Discussion
Many women had health problems that are relatively easily treatable and may have been present from childhood or adolescence. In view of this, the median age of 26 years was high. In combination with the large proportion of pregnant women who had been referred in connection with pregnancy check-ups, this may indicate an unmet need for treatment among young and non-pregnant women. This is consistent with qualitative studies that have reported a lack of awareness of the outpatient services among women in the user groups, and low social acceptance of deinfibulation outside the context of pregnancy and childbirth (14–16) .
The proportion that contacted the outpatient clinic was higher among non-pregnant than pregnant women. This may reflect a lower awareness of health problems than of issues associated with birth among the referring agencies, and that women seek them out less frequently for urogenital problems. Norwegian guidelines recommend that both pregnant and non-pregnant women who have undergone FGC be examined and/or referred in order to determine their FGC status and need for surgical treatment (17) .
The anatomical extent of FGC type III varied considerably. Such variations are described in the World Health Organization's (WHO) revised classification, which no longer assumes that the amount of genital tissue removed increases from type I to types II and III (3, 18) . While 17 % of the infibulated women had a vaginal opening measuring less than 1 cm, the clitoris was described as unaffected in 31 %. One study reports that nearly one-half of the women had an intact clitoris below the infibulation seal (19) . In comparison, the clitoris was unaffected in 14 % of the women with type I or type II in our material.
Consistent with previous studies, women with FGC type III had more gynaecological health problems than those with type I or type II (7) . Up to one-half of the women with type I or type II FGC nevertheless reported pain, dyspareunia, problems with urination and other gynaecological problems. The outpatient clinics have no specific treatment options for these women. Psychosexual counselling services for women who have undergone FGC, as recommended by the WHO guidelines, are not provided in Norway (9) .
As expected, deinfibulation was the predominant form of treatment. Opening the infibulation is necessary to re-establish vaginal functioning, but there is little evidence available on the effect of deinfibulation on health outcomes (12) . The risk of caesarean section and perineal ruptures has been found to be lower in women who have been deinfibulated when compared to those who have not. It is unclear if it makes any difference whether the deinfibulation procedure is performed before pregnancy, during pregnancy or during childbirth (9, 12) . One recent study found that deinfibulation prior to or during pregnancy did not protect against emergency C-section in primiparous Somali women in Norway (20) . One explanation could be that women who were opened prepartum had a more extensive infibulation that implied a more persistent risk of C-section. Alternatively, in line with the hypothesis that scar tissue from the FGC negatively affects both the progression of the birth and the perineum (10) , new scar tissue from the deinfibulation could be deemed to increase this effect. In this study, primiparous Somali women had a greater risk of C-section than primiparous women in general, irrespective of FGC status. Many women prefer to be deinfibulated during childbirth rather than during pregnancy (14, 21) . Norwegian guidelines recommend that the woman's preferences be given weight. In our material, however, nearly three out of four pregnant women were deinfibulated during pregnancy. In the majority, the infibulation was opened up as far as the clitoris, in line with the recommendation when the woman is not in labour (17) . We believe that the different practices between the hospitals in terms of whether the deinfibulation was performed under local anaesthetic, regional analgesia or general anaesthesia are due to custom. The fact that gynaecological examination and treatment can be especially painful and stressful for women who have undergone FGC has often been referred to, since this may cause them to have flashbacks to the procedure. In our material, one 1 % had their examination halted because it was too painful or mentally stressful.
FGC-related cysts are mainly epidermal inclusion cysts (12, 22) . We found that nearly one-half of the women with such cysts did not report any symptoms. Most articles on cystectomy are case reports, and no complications have been reported (12) . However, a study on clitoris reconstruction included a group of women who had clitoral cysts and underwent a parallel cyst excision. Post-operatively these women reported reduced sexual functioning scores, while the women without cysts reported an improvement (22) . Caution could hence be advised with regard to removing small, asymptomatic periclitoral cysts.
As far as we are aware, no clitoral reconstructions have been performed in Norway, but are offered in some other European countries, including Sweden (23) . They have been performed especially in women with FGC type II, and the motivation has been to restore identity, improve sexual functioning and reduce pain (14) . Most patients are satisfied, but up to 20 % have reported a deterioration in sexual functioning (12, 14) . The intervention is controversial, and is advised against in the UK and by the World Health Organization until better documentation becomes available (9, 24) .
A number of qualitative studies have reported weaknesses in the Norwegian treatment services for women living with FCC. For example, women report insufficient knowledge among healthcare personnel, and there is no psychosexual counselling service (25, 26) . The regional outpatient clinics appear to fulfil a need for treatment. A psychosexual counselling service can be linked with the outpatient clinics, and they will be an easily accessible resource for healthcare personnel. If clitoris reconstruction is to be provided, we recommend that it be undertaken as an interdisciplinary research project in collaboration with one or more of the outpatient clinics.
The strength of our study lies in its descriptive data from all the specialised outpatient services. The weaknesses inherent in the use of retrospective patient records data include data quality and missing information. The sample means that findings cannot be generalised to all those who have undergone FGC, and systematic data on complications are missing.
Conclusion
Conclusion
Many women had symptoms that may have been present since childhood and adolescence and that are easily treatable by deinfibulation. There should be a low threshold for referral. Doctors should pay attention to pain and other urogenital problems in girls and women who have undergone FGC, irrespective of type.