Diagnoses and procedures of inpatients with female genital mutilation/cutting in Swiss University Hospitals: a cross-sectional study

Background Female genital mutilation/cutting (FGM/C) can result in short and long-term complications, which can impact physical, psychological and sexual health. Our objective was to obtain descriptive data about the most frequent health conditions and procedures associated with FGM/C in Swiss university hospitals inpatient women and girls with a condition/diagnosis of FGM/C. Our research focused on the gynaecology and obstetrics departments. Methods We conducted an exploratory descriptive study to identify the health outcomes of women and girls with a coded FGM/C diagnose who had been admitted to Swiss university hospitals between 2016 and 2018. Four of the five Swiss university hospitals provided anonymized data on primary and secondary diagnoses coded with the International Classification of Diseases (ICD) and interventions coded in their medical files. Results Between 2016 and 2018, 207 inpatients had a condition/diagnosis of FGM/C. The majority (96%) were admitted either to gynaecology or obstetrics divisions with few genito-urinary and psychosexual conditions coded. Conclusions FGM/C coding capacities in Swiss university hospitals are low, and some complications of FGM/C are probably not diagnosed. Pregnancy and delivery represent key moments to identify and offer medical care to women and girls who live with FGM/C. Trial registration: This cross-sectional study (protocol number 2018-01851) was conducted in 2019, and approved by the Swiss ethics committee.


Introduction
Female Genital Mutilation/Cutting (FGM/C) comprises all procedures involving partial or total removal of the external female genitalia without medical indication [1]. The World Health Organization (WHO) defines four main types of FGM/C (Table 1) [2]. 200 million women and girls have undergone the practice in 31 countries according to nationally representative household surveys, without counting female migrants with FGM/C who live high-income countries [3,4]. According to estimates, almost 600,000 individuals living in the European Union are believed to have been exposed to ritual genital cutting (2016) [5], and in Switzerland, approximately 21,706 women and girls are estimated to have been exposed to this practice (2018) [6]. These estimates were obtained by indirect measures: multiplying the number of female migrants from an FGM/C practicing country with the FGM/C prevalence rate from the same country. This method does not account for regional and ethnic variations of the practice within countries, and does not include corrections for any changes in attitudes towards FGM/C, which have been described among migrants [7][8][9][10][11], nor include other female genital modifications such as female genital cosmetic surgeries. The actual prevalence of FGM/C among communities of migrants remains unknown [12,13]. Recent studies conducted in the United Kingdom (UK) showed significantly fewer cases of FGM/C than expected among minors according to prevalence estimates [14,15]. Nevertheless, the total number of women and girls who have undergone FGM/C is expected to grow in high-income countries because of increasing migration from countries where FGM/C prevalence remains high [16]. Although several interventions effectively promote the abandonment of FGM/C, many countries are simultaneously facing population growth, with consequent increase in the absolute number of girls exposed to FGM/C [17].
It has been widely studied that FGM/C, particularly type III, can result in short and long-term complications, which can impact physical, psychological and sexual health [1]. Systematic reviews and meta-analyses show that female individuals with FGM/C are at higher risk of dyspareunia, genito-urinary complications, prolonged labour, episiotomies, and birth complications [18][19][20][21]. Frequently cited as a limitation, the lack of high-quality studies makes it difficult to reach consensus surrounding the association between FGM/C and caesarean section, infertility and HIV [18][19][20]. Depending on the study design, some of the available data about FGM/C complications and their clinical management may be subject to self-report and recall bias [22]. Inappropriate health management due to the lacking training surrounding FGM/C may further bias the existing data. To our knowledge, no study has yet described FGM/C complications and childbirth represent key moments to care for and counsel a population that might not consult or be identified otherwise.
Keywords: Female genital mutilation, Female genital cutting, Female genital mutilation/cutting, International classification of diseases, ICD, Coding, Switzerland Table 1 Classification of FGM/C types and subtypes according to WHO [2] Type I Partial or total removal of the clitoral glans (the external and visible part of the clitoris, which is a sensitive part of the female genitals, with the function of providing sexual pleasure to the woman), and/or the prepuce/clitoral hood (the fold of skin surrounding the clitoral glans) and associated procedures using hospital inpatient data coded with the International Classification of Diseases (ICD). We sought to describe the most frequent health conditions and procedures associated with FGM/C in inpatient women and girls identified from ICD diagnoses of FGM/C from five Swiss university hospitals.

