Dispensing of clomiphene citrate to

treat infertility: medication supplied

and population prevalence of
assisted pregnancies and
multiple births
Vivienne Moore, Ph.D.,a,b Alice Rumbold, Ph.D.,b,c Renae Fernandez, Ph.D.,b,d
Heather McElroy, M.Med.Stat.,d Lynette Moore, M.B.B.S., M.B.A.,e,f Lynne Giles, Ph.D.,a,b
Luke Grzeskowiak, Ph.D.,b,c,d,g Elizabeth Roughead, Ph.D.,h Michael Stark, M.B.Ch.B., Ph.D.,b,d
and Michael Davies, Ph.D.b,d
a
School of Public Health, and b Robinson Research Institute, The University of Adelaide, South Australia; c South Australian
Health and Medical Research Institute, South Australia; d Discipline of Obstetrics and Gynaecology, The University of
Adelaide, South Australia; e SA Pathology, Women’s and Children’s Hospital, Adelaide, South Australia; f Discipline of
Medical Sciences, The University of Adelaide, South Australia; g College of Medicine and Public Health, Flinders
University, South Australia; and h Quality Use of Medicines and Pharmacy Research Centre, University of South Australia,
South Australia

Objective: To determine the proportion of pregnancies resulting in birth that were conceived with the use of clomiphene citrate (CC)
and the frequency of multiple pregnancy.
Design: Whole-of-population cohort study, constructed through data linkage. Comprehensive Australian Government records of
dispensed medications were linked to state Perinatal Registry records for all births of at least 20 weeks’ gestation.
Setting: The state of South Australia.
Patient(s): Women who maintained pregnancy for at least 20 weeks and gave birth between July 2003 and December 2015, a total of
150,713 women with 241,561 pregnancies.
Intervention(s): Not applicable.
Main Outcome Measure(s): Ongoing pregnancy occurring in proximity to CC, defined as dispensing from 90 days before to the end of
a conception window derived from newborn date of birth and gestational age.
Result(s): Linkage to dispensed prescription records was achieved for 97.9% of women. Women who conceived with CC tended to be
older and socioeconomically advantaged and more likely than other women to have a history of miscarriage. Ongoing pregnancies
associated with CC comprised 1.6% of the total; 5.7% were multiple births (mostly twins, 94.6%) compared with 1.5% in the remainder
(98.5% twins).
Conclusion(s): In South Australia, 1.6% of pregnancies (1 in 60) of at least 20 weeks’ gestation were conceived proximal to CC
dispensing. Of these, 5.7% were multiple pregnancies. This takes the proportion of women who achieved an ongoing pregnancy
with medical assistance from 4.4%, based on reports from assisted reproductive technology clinics, to 6% in total. (Fertil SterilÒ
2022;117:202–12. Ó2021 by American Society for Reproductive Medicine.)
El resumen está disponible en Español al final del artículo.
Key Words: Infertility, medically assisted reproduction, ovarian stimulation, clomiphene citrate, multiple pregnancy

DIALOG: You can discuss this article with its authors and other readers at https://www.fertstertdialog.com/posts/32890

Received April 15, 2021; revised June 30, 2021; accepted August 16, 2021; published online October 13, 2021.
H.M. reports personal fees from Sanofi, outside the submitted work. V.M. has nothing to disclose. A.R. has nothing to disclose. R.F. has nothing to disclose.
L.M. has nothing to disclose. L.Giles has nothing to disclose. L.Grzeskowiak has nothing to disclose. E.R. has nothing to disclose. M.S. has nothing to
disclose. M.D. has nothing to disclose.
Supported by The Hospital Research Foundation of South Australia (grant 1147683). L. Grzeskowiak was supported by a Mid-Career Fellowship from The
Hospital Research Foundation (C-MCF-10-2019).
Reprint requests: Vivienne Moore, Ph.D., School of Public Health, The University of Adelaide, Adelaide SA 5005, Australia (E-mail: vivienne.moore@adelaide.
edu.au).

Fertility and Sterility® Vol. 117, No. 1, January 2022 0015-0282/$36.00

Copyright ©2021 American Society for Reproductive Medicine, Published by Elsevier Inc.
https://doi.org/10.1016/j.fertnstert.2021.08.030

