Abstract
Following recent debate on prenatal sex selection, the Council of Europe has resolved to encourage national ethics bodies to develop and introduce guidelines for medical staff discouraging prenatal sex selection by whatever means unless justified for the prevention of serious sex-linked genetic diseases. Ultrasound is the most common method for determining sex prenatally and is used for medical and social reasons. In some countries where male offspring are more desirable, fetal sexing during ultrasound examinations has been outlawed due to growing concern that large numbers of female fetuses are being terminated. Although Council of Europe decisions are not binding on the UK, it is timely to examine the use of ultrasound for the determination of fetal sex in the UK and to discuss some of the reasons behind why ultrasound cannot always get it right. We describe sex-linked conditions where sex identification is especially important and conditions and circumstances which make sexing difficult, including ambiguous genitalia. The social aspects of determining the sex of the fetus during an ultrasound examination are also explored as there is no doubt that many women want to know either from curiosity or to help plan for the new arrival. We recommend that ultrasound practitioners remain aware of the accuracy and limitations of ultrasound for the determination of fetal sex and should therefore exercise due caution when imparting sex information to parents.
In the late first, second and third trimester of pregnancy ultrasound is a reliable method for determining the sex of the fetus, both for medical reasons and for social reasons such as parental curiosity. However, accuracy is not always 100%. Therefore, if the sex needs to be determined without doubt chromosomal analysis is usually required. Using ultrasound, more than 95% of fetuses are correctly sexed at 20 weeks. 1,2 This figure sounds impressive but considering an average UK district general hospital has an annual birth rate of 3000, there will inevitably be a few mothers who are ‘surprised’ each year. Does it matter? What are the implications?
For many parents, an important aspect of the anomaly scan is the opportunity to find out the sex of their unborn child. Currently, in the UK most NHS Trusts and Health Boards, although not all, have a policy whereby the ultrasound practitioner is permitted to give their opinion on fetal sex during this examination. We assume that in the UK for most the desire to find out the sex of the fetus is only to satisfy curiosity or for other practical reasons. However, there is growing concern globally that male offspring are more desirable in some cultures and, as a consequence, large numbers of female fetuses are being terminated. This has been described as ‘gendercide’, a term first used in 1985 by the American philosopher Mary Anne Warren. 3 A sad irony is that undoubtedly, due to the error rate, some of those terminated will be male.
The morality and acceptability of using ultrasound for sex selection has been debated recently at a European level. 4,5 Here, we examine the reasons behind why ultrasound cannot always get it right. We describe briefly some sex-linked conditions where sex identification is important, and discuss the social aspects of sexing fetuses during ultrasound examinations.
Development of primary sex characteristics
The sex of a person is determined by the karyotype (normally 46XX for females and 46XY for males) and by hormonal influences in embryonic development at around six to seven weeks post fertilization. A critical event that determines whether the embryo will develop into a boy or girl occurs in the second half of week six. If the Y chromosome is present in the embryo's cells, a gene within this chromosome will turn on the sex-determining region on the Y gene (SRY gene) initiating the production of male hormones (androgens), which are necessary for the embryo to develop into a boy. 6 If two X chromosomes are present or if this particular gene is missing from the Y chromosome the embryo will develop into a girl.
Babies can be born with ambiguous or absent genitalia. This is a very complex area with regard to both subsequent evaluation and management after birth and specialist referral will be required. The main classifications are:
Female pseudohermaphroditism. This is a disorder in which girls with otherwise normal 46XX chromosomes are masculinized because of exposure to an excess of androgens during prenatal development. In baby girls it can enlarge the clitoris so that it resembles a penis. The internal female genital anatomy is otherwise normal;
7
Persistent Müllerian Duct Syndrome is a rare type of pseudohermaphroditism that occurs in 46XY boys. This results in a persistence of a rudimentary uterus and fallopian tubes in an otherwise normal boy. Unilateral or bilateral undescended testes sometimes cause infertility; Male pseudohermaphroditism is a failure of androgen production or inadequate androgen response (androgen insensitivity syndrome), which can cause feminization in XY boys. It can vary from mild failure of masculinization with undescended testes to complete reversal of sex and female phenotype; Mixed gonadal dysgenesis is a form of complete gonadal dysgenesis, mostly due to mutations of the SRY gene on the Y chromosome. Babies with this disorder have female external genitalia and a normal uterus and fallopian tubes. However, they do not have functional gonads (ovaries or testes). Instead, they have undeveloped clumps of tissue called streak gonads; True hermaphrodites usually have an ovary on one side and a testis on the other side, or the gonads may be fused as ovotestes. A uterus is often present but may be hypoplastic. External genitalia may be ambiguous;
8
Abnormal karyotype. Examples are Klinefelter's syndrome (46/47, XXY) and Turner's syndrome (45, XO). In the latter example only rudimentary ovaries are present; there will also be a varying range of other physical features. An abnormal karyotype may also be associated with (4) and (5) above; Exstrophy-epispadias complex. This rare and complicated condition represents a spectrum of genitourinary malformations which may involve the bladder, abdominal wall, pelvic floor, genitalia and occasionally the spine and anus.