Materials and methods
This cross-sectional study (protocol number 2018-01851) was conducted in 2019, and approved by the Swiss ethics committee. We invited all five Swiss university hospitals (Geneva, Lausanne, Bern, Basel and Zürich) to provide anonymized data for all inpatient adult women and girls (< 18 years) with a nationality from any of the 30 FGM/C practicing countries [3] in addition to all inpatients who had a coded condition/diagnosis of FGM/C between January 1, 2016 and December 31, 2018. We did not include inpatients from the Maldives, where FGM/C has been recently reported [23], because no nationally representative survey was available when the study began. Please note that we talk about a "condition/diagnosis" of FGM/C as the ICD contains specific codes for FGM/C, which are also used to justify reimbursement of healthcare provided in case of need by health insurances. We also use the term condition, to acknowledge the fact that not all women and girls with FGM/C are sick.
In Swiss university hospitals, healthcare professionals record the diagnosis responsible for the hospitalization (primary diagnosis); eventual complications that arise during the patient's hospital stay, as well as any additional diseases treated (secondary diagnoses) in the patients' electronic medical charts. Professional coders in Switzerland code this information with the German Modification of the tenth edition of the ICD (ICD-10-GM), and interventions are coded with the Swiss Classification of Surgical Interventions (CHOP) [24].
We received the requested data from four university hospitals: Geneva (HUG), Lausanne (CHUV), Bern (Inselspital), and Zürich (USZ). The university hospital of Basel (USB) did not participate due to logistical difficulties in data provision. All data were then merged in a single database using STATA version 15.
The data for all inpatient women and girls from the 30 targeted FGM/C countries and all primary and secondary diagnoses of FGM/C coded between January 1, 2016 and December 31, 2018 was anonymized. The university hospital of Bern did not provide data on the interventions performed. Lausanne and Zürich provided CHOP codes of the interventions performed, and Geneva provided the name of the CHOP interventions. We analyzed all diagnoses and interventions in patients' records with a coded primary or secondary diagnosis of FGM/C. We provided descriptive statistics with mean, ± standard deviation, and median for continuous variables, numbers by categorical variables. We compared all diagnoses from our sample with the FGM/C ICD "tip-sheet" for FGM/C associated health conditions (full methods available in another manuscript) [25]. We focused our analysis on the gynaecology and obstetrics divisions, where most of the inpatients with an FGM/C code were admitted.
The Swiss Federal Office of Public Health, the Swiss Network against Female Circumcision, and Caritas Switzerland funded the study. They had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Results
In four of the five Swiss university hospitals, 207 inpatients received a primary (n = 22, 10.6%) or a secondary (n = 185) diagnosis of FGM/C during the study period ( Table 2). Of these 207 women and girls, 199 (96%, 89.4%) were admitted either to gynaecology or obstetrics divisions. The remaining women and girls were admitted to other departments (surgery, internal medicine, emergency, and paediatrics).
The mean number of secondary diagnoses coded among women with a primary or secondary diagnosis of FGM/C was 2.59 (median 2, range 0-15), spanning 16 chapters of the ICD-10 (Table 4). There were 281 secondary diagnoses related to pregnancy and childbirth, including 114 codes describing duration of pregnancy (O09.1-O09.7, O48). Other frequent codes were perineal laceration during delivery (n = 21), prolonged second stage of labour (n = 8), and anaemia complicating pregnancy, childbirth and the puerperium (n = 24).
Among diseases of the genitourinary system, coded diagnoses featured vulvar cyst (n = 1), urinary tract infection (n = 1) and mild cervical dysplasia (n = 1). Other secondary diagnoses related to infections were Streptococcus group B (n = 17), possibly describing a carrier-state in pregnant women, and carrier of other specified bacterial or infectious diseases (n = 17), and asymptomatic HIV status (n = 1). Eight women required immunization against viral diseases such as measles, diphtheria, and other viral diseases.
Mental disorders and sexual health conditions were rarely coded as either primary or secondary conditions. "Problems related to psychosocial and/or economic circumstances" appeared five times as secondary diagnosis, and once as a primary diagnosis for a minor inpatient that was admitted in paediatrics. Out of the other four minors with a code of FGM/C (n = 5), another was admitted in paediatrics to undergo surgery for mitral valve stenosis, and the remaining two were admitted in gynaecology for surgical treatment of a vulvar cyst. The only minor inpatient with a primary diagnosis of FGM/C underwent defibulation and had secondary codes related to pregnancy.
In total, there were 62 primary and secondary diagnoses of anaemia in 36 patients admitted in gynaecology or obstetrics. Among them, six had third-stage haemorrhage, six a first-or second-degree perineal tear, and nine underwent caesarean section 27 of 135 patients admitted Table 2 Description of inpatients with a FGM/C (n = 207) as primary or secondary diagnosis between 2016 and 2018 followed in one of four Swiss university hospitals (Geneva, Lausanne, Bern and Zürich) a Data obtained from Bern did not specify whether patients were admitted in gynecology or obstetrics Country of origin, n (%) Benin  in obstetrics (19%), had a primary or secondary diagnosis of anemia complicating pregnancy and childbirth. Several coded diagnoses in our sample might be possible long-term complications of FGM/C found in the FGM/C "tip-sheet" [25] ( Table 5). The most frequently coded diagnoses (primary and secondary combined) were: perineal laceration during delivery (n = 50, 37.5% of FGM/C type III), prolonged second stage of labour (n = 21, 28.6% of FGM/C type III), postpartum     Table 6). The most frequent obstetrical intervention was caesarean section (n = 29, 48.3% of FGM/C type III). 14 patients had an episiotomy (35.7% of FGM/C type III) and 15 required unspecified manual assistance during delivery (20% of FGM/C type III). The most frequent intervention aimed at treating complications of FGM/C was surgery of the clitoris (n = 11, 36.4% of FGM/C type III). In Geneva, four inpatients underwent defibulation.