202 VOL. 117 NO. 1 / JANUARY 2022

 Fertility and Sterility®

C
lomiphene citrate (CC) is an antiestrogen medication as diabetes and hypertension are recorded in view of their
taken by women as a treatment for infertility (1). It relevance to health in pregnancy. A registry code may flag a
was approved by the United States Food and Drug history of infertility or recourse to treatment, but for many
Administration in 1967 and has remained in use for over 50 women, this will not be ascertained.
years partly because it is relatively cheap and is taken orally Multiple pregnancy is a prominent source of pregnancy
as a tablet, rather than requiring injection (like gonadotro- complications and adverse fetal outcomes (21). The prevalence
pins) (2). Although CC may only be prescribed by medical spe- of multiple pregnancy after CC has been estimated at 8%–10%
cialists in many countries, access is still relatively easy (1, 3), but large variation has been reported across clinical
because these specialists are not confined to clinics licensed studies, from approximately 4% to over 20% (22, 23). Variation
to perform assisted reproductive technology (ART) treatment. may reflect differences in patient characteristics (including
Clomiphene citrate is a recommended first-line treatment infertility diagnosis) because these affect the likelihood of mul-
for World Health Organization group II ovulation disorders tifollicular development (24), as well as differences in dose and
(1, 3, 4). The most common of these disorders is polycystic other aspects of treatment protocol, including combinations of
ovary syndrome, which affects an estimated 8%–12% of medication (e.g., gonadotropins).
women (5). Clomiphene citrate has also been used for unex- Compared with singletons, twins are at increased risk of
plained infertility, in the absence of a sound evidence base serious health problems, principally due to premature birth,
about effective and ineffective treatment for this condition which affects half of twin pregnancies (17, 25, 26). Apprecia-
(6). The National Institute for Health and Care Excellence in tion of the harms of iatrogenic multiple pregnancy has revolu-
the United Kingdom recommended against CC for unex- tionized practice in ART, with concerted adoption of single
plained infertility in 2013 (3), but CC continued to be sought embryo transfer in the United States, Canada, the United
by couples there and was viewed as an option by clinicians Kingdom, several European countries, Australia, and New Zea-
who preferred to provide minimal treatment rather than land (19, 27). Ultrasound monitoring of follicle development
expectant management (7–9). The new American Society after ovarian stimulation has long been recognized as a means
for Reproductive Medicine (ASRM) guidelines now to reduce multiples (12, 28–30). However, it is unclear that this
recommend against the use of CC as a stand-alone treatment practice occurs routinely with CC, and there are reports in
for unexplained infertility but endorse use in this circum- which it is considered unnecessary or infeasible (2, 13, 31).
stance in conjunction with intrauterine insemination (10). Reinforcement of this step appears to have received little
Prescription data and surveys of women indicate that CC attention outside the United Kingdom (3, 32).
is widely used. In the United States Nurses’ Health Study II, The optimal duration of the use of CC for anovulatory
approximately 80% of women who reported infertility in infertility is uncertain, and practice has varied over time. For
the period 2000–2011 had used CC, usually as the first and example, in a sample of US women evaluated for infertility
often the only treatment (11). before 1986, 22% had used CC for more than 12 months (33).
In Australia, CC is the only oral agent for ovulation stim- Subsequently, it was suggested that use should be limited to
ulation that is approved under the Pharmaceutical Benefits 12 months (12, 28, 34). In 2013, the ASRM guidelines stated
Scheme, thus attracting a government subsidy. From the early that ‘‘pregnancy is most likely to occur in the first 3 to 6 cycles,
1990s, CC prescribing has predominantly occurred outside of and therapy beyond 6 cycles is generally not recommended’’
Australian clinics licensed to undertake ART (12, 13), so it is (1:343). The 2013 UK guidelines included the new statement
rarely used in combination with other fertility treatments. that CC should not be continued for longer than 6 months
(The use of off-label letrozole on a private prescription is (3); CC is licensed for use for up to 6 months, and CC resistance
limited and may occur within ART clinics.) In 2020, there should be evident by then. There are long-standing concerns
were almost 20,000 prescriptions for CC dispensed in about cancer after fertility treatment. A recent systematic re-
Australia (14), corresponding to approximately one prescrip- view showed that CC use was associated with an increased
tion for every 130 women aged 30–45 years or 2 prescriptions risk of ovarian cancer (odds ratio, 1.4; 95% confidence interval,
for every 260 women in this age group (15). 1.1–1.8) (35) although potential confounding could not be
The extent to which CC contributes to births in any popu- excluded; duration of use may be relevant.
lation is largely unknown. Internationally, there is no routine In this study, we used data from a state birth registry for
capture of this information (16–19). To explain, in some July 2003 to December 2015 linked to government data on
countries, data about treatment with ART and resulting prescription medications. Our first aim was to establish a
births are collected in dedicated registries, but this method for defining a probable conception window for each
information is derived from clinics registered to undertake birth, to identify pregnancies where conception was proximal
gamete manipulation using in vitro fertilization (IVF) and to CC dispensing. This is necessary because although mother’s
related procedures (17, 20). Such registries do not include date of last menstrual period (LMP) was contained in the birth
births arising from CC prescribed in other health care settings registry, it was missing for up to a quarter of women. Impor-
(as a sole therapy). Population registries pertaining to tantly, many women with ovulation disorders menstruate
pregnancy and birth do not contain detailed data on the irregularly, rarely, or not at all and, thus, may not have an
method of becoming pregnant. Information is based on LMP date recorded. Subsequent aims were to determine the
antenatal records, which usually commence toward the end proportion of ongoing pregnancies conceived proximal to
of the first trimester of pregnancy with minimal information CC dispensing and the proportion of these that were multiple
on health before pregnancy. Typically, only conditions such gestation. Lastly, we aimed to describe the supply patterns

VOL. 117 NO. 1 / JANUARY 2022 203

ORIGINAL ARTICLE: INFERTILITY

and inferred duration of use that preceded success in datasets so that they could be joined in the absence of identi-
conceiving. fying information.
The data were made available to us through the Secure
Unified Research Environment (SURE) facility, auspiced by
MATERIALS AND METHODS the Sax Institute in Sydney, Australia. The SURE is a curated
Setting and Study Population remote-access computing environment that allows re-
A population-based cohort study was undertaken based on searchers to manage and analyze linked health-related data
births in South Australia from July 2003 to December 2015. files over encrypted Internet and Australian Academic
Approximately 1.7 million people live in South Australia, Research Network connections, replacing a user’s local
and some 20,000 births are recorded every year, almost all computing environment. Only the researchers listed on the
occurring in a hospital (36). ethics and data custodian approvals have access to data.
By law, all births of at least 20 weeks’ gestation or with a Curation takes place within the secured gateway, which is a
birth weight of at least 400 g must be notified to the State gov- discrete staging space for all files inbound and outbound to
ernment, which maintains the Perinatal Registry concerning a specific study.
pregnancy and birth outcomes. The records of dispensed med-
ications, maintained by the Australian Government, were Outcome
linked to women and their births.
The outcome was pregnancy of at least 20 weeks’ gestation
that resulted in a birth occurring between July 2003 and
Data Sources December 2015, as recorded in the South Australian Perinatal
The Perinatal Registry holds records of pregnancy and peri- Registry. Additionally, whether the pregnancy was singleton
natal outcomes. A standard notification form is used, usually or multiple was considered.
completed by staff of hospital labor wards. The form covers
maternal sociodemographic background, pregnancy history, Exposure
selected aspects of maternal health, pregnancy complications, Conception was designated as occurring proximal to CC
and birth details. (exposed) if a prescription for CC was dispensed from 90
Under the Pharmaceutical Benefits Scheme, the Austra- days before to the end of the conception window (defined
lian Government keeps electronic records of dispensing hereafter). The first dispensing provides a woman with
because it subsidizes the costs of approved medicines (37). enough medication for 2 cycles, assuming the dose is 50
The amount of the subsidy depends on whether an individual mg, and a 90-day time period allows 2 cycles to elapse. We
is a general beneficiary or has concessional (i.e., welfare) sta- did not have information on the dose prescribed. We ascer-
tus, with a substantially reduced copayment applying in the tained whether women were prescribed other fertility medica-
latter circumstance. Where a medicine costs less than the rele- tions (in the same class as per the Anatomical Therapeutic
vant copayment (i.e., under copayment), full payment is made Classification System) in the exposure period.
by the recipient. For term births that occurred in July 2003, there were 3
Clomid was AUD$40.00 in 2003 (for a pack of 10 tablets months of dispensing data before the conception window.
of 50 mg) and AUD$34.85 in 2015. For general recipients, the For births occurring subsequently, there was progressively
copayment was AUD$23.10 in 2003 and AUD$37.70 in 2015. more dispensing data before becoming pregnant for a given
Before April 2012, the Pharmaceutical Benefits Scheme data- woman.
base did not contain records for the supply of medicines that For each birth, the conception window was defined as the
did not attract a government subsidy; this affected recording week containing the probable conception date, on the basis of
of some brands of CC between January and June 2009 and the recorded date of birth and gestational age at birth (in
again from January 2012 (and we consider implications of weeks), with an additional 7 days either side to allow for
underascertainment in these periods). From July 2012, com- biological variability and inaccuracy of routine ultrasound
plete dispensing records were kept regardless of subsidy. (38, 39). To explain further, gestational age at birth is deter-
For the present study, the Australian Institute of Health mined by a clinician, usually with reference to the expected
and Welfare undertook data linkage between South Australian date of delivery (not included in the Registry) that was set early
perinatal records and Pharmaceutical Benefits Scheme records. in pregnancy, on the basis of LMP date, if available, and ultra-
The South Australian Department of Health (SA Health) pro- sound measurements of the fetus (Fig. 1). While ovulation is
vided identifiers for each mother (name, date of birth) and in- often assumed to occur 14 days after LMP, there is considerable
fant (date of birth) to the Australian Institute of Health and biological variability (40, 41) for which allowance was made by
Welfare. In the first round of the linkage process, identifiers the use of this window. One missing gestational age was
for mothers were matched to Medicare enrollment records for imputed as the 50th centile for the relevant birth weight and
95%, leaving 5% of mothers unlinked. To improve the linkage sex using Australian data from Dobbins et al. (42).
rate, in a second round, the date of birth of the infant was used
because this information is encrypted with mother’s Medicare
record. This enabled linkage for more than half (60%) of those Covariates
unlinked after the first round. A linkage key with an identifier Maternal characteristics were obtained from the Perinatal
unique to each mother was then created and added to both core Registry records. These included age, country of birth,

