Fetal macrosomia: a problem in need of a policy



Fetal macrosomia is associated with increased risks for the mother, including emergency Cesarean section (CS), instrumental delivery, shoulder dystocia and trauma to the birth canal, bladder, perineum and anal sphincter; for the baby, complications include increased mortality, brachial plexus or facial nerve injuries, fractures of the humerus or clavicle and birth asphyxia[1-7]. Fetal macrosomia now is usually defined as a neonate with a birth weight above 4.5 kg[2-4] and has a prevalence in developed countries of 1.3–1.5% of all births. A gestational-age-dependent definition, the 97th centile, is sometimes used and has a similar prevalence. Other definitions, such as birth weight > 4 kg (prevalence, 7%) or birth weight > 90th centile[5], include many babies with a lower order of risk for adverse outcome. The 4.5-kg limit is only appropriate for term births and can be criticized for being rather arbitrary; for example, the incidence of shoulder dystocia rises steeply between birth weights of 4.0 and 4.25 kg[6], so 4.25 kg could legitimately be used as the cut-off. However, the 4.5-kg limit defines unequivocally a high-risk group of babies that requires accurate diagnosis and intelligent management to minimize the risks to mother and baby. Fetuses growing above the 97th centile during the third trimester can also be regarded as macrosomic, but, although at increased risk from metabolic problems[7], they do not contribute significantly to the classical injuries associated with fetal macrosomia until term is reached.

Three per cent of all births and 5–10% of macrosomic fetuses[8] are associated with maternal diabetes and the management of this subset will not be discussed in detail in this Editorial. Women with diabetes or gestational diabetes are usually identified prenatally and monitored and cared for by a specialist team. Induction of labor and elective CS is recommended according to individual circumstances[2]. Macrosomic fetuses of non-diabetic women can be identified as being at risk by factors such as maternal obesity and family history but are generally unsuspected until the possibility of a big baby is raised by antenatal clinical or ultrasound examination. Management of this group of pregnant women has been determined largely by two influential publications. In the first, Rouse et al.[9] performed a decision analysis on the monetary costs of three different management policies, i.e. management without ultrasound; ultrasound and elective CS delivery for a predicted birth weight of 4.0 kg; and ultrasound and CS delivery for a predicted birth weight of 4.5 kg. On the basis of an ultrasound sensitivity of 60% and specificity of 90% for prediction of a macrosomic fetus, they calculated that in the USA the 4.5-kg management policy would entail carrying out 3695 CSs or spending an extra 8.7 million dollars for each brachial plexus injury prevented in non-diabetic gestations. Their conclusion was that, ‘for the 97% of women who are non-diabetic, elective CS for ultrasonically diagnosed macrosomia was economically unsound.’ This study was quoted extensively in the second influential publication, the American College of Obstetricians and Gynecologists (ACOG) Practice Bulletin[2], which provided a detailed quantification of the risks of macrosomia and recommended clinical management when ultrasound predicted this condition. The ACOG concluded that although the diagnosis of fetal macrosomia was imprecise, prophylactic CS ‘may be considered’ for suspected fetal macrosomia with estimated fetal weights (EFW) ≥ 5.0 kg (i.e. 11 pounds) in women without diabetes. The implication, of course, is that for weights < 5.0 kg, elective CS should not be considered. This recommendation has since been echoed by several authors[8, 10, 11]. In the UK, the Royal College of Obstetricians and Gynaecologists (RCOG), in its Shoulder Dystocia Guideline in 2012[12], refers to the Rouse paper findings and notes the ACOG recommendations, but does not provide any management guidelines for suspected macrosomia in non-diabetic women. However, as the RCOG/National Institute for Health and Care Excellence (NICE) guidelines on antenatal care (2012)[13] stipulate that ‘Ultrasound estimation of fetal size for suspected large for gestational age unborn babies should not be undertaken in a low risk population’ and the RCOG/NICE guideline on CS[14] does not list fetal macrosomia amongst the indications for elective CS, the view of RCOG/NICE on this subject is clear: elective CS should not be performed for this indication.