Main findings
In four Swiss university hospitals, 207 inpatients had a primary (n = 22, 10.6%) or secondary (n = 185, 89.4%) diagnosis of FGM/C coded at admission between 2016 and 2018 [26]. As discussed in our related paper on Swiss university hospitals' capacities of coding FGM/C, this was much less than expected when compared with the number of inpatients who could have undergone FGM/C based on their nationality and indirect estimates (n = 4947) [26]. Either fewer women than expected have undergone FGM/C, or healthcare professionals did not identify and/or record it, or professional coders failed to code FGM/C, resulting in suboptimal coding. Nearly all patients with a coded condition/diagnosis of FGM/C were admitted to an obstetrics and/or gynaecology division, and most of their primary and secondary diagnoses were related to pregnancy and delivery.

Limitations and strengths
Limitations included the absence of participation from Basel; of interventions' data from Bern; the exclusion of outpatients, which would inform on the health conditions treated and interventions performed (e.g. defibulation) in ambulatory care; and of non-university Table 5 Specific codes for long-term complications to FGM/C when FGM/C was coded as primary or secondary diagnosis Future studies could assess the prevalence of FGM/C and associated health outcomes in all hospitals, and study regional variations, such as in areas near asylum centres. Application of our method is mostly limited by undercoding of FGM/C, which most likely results from insufficient training about FGM/C [26]. Besides gynaecology and obstetrics, health professionals working in paediatrics, travel medicine, infectious diseases, primary care, and migrant health programmes, could benefit from such training. This study's main strength was the use of ICD-10 codes to identify health complications of FGM/C, an affordable and objective method, easily reproducible over time, and at national and international level, with good comparability of data. Impact of training, specific care, as well as financial costs resulting from health complications of FGM/C might also be assessed using ICD codes. They could be used in both diaspora and Table 6 Main intervention reported among patients with FGM/C according to hospital  FGM/C high prevalence countries, as an alternative to the FGM/C cost calculator developed by WHO only for high prevalence countries [35].