204 VOL. 117 NO. 1 / JANUARY 2022

 Fertility and Sterility®

FIGURE 1

Construction of conception window on the basis of the date of birth and gestational age.
Moore. Prevalence of pregnancy from clomiphene. Fertil Steril 2021.

ethnicity, pregnancy history, body mass index (collected from from the conception window and, separately, those with a
2007), smoking, preexisting hypertension, and diabetes. Most missing LMP date.
of these variables were used in the format provided. However, The proportions of pregnancies and of multiple pregnan-
age of each woman at the time of birth was categorized as cies exposed to CC were calculated. The prevalence ratio was
<25, 25 to <30, 30 to <35, 35 to <40, and R40 years graphed over the time frame of the study to depict the two in-
(with the one woman for whom this was unknown assigned tervals where dispensing was not recorded in full to broadly
to 30 to <35 years). Region of birth was classified as gauge the magnitude of underascertainment arising from this
Australia, Europe, Asia, or ‘‘other’’ (with the one woman for source as well as to illustrate any secular trend. The character-
whom this was unknown assigned to other). Ethnicity was istics of women who conceived with and without CC exposure
classified as Aboriginal or Torres Strait Islander, Australian, were compared as well as the prevalence of multiple births.
Asian, or ‘‘other’’ (with the one woman for whom this was un- Dispensing patterns for CC were described in terms of the num-
known assigned to other). More than 1% of data was missing ber of supplies and relevant time intervals, summarized as
for body mass index and smoking; thus, for these variables, a medians and interquartile ranges. All statistical analyses
separate category for ‘‘unknown’’ was created. were performed using SAS 9.4 (SAS Institute Inc., Cary, NC).
The Perinatal Registry record included postcode of resi-
dence, which was converted to the corresponding geographic Ethics Approval
statistical area and assigned a socioeconomic Index of Rela-
tive Disadvantage (43). We categorized this variable in quin- Approval for the study was obtained from the ethics commit-
tiles using the Australian distribution as the reference (rather tees of the South Australian Department of Health [HREC/15/
than the State distribution) so as to allow classification of SAH/80] and the Australian Institute of Health and Welfare
women who lived outside the South Australian borders but [EO2013/3/51]. Individual level consent was not required
gave birth within the State. (This usually occurred because because the study entailed the use of deidentified records
the maternity hospital closest to the place of residence was (managed through the SURE facility, described earlier).
located in South Australia.) The eight women for whom this
information was missing were assigned to quintile 3. RESULTS
From July 2003 to December 2015, there were 245,439 births
associated with 241,561 pregnancies of at least 20 weeks’
Statistical Analysis gestation to 150,713 mothers. Hereafter, ‘‘pregnancy’’ will
Maternal characteristics for each ongoing pregnancy result- be used to refer to pregnancies maintained for at least 20
ing in birth from July 2003 to December 2015 were tabulated. weeks. The results are reported at the pregnancy level because
For pregnancies with an LMP date, the proportion with a date multiple births have the same exposures.
that was up to 14 days before or within the conception win- Linkage to prescription records was achieved for 97.9% of
dow was calculated. The characteristics of pregnancies where women (and 98.6% of pregnancies) (Fig. 2). Women with un-
the LMP date was concordant with the conception window linked records were included in data presented and assumed
were compared with those where the LMP date was discrepant to be unexposed to CC.

VOL. 117 NO. 1 / JANUARY 2022 205

ORIGINAL ARTICLE: INFERTILITY

FIGURE 2 TABLE 1

Maternal characteristics for each pregnancy (births in South

Australia between July 2003 and December 2015).
All pregnancies
(N[241,561)
Characteristic N %
Age (y)
<25 46,018 19.1
25 to <30 69,197 28.6
30 to <35 77,860 32.2
35 to <40 39,831 16.5
R40 8,655 3.6
Ethnicity
Caucasian 201,817 83.5
Aboriginal and/or Torres Strait 7,686 3.2
Islander
Asian 21,688 9.0
Other 10,370 4.3
Region of birth
Australia 192,919 79.9
Europe 12,157 5.0
Asia 23,926 9.9
Other 12,559 5.2
Area index of relative
socioeconomic
disadvantage quintile,
Australia
1 (most disadvantaged) 62,327 25.8
2 73,921 30.6
3 26,293 10.9
4 51,320 21.2
5 (least disadvantaged) 27,700 11.5
Number of previous
pregnancies resulting in
birth ‡20 weeks
0 101,698 42.1
Flow diagram for data linkage. 1 84,827 35.1
Moore. Prevalence of pregnancy from clomiphene. Fertil Steril 2021. 2 35,186 14.6
R3 19,850 8.2
Moore. Prevalence of pregnancy from clomiphene. Fertil Steril 2021.

The characteristics of the 3,189 women (2.1%) with 3,306

pregnancies (1.4%) that were not able to be linked were exam-
ined (data not shown). A considerable proportion (37.0%) of for 55,032 pregnancies (22.8%). Where the LMP was not
these women were born in Asia and may not have been per- missing, it was consistent with the conception window in
manent residents of Australia, therefore not eligible to access the great majority of cases (94.5%).
medicines under the Pharmaceutical Benefits Scheme. (It is Comparisons were made between 3 groups of preg-
common, for example, for individuals arriving on a student nancies: those where the LMP date concorded with the
visa to gain temporary residency for a defined period after conception window; those where the recorded LMP date
completing the program of study.) A similar proportion was discrepant from the conception window; and those
(39.1%) were women who were born in Australia; these had with a missing LMP date (Supplemental Table 1, available
broadly similar sociodemographic profiles to women for online). Pregnancies where the LMP date concorded with
whom linkage was achieved, suggesting that any errors that the conception window tended to involve older, advan-
affected matching were random. taged women who did not smoke. Pregnancies involving
Ongoing pregnancies corresponding to births in the study women aged <25 years and those of Aboriginal or Torres
period are characterized in Table 1. Consistent with the predom- Strait Islander ethnicity were overrepresented among
inant age group being 30 to <35 years, the median age of those where the LMP date was discrepant or missing,
women at the time of birth was 30.3 years (interquartile range, as were those involving women who had at least 3 pre-
26.3–34.1 years). The great majority of women were born in vious births.
Australia and reported Caucasian ethnicity. For approximately In all, 3,853 pregnancies were exposed to CC, comprising
40% of pregnancies, women had no previous pregnancy of at 1.6% of the total. A prescription for gonadotropin (injectable)
least 20 weeks’ gestation, whereas a slightly lower proportion was also dispensed to 5.7% of these women in the exposure
involved women with one such previous pregnancy. period.
As described in the methods, a 21-day conception win- The proportion of births exposed to CC in each month was
dow was derived for each pregnancy. The LMP was missing examined. The range was from 1.0% to 1.9%, with the lowest

