professional to obtain EC were reluctant to admit to having unprotected sex and so claimed that a condom had been used even when it had not.
The Hong Kong study is the first to give advanced supplies of levonorgestrel; all the earlier studies offered the Yuzpe regimen of EC. Since the latter contains estrogen and is associated with a much higher incidence of side-effects, particularly nausea and vomiting (Task Force on Postovulatory Methods of Fertility Regulation, 1998), there was some concern among the investigators that the use of a better-tolerated regimen of EC might be associated with more repeated use. This does not seem to have happened. In the Hong Kong study, 30% women from the intervention group used EC at least once compared with 47% in the Scotland study (Glasier and Baird, 1998
), which involved a similar number of women.
Giving women a supply of EC to keep at home did not appear to affect their use of more effective methods of regular contraception. This was the case even for the women aged <26 years. This is in contrast to the findings of a study undertaken among women aged 1624 years in California (Raine et al., 2000) where women given EC in advance of need tended to use less effective regular methods of contraception and to use their method less consistently. While both of the Scottish studies (Glasier and Baird, 1998
; Glasier et al., 2004
) showed a clear trend with time for women to change from condoms to hormonal contraception, women in Hong Kong stuck with condoms. Indeed the consistency of method use during the year of follow-up is quite remarkable and contrasts with patterns of method-switching in other populations (Grady et al., 2002
). Condoms are a very popular method of contraception in Hong Kong (Glasier et al., 2003
). In this setting, improved access to EC does not appear to undermine the use of barrier methods of contraception. This finding will be reassuring to those who express concern that easier access to EC will lead to an increase in unsafe sex, and thereby an increase in sexually transmitted infections including HIV/AIDS (Stammers, 2001
; Bissell and Anderson, 2003
).
Although a randomized trial, the study has several limitations. Young unmarried women attending the clinic were more likely to agree to participate in the study than the older women. This may reflect the fact that different investigators were responsible for recruitment in the two clinics. It is more likely, however, that the older women did not perceive themselves as likely to need to use EC and therefore did not want it to keep at home. In assessing coital frequency, contraceptive behaviour and correct use of EC, we tried to minimize recall bias by asking women to keep a record of their behaviour. The data were gathered during the telephone follow-up but very few record forms were returned to the investigators. We designed the follow-up record to include some questions which would allow cross-validation of the responses. We found a reassuringly high reliability of responses to all the questions, but we do acknowledge that probably not all women kept complete or accurate records. Some will have relied on recall of events, but as follow-up was every 3 months it is likely that the data, at least for contraceptive use, were reasonably reliable. We were also aware of the possibility that behavioural change in this kind of observational study may be attributable to confounding factors such as participation in the study. It is almost certainly for this reason that contraceptive use was reported to be more consistent during the study than before it began. Moreover, tightly controlled studies of this nature may not truly represent how women would behave were they not participating in a research project. A strength of the study, however, lies in the fact that all the women were recruited, informed about EC, and followed up, by one of only two investigators.
Only two other studies of advanced provision of EC have looked at the effect on abortion. The first (Glasier and Baird, 1998) suggested a non-significant trend towards a reduction in unintended pregnancies. A much larger (but non-randomized) intervention study in the same setting (Glasier et al., 2004
), however, failed to show any effect on abortion rates. Both Scottish studies used the Yuzpe regimen of EC. Levonorgestrel is thought to be more effective, but we too were unable in this study to demonstrate any difference in pregnancy rates between the control and intervention groups. As in the Scottish study (Glasier et al., 2004
), most of the women who conceived despite having a supply of EC to use, failed to recognize any risk of pregnancy and so failed to use EC. Out of the 17 unplanned pregnancies, only two had used EC.
In conclusion, advanced provision of EC to women attending a family planning service in Hong Kong demonstrated increased use of EC, without appearing to encourage repeated use or the abandonment of their regular contraceptives. Despite having a supply at home, most women who had an unintended pregnancy during the study did not use EC. The results suggest, however, that advanced provision of EC as a public health strategy to reduce unintended pregnancy may prove disappointing.
![]() |
Acknowledgements |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
![]() |
References |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Ellertson C, Ambardekar S, Hedley A, Coyaji K, Trussell J and Blanchard K (2001) Emergency contraception: randomized comparison of advanced provision and information only. Obstet Gynecol 98, 570575.
Glasier AF and Baird DT (1998) The effects of self-administering emergency contraception. N Engl J Med 339, 14.
Glasier AF, Smith KB, van der Spuy ZM, Ho PC, Cheng L, Dada K, Wellings K and Baird DT (2003) Amenorrhoea associated with contraceptionan international study on acceptability. Contraception 67, 18.[CrossRef][ISI][Medline]
Glasier AF, Fairhurst K, Wyke S, Ziebland S, Seaman P, Walker J and Lakha F (2004) Advanced provision of emergency contraception has not reduced abortion rates. Contraception 69, 361366.[CrossRef][ISI][Medline]
Grady WR, Billy JOG and Klepinger DH (2002) Contraceptive method switching in the United States. Perspect Sex Reprod Health 34(3), 135145.
Hayes M, Hutchings J and Hayes P (2000) Reducing unintended pregnancy by increasing access to emergency contraceptive pills. Matern Child Health J 4, 203208.[CrossRef][Medline]
Jackson RA, Schwarz EB, Freedman L and Darney P (2003) Advance supply of emergency contraception: effect on use and usual contraceptiona randomized trial. Obstet Gynecol 102, 816.
Jones RK, Darroch JE and Henshaw SK (2002) Contraceptive use among US women having abortions in 20002001. Perspect Sex Reprod Hlth 34, 294303.
Lovvorn A, Nerquaye-Tetteh J, Glover EK, Amankwah-Poku A, Haya M and Raymond E (2000) Provision of emergency contraceptive pills to spermicide users in Ghana. Contraception 61, 287293.[CrossRef][ISI][Medline]
Raine T, Harper C, Leon K and Darney P (2000) Emergency contraception: advanced provision in a young, high-risk clinic population. Obstet Gynecol 96, 17.
Stammers T (2001) Pharmacists and STIs. Br Med J 323, 1416.
Task Force on Postovulatory Methods of Fertility Regulation (1998) Randomised controlled trial of levonorgestrel versus the Yuzpe regimen of combined oral contraceptives for emergency contraception. Lancet 352, 428433.[CrossRef][ISI][Medline]
von Hertzen H, Piaggio G, Ding J, Chen J, Song S, Bartfai G, Ng E, Gemzell-Danielsson K, Oyunbileg A, Wu S et al. (2002) Low dose mifepristone and two regimens of levonorgestrel for emergency contraception: a WHO multicentre randomized trial. Lancet 360, 18031810.[CrossRef][ISI][Medline]
World Health Organization (1998) Emergency contraception: a guide for service delivery. WHO, Geneva, WHO/FRH/FPP/9819.
Submitted on April 2, 2004; accepted on June 25, 2004.
|