Obstetric outcomes for women with female genital mutilation at an Australian hospital, 2006–2012: a descriptive study

We cross-tabulated the FGM type recorded in the ObstetriX Database with the FGM type recorded in the Midwives Database for all women’s records between July 2006 and December 2012. We calculated the proportions where FGM was recorded in both datasets and where the FGM type agreed. We calculated the proportion of women with FGM among women who gave birth in the years 2006 to 2012. Among women with and without FGM, we summarised categorical variables using proportions. We fitted Poisson regression models, with general estimating equations to account for subsequent births for women over the years, to estimate the rate of obstetric complications among women with FGM compared with women without FGM. Additionally, for nulliparous women, we fitted a Poisson regression model to assess the association between FGM type and caesarean section rates after adjusting for the potential confounders year of birth, age group, body mass index group, presence of hypertension and diabetes, and smoking status. This analysis was performed only among nulliparous women to remove any effect of previous obstetric history. We were unable to fit adjusted models for the other obstetric outcomes due to their small numbers among nulliparous women.

There were 9048 maternity records for the period July 2006 to December 2012 for women with singleton pregnancies. ObstetriX recorded FGM in 139 records, whilst the Midwives Database indicated that 180 records were for women with FGM (Table 2). In this period, ObstetriX documented FGM in 73 % (139/191) of all records where FGM was recorded (ObstetriX or Midwives Databases). This proportion increased from 14 % in 2006 to 90 % in 2012.

Among the records for women with FGM documented in either database, the ObstretriX record agreed with the Midwives database in only 35 % of cases and the Kappa statistic was 0.28 (95 % CI: 0.18, 0.38). Approximately 39 % of FGM type III recordings in the Midwives Database were recorded incorrectly as grade I-II or “unknown” in ObstetriX. Among women with FGM recorded in ObstetriX, the proportion of “unknown” classification was high at 42 % (Table 2).

There were 196 births from women with FGM at the study hospital between 2006 and 2012 (Table 3). The prevalence of women with FGM varied in year from 1.8 to 2.8 %. The highest prevalence occurred in 2009. The percentages of women with FGM types I, II and III among the known types were 33, 33, and 26 %, respectively.

The majority (67 %) of women with FGM were born in East Africa. Almost a quarter (23 %) were from the western and southern regions of Africa, and the remainder from North Africa, the Middle East and Asia. The majority (89 %) of women with FGM type III were from East Africa. Approximately half of the women with FGM from Sudan and Somalia had FGM type III. The prevalence of FGM was the same among women from East Africa and those from western and southern regions of Africa, i.e., 45 %. The prevalence among women from the other countries was less than 1 %.

The age distribution for women with and without FGM was similar, with most women aged between 20 and 34 years (Table 4). A third of the births among women with FGM were to women who had three or more previous deliveries of babies, compared to 17 % among women without FGM. Women with and without FGM were similar in body mass index and the diagnosis of hypertension. Gestational diabetes mellitus was more common among women with FGM type III. Only 1 % of women with FGM were smokers during pregnancy as compared to 5.6 % of women without FGM.

Table 5 displays the rates of maternal and neonatal outcomes for women with different types of FGM and for women without FGM. Almost all women with FGM type III had deinfibulation performed either prior to or during pregnancy. Only 9 % were deinfibulated during labour.

Maternal and neonatal outcomes for women with FGM were similar to those without FGM, except for statistically significant higher rates of caesarean section and first and second degree perineal tears. The latter was more common among women with FGM type III as compared to types I and II (Table 5).

Nulliparous women with FGM types III and I/II, and those without FGM had a caesarean section rate of 46 % (n = 6), 40 % (n = 15), and 20 % (n = 697), respectively (Table 6).

Despite the statistically significant higher rates of caesarean section among women with FGM, none of the caesarean sections had FGM as an indication.