206 VOL. 117 NO. 1 / JANUARY 2022

 Fertility and Sterility®

TABLE 2

Maternal characteristics for pregnancies (births in South Australia, July 2003 to December 2015), according to clomiphene citrate exposure.
Exposed to CC
No Yes
(N[237,708) (N[3,853)
Characteristic N % N % P valuea
Age (y) < .001
Under 25 45,727 19.2 291 7.6
25 to <30 67,933 28.6 1,264 32.8
30 to <35 76,345 32.1 1,514 39.3
35 to <40 39,174 16.5 657 17.1
R40 8,528 3.6 127 3.3
Ethnicity < .001
Caucasian 198,337 83.4 3,480 90.3
Aboriginal and/or Torres Strait Islander 7,649 3.2 37 1.0
Asian 21,446 9.0 242 6.3
Other 10,276 4.3 94 2.4
Region of birth < .001
Australia 189,627 79.8 3,292 85.4
Europe 11,982 5.0 175 4.5
Asia 23,657 10.0 269 7.0
Other 12,442 5.2 117 3.0
Area index of relative socioeconomic < .001
disadvantage quintile, Australia
1 (most disadvantaged) 61,568 25.9 759 19.7
2 72,763 30.6 1,158 30.1
3 25,866 10.9 419 10.9
4 50,368 21.2 952 24.7
5 (least disadvantaged) 27,135 11.4 565 14.7
Number of previous pregnancies resulting in ‡ 20 < .001
weeks
0 99,476 41.8 2,222 57.7
1 83,567 35.2 1,260 32.7
2 34,902 14.7 284 7.4
R3 19,763 8.3 87 2.3
Number of previous miscarriages < .001
0 181,908 76.5 2,803 72.7
1 39,992 16.8 709 18.4
2 10,490 4.4 222 5.8
R3 5,318 2.2 119 3.1
Smoking status at first antenatal visit < .001
Smoker 35,184 14.8 202 5.2
Quit in pregnancy before first visit 8,665 3.6 86 2.2
Nonsmoker 190,935 80.3 3,497 90.8
Unknown 2,924 1.2 68 1.8
Body mass index (kg/m2) < .001
Underweight (<18.5) 3,939 1.7 41 1.1
Healthy weight (18.5–24.9) 62,504 26.3 810 21.0
Overweight (25–29.9) 37,681 15.9 531 13.8
Obese (R30) 32,946 13.9 625 16.2
Unknown 100,638 42.3 1,846 47.9
Preexisting diabetes 1,576 0.7 59 1.5 < .001
Preexisting hypertension 2,723 1.1 61 1.6 .012
Missing or discrepant date of last menstrual 64,235 27.0 989 25.7 .060
period
Plurality < .001
1 234,119 98.5 3,632 94.3
2 3,534 1.5 209 5.4
R3 55 0.0 12 0.3
Note: CC ¼ clomiphene citrate.
a
P values from the c2 test including ‘‘other’’ but excluding ‘‘unknown.’’
Moore. Prevalence of pregnancy from clomiphene. Fertil Steril 2021.

prevalence corresponding to the times at which the dispensing presented as a moving average over 3 successive months,
price fell below copayment. In Supplemental Fig. 1 (available providing visual interpolation of the likely pattern across the
online), the periods in which dispensing data were not fully periods where some data were missing. There was no evidence
captured are shown with dashed lines. The proportion is also of a secular trend.

VOL. 117 NO. 1 / JANUARY 2022 207

ORIGINAL ARTICLE: INFERTILITY

The prevalence of multiple births was 5.7% among CC- Almost half of women who achieved pregnancy of at least
exposed pregnancies, whereas in the remainder of the cohort, 20 weeks in proximity to dispensed CC had only one supply
it was 1.5%. For the 219 pregnancies in which women were (corresponding to 2 courses at the starting dose of 50 mg).
also prescribed gonadotropin, the prevalence was 5.8%. The proportion of pregnancies occurring after further supplies
(Among women whose records could not be linked, the prev- declined progressively, roughly halving with each additional
alence was 1.6%.) Multiple pregnancies exposed to CC supply.
comprised 209 twin pregnancies (94.6%) and 12 higher- A date for the LMP was not available for approximately a
order multiple pregnancies. In the remainder of the cohort, quarter of pregnancies in this cohort, underscoring the need
almost all multiple pregnancies were twins (98.5%), with 55 to derive a conception window from the recorded date of birth
higher-order multiples recorded. and gestational age. The conception window was consistent
The profiles of women who conceived and maintained a with the LMP date, where this date was available, in the great
pregnancy for at least 20 weeks in the presence or absence majority of cases; approximately 5% of recorded dates for the
of CC exposure are presented in Table 2. Compared with un- LMP were sufficiently discrepant that they were infeasible. In
exposed pregnancies, those exposed to CC tended to be older general, pregnancies where the date of LMP was missing or
women who were Australian born, Caucasian, and from the infeasible tended to involve women aged <25 years, those
most advantaged socioeconomic group. These women were residing in disadvantaged areas, smoking in pregnancy, and
less likely to smoke but more likely to be obese and to have high parity. In Australia, apart from high parity, these charac-
preexisting hypertension or diabetes compared with other teristics are also associated with the use of less reliable contra-
women. Close to 60% had never previously had a pregnancy ceptive methods (46) and with unintended pregnancy (47).
maintained for 20 weeks, compared with approximately 40% Women who conceived and maintained a pregnancy for at
of other women. However, 27% had a history of miscarriage, least 20 weeks with CC exposure had characteristics broadly
compared with 23% of other women. compatible with infertility, including older age, high body
Because CC is the first-line treatment for anovulatory mass index, and relatively high prevalence of preexisting dia-
infertility, we expected that the LMP would be missing for a betes or hypertension. The proportion missing LMP date was
substantial proportion of these pregnancies. However, similar to that for other women, which was unexpected
approximately a quarter were missing LMP, no different because CC is the recommended first-line treatment for anovu-
from other pregnancies. latory infertility (1, 3, 4). While these women may keep careful
Of the ongoing pregnancies exposed to CC, 2 in 5 (42.8%) account of menstruation, reporting by clinical staff is likely to
occurred after dispensing of one supply of the medication. reflect whether this information is relevant to calculating
Progressively smaller proportions occurred after dispensing gestational age and planning the first ultrasound, with infor-
of 2 (22.2%) or 3 supplies (13.5%). Most pregnancies (71%) af- mation pertaining to irregular (occasional, implausible) dates
ter 2 supplies occurred within 3 months of first dispensing discounted. Additionally, CC is thought to have been widely
(Supplemental Fig. 2). Some pregnancies were associated used for unexplained infertility in the time frame of the study,
with sporadic dispensing over a longer period, for example, with these women not likely to be missing the date of LMP.
6 to 12 months (12.4%) or more than 12 months (7.9%). The use of CC for unexplained infertility was discouraged
in the UK guidelines from 2013 although the practice was
continued by some clinicians in a context of clinical debate
DISCUSSION and pressure from couples (9). The ASRM guidelines now
Overall, 1.6% of pregnancies of at least 20 weeks’ gestation concur with the UK guidelines that CC should not be used as
were conceived in proximity to CC dispensing (defined as a stand-alone treatment for unexplained infertility (3, 10).
90 days before to the end of a 21-day conception window). Accumulation of evidence about appropriate treatment for un-
There was little variation in this prevalence over the time explained infertility has seen a revival of interest in intrauter-
frame of the study. The prevalence of multiple gestation ine insemination in combination with CC as a relatively
(mostly twins) was 5.7% among CC-exposed pregnancies, inexpensive and less invasive option than IVF (10, 48–50).
whereas in the remainder of the cohort, it was 1.5%, suggest- The prevalence of births likely to have been conceived
ing that dispensed medicine was frequently used. Dispensing with CC in South Australia was the same as a report for the
of gonadotropins in addition to CC was uncommon, a feature United States but higher than for the United Kingdom.
of 5.9% of exposed pregnancies, and did not contribute Duwe et al. (16) reported that the proportion of live births in
disproportionately to twins. the United States conceived with CC from 1997 to 2004 was
The study population is distinct through having universal 1.6%, based on information provided by women in a control
health care cover and being highly urbanized, but in these re- group (n>5,000) of the National Birth Defects Prevention
spects the other five states of Australia are similar. Only Study. In the United Kingdom, a survey by Bardis et al. (51)
authorized practitioners may prescribe CC, usually specialists of births occurring during 1 week in 2003 (n>6,000) showed
in endocrinology or in obstetrics and gynecology. Some Euro- that 1.9% were conceived with some form of medical assis-
pean countries seem to have similar access and prescribing tance, and of these, 16.7% resulted from CC, corresponding
patterns, despite geographic differences (44, 45). Furthermore, to 0.3% of all births. Bardis et al. (51) suggested that this re-
as a depiction of a low-cost first-line treatment, the findings flected diminished treatment outside of fertility clinics in
are arguably relevant to a range of other countries where this the United Kingdom, at the same time as clinics prioritized
may be the only option because of the high cost of ART. IVF over less invasive options.