Interpretation
Women with FGM/C might consult, be admitted or referred more frequently when pregnant, resulting in better FGM/C coding in obstetrics divisions. Furthermore, Swiss basic health insurance covers most pregnancyrelated costs, facilitating access to healthcare [36]. Obstetricians and gynaecologists routinely perform genital examinations and are more likely trained to diagnose FGM/C [26]. FGM/C is also more likely to be recorded in obstetrics charts, because it can influence childbirth [1]. For instance, UK's report on FGM/C prevalence in the National Health System (NHS) showed that 1630 women and girls had a consultation where FGM/C was recorded between October and December 2020, with 74.9% of attendances in midwifery or obstetrical units [37]. Antenatal consultations provide major opportunities to identify and care for individuals with FGM/C who might not seek or receive medical attention otherwise [1,38]. Meta-analyses including studies from FGM/C practicing countries, and diaspora countries showed that FGM/C was significantly associated with prolonged labour, perineal tears, episiotomy, and non-significantly associated with caesarean section [19,20]. Obstetric outcomes coded in our study were mainly prolonged second stage of labour (n = 21) and perineal lacerations (n = 50) especially of first-or second-degree (90%). 29 inpatients required a caesarean section, 14 episiotomy, and 15 assistance during delivery. We were not able to calculate the prevalence of complications from FGM/C for several reasons. Our data was fully anonymized, and thus some records could potentially be returning patients, so we cannot know the exact denominator of pregnant women in our sample. Second, the study was cross-sectional, and some pregnant women might have delivered after the end of the study, leaving their birth outcomes unknown.
Among 85,990 deliveries in 2017 in Swiss medical institutions, 54.7% of women had a perineal tear mainly of first-or second-degree (94.7%); 32.3% a caesarean section; 11.1% an assisted delivery, and 17% an episiotomy [39]. Considering that at least 135 women were pregnant (135 inpatients admitted in obstetrics, and 30 in gynaecology and/or obstetrics), and subject to the limitations stated above, our data do not suggest high rates of obstetric complications.
Studies about obstetric complications of FGM/C sometimes show diverging results. A prospective study conducted in six African countries found a significant association between obstetric complications and FGM/C, especially type III [40], whereas retrospective studies from high-income countries such as Sweden, the UK, and Switzerland showed similar obstetric outcomes among women with and without FGM/C [41][42][43]. FGM/C has been significantly associated with higher rates of caesarean sections in studies conducted in both practicing and diaspora countries [40, 44,45], and meta-analyses show a non-significant trend towards higher rates [19,20]. Future studies could assess if training of health professionals and access to interpreters could improve obstetric outcomes of individuals with FGM/C. Indeed, health professionals unfamiliar with FGM/C might perform caesarean sections for inappropriate reasons, especially in cases of infibulation [46]. Moreover, migrant women in high-income countries often have higher rates of caesarean sections than non-migrants [47]. Communication barriers, economic difficulties, and exposure to violence can result in poor maternal health and/or care quality for some migrants regardless of FGM/C [48][49][50][51][52].
Only five minor inpatients had an FGM/C code. Outpatient clinics may attend more children with FGM/C than hospitals, but paediatricians may also not know when and how to discuss FGM/C with parents and their children, not recognize it if they perform a genital examination, or simply not record it [53][54][55]. Alternatively, they could be second-generation migrants and beyond, and therefore less exposed to the practice. A UK study showed that among 55 children with FGM/C referred to specialized clinics, 21% suffered from mental health symptoms such as anxiety, sleep and behaviour disorders, and 13% from physical symptoms such as problems with micturition, menstruation and genital pain [14]. Except one post-traumatic stress disorder, psychological symptoms were not coded in our minor population, and rarely among adults. Swiss university hospitals' health professionals may lack time or training on how to detect and treat such symptoms and other FGM/C complications. Or, they may identify and manage psychological complications, without however identifying or documenting the FGM/C as an associated condition [54][55][56][57][58][59][60].
Coding of surgical interventions was incomplete. Perineal tears were more coded (n = 50) than perineal tears repairs (n = 8). Other repairs were either not coded, or coded as secondary interventions, which were not provided. Because no CHOP codes exist for defibulation and clitoral reconstruction, we had to hypothesize that codes such as repair (n = 5), or incision (n = 4) of vulva and perineum had been used to indicate these surgeries. Geneva provided the interventions' names instead of codes, and reported 8 clitoral surgeries and 4 defibulations among inpatients, and additionally reported 12 clitoral surgeries, 25 defibulations and 8 other surgeries for scar complications of FGM/C in outpatient care. Some Swiss insurance companies have tried to refuse to reimburse these surgeries. Specific CHOP codes would facilitate medical coding and reimbursement.
Sensitisation and training of healthcare professionals and professional coders on FGM/C could improve identification, documentation and coding of FGM/C and its complications in Swiss university hospitals; inform and improve the quality of future policies, services and interventions. Future prospective and case-control studies could assess coding of FGM/C and associated health outcomes according to training and specialised care resources.

Conclusion
Most of the 207 women and girls admitted to Swiss university hospitals between 2016 and 2018 with a primary or secondary diagnosis of FGM/C were admitted to obstetrics divisions. Pregnancy and delivery seem to be key moments to care for and counsel a population that might not consult or be identified otherwise. FGM/C coding capacities in Swiss university hospitals are low, and some complications of FGM/C are probably not diagnosed, or diagnosed alone, without FGM/C.