9
Surgical reconstruction of genitals may be required but in cases of inadequate or absent penis in genetic men, neonatal assignment to female sex is common;
10
Maternal factors. In rare cases pathology affecting the mother such as androgenizing tumours may affect the development of primary sex characteristics as can the misuse of androgenising drugs in pregnancy.
Although individually the above conditions are rare (some extremely rare), taken together the incidence of ambiguous or absent external genitalia at birth is in the order of 1:2000 to 1:4500 depending on how the condition is defined. One team suggests that gonadal abnormality of some degree may be present in up to 1% of all live births.
6
These cases are likely to cause the parents considerable distress,
11
and present them with a major dilemma. Most parents will be unaware even of the possibility of having an infant with ambiguous genitalia and are usually totally unprepared. Having perhaps been told that it is ‘definitely’ a girl or a boy during an ultrasound scan may well add to this distress. Given the relative rarity it is not suggested that the existence of ambiguous genitalia is discussed unless the sonographer really does have a concern, in which case appropriate onward referral for a specialist opinion should be made. A useful paper by Pajkt and Chitty
12
reviews the role of ultrasound in the diagnosis of ambiguous genitalia, and is readily available online.
Sex-linked disorders requiring accurate assessment of fetal sex
Any disease or abnormal condition that is determined by the sex chromosomes or by an anomalous gene on a sex chromosome is defined as being sex-linked. As described above syndromes such as Turner's and Klinefelter's result from absence or multiplicity of the X chromosome. However, other sex-linked conditions such as androgen insensitivity syndrome, haemophilia, fragile X and Duchenne muscular dystrophy (DMD) are due to a faulty gene on the X chromosome. Many of these conditions are described as X-linked recessive since they do not affect the individual in the presence of a normal X chromosome. Therefore these conditions are always carried and passed on by unaffected women (Xx) and affect only men (xY). Although some are debilitating and/or fatal like DMD, which is a progressive wasting condition, others like colour blindness, have less serious consequences.
There are no known abnormal conditions carried by the Y chromosome. Indeed its function is primarily to instigate male secondary sexual characteristics.
Clearly, families affected by a sex-linked condition are faced with serious decisions, both before and during pregnancy. Ultrasound alone is often not adequate for sexing these embryos or fetuses. Invasive testing such as chorionic villus sampling or amniocentesis used to be the only way to determine reliably fetal DNA and therefore sex. However, risk of miscarriage is approximately 1%. 13 Now, non-invasive testing has been developed to detect fragments of free fetal DNA (ffDNA) in maternal plasma early in pregnancy. The test works on the principle of detecting fragments of Y chromosome-specific sequences. Female fetuses are not detected directly but only inferred by a negative result for fragments of Y chromosome. With 100% accuracy in some studies, it is a reliable and safe alternative. 14–18
Why ultrasound is not 100% accurate
In general terms the accuracy with which the fetal sex can be determined by ultrasound increases with gestation. Many studies have been published that have assessed its accuracy from 11 weeks onwards. In the first trimester Efrat et al. 19 reported that the accuracy of sex determination increased from 70% at 11 weeks to 98.7% at 12 weeks and 100% at 13 weeks gestation. It was not possible to determine the fetal sex in 8.7% of cases. Pajkrt and Chitty 12 reviewed the role of ultrasound in the determination of fetal sex and evaluation of ambiguous genitalia and compared accuracy in a number of studies in both the first and second trimesters. Harrington et al. 2 reported that at the 20-week anomaly scan 96.6% of fetuses were sexed correctly and 3.3% incorrectly. It was not possible to determine the fetal sex in 10%.