208 VOL. 117 NO. 1 / JANUARY 2022

 Fertility and Sterility®

Multiple pregnancy after assisted conception is a long- with the National Institute for Health and Care Excellence
standing concern because of the increased risks of adverse guidelines (3), the highest frequency was 7.4% in studies in
health outcomes for women and infants (25, 28, 52). With the which ultrasound tracking in the first cycle was not specified
swift adoption of single embryo transfer in Australia, multiples as part of the protocol. The investigators suggest that older es-
among ART pregnancies fell from over 20% in 2000 to 7.8% in timates are inflated through including women with a variety of
2010, 4.4% in 2015, and 3.2% in 2018 (53, 54). Elsewhere, infertility diagnoses and adjunctive treatments. Moreover, we
notably the Netherlands, Nordic countries, the United Kingdom, suggest that the general increase in body weight of women
the United States, and Canada, there have also been reductions since the 1960s means that the 50-mg starting dose is effec-
in multiple pregnancy (55), but action has not been universal. tively lower, reducing multiples. Finally, some Australian pre-
Evidence continues to underscore the risks associated with scribers may undertake ultrasound monitoring because there is
multiple pregnancy. For example, a recent systematic review of some awareness of the matter (13).
studies comparing outcomes of singleton and multiple preg- Nevertheless, our finding that 1 in 20 CC-exposed preg-
nancies conceived with IVF or intracytoplasmic sperm injec- nancies was a multiple pregnancy suggests that there is an
tion showed that the risk of preterm birth was eightfold opportunity to improve practice in Australia. Monitoring
higher in twins compared with singletons (odds ratio, 8.3; twinning is the principal reason why there have been calls
95% confidence interval, 7.8–8.9, on the basis of data for to include births conceived with ovulation stimulation as a
285,078 twins and 853,038 singletons) (56). Similar compari- sole therapy in national registers of births occurring after
sons for pregnancies where conception was medically assisted ART (18, 51). A register would also facilitate the investigation
but did not involve IVF/intracytoplasmic sperm injection are of possible longer-term health issues associated with hormon-
limited, but results are broadly consistent (57). Per infant, costs al stimulation in women and children (35, 64).
of twin pregnancy and birth are approximately 2.5 times In Australia in 2015, 4.4% of women who gave birth
higher than those of singletons (58) with hospital costs in the became pregnant with ART treatment provided by a specialist
first 5 years of life at least 3 times higher (59). reproductive technology clinic (65). Assuming that the prev-
Recently, Evans et al. (60) confirmed that the risk of mul- alence of 1.6% for CC conceptions applies across Australia,
tiple pregnancy increased substantially where women aged the proportion of women who required medical assistance
<40 years had more than two mature follicles after ovarian to achieve an ongoing pregnancy amounts to 6%, corre-
stimulation (by CC or other drugs, before intrauterine insem- sponding to one in 16 pregnancies resulting in birth.
ination) without improving the chance of a singleton clinical
pregnancy. Clinical guidelines recommend ultrasound in the
first CC cycle to minimize the risk of multiple pregnancy by Strengths and Limitations
enabling the number of developing follicles to be determined, Our study uses high-quality data in terms of population
and, where necessary, couples are advised to avoid inter- coverage, reporting consistency, and linkage rate. Our deter-
course so that pregnancy does not occur. There is some mination of conception window is more precise than many
complexity involved, but limiting to two preovulatory folli- other studies of medication use before or during pregnancy,
cles has been recommended (17). In the United Kingdom, especially those based on health insurance data (66). We
approximately half of clinicians now adhere to the recom- considered births from 20 weeks of gestation because com-
mendation for first cycle ultrasound with CC treatment (32), plete data were available from this point; the results were
but the extent of this practice elsewhere is uncertain. very similar when restricted to births from 24 weeks of gesta-
Bhattacharya et al. (61) demonstrated that with ultrasound tion (the cutoff for viability).
monitoring and tailored doses of CC (for unexplained infer- We have information on dispensing of CC, not consump-
tility), it was possible to achieve a frequency of twins from tion. However, as women have been diagnosed with infertility,
CC of 1%. The review of Galazis et al. (62) has been cited as their ability to become pregnant in the absence of treatment is
finding insufficient evidence to suggest that ultrasound moni- low. Further, adherence is likely to be high because this medi-
toring reduces multiple pregnancy rates resulting from ovarian cine is short-term and goal-oriented and is prescribed under
stimulation (63). However, Galazis et al. (62) did not identify conditions optimal for adherence (67–69). One study has
any trials evaluating the efficacy of ultrasound to reduce twins shown that adherence with fertility medication is indeed high
after CC. Instead, these investigators considered findings of (70). Finally, time between a prescription being issued and
two trials in which ultrasound had been used to monitor ovula- dispensed has been used as an indicator of adherence (70); in
tion induction in comparison with other approaches such as our data, the median time to dispensing was 3 days, and
measuring urinary luteinizing hormone, and three clinical co- 70% of women had their prescription filled within 14 days.
horts in which approaches included no monitoring. Only one of Linkage to prescription data was not possible for 1.4% of
the studies (24) considered the outcome of multiple pregnancy ongoing pregnancies. We assumed that these were not
(and it recommended ultrasound monitoring). Further, selec- exposed to CC, which was consistent with the prevalence of
tion criteria were applied such that the study of Bhattacharya twins in this group. However, if CC exposure was similar to
et al. (61) in which ultrasound monitoring was part of the pro- that for the rest of the cohort, there would be an additional
tocol, but not under evaluation, was excluded. 53 exposed pregnancies, with a small increase in the overall
Our finding of 5.7% multiple births associated with CC is prevalence (from 1.60% to 1.62%).
lower than reported historically (1, 3). However, in a recent sys- There were two 6-month periods in the 13 years of expo-
tematic review (23) of studies in which CC use was consistent sure data when dispensing was underascertained; women