Thus, while it is universally accepted that elective CS can virtually eliminate shoulder dystocia and brachial plexus injury without increasing maternal mortality, the current mantra is that ultrasound prediction of macrosomia is inaccurate and brachial plexus injuries are rarely permanent, so justification fails both on clinical and on monetary grounds. However, as the ACOG and RCOG recommend elective CS in women with diabetes with a predicted birth weight > 4.5 kg, they appear to be acknowledging tacitly that ultrasound estimates of birth weight are not valueless. The findings of the Rouse paper have been challenged recently by Culligan et al.[4], who carried out a similar decision analysis comparing current ‘standard care’ with an ultrasound scan at 39 weeks and elective CS for predicted birth weight ≥ 4.5 kg. This analysis included new longitudinal data on brachial plexus injury to the newborn and also risks of maternal anal and urinary incontinence. Their outcome measures included the number of brachial injuries and cases of incontinence averted, incremental monetary costs per 100 000 deliveries, expected quality of life of the mother and child and the ‘quality adjusted years’ (QUALY) associated with each of the two policies. They conclude that a policy whereby primigravid patients in the USA have a 39-week scan to estimate fetal weight followed by elective CS for predicted birth weights ≥ 4.5 kg was cost effective. This paper has had minimal impact on public health care policy, perhaps because it was published in a urogynecological journal. This Editorial will examine the issues surrounding the debate over elective CS for fetal macrosomia and make suggestions for a policy fit for the 21st century.

Risks of macrosomia

Fetal injury

Published graphs of neonatal mortality against birth weight show a sharp rise in babies > 4.5 kg[15, 16]. While this may in part be due to the association of macrosomia with post-dates gestations and maternal diabetes, it also reflects the high rate of emergency CS and instrumental delivery and subsequent birth trauma in women who labor with a large fetus. For babies weighing 4.5 kg or more, the emergency CS rate is 45%[9] and the instrumental delivery rate 19%[17]. Long-term damage to the infant is most commonly discussed in relation to shoulder dystocia. The average risk of shoulder dystocia in the obstetric population is 1.4%, but when birth weight is greater than 4500 g the risk varies from 9 to 24%[2]. The large study from California[6], which included 6238 infants with shoulder dystocia, put the risk at 14.3% for non-diabetic infants weighing between 4500 and 4750 g and 21.1% for infants weighing 4750–5000 g. The incidence of shoulder dystocia in vacuum- or forceps-assisted births was 23% for infants with birth weight between 4500 and 4750 g and 29% for those between 4750 and 5000 g. Similar findings were reported by Acker et al.[18]. Rates of shoulder dystocia in mid-forceps deliveries of infants heavier than 4500 g have been reported to be above 50%[2], so, barring extreme emergencies, CS delivery should be performed for mid-pelvic arrest of the fetus with suspected macrosomia. Induction of labor is associated with a doubling of the risk of CS delivery without reducing the risk of shoulder dystocia[19, 20].

Obstetric brachial palsy injury (OBPI) is strongly associated with macrosomia and shoulder dystocia. The average rate of OBPI in the obstetric population is 0.5–1.9/1000, while for babies with a birth weight > 4.5 kg there is an 18–21-fold increased risk[2]. Macrosomic compared with non-macrosomic infants delivered after shoulder dystocia are at a higher risk of OBPI. Rouse et al.[9], in their analysis, estimated that 26% of babies with birth weights ≥ 4.5 kg delivered following shoulder dystocia will have a brachial plexus injury. Overall, the risk of OBPI for macrosomic infants delivered vaginally is 4–8%[2]. Raio et al.[21] stressed the importance of short maternal stature as an associated risk factor in predicting birth injury. They calculated that the number of CS deliveries required to prevent a single case of permanent OBPI in fetuses predicted to be ≥ 4.5 kg would be 30 in women shorter than 155 cm and 715 in those taller than 180 cm. Gudmundsson et al.[22] confirmed this and constructed risk estimation curves based on birth weight and maternal height.