Although some have reported 100% accuracy the general consensus is that, using modern ultrasound units, it is greater than 95% at or after the 18+0−20+6 weeks anomaly scan. 1,17,20 This is when sonographers are asked most often to give an opinion on fetal sex. There are consistently better accuracy rates reported for male fetuses in the second and third trimester.
Despite studies determining high levels of accuracy, there are still factors which will always prevent 100% success. These include fetal lie, maternal obesity, poor quality ultrasound equipment and inexperienced operators. Furthermore, even if all the above factors are optimized, pseudohermaphrodite conditions such as congenital adrenal hyperplasia and androgen insensitivity syndrome, discussed earlier, will continue to cause problems for sonographers.
For all the reasons given, it is vital that sonographers and midwives emphasize the error rate of ultrasound, albeit small, when using it to attempt sexing. 2 Parents have high expectation levels and may need reminding that ultrasound is not a perfect science. For example, one woman tried to sue an English NHS Trust over inaccurate sexing in utero since she claimed it prevented her from bonding effectively with her baby after birth. Another woman at a different Trust took trouble to write a complaint against a sonographer stating that she had just endured a very difficult birth which was made all the worse by the baby not being the sex the sonographer had told her it was going to be. The sonographer had explained at the time that inaccuracies occur, but clearly the warnings went unheeded.
It is, therefore, unfortunate that some advertisements in mainstream radiology magazines and online sites perpetuate the myth that ultrasound is 100% accurate. A quick search of the Internet reveals dozens of adverts from three/four-dimensional scanning companies offering 100% accuracy. Such claims are misleading and only add to the appetite among parents and their relatives for wanting cast iron guarantees on the sex of their unborn child.
What then can ultrasound practitioners do to ensure parents receive and understand information on fetal sexing? Arguably, very little apart from including basic information in patient leaflets prior to the examination and reminding them of the error rate at the time of the examination. The Society and College of Radiographers have also produced recent guidance which is freely available to both members and non-members from its website. 21 Certainly it is unprofessional to say the fetus is ‘definitely a boy’ when, put simply, it may not be. The two brief cameos outlined above demonstrate how deeply some are affected when we get it wrong.
It has been suggested that opinions on fetal sex at ultrasound should be withheld, and some NHS Trusts and Health Boards do, but parents have the freedom to access services provided by another Trust that will attempt to sex their baby or they can easily buy a private scan, so this is not the answer. Furthermore, women are entitled to know what the scan findings are and it may be viewed as discriminatory to be selective on who to tell or not tell. In the hopefully rare incidences where termination or postnatal neglect results from the information, this is not the responsibility of the sonographer or service provider.
The need to know
Imparting information to parents on probable gender for sex-linked conditions can bring either relief and joy or devastation and grief but few would argue that, in these cases, sex identification is important. Thankfully, these women represent only a tiny proportion of the pregnant population. However, many women who are not at risk of a sex-linked condition also want to know the sex of their child in advance of the birth. 2 Anecdotally, only a small percentage seems to want a ‘surprise’. There is little robust research in this area but it seems that some feel this knowledge will help them better prepare for the arrival of their baby. Reasons given to sonographers include being able to buy the ‘right’ colour clothes and paint the nursery, or to prepare siblings for the new arrival. For others it may offer comfort after a recent miscarriage or provide information for a terminally ill relative for whom knowing the sex of the fetus is important. For some it is simply curiosity. The NHS Fetal Anomaly Screening Programme (FASP) has produced National Standards and Guidance for the 18+0−20+6 weeks fetal anomaly scan in England and for screening in the first trimester for Trisomy 21. 22 There are similar documents that relate to screening for fetal abnormality produced by the equivalent organizations to NHS FASP in Scotland 23 and Wales. There is no requirement to determine the fetal sex in these programmes. The information provided on fetal sex is therefore to a locally agreed protocol if it is undertaken and may have a resource issue in view of the additional, albeit small, time required.
For some parents, the need to know is driven by a strong preference for one sex or the other and this is often clear from the parents' behaviour once they have received the information during the scan. That said, it is accepted that the ultrasound practitioner is simply providing sexing information at the request of the woman and does not have a further responsibility with respect to how that information is subsequently used. Not all ultrasound practitioners feel entirely comfortable with this, however. If parents are told the ‘wrong’ sex one may assume they quickly come to terms with the disappointment and are philosophical towards the future but this may not always be the case.