Sixty-three percent of all caesarean sections for women with FGM (multiparae and nulliparae) were emergency and the remainder elective. Amongst emergency operations, 78 % had FGM type I/II. The indications were non-reassuring fetal heart rate trace with or without abnormal fetal scalp lactate (65 %), failure to progress (19 %), pre-eclampsia (9 %), placental abruption (3 %) and placenta praevia (3 %). Sixty-three percent underwent a caesarean section at cervical dilatation of up to 4 cm.

The indications for all elective caesarean sections were previous caesarean section (74 %, n = 14), malpresentation (21 %, n = 4), and meconium-stained liquor with no labour (5 %, n = 1).

Amongst nulliparae, 81 % were emergency caesarean sections. Most of them (71 %) were for non-reassuring fetal heart rate trace. Two-thirds of these were inductions of labour for postdates, 75 % of whom underwent caesarean section at cervical dilatation of 2-4 cm. See (Additional file 1: Table S7) for the original data used to perform the above analysis.

The prevalence of women with FGM who gave birth at the metropolitan hospital in Australia between 2006 and 2012, was 2 to 3 %. The accuracy of ObstetriX database in collecting data on the presence of FGM increased from 14 to 90 % over six years. However, it was only 35 % correct in identifying the type of FGM. Ninety percent of women with FGM were born in the countries of Africa. Two thirds of all women with FGM had types II and III. Ninety percent of women with type III were from East Africa. Half of the women from Somalia and Sudan had FGM type III.

All women with FGM III were deinfibulated for the birth of their baby unless an elective caesarean section was performed. Caesarean section and genital tract trauma were more common in women with FGM. The latter was higher amongst women with FGM type III as compared to types I and II. None of the women’s caesarean sections were directly indicated for FGM.

The main strength of the study is that it is the first available data on the number of women with FGM giving birth at a hospital in Australia. It is also the first information on obstetric and neonatal outcomes in women with FGM in this country. Moreover, our study is one of few in the literature to stratify obstetric complications for FGM type. Our study highlights the importance of the need for hospitals and healthcare professionals with expertise in caring for women with FGM.

The small number of women with FGM in the obstetric outcomes categories limited statistical analysis. Combining datasets from multiple hospitals across Australia and internationally may address this issue.

In 2004, 1.4 % of babies born in all maternity services in England and Wales were to women affected by FGM [19]. Our figure of 2 to 3 % for the study hospital would be lower nationally, as this hospital is known for its FGM expertise. Women born in countries where FGM is prevalent, live near it, and hence are referred to this hospital. Eighty-nine percent of women who were born in these countries, lived in major cities in Australia in 2011 [20].

In keeping with our findings of ObstetriX recording of FGM type, a study in Switzerland reported that FGM identification were missed in 37 % and the type misclassified in 23 % [21]. There is no data on FGM prevalence in Australia. Similarly, in the European Union, data relevant to FGM is neither systematically collected nor centrally stored [19]. Accurate data is important in developing evidence-based policies to appropriately implement prevention programmes, health services, monitor their effectiveness, and assess how integration processes affect the practice [19]. It is important to again stress the fact that a radical change occurs in families to oppose and/or abandon FGM after migration [13], and we can expect a very low number of children being cut in their new countries. Hence, while it is paramount to prevent and protect girls from FGM in a similar way as we do with child abuse, it is also important to focus on the issues surrounding social integration and education to aid the cultural change.

The WHO study found that obstetric complications increased with severity of FGM type [9]. The study centres varied from isolated rural hospitals to tertiary teaching hospitals in capital cities. In our study, most of the women had the more severe types II and III and obstetric outcomes were similar to women without FGM except for caesarean section and first and second degree perineal tears. The findings of our study are in keeping with those of other high-income countries, which show that high quality obstetric care with expertise in FGM can minimise obstetric complications [22‐25].