VOL. 117 NO. 1 / JANUARY 2022 209

ORIGINAL ARTICLE: INFERTILITY

purchasing CC during these months (other than those with 6. Ray A, Shah A, Gudi A, Homburg R. Unexplained infertility: an update and
concession cards, typically only one in five women) may review of practice. Reprod Biomed Online 2012;24:591–602.
7. Nadeem R. Unexplained infertility: an exploration of women's perceptions of
not have had all supplies recorded in the Government data-
diagnosis, treatment and decision-making. J Obstet Gynaecol 2010;30:83–4.
base. The prevalence of CC-exposed pregnancies in the years 8. Kamath MS, Bhattacharya S. Demographics of infertility and management
not affected by this circumstance suggests that there would be of unexplained infertility. Best Pract Res Clin Obstet Gynaecol 2012;26:
at most 200 additional exposed pregnancies (with overall 729–38.
prevalence at most 1.67%). More generally, if women were 9. Nandi A, Gudi A, Shah A, Homburg R. An online survey of specialists'
dispensed CC but delayed use for more than 3 months, the opinion on first line management options for unexplained subfertility.
pregnancy would be designated as unexposed. Hum Fertil (Camb) 2015;18:48–53.
10. Practice Committee of the American Society for Reproductive Medicine. Ev-
We do not have information to comprehensively identify
idence-based treatments for couples with unexplained infertility: a guide-
pregnancies conceived using clinic-based ART (which rarely line. Fertil Steril 2020;113:305–22.
involved CC in this time frame), and these pregnancies are 11. Farland LV, Missmer SA, Rich-Edwards J, Chavarro JE, Barbieri RL,
included in the comparison group. Grodstein F. Use of fertility treatment modalities in a large United States
cohort of professional women. Fertil Steril 2014;101:1705–10.
12. Nasseri S, Ledger WL. Clomiphene citrate in the twenty-first century. Hum
CONCLUSION Fertil (Camb) 2001;4:145–51.
One in every 60 births in South Australia was conceived prox- 13. McDowell S, Kroon B, Yazdani A. Clomiphene ovulation induction and
imal to CC dispensing. Births arising in conjunction with higher-order multiple pregnancy. Aust N Z J Obstet Gynaecol 2013;53:
395–8.
dispensed CC increased those documented as arising from
14. Australian Government Department of Health. The Pharmaceutical Benefits
ART by approximately a third, taking the total prevalence Scheme date of supply report. Available at: https://www.pbs.gov.au/info/
in 2015 to 6% of all births (1 in 16). The frequency of multiple statistics/dos-and-dop/dos-and-dop. Accessed May 21, 2021.
pregnancy arising from CC is now higher than that for ART 15. Australian Bureau of Statistics. 3101.0 Australian demographic statistics, Jun
conceptions. There is an opportunity to provide couples 2019. Available at: https://www.abs.gov.au/AUSSTATS/[email protected]/Details
with better information about the risk of multiple pregnancy Page/3101.0Jun%202019?OpenDocument. Accessed March 5, 2021.
and to consider whether current clinical practice is optimal. 16. Duwe KN, Reefhuis J, Honein MA, Schieve LA, Rasmussen SA. Epidemiology
of fertility treatment use among US women with liveborn infants, 1997-
2004. J Womens Health (Larchmt) 2010;19:407–16.
Data Availability 17. Black M, Bhattacharya S. Epidemiology of multiple pregnancy and the effect
of assisted conception. Semin Fetal Neonatal Med 2010;15:306–12.
The authors do not have permission to share the data because 18. Chaabane S, Sheehy O, Monnier P, Bissonnette F, Trasler JM, Fraser W, et al.
this was provided specifically for the scope of research as Association between ovarian stimulators with or without intrauterine insem-
approved by the ethics committees. Requests to access these ination, and assisted reproductive technologies on multiple births. Am J Ob-
datasets should be directed to the data custodians. stet Gynecol 2015;213:511.e1–14.
19. Adamson GD, de Mouzon J, Chambers GM, Zegers-Hochschild F,
Acknowledgments: The authors thank Kristyn Willson for Mansour R, Ishihara O, et al. International Committee for Monitoring Assis-
extensive contributions to the background research and plan- ted Reproductive Technology: world report on assisted reproductive tech-
nology, 2011. Fertil Steril 2018;110:1067–80.
ning for this study, Wendy Scheil for her insight and support
20. Nyboe Andersen A, Erb K. Register data on Assisted Reproductive Technol-
while Director of the Pregnancy Outcome Unit, and Anthea ogy (ART) in Europe including a detailed description of ART in Denmark. Int J
Hutchison for her valuable work as Project and Data Manager Androl 2006;29:12–6.
of the Life Course and Intergenerational Research Group. 21. Blondel B, Kaminski M. Trends in the occurrence, determinants, and conse-
quences of multiple births. Semin Perinatol 2002;26:239–49.
DIALOG: You can discuss this article with its authors and 22. Chaabane S, Sheehy O, Monnier P, Fraser W, Bissonnette F, Trasler JM, et al.
other readers at https://www.fertstertdialog.com/posts/ Ovarian stimulation, intrauterine insemination, multiple pregnancy and ma-
32890 jor congenital malformations: a systematic review and meta-analysis- The
ART_Rev Study. Curr Drug Saf 2016;11:222–61.
23. Garthwaite H, Stewart J, Wilkes S. Multiple pregnancy rate in patients un-
REFERENCES dergoing treatment with clomifene citrate for WHO group II ovulatory disor-
1. Practice Committee of the American Society for Reproductive Medicine. Use ders: a systematic review. Hum Fertil (Camb) 2021:1–10.
of clomiphene citrate in infertile women: a committee opinion. Fertil Steril 24. Coughlan C, Fitzgerald J, Milne P, Wingfield M. Is it safe to prescribe clomi-
2013;100:341–8. phene citrate without ultrasound monitoring facilities? J Obstet Gynaecol
2. Wilkes S, Murdoch A. Ovulation induction with clomifene: a primary care 2010;30:393–6.
perspective. J Fam Plann Reprod Health Care 2012;38:48–52. 25. Braude PE, ed. One child at a time: reducing multiple births after IVF. Report
3. National Institute for Health and Care Excellence (NICE). Fertility: assessment of the Expert Group on Multiple Births after IVF. London: Human Fertilization
and treatment for people with fertility problems. London: Royal College of and Embryology Authority, 2006.
Obstetricians and Gynaecologists; 2013. 26. Russo FM, Pozzi E, Pelizzoni F, Todyrenchuk L, Bernasconi DP, Cozzolino S,
4. Teede HJ, Misso ML, Boyle JA, Garad RM, McAllister V, Downes L, et al. et al. Stillbirths in singletons, dichorionic and monochorionic twins: a com-
Translation and implementation of the Australian-led PCOS guideline: clin- parison of risks and causes. Eur J Obstet Gynecol Reprod Biol 2013;170:
ical summary and translation resources from the International Evidence- 131–6.
based Guideline for the Assessment and Management of polycystic ovary 27. International Federation of Fertility Societies’ Surveillance. Global Trends in
syndrome. Med J Aust 2018;209:S3–8. Reproductive Policy and Practice. 8th ed. Glob Reprod Health 2019;4:e29.
5. March WA, Moore VM, Willson KJ, Phillips DI, Norman RJ, Davies MJ. The 28. Royal College of Obstetricians and Gynaecologists (RCOG). The manage-
prevalence of polycystic ovary syndrome in a community sample assessed ment of infertility in secondary care. Evidence-based Guideline No. 3. Lon-
under contrasting diagnostic criteria. Hum Reprod 2010;25:544–51. don: RCOG; 1998.