Many cultures and populations in, for example, India, Nepal, China, South Korea, Vietnam, Jordan, Albania and some former Soviet states including Azerbaijan, Armenia and Georgia have a preference for boys. 4,24 Termination of female fetuses is occurring and, as a consequence, some of these countries are beginning to record a departure from the normal child sex ratio. Some prosperous Indian districts have as few as 678 girls per 1000 boys thought to be due directly to misuse of prenatal sexing. 24 In fact, with elimination rates of female fetuses exceeding one in seven, fetal life in some cultures may be the riskiest time for a woman according to George. 24 The potential scale of the problem is illustrated in a small contemporary American study, which showed that 40% of Indian women immigrants admitted to terminating a pregnancy on the basis of it being a girl, and 89% currently pregnant with a female fetus had tried to obtain a termination. 25 Evidence that child sex ratios are changing in some UK immigrant populations is also beginning to emerge, although this is not the case among Bangladeshi and Pakistani UK communities. 4
There are many methods that can be used to determine and select the fetal sex that may have ethical and moral considerations. For example, preimplantation genetic diagnosis makes it possible to identify chromosomes in embryos created by in vitro fertilization. This is prohibited in the UK except when it is used to screen for genetic diseases. The Human Fertilisation and Embryology Act 1990, as amended, prohibits the use of techniques designed to select an embryo of a desired sex for social reasons or family balancing. However, it is acceptable for preventing the transmission of serious medical conditions, illness or physical or mental disability. 26 Laws in the USA and some other countries are more relaxed. In early pregnancy commercially available tests for ffDNA claim to give 95% accurate results as early as seven weeks. These are not yet legal in the UK but are available on the Internet, 27 and may lead to more women seeking to terminate the pregnancy in the first trimester when they might otherwise not have done so.
As discussed earlier, ultrasound can also determine fetal sex accurately from as early as 12–13 weeks. 12 In the UK termination of pregnancy for ‘wrong’ fetal sex is illegal. The Abortion Act (1967) 28 does, however, allow for termination for psychological reasons including maternal anxiety and depression and it is likely that some women will seek an early termination for ‘wrong sex’ although not cite this as the reason. Anecdotally, some ultrasound practitioners have reported that during an anomaly scan they have been asked whether it will be possible to terminate a pregnancy for a ‘wrong sex’ fetus. Some ultrasound practitioners have also reported that they have felt uncomfortable about providing information on the sex of the fetus when they suspected that a termination might be sought, although this would be impossible to obtain in the UK without medical grounds. 27 These instances are of course rare in comparison to the great majority of ultrasound examinations on women who simply want to know for curiosity or practical reasons.
The Council of Europe has recently debated issues around prenatal sex selection in response to the skewed ratios now emerging from some populations; globally the most common method is termination after ultrasound. 5 They are encouraging national ethics bodies to develop and introduce guidelines for medical staff discouraging prenatal sex selection by whatever method unless justified for the prevention of serious sex-linked genetic diseases. Council of Europe recommendations would not, however, be binding on the UK but it is a highly influential body and has often seen its decisions enacted through conventions and treaties. The Department of Health (England) has so far reserved judgement stating that the primary purpose of an ultrasound examination is to date the pregnancy and to look for fetal abnormality. 29 The influential Internet group ‘Mumsnet’ is also highly supportive of women being able to be informed of the sex of their baby at an ultrasound scan. 27
Conclusion
In the UK many Trusts and Health Boards allow ultrasound practitioners to give an opinion of the fetal sex if requested; although providing this information does not form part of any of the national fetal anomaly screening programmes. The accuracy of fetal sex determination by ultrasound is greater than 95% but it is important that women are informed at the time of scan that errors can occur. In some countries termination for ‘wrong sex’ has resulted in skewed male/female ratios among the population, and providing fetal sex information during an ultrasound scan has subsequently been made unlawful (Figure 1). The Council of Europe has recently debated issues surrounding prenatal sex selection although its decisions and recommendations are not necessarily binding on the UK. 5 The great majority of women in the UK want to know just for curiosity or practical reasons. Occasionally, however, ultrasound practitioners in the UK will encounter situations where it is apparent that a woman may seek a termination of a ‘wrong sex’ pregnancy thus raising issues of moral and ethical concern both for the ultrasound practitioner and society in general.

A sign in a New Delhi hospital stating that prenatal sexing is a crime. Freely available at
DECLARATIONS