The risk of permanent injury is frequently reported to be about 10% among all cases of OBPI[2] and Rouse et al.[9] estimated for their cost analysis that the risk of permanent injury from OBPI was only 6.7%. Most of these estimates were based on short-term follow-up and few gave details of the pediatric examinations. Subsequent to the Rouse paper, two substantial longitudinal studies, each of over 60 infants with OBPI carried out in orthopedic surgery and rehabilitative medicine departments, were published. Waters[23], in an extremely detailed study, found that only 16.7% of OBPI cases spontaneously resolved, 27% were permanent and severe (arm useless throughout life) and 56% were permanent and moderate (abduction and rotation limited to < 30°). Hoeksma et al.[24] reported that only 51% of OBPI children underwent complete recovery because shoulder contractures with functional deterioration developed in 30% of infants who had apparently made a full neurological recovery. It should also be noted that persistent OBPI is six times more common when the birth weight is ≥ 4.0 kg compared with < 4.0 kg6. Culligan et al.[4] in his cost analysis factored in the data from the Waters study.

More serious are the other problems associated with the birth of a large baby. The California study[6] found that birth injuries in general as well as birth asphyxia were significantly more frequent in infants born following labors complicated by shoulder dystocia. There was also an association between neonatal length of stay and ‘non-normal’ neonate at discharge. Iffy et al.[25] studied retrospectively 316 fetal neurological injuries associated with arrest of the shoulders, across the USA. He found that as many as 64 (20%) children displayed manifestations of central nervous system (CNS) damage, such as cerebral palsy, seizures and cognitive defects, 6 or more months after their birth. They estimated that the risk of irreversible fetal damage in cases of attempted vaginal delivery exceeds 2.5% for fetuses ≥ 4.5 kg and 5% for birth weights above 5 kg. It should be noted that these asphyxia problems were only recorded in association with shoulder dystocia. The overall risk of CNS problems associated with macrosomia is likely to be higher.

Maternal injury

There is a strong correlation between macrosomia and pelvic floor damage and the development of anal and urinary stress incontinence and prolapse. Birth weight > 4 kg imposes risk of perineal injury, especially third- and fourth-degree tears due to larger head circumference (HC), prolonged labor and difficult delivery, especially if instrumental birth is performed.

Anal incontinence represents a distressing social handicap and vaginal delivery is now recognized as its principal cause. Three large studies[26-28] using multivariate regression models found that macrosomia was a strong independent risk factor for anal sphincter damage. The other consistent factor was assisted vaginal delivery, especially with forceps. Even after safe delivery of the head, shoulder dystocia – more common in macrosomic infants – may contribute to perineal and anal sphincter trauma. There is a strong association between anal sphincter injury and maternal fecal incontinence. A meta-analysis of five studies in which endoanal sonography was performed after vaginal birth found a 26.9% incidence of anal sphincter defects, one third of which were symptomatic[29]. Culligan et al.[4] calculated the likelihood of chronic anal incontinence following third- or fourth-degree anal sphincter disruption to be 23%, but this seems high and a figure of 9–11% is more likely[30-32]. Anal incontinence has a significant negative effect on sexuality, exercise, social activities and work activities and is directly associated with depression[33].

A large baby is also likely to disrupt the fascial supports of the pelvic floor and cause a stretch injury to the pelvic and pudendal nerves, leading to vaginal prolapse. Handa et al.[30] followed up over a thousand women for a longitudinal cohort study 5–10 years after first delivery. Compared with elective CS delivery, spontaneous vaginal birth was associated with a significantly greater odds of stress incontinence and prolapse to or beyond the hymen. Operative vaginal delivery significantly increased the odds for urinary stress incontinence (odds ratio (OR), 4.45), anal incontinence (OR, 2.22) and especially prolapse (OR, 7.5). What was important was that CS was protective even when performed in active labor up to complete cervical dilatation.

Vaginal birth predisposes to genuine stress incontinence (GSI). Urethral closure pressure and functional length are reduced following vaginal delivery, but this does not occur after CS[34]. The association between a large fetus, prolonged second stage and perineal nerve damage has also been clearly demonstrated[35]. Farrell et al.[36], in a follow-up study of healthy primiparae, found urinary incontinence rates at 6 months of 10% following CS, 22% following spontaneous vaginal delivery and 33% following forceps delivery. CS at any stage was protective. Viktrup et al.[37], in a prospective study, found that postpartum GSI was independently related to birth weight, episiotomy and HC. In a follow-up survey, Viktrup and Lose[38] reported a prevalence of GSI of 30%, 5 years after first delivery. Like Handa et al.[30], they found CS birth to significantly reduce the risk of GSI. Meyer et al.[39] found similar results with 21% of women having GSI persisting after vaginal birth but only 3% following CS.