210 VOL. 117 NO. 1 / JANUARY 2022

 Fertility and Sterility®

29. Hughes J. Ovarian stimulation for assisted reproductive technologies. In: 51. Bardis N, Maruthini D, Balen AH. Modes of conception and multiple preg-
Vayena E, Rowe PJ, Griffin PD, editors. Current practices and controversies in as- nancy: a national survey of babies born during one week in 2003 in the
sisted reproduction: report of a WHO meeting. Geneva: WHO; 2002:102–23. United Kingdom. Fertil Steril 2005;84:1727–32.
30. Balen AH, Braat DD, West C, Patel A, Jacobs HS. Cumulative conception and 52. Ward HR, Whittle ME. Multiple pregnancy. London: RCOG Press; 1995.
live birth rates after the treatment of anovulatory infertility: safety and effi- 53. Wang Y, Dean J, Badgery-Parker T, Sullivan E. Assisted reproductive technol-
cacy of ovulation induction in 200 patients. Hum Reprod 1994;9:1563–70. ogy in Australia and New Zealand 2006. Sydney: AIHW National Perinatal
31. Legro RS, Barnhart HX, Schlaff WD, Carr BR, Diamond MP, Carson SA, et al. Statistics Unit, University of New South Wales; 2008.
Clomiphene, metformin, or both for infertility in the polycystic ovary syn- 54. Newman J, Paul RC, Chambers GM. Assisted reproductive technology in
drome. N Engl J Med 2007;356:551–6. Australia and New Zealand 2018. Sydney: National Perinatal Epidemiology
32. Garthwaite H, Stewart J, King K, McGarry K, Wilkes S. Ultrasound moni- and Statistics Unit, University of New South Wales; 2020.
toring during first-cycle treatment with clomifene citrate: a national survey 55. De Geyter C, Calhaz-Jorge C, Kupka MS, Wyns C, Mocanu E, Motrenko T, et al.
of compliance with NICE. Hum Fertil (Camb) 2020;23:193–9. ART in Europe, 2014: results generated from European registries by ESHRE: The
33. Rossing MA, Daling JR, Weiss NS, Moore DE, Self SG. Ovarian tumors in a European IVF-monitoring Consortium (EIM) for the European Society of Human
cohort of infertile women. N Engl J Med 1994;331:771–6. Reproduction and Embryology (ESHRE). Hum Reprod 2018;33:1586–601.
34. Kousta E, White DM, Franks S. Modern use of clomiphene citrate in induc- 56. Eapen A, Ryan GL, Ten Eyck P, Van Voorhis BJ. Current evidence supporting
tion of ovulation. Hum Reprod Update 1997;3:359–65. a goal of singletons: a review of maternal and perinatal outcomes associated
35. Barcroft JF, Galazis N, Jones BP, Getreu N, Bracewell-Milnes T, Grewal KJ, with twin versus singleton pregnancies after in vitro fertilization and intracy-
et al. Fertility treatment and cancers-the eternal conundrum: a systematic re- toplasmic sperm injection. Fertil Steril 2020;114:690–714.
view and meta-analysis. Hum Reprod 2021;36:1093–107. 57. Ombelet W, Martens G, De Sutter P, Gerris J, Bosmans E, Ruyssinck G, et al.
36. Scheil W, Jolly K, Scott J, Catcheside B, Sage L, Kennare R. Pregnancy Perinatal outcome of 12,021 singleton and 3108 twin births after non-IVF-
outcome in South Australia 2015. Adelaide: SA Health, Government of assisted reproduction: a cohort study. Hum Reprod 2006;21:1025–32.
South Australia; 2017. 58. Howell EM, Johnson P, Cross-Barnet C. Twin births in Medicaid: prevalence,
37. Mellish L, Karanges EA, Litchfield MJ, Schaffer AL, Blanch B, Daniels BJ, et al. outcomes, utilization, and cost in four states, 2014-2015. Matern Child
The Australian Pharmaceutical Benefits Scheme data collection: a practical Health J 2020;24:546–51.
guide for researchers. BMC Res Notes 2015;8:634. 59. van Heesch MM, Evers JL, van der Hoeven MA, Dumoulin JC, van
38. Bottomley C, Bourne T. Dating and growth in the first trimester. Best Pract Beijsterveldt CE, Bonsel GJ, et al. Hospital costs during the first 5 years of
Res Clin Obstet Gynaecol 2009;23:439–52. life for multiples compared with singletons born after IVF or ICSI. Hum Re-
39. Sarris I, Ioannou C, Dighe M, Mitidieri A, Oberto M, Qingqing W, et al. Standard- prod 2015;30:1481–90.
ization of fetal ultrasound biometry measurements: improving the quality and 60. Evans MB, Stentz NC, Richter KS, Schexnayder B, Connell M, Healy MW,
consistency of measurements. Ultrasound Obstet Gynecol 2011;38:681–7. et al. Mature follicle count and multiple gestation risk based on patient
40. Wilcox AJ, Dunson D, Baird DD. The timing of the "fertile window" in the age in intrauterine insemination cycles with ovarian stimulation. Obstet Gy-
menstrual cycle: day specific estimates from a prospective study. Br Med J necol 2020;135:1005–14.
2000;321:1259–62. 61. Bhattacharya S, Harrild K, Mollison J, Wordsworth S, Tay C, Harrold A, et al.
41. Nakling J, Buhaug H, Backe B. The biologic error in gestational length related Clomifene citrate or unstimulated intrauterine insemination compared with
to the use of the first day of last menstrual period as a proxy for the start of expectant management for unexplained infertility: pragmatic randomised
pregnancy. Early Hum Dev 2005;81:833–9. controlled trial. Br Med J 2008;337:a716.
42. Dobbins TA, Sullivan EA, Roberts CL, Simpson JM. Australian national birth- 62. Galazis N, Zertalis M, Haoula Z, Atiomo W. Is ultrasound monitoring of the
weight percentiles by sex and gestational age, 1998-2007. Med J Aust ovaries during ovulation induction by clomiphene citrate essential? A sys-
2012;197:291–4. tematic review. J Obstet Gynaecol 2011;31:566–71.
43. Trewin D. 2039.0 Information paper: census of population and housing 63. Bergh C, Kamath MS, Wang R, Lensen S. Strategies to reduce multiple preg-
socio-economic indexes for areas, Australia, 2001. Canberra: Australian Bu- nancies during medically assisted reproduction. Fertil Steril 2020;114:673–
reau of Statistics; 2003. 9.
44. Wang H, Land JA, Bos HJ, Bakker MK, de Jong-van den Berg LT. Clomiphene 64. Schieve LA, Devine O, Boyle CA, Petrini JR, Warner L. Estimation of the
citrate utilization in the Netherlands 1998-2007. Hum Reprod 2011;26: contribution of non-assisted reproductive technology ovulation stimulation
1227–31. fertility treatments to US singleton and multiple births. Am J Epidemiol 2009;
45. Ben Messaoud K, Bouyer J, de La Rochebrochard E. Infertility treatment in 170:1396–407.
France, 2008-2017: a challenge of growing treatment needs at older 65. Fitzgerald O, Paul RC, Harris K, Chambers GM. Assisted reproductive tech-
ages. Am J Public Health 2020;110:1418–20. nology in Australia and New Zealand 2016. Sydney: National Perinatal
46. Coombe J, Harris ML, Wigginton B, Lucke J, Loxton D. Contraceptive use at the Epidemiology and Statistics Unit, University of New South Wales; 2018.
time of unintended pregnancy: findings from the Contraceptive Use, Preg- 66. Margulis AV, Palmsten K, Andrade SE, Charlton RA, Hardy JR, Cooper WO,
nancy Intention and Decisions study. Aust Fam Physician 2016;45:842–8. et al. Beginning and duration of pregnancy in automated health care data-
47. Rowe H, Holton S, Kirkman M, Bayly C, Jordan L, McNamee K, et al. Prevalence bases: review of estimation methods and validation results. Pharmacoepide-
and distribution of unintended pregnancy: the Understanding Fertility Man- miol Drug Saf 2015;24:335–42.
agement in Australia National Survey. Aust N Z J Public Health 2016;40:104–9. 67. Osterberg L, Blaschke T. Adherence to medication. N Engl J Med 2005;4:
48. Bahadur G, Homburg R, Muneer A, Racich P, Alangaden T, Al-Habib A, et al. 487–97.
First line fertility treatment strategies regarding IUI and IVF require clinical ev- 68. DiMatteo MR, Haskard-Zolnierek KB, Martin LR. Improving patient adherence:
idence. Hum Reprod 2016;31:1141–6. a three-factor model to guide practice. Health Psychol Rev 2012;6:74–91.
49. Wang R, Danhof NA, Tjon-Kon-Fat RI, Eijkemans MJ, Bossuyt PM, 69. Aznar-Lou I, Potteg
ard A, Fernandez A, Pen ~arrubia-María MT, Serrano-
Mochtar MH, et al. Interventions for unexplained infertility: a systematic re- Blanco A, Sabes-Figuera R, et al. Effect of copayment policies on initial medi-
view and network meta-analysis. Cochrane Database Syst Rev 2019;9: cation non-adherence according to income: a population-based study. BMJ
Cd012692. Qual Saf 2018;27:878–91.
50. Ayeleke RO, Asseler JD, Cohlen BJ, Veltman-Verhulst SM. Intra-uterine 70. Heredia M, Tenías JM, Ruiz R, Sanchez M, Calleja MA, Martinez F. Pharma-
insemination for unexplained subfertility. Cochrane Database Syst Rev ceutical care for patients undergoing assisted reproduction techniques. Eur J
2020;3:Cd001838. Hosp Pharm 2012;19:289–92.

