Many case series have insufficient sample sizes to draw conclusions about rare fistula causes. This relatively large dataset offers a unique opportunity to assess the etiology of fistula beyond the dominant causes of childbirth and obstetric and gynecological surgery. Fistulas attributable to rare causes illuminate a variety of risks confronting women.
In this series the most common rare causes were traumatic, whether attributable to accidents, sexual violence, or traditional healers. Among these, traumatic fistulas from accidents were most common, especially due to falls and road traffic accidents.
Some rare causes involve a person directly doing something harmful to women’s bodies, including unsafe abortion, sexual violence, and treatments from traditional healers. In our experience, fistulas following these circumstances tend to be larger and more difficult to repair than the other rare fistulas. In particular, the average size of fistulas following unsafe abortion is similar to fistulas that develop from pressure necrosis during prolonged, obstructed labor. Following in average size are fistulas attributable to traditional healers, which often involve stenosis and scarring, and fistulas following sexual violence. Fistulas from other causes are typically smaller and thus more similar in size to iatrogenic fistulas from surgical complications. Though generally not large, fistulas from radiation can be difficult to repair because of damage to the surrounding tissue.
Traumatic fistula resulting from sexual violence has been of particular interest in the Democratic Republic of the Congo, where it accounts for 2–3% of fistulas [8,9,10]. Some hypothesize that female genital cutting predisposes women to fistula given population-level statistical associations [11]. Although it is included as a potential fistula cause in the FIGO fistula training manual, we did not identify any woman with fistula associated with female genital cutting. Ritual female genital cutting such as clitoridectomy or excision does not generally create fistulas [12, 13]. Traditional treatments that cut the anterior vaginal wall can cause fistulas, however. Known as the gichiri or gurya cut, it has been documented in Nigeria and Niger, where it accounts for up to 18% of treated genitourinary fistulas [14,15,16].
Other fistulas caused by traditional treatments have not received similar recognition, such as the burns from hot iron rods seen in Somalia. Such burns are a specific form of “traditional, damaging, and dangerous [Somali] practices” of fire-burning body parts with coal, iron rods, or burning sticks [17]. Traditional healers may create fistulas when they burn the vagina with hot iron rods to address urinary tract infections or vaginal itchiness [18]. Our review can add other reasons for intentional vaginal burning to this list, including hemorrhoids, swelling, infertility, and various pelvic or genital pains. In Somalia and Kenya, health education and outreach should stress that intentional vaginal burning is not an effective therapy and in fact causes substantial harm.
Traditional therapeutic interventions were not listed among possible fistula causes in the latest fistula surgery training manual [13]. We recommend inclusion in future editions. Otherwise, fistulas caused by traditional healers may be combined with accidental cases that occur in health facilities. Trauma inflicted by traditional healers is clearly distinct from surgical complications, and separate counting will enable appropriately targeted responses.
The FIGO fistula surgery training curriculum groups all iatrogenic fistulas resulting from pelvic surgery. We see value in distinguishing fistulas following gynecological surgery from fistulas attributable to unsafe abortion. Reports have identified cervico-vaginal fistulas as a complication of induced abortion, with implications for women’s subsequent fertility and risk of obstetric complications [19, 20]. This series includes two fecal fistulas through the uterus after dilatation and curettage for incomplete abortions, as well as one case of combined vesico-vaginal and recto-vaginal fistula following unsafe abortion.
HIV infection can lead to fistula formation in rare cases. Acquired rectovaginal fistulas shortly after birth have been documented as an early manifestation of HIV infection in girls, and most acquired recto-vaginal fistulas in children can be attributed to HIV infection [21, 22]. In adults, seropositive women can develop anorectal sepsis, as seen with the four women in this series who had acquired small recto-vaginal fistulas without relation to pregnancy or birth [23].
Radiation therapy for cervical, endometrial, or multifocal cancer can induce fistula and decrease the elasticity of surrounding tissues [2, 24, 25]. Radiotherapy caused 4.6% (4/87) of fistulas reported by one tertiary fistula center in South Africa; the much lower proportion in our population likely reflects comparatively low access to cancer treatment [26].
This large retrospective review has limitations. Sample sizes reflect variation in where the second and third authors and colleagues repaired fistulas. Countries typically included data from multiple hospitals with wide catchment areas. Countries with large sample sizes, however, may be more representative of all women seeking fistula repair than those with smaller samples. This series includes women who sought surgical treatment for their fistulas. It cannot include women who developed fistula but did not reach treatment centers. Overall etiology may be different among all women with fistula if women with particular fistula causes are more likely than others to reach treatment facilities. We cannot rule out the risk of response bias and recall bias given our dependence on women’s accounts of past events.