The aim of OI is to use the lowest effective dose of fertility drug in order to achieve monofollicular ovulation for patients with anovulatory infertility [12]. This is then repeated monthly until pregnancy is achieved for up to six to nine cycles. The method of ovulation induction depends on the ovulatory disorder, classified by the World Health Organization [13]. The two groups that would benefit from ovulation induction are those with hypothalamic pituitary failure (Group I: hypothalamic amenorrhoea or hypogonadotrophic hypogonadism) and hypothalamic pituitary dysfunction (Group II: normogonadotrophic, predominantly polycystic ovary syndrome).

In WHO Group I, women may have low or normal serum FSH and LH, with low estradiol concentrations and normal or low testosterone. They do not have a withdrawal bleed with a progesterone challenge test [13]. Ovulation is induced either with pulsatile gonadotrophin-releasing hormone via a pump or with urinary (FSH and LH) or recombinant gonadotrophin (FSH) therapy. The aim is to support the growth of a single follicle until it reaches 16–18 mm size when hCG is administered to trigger ovulation. Usually a low-dose step up regime of gonadotrophins is used to minimise multifollicular development, reducing rates of multiple pregnancy and ovarian hyperstimulation [12]. If the trigger of hCG is administered in the presence of more than one large follicle the rates of multiple pregnancies exponentially increase, with reported rates of 50 % with greater than 3 large pre-ovulatory follicles [10]. It is therefore recommended to cancel the hCG trigger and to advise the couple to avoid unprotected sexual intercourse in that cycle if there are >3 pre-ovulatory follicles developed.

In WHO Group II disorders, clomifene citrate is used for stimulation of ovulation by blocking the estrogen receptors in the hypothalamus and blocking the negative feedback effect of estradiol [12], leading to increased endogenous FSH secretion and stimulating follicular development. Again the aim is to use the lowest necessary dose of clomifene in order to nurture one follicle. The risk of multiple pregnancy rises from the background rate of 1 in 80 to 1 in 10–20 with clomifene use, becoming more common with the use of higher doses of clomifene in those with PCOS [14]. Side effects of clomifene include Ovarian Hyperstimulation Syndrome (OHSS) (1–6 %) [12], visual disturbances, nausea, vomiting, dizziness and in some cases seizure activity.

In this same group of ovulation disorders, FSH can be used for women resistant to clomifene to achieve ovulation. As for the Group I disorders a low-dose step up regime is employed to reduce the risks of multiple pregnancy and OHSS. In some cases aromatase inhibitors such as letrozole have been used. They work by decreasing the aromatization of androgens to estrogens, decreasing the negative feedback cycle of estradiol and increasing follicular growth [12]. Pregnancy rates are promising with a lower incidence of multiple pregnancies, and a more favorable effect on the endometrium compared with clomifene [15].

IUI with controlled ovarian stimulation is widely used in cases of unexplained subfertility and mild male-factor infertility before resorting to more invasive options like IVF [16]. In contrast to older studies, more recent evidence has suggested that using IUI with gonadotrophin stimulation may correct subtle ovulation issues, leading to a greater number of oocytes and consequently a higher live pregnancy rate. The offset is of course multiple pregnancies, rates of which have been reported as high as 20–30 % in some centres regardless of the infertility cause [10, 11]. Reduction of the risk of a multiple pregnancy can be obtained by either avoiding any gonadotrophic stimulation (a ‘natural cycle’), using strict cancellation regimes or using a low-dose step up regime similar to that described in the last section of this chapter. This can result in a reduction to 10 % multiple pregnancy rate without an overall impact on live birth rates [17]. Recent NICE guidelines [1] do not support using this method in those with unexplained infertility and instead suggest that it is restricted to those who are unable to have vaginal intercourse due to a disability or psychosexual problem, those in whom sperm washing is appropriate (such as HIV positive men) or those in same sex relationships. However, NICE are currently reviewing this recommendation which is therefore likely to change at the next update.

In IVF procedures, controlled ovarian hyperstimulation can also be employed to generate the follicles for embryo creation in vitro. It is the number of embryos transferred which has a direct bearing on the chance of a multiple pregnancy. The risk of twins after double embryo transfer (DET) is 23.5 % for cleavage stage embryos (day 2–3 of development) and 36.4 % for blastocysts (day 5 of development). Elective single embryo transfer (eSET) in selected patient groups has shown promising success with clinical and live pregnancy rates, which are not dissimilar to those for double or higher order embryo transfers.