VOL. 117 NO. 1 / JANUARY 2022 211

ORIGINAL ARTICLE: INFERTILITY

Dispensacion de citrato de clomifeno para tratar la infertilidad: medicaci

on suministrada y prevalencia poblacional de embarazos asis-
tidos y partos multiples.
Objetivo: Determinar la proporcion de embarazos resultantes en nacimientos que fueron concebidos con el uso de citrato de clomifeno
(CC) y la frecuencia de embarazos multiples.
~o: Estudio de cohorte de poblacion completa, construido a traves de enlace de datos. Los registros completos del Gobierno de
Disen
Australia de los medicamentos dispensados se vincularon a los datos del Registro Perinatal del estado para todos los nacimientos de
al menos 20 semanas de gestaci
on.
Entorno: Estado de Australia Meridional.
Paciente(s): Mujeres que mantuvieron el embarazo durante al menos 20 semanas y dieron a luz entre julio de 2003 y diciembre de
2015, un total de 150.713 mujeres con 241.561 embarazos.
Intervencion(es): No aplica.
Principales medidas de resultado: embarazo en curso que ocurre cerca del CC, definido como dispensacion desde 90 días antes hasta el
final de una ventana de concepci
on derivada de la fecha de nacimiento del recien nacido y la edad gestacional.
Resultado(s): Se logro la vinculacion con los registros de prescripci
on dispensada para el 97,9% de las mujeres. Las mujeres que con-
cibieron con CC tendían a ser mayores y con ventajas socioecon omicas y tenían mas probabilidades que otras mujeres de tener ante-
cedentes de aborto espontaneo. Los embarazos en curso asociados a CC representaron el 1,6% del total; El 5,7% fueron partos m ultiples
(principalmente gemelos, 94,6%) en comparaci on con el 1,5% en el resto (98,5% gemelos).
Conclusion(es): En Australia del Sur, el 1,6% de los embarazos (1 de cada 60) de al menos 20 semanas de gestaci on se concibieron cerca
de la dispensaci
on de CC. De estos, el 5,7% fueron embarazos m ultiples. Esto lleva la proporci
on de mujeres que lograron un embarazo
en curso con asistencia medica del 4,4%, segun los informes de las clínicas de tecnicas de reproduccion asistida, al 6% en total.

212 VOL. 117 NO. 1 / JANUARY 2022