A study in Switzerland found statistically higher incidences of caesarean section and third degree perineal tears in women with FGM III [24]. The majority of their 122 patients had been deinfibulated. The main reason for the emergency caesarean sections, however, was the inability to perform a vaginal examination in labour in women with FGM III who had not been deinfibulated. We do not know whether the caesarean section rate would still have been higher if all women had been deinfibulated and allowed to continue to labour. There was no difference in the length of labour between women with and without FGM. The latter finding was supported by a study of women who underwent deinfibulation at vaginal delivery at a tertiary referral hospital in Saudi Arabia [26]. There were no differences in obstetric and perinatal outcomes. No caesarean section was performed for the indication of FGM. The strength of that study was that the women were matched to those of the same nationalities. A UK study showed that reversal of FGM III significantly reduced the increased risk of caesarean delivery seen with multiparae who have FGM III [27]. A Swedish study examined 68 nulliparous women who had mainly FGM III and were deinfibulated in labour [23]. It found that women with FGM even had a lower risk of prolonged labour.

On the other hand, in keeping with findings from the WHO study [9], the above obstetric outcomes have been found to be different in low-income countries. A study of 85 women in Burkina Faso again showed that women with FGM had a higher risk of caesarean section [28]. The indications were cephalopelvic disproportion and prolonged labour with fetal distress. The study also found higher risks of episiotomy, neonatal resuscitation and stillbirth in women with FGM. These findings were supported by a prospective study in The Gambia [29]. FGM did not affect the main reason of cephalopelvic disproportion for caesarean section. The authors posited that caesarean sections were performed preventively for severe abnormal scarring and/or synaechiae in women with FGM II [29].

Studies of Somali-born women in high-income countries (HIC) of migration reveal significantly adverse perinatal outcomes as compared to receiving country-born women [17, 19, 20, 30]. A meta-analysis of these women in six countries showed that they were more likely to labour without any analgesia or epidural [17]. Language and communication barriers were considered to be contributing factors to the differences in the use of pain relief [17]. These may then also affect the increased number of caesarean section, which occurred despite the fact that Somali women prefer vaginal birth and have an aversion to caesarean section [19, 26, 27]. This fear of caesarean section understandably arises from having lived in a country with a high maternal mortality rate. One of the reasons could be late presentation of women to hospital when a caesarean section is performed but the complications are too advanced to safe a woman’s life. Moreover, women may die from uterine rupture with subsequent pregnancies following a caesarean section, as there may be no access to appropriate obstetric care. These women’s beliefs regarding safe birth need to be explored and addressed early in their antenatal care. A qualitative study of pregnancy and childbirth experiences in Somalian women in Sweden proposes a relationship between adverse perinatal outcome and cultural factors [31]. A comparison of these studies from HIC with otherwise high-quality obstetric care with those of the HIC above suggests that culturally sensitive care and FGM expertise by the obstetrician and midwife are likely to play important roles in providing optimal perinatal care. The higher caesarean section rate in our study is unclear. A prospective multicentre study in countries of migration comparing women with and without FGM of same ethnicities would be required to provide that information. Moreover, the study hospitals would need to provide similar standards of obstetric care and FGM expertise.

Our study hospital underlines the importance of integrating specialised FGM units into hospitals that care for women with FGM. They would provide holistic healthcare, including gynaecological and obstetric care with deinfibulation procedures, integrated with culturally appropriate counseling, psychological, sexual and social support services. There need to be referral pathways for healthcare professionals, hospitals, schools, social work, child protection, immigration and judicial sectors, the police and nongovernment organisations working with affected communities. Training in FGM and its legal and health implications for women and their families can be integrated in these sectors.

In Australia and other countries where women with FGM live, there is a concern about the cultural competence, experience and training of healthcare professionals in the management of FGM [32‐36]. The comparison of the ObstetriX and specialised Midwives Databases in our study show a significant improvement in the accuracy of FGM recording from 14 to 90 % over six years, indicating increasing knowledge of midwives through teaching of the Clinical Midwifery Consultant. Our study highlights the importance of education and training of healthcare providers in FGM diagnosis and care. These may be formally incorporated into undergraduate and postgraduate nursing and medical programmes.