IVF itself appears to increase the risk of monozygotic twins by twofold compared with natural conception (0.8 % vs 0.4 %) although the overall incidence is low [6, 18]. The HFEA routinely collects outcome data from all IVF/ICSI treatment cycles across the UK and have reported that the incidence of twin pregnancy after eSET of a cleavage stage embryo is 0.6 %, identical to natural conception whereas eSET of a blastocyst embryo results in a more than doubling of the twin pregnancy rate to 1.9 % [3]. These are presumed to be monozygous pregnancies as the chance of simultaneous natural conception is thought to be very low.

Monozygosity itself is associated with higher adverse outcomes as two thirds of monozygous twins are monochorionic [7]. A twin pregnancy with a shared chorion is at increased risk of complications due to the vascular placental anastomoses that connect the umbilical circulations of both twins, leading to twin-to-twin transfusion syndrome (TTTS), which complicates 10–15 % of monochorionic pregnancies. This leads to haemodynamic and liquor discordance in the “donor” and “recipient” twin and in severe cases death of the recipient twin due to high output cardiac failure. In these cases death of the surviving twin can be as high as 12 % with the risk of neurological abnormalities in those that do survive being approximately 18 % [19]. Monochorionic pregnancies also have a higher chance of fetal loss greater than 24 weeks of gestation (3.3 % fetuses) compared with dichorionic pregnancies [19]. Overall these babies may also be more at risk of neurodevelopmental abnormalities.

IVF twins also seem to have a small but statistically significant increase in the risk of preterm labour, approximately by 23 % [6] compared with spontaneously conceived twins. Early fetal loss of one twin can lead to premature delivery of the remaining twin. Similarly IVF twins have shown an increased risk of low birth weight in IVF twins [6]. Rates of congenital anomalies are known to be 30–40 % higher in IVF pregnancies (septal heart defects, cleft lip, oesophageal atresia, anorectal atresia). The risk of anomalies after conventional ART is the same as natural conception but is higher after ICSI. This is much like a “chicken and egg” problem with ICSI. It is not completely clear if it is the process of ICSI (i.e., stripping cumulus cells and longer exposure to light and oxygen, which causes the anomalies) or if it is due to the underlying sperm dysfunction which provokes the need for ICSI in the first place. Most severe sperm dysfunction (<5 million sperm/ml) is probably genetic involving the Y chromosome, although there is no evidence to suggest that in multiple pregnancies this rate would be higher, especially for hypospadias [20, 21]. The risk of congenital heart disease in monochorionic twins has been shown to be higher [19, 22].

Overall, it should be acknowledged that factors that predispose to infertility are also linked with adverse perinatal outcomes. To determine whether a particular ART is leading to an adverse outcome or whether it is a consequence of other infertility causes and complex factors between the couple needs further investigation [6, 7].

The ethical rationale in relation to fetal reduction is that of a “consequentialist” approach, in which the parents and the clinician weigh the benefits and risks of the pregnancy continuing and make a “best interest” decision for the remaining fetus(es) and for the mother’s health [24, 25].

Multifetal pregnancy reduction (MFPR) attempts to ameliorate the maternal and fetal risks of higher order pregnancies by reducing the number of fetuses to a more manageable number [8]. Epidemiological studies have shown that twin pregnancies produced a child with cerebral palsy 8 times more often than singletons and for triplet pregnancies this rate was 47 times higher [26]. For example, 8–12 % of triplet pregnancies will experience some kind of neurodevelopmental sequelae compared with twin pregnancies. This is likely to be even higher if the triplet pregnancy contains a monoamniotic pair [19]. Reducing the triplet pregnancy to twins significantly reduces the risk of preterm delivery without an increase in miscarriage rates [27]. Full fetal medicine assessment should be carried out before deciding on which fetus(es) to terminate. This is best carried out between 11 and 14 weeks gestation when the risk of spontaneous reduction has passed and in order to identify features of aneuploidy (i.e., nuchal translucency) [27]. Fetuses at lowest risk of aneuploidy, determined by nuchal translucency should be left intact as should those implanted closest to the cervix so as not to increase the risk of miscarriage of the entire pregnancy should the fetus closest to the cervix miscarry following MFPR. Studies to date do have major limitations, however, as many do not differentiate between trichorionic and non-trichorionic pregnancies, the latter in which a monochorionic pair exists, which of course will have a bearing on fetal outcomes (which will be discussed below). However, despite the controversies, reducing triplets to twins suggests that the chance of preterm labour before 32 weeks gestation drops by around 55 %, with very little increase in miscarriage [27], and the potential to take a live born baby home increases from 80 to 90 % [8]. However, it is clear that expectant management of a trichorionic triplet pregnancy does have a reasonable perinatal outcome.