In summary, there is a strong relationship between the delivery of a macrosomic fetus, instrumental delivery, anal sphincter disruption and the development of anal and urinary stress incontinence and uterovaginal prolapse. What is also clear is that most of the serious effects for both child and mother are related to vaginal birth and that CS at any stage of labor up to full dilatation is largely protective for these complications. With this knowledge it has to be asked why the obstetrician proceeds to a difficult delivery when a decision to perform CS would avoid these problems.

Antenatal prediction of macrosomia

It is highly likely that the current laissez-faire attitude towards management of the suspected macrosomic fetus would be different if a very precise method of estimating fetal weight were available. It is stated not infrequently that ultrasound is no better than clinical examination[40-42] or the patient's own estimation of the weight of her baby[43, 44]. There is now, however, compelling evidence that ultrasound is superior to all other methods at the extremes of the birth-weight range[45-47], with, perhaps, the exception of magnetic resonance imaging (MRI)[48]. Therefore, ultrasound is the only practical method with which to screen pregnant women for fetal macrosomia. The almost universal practice is to put the measurements of biparietal diameter (BPD), HC, abdominal circumference (AC) and femur length into a regression model which calculates EFW. A few of the equations that are currently used remove the BPD component (to remove the bias caused by dolichocephaly) and a few use the alternative of the mean abdominal diameter (AD). Ultrasound has a maximum random error of approximately 100 g/kg, which means that weight predictions in grams for small fetuses appear to be more clinically useful. Indeed, many researchers have found ultrasound prediction of macrosomia to be inaccurate and of little clinical value[49, 50].

There are two principal components contributing to inaccuracy in ultrasound predictions, i.e. the systematic error, which is dependent on the appropriateness of the prediction formula, and the random error, which is dependent on the inherent inaccuracies of the technique and can be lessened by taking repeated and multiple fetal measurements[51] and making certain that the sections of the fetal anatomy have been obtained accurately. Other factors that increase random error are maternal obesity, oligohydramnios, poor-quality equipment and an inexperienced operator. In Dudley's[52] comparative analysis of 11 different formulae, it is clear that the systematic errors of the various formulae are close to the mean for normal-weight fetuses but most underestimate the weight of large and overestimate the weight of small fetuses. What is also striking from Dudley's analysis is that the percentage random error for large fetuses is smaller than that for normal or small fetuses, with 95% confidence limits consistently just below 10%. Dudley's findings were confirmed by Melamed et al.[51], who compared 26 different formulae on 3705 fetal weight estimations within 3 days of birth. For fetuses ≥ 4.5 kg, the random error was 10% or less in 21 formulae, with an overall mean of 8.1%. The systematic underestimation of birth weights in this category was –6.2%. The authors postulated that a strategy of replacing the original coefficients of the models by coefficients derived for the specific study population could theoretically have reduced the systematic bias in the macrosomic group. This makes it important to have a targeted formula for fetuses over 4.5 kg, to reduce systematic errors to a minimum[53]. Most fetal weight prediction studies for macrosomia choose 4 kg and above as the definition of macrosomia or select high-prevalence populations, such as women with diabetes mellitus[54] or who are post-dates[55], thus improving the positive predictive value (PPV) of the test. As demonstrated by Rouse et al.[9], a high PPV is essential to reduce costs from unnecessary CSs.

If screening for the large-for-gestational age infant is to be attempted, it might be a good option to undertake a two-stage operation, i.e. a screening scan at around 32–34 weeks' gestation to identify a high-risk group, followed by a detailed scan at 39 weeks in those identified as being large. The object of the earlier or triage scan would be to achieve as high a sensitivity as possible so that most large fetuses are in the screen-positive group. The aim of the second diagnostic scan would be to achieve a high predictive value so that the woman would know that her chances of delivering a fetus weighing more than 4.5 kg are high. The 39-week scan being performed on a high-risk population with a higher prevalence of large fetuses would therefore be more likely to achieve clinically useful PPVs, especially if a targeted formula for macrosomia and the latest ultrasound three-dimensional (3D) volumetric studies or even MRI were utilized[56-58]. Unlike MRI, the disadvantage of 3D volumetric studies is that the constraints of transducer size limit the region of interest that can be examined, so the current idea is to include the volume of a short cylinder rather than a two-dimensional (2D) slice in the typical 2D EFW equation. For example, recently Lindell et al.[58] used a formula combining 2D measurements of fetal head, abdomen and femur and 3D volumetry of fetal abdomen and thigh to assess the weight of 114 clinically and ultrasonically large-for-dates fetuses. For fetuses > 4.5 kg (with the formula set at a cut-off of 4.3 kg), they were able to identify 93% of macrosomic fetuses for a false-positive rate of 38%. This 2D/3D formula was an improvement on existing 2D equations for the prediction of macrosomia. At the moment it is a time-consuming technique as volume measurements are made offline, but, with the development of automated volume estimation, 3D volumetry may add considerably to the precision of sonographic weight estimates. Another approach is to attempt to predict shoulder dystocia on the premise that it is more likely to develop in fetuses with disproportionately large abdominal girth. This is particularly true of the fetus of the diabetic mother. A paper in this issue of the Journal[59] explores this concept; although AC and EFW had better sensitivities and specificities, an AD − BPD difference > 26 mm had a high OR of 7.6 for predicting shoulder dystocia.

Another benefit of the early third-trimester scan would be to provide an opportunity to identify and treat women with abnormal glucose tolerance and address any problems associated with obesity. The scan would also reveal an increase in amniotic fluid volume, which is also associated with macrosomia[60], and, at the other end of the scale, any unexpectedly small fetuses.

Several authors have found AC measurement alone to be as accurate as more complex formulae for predicting fetal weight or macrosomia[46, 52, 61, 62]. While multiple-parameter EFW would be ideal, the logistics of screening every woman might favor the simple approach, especially in low-risk settings. For example, in a study from a Dutch primary care midwifery practice[63], routine AC measurements performed between 27 and 33 weeks enabled the detection of two thirds of cases of macrosomic fetuses with a PPV of 23%. If all the scans had been performed after 30 weeks, the results would likely have improved. Several authors have reported that AC is effective in identifying the large fetus in the early third trimester, although many have used a birth weight of 4 kg as the definition of macrosomia. Pilalis et al.[64] found that screening with AC alone between 30 and 34 weeks identified 70% of babies with birth weight over 4 kg for a screen-positive rate of 25%. A full EFW with four parameters contributed only a 3% improvement in the detection rate. Lindell et al.[65], using a multiparameter equation, reported sensitivity, specificity and predictive values of screening in an unselected Swedish population at 32–34 weeks. Using a z-score of 0.5 as the cut-off, they achieved a sensitivity of 88% and specificity of 73% for the prediction of birth weight > 4.5 kg. Although the PPV was only 14% it is likely that this would rise considerably if the 27% of women identified as high risk were rescanned at 39 weeks.

The sensitivities and specificities chosen by Rouse et al.[9] and Culligan et al.[4] to input into their decision analyses were almost identical but were based mostly on studies of diabetic or post-dates gestations or babies with birth weights over 4 kg, and are therefore not relevant to the screening of an unselected population to detect birth weights over 4.5 kg. Furthermore, Rouse et al.[9] chose a very low PPV for ultrasound and calculated the cost of a seven-fold excess of CSs attributable to this. It should be remembered that most equations show that ultrasound underestimates the birth weight of macrosomic fetuses. If the systematic error were minimal, with 95% confidence limits of 10% for random errors, nearly all predictions would be between 4 and 5 kg. Shoulder dystocia rises sharply between 4 and 4.5 kg which means that an overestimate would still identify a high-risk population. This is not so for fetal weight predictions for a 4-kg birth weight where many low-risk fetuses would be included in the ultrasound estimate.

In summary, ultrasound estimations of fetal weight to predict macrosomia, although imprecise, are not valueless. It requires, however, a targeted approach and a realization of the seriousness of the condition and the importance of making predictions as accurate as possible.

Patient autonomy

Before considering factors affecting the delivery of the macrosomic fetus, it is worth examining recent developments in relation to maternal autonomy and birth in normal gestation. At the turn of this century, there was an intense debate in the literature as to whether elective CS should be performed at a woman's request in the absence of an obstetric indication. What is clear is that there was no dispute that elective CS and vaginal birth were equally safe in terms of mortality for the mother, but that both had different types of complications, of which prospective parents should be made aware[66]. In a widely quoted study[67], a postal survey asked British obstetricians to consider whether they or their partners would choose either elective CS or vaginal delivery if they had a hypothetical singleton pregnancy with cephalic presentation at term. The response rate was 73%. Overall, 17% chose CS, with 31% of female obstetricians choosing this option over vaginal delivery. Among the reasons given for choosing elective CS was fear of long term sequelae of vaginal birth, specifically stress incontinence and anal sphincter damage, fear of perineal damage from vaginal birth and its effect on sexual function and fear of damage to the baby. A Lancet editorial in 1997[68] stated: ‘The trend for use of CS coupled with greater emphasis on individual autonomy in medical decision making has clearly progressed too far for a return to paternalistic directions to women on how they should give birth. Instead the emphasis should be on comparisons of the implications of vaginal versus CS delivery. The uptake of CS in women made aware of such information will clearly be more appropriate than any of the current “desirable targets”.’ The ACOG[69] in 2003 published a Committee Opinion stating: ‘If taken in a vacuum the principle of patient autonomy would lend support to the permissibility of elective CS delivery in a normal pregnancy after adequate informed consent.’ These sentiments were challenged in combative papers by Bewley and Cockburn[70, 71], who argued that the sizeable minority of female obstetricians choosing elective CS as a first option ‘may be biased by their exposure to the complications of childbirth.’ Their response to a CS delivery on maternal request included referral to a psychiatrist, a mandatory second opinion and a series of ‘checks and hurdles’ in a continuing dialogue with the woman which might be mistaken for coercion. Subsequently, a similar USA survey[72] confirmed a high (21%) preference for maternal request among American obstetricians, listing urinary and anal incontinence and concern for fetal death or injury as reasons. A further survey in New Zealand[73] found that 11% of midwives, 21% of obstetricians, 42% of urogynecologists and 41% of colorectal surgeons preferred the option of elective CS.

What is particularly relevant to this Editorial is that when obstetricians in the original survey were presented with the scenario that the baby was in cephalic presentation but the EFW was > 4.5 kg, 66% of male and 60% of female obstetricians chose elective CS. Note that male obstetricians had gone from 8% to 66% in favor of elective CS for their partners when informed of the fact that the fetus was macrosomic. The Lancet paper concludes with the question: ‘In this era of patient choice should information regarding the potential benefits of elective CS delivery be given to all women?’ With suspected fetal macrosomia we are dealing not with a normal situation but with one in which the predicted weight of the fetus is 4.5 kg or more. When presented with this scenario, most obstetricians would choose elective CS for themselves or their partners, so surely it is unthinkable that any obstetrician armed with this knowledge would not counsel his (or her) patient on the risks of vaginal birth associated with macrosomia and at least offer a choice of elective CS?

Balance of risks

Most women want a vaginal birth. Turner et al.[73] reported that only 2% of women in their survey of normal primiparous women wanted an elective CS, but women know instinctively about the risks of birth injury and to their pelvic floor if their baby is large. If you want proof, just look at the expression of concern on the mother's face when she is told she is having a big baby. Nevertheless, the instinct of the mother will be towards having a vaginal birth and it will be up to the obstetrician to provide objective information regarding the balance of risks so that she and her partner can make an informed choice about the method of delivery. The risks of elective CS, such as neonatal respiratory distress, tachypnea, repeat CS and placenta accreta, which form part of the discussion over maternal request for CS, pale into insignificance when the risks of vaginal birth with a macrosomic fetus are now on the other side of the equation. The maternal mortality rate with elective CS in a healthy woman in uncomplicated pregnancy is now widely recognized as being no greater than that of vaginal birth[74], although data are not robust and, as the RCOG state in their maternal mortality report, ‘it is virtually impossible to disentangle the fetal and maternal reasons for most of the operations to make a meaningful comparison.’ CS at any stage of labor is protective of most of the serious effects to mother and child that occur as a result of shoulder dystocia, instrumental delivery and vaginal delivery. The question that must be asked is why, if this is known, do these injuries continue to occur? Shoulder dystocia can occur unexpectedly at the end of a normally progressing labor, but this is more likely when it is associated with a normal-weight baby. Could the current pressure on obstetricians to reduce the number of CSs performed be a contributing factor? Or is it that the heat of the labor ward is not conducive to considered thought? Predicting fetal macrosomia does not imply that elective CS is the method of choice, although it should be made clear to the couple that elective CS is the low-risk option. What is important is that each woman and her partner should be informed of the particular risks associated with a macrosomic fetus and shoulder dystocia. The accuracy of ultrasound should be part of the discussion as well as the protective effect of CS in labor on both pelvic floor disorders and fetal trauma. Factors that will bias the advice towards elective CS will be a small maternal stature, and other relative risk factors such as age over 30 and history of infertility. A woman with a height over 175 cm has a significantly lower chance of shoulder dystocia and trial of labor might be chosen. No pressure should be placed on the couple and, once they have decided, there is no need to revisit the issue unless the couple request this. It is important that the couple feel supported in the decision they make. This is a situation in which maternal autonomy is paramount.


Fetal macrosomia should be defined as a birth weight ≥ 4.5 kg. It is a high-risk condition which should be taken just as seriously by obstetricians as intrauterine growth restriction. Neither ACOG nor RCOG/NICE, in their guidance, appear to be proactive in dealing with this problem. The cost analysis by Rouse et al.[9], on which they lean heavily, does not factor in correct data on long-term effects of OBPI, intrapartum asphyxia and maternal pelvic floor disruption, and it is outdated in many of its assumptions. The cost analysis by Culligan et al.[4] is robust, although some of their assumptions can be disputed. I believe that the long-term risk of anal incontinence has been exaggerated, but no account has been made for the staggering healthcare costs associated with birth asphyxia, which is a major component of shoulder dystocia. There is, however, sufficient flexibility in the cost estimates based on QUALYs[75] to allow us to be confident that elective CS at 39 weeks in primigravidae with an ultrasound EFW > 4.5 kg is cost-effective. Awareness of the possibility of macrosomia should begin at the first antenatal visit. Obesity is a major risk factor and a body mass index > 30 kg/m2 should prompt awareness of a potential problem[76]. Macrosomia is also associated with excessive maternal weight gain.

Awareness of the problem is the key to better management and avoidance of risk. Obstetricians and midwives must be as aware of the dangers of a large fetus as of a small one. Ultrasound is imprecise, but it is better than other methods in predicting fetal macrosomia. If performed assiduously, over 90% of ultrasound predictions of birth weight ≥ 4.5 kg will be over 4 kg and this will therefore define a group at high risk for obstructed labor, emergency CS, assisted vaginal delivery and maternal and fetal damage. Taking fetal macrosomia seriously means recommending routine ultrasound assessment of fetal size (the macrosomia scan) in the third trimester. A triage scan between 32 and 34 weeks, with a diagnostic scan at 39 weeks, has much to commend it.

When a diagnosis of fetal macrosomia is made, the mother and her partner must be made aware of the risks of vaginal birth and be free to choose elective CS or trial of labor without coercion. Induction of labor is contraindicated as it increases the risk of shoulder dystocia. Flagging up all cases of predicted fetal macrosomia is vitally important, so that the attendants in the labor suite will recommend CS if there is any delay in cervical dilatation or arrest of head rotation or descent. CS should also be the preferred option if an abnormal fetal heart tracing develops. Protection of the pelvic floor will prevent years of suffering from anal and stress incontinence and prolapse. Avoidance of a difficult birth will prevent fetal trauma and long-term handicap, especially from brachial plexus injury and fetal asphyxia. To paraphrase the Lancet editorial: we are not interested in targets for CS births; we are interested in healthy mothers and babies.