During the study period, 346 patients affected with fistula were operated in 14 caravans, an average of 24.8 patients (range: 14 and 34) per caravan. Their average age was 33.11 (range: 12 - 70 years). The majority (73.4%) lived in rural areas. Illiterates and housewives accounted for 63.3% (n = 219) and 80% (n = 277) of the cohort 49.1% of the patients were divorced or abandoned by their families. One hundred and forty-five patients (41.9%) were residing in the host city of the caravan. The other 201 (58.1%) had to travel an average of 192 km (extremes: 39 and 696 km) from their place of residence to access the caravan site.

The average duration of fistula progression was 6.08 years (range: 0 to 42 years). Fistula has been evolving for more than 10 years in 265 patients (76.6%). 245 patients (70.8%) had no prior treatment. The etiology of the fistula was obstetricin 304 patients (Table 1). The average duration of labor was 2.3 days (range 0 - 5 days) and had a low vaginal delivery in 31.9%, caesarean section in 62.2% and hysterectomy in 5.9%.

- Origin of the fistula

- Classification of the fistulae

Table 2 gives the type of UGF according to the Kees Waaldijk classification. The average fistula size was 1.2 cm (range: 0.1 - 6 cm).

At the therapeutic level, spinal anesthesia was performed in 329 patients (95.10%). The surgical route was vaginal in 82.1% of cases (284 patients), abdominal in 11.6% of cases (40 patients) and mixed in 6.3% of cases (22 patients). The vesico-vaginal splitting with suture separated from the bladder and vagina was the basic technique (94.7%; n = 328); 25 patients had a colposuspension. In 18 patients (5.3%), an exclusive ureteral reimplantation was performed to treat uretero-vaginal fistulae.

The average urinary catheter ablation time was 27.27 days (range: 14 - 30 days).

Of the total cohort, 80 (23.1%) patients were lost of sight to follow-up at 1 month and 245 (70%) at 3 months. The cure rates evaluated in patients reviewed at 1 month and 3 months were respectively 70% and 64%. Figure 2 shows the operative results according to the type of fistula at 1 month.

The cohort of this study was composed of 346 fistula patients operated in one

*FGM: female genital mutilation.

year in seven towns (Figure 1). These figures, which appear to be high, are not very representative of the real needs to be met in terms of repairing UGF in the subregion countries in general. Indeed in these countries which already had the experience of the fistula caravan, many fistula patients are not treated at every caravan either for lack of information or difficulties to access to the caravan site [7] [8] [9] . The surgical caravans have had the merit of widening the provision of care. Therefore they must be perpetuated and above all extended to all the regions of Côte d’Ivoire. However, they must quickly give way to the establish- ment of permanent services able to manage a greater number of fistula patients [10] . However, the biggest challenge remains the active identification of the many fistula patients living in the community who are hiding and who are unaware of the existence of treatment [1] [8] [11] . They should be informed of thehealing possibilities and directed towards the care sites. A large-scale awareness of policy makers, health professionals, non-governmental organizations, religious and community leaders would make it possible to find one large number.

Regarding the epidemiological profile of the fistula patients found in the series, it was not different from that reported by most African authors [2] [7] [8] [12] [13] [14] . They were young, poor, uneducated, residing in rural areas and abandoned by their spouses or their entourage.

The majority of the patients of this study had a fistula which evolved over 10 years (76.6%), and 70.8% of cases didn’t received any care. This observation shows the importance of uncovered needs for fistula repair in Côte d’Ivoire.

As for the occurrence mechanism, obstetric factors were the main causes (87.9%) of the cases we had cared for (Table 1). Many authors have also reported that obstetric factors were the main causes of fistula (around 90%) in developing countries because of the difficulties to access to emergency obstetric care [2] [7] [8] [11] [12] [13] [14] [15] .

According to the seriousness, Table 2 shows that complex fistulae were common in the study series (21.1%) and many african authors also described these complex forms in their articles [2] [8] [12] .

Moreover 67.4% of simple fistulae were found in this study (Table 2) which could be linked to selection bias. Indeed, the punctual nature of the caravans and the lack of equipment at the sites proved to be unsuitable for the management of complex forms requiring more means, more important surgery and prolonged follow-up.

The majority of fistulas (82.1%) were surgically accessible through vaginal route and spinal anesthesia (95.10%) was the most common type of anesthesia which permitted to minimize the cost of care. The most performed surgical technique was the vesicovaginal splitting with separated suture of the bladder and the vagina, without the addition of neighborhood structure (Figure 2). This techni- que was carried out on fistulous orifices of size varying between 0.1 and 6 cm. It has also been performed by Wall and Dekou when the orifices had a diameter of between 0.2 and 8 cm [16] [17] . Therapeutic results were judged after one month and three months follow-up. They are characterized by a significant number of lost of sight. These facts point to another challenge in evaluating therapeutic outcomes: the difficulty of monitoring the patients undergoing surgery. At 1 month the results were characterized by a high rate of failure in type II fistulas and a 70% success rate in types I and III, and were consistent with data from the literature [15] [18] [19] . The care of type II fistulae is marked by a high rate of failure in relation to post-cure urinary incontinence even when sealing is achieved [10] [12] [18] [19] . Different techniques are proposed to treat post fistula urinary incontinence [14] [18] [19] . Among these techniques, we performed colposuspension (25 cases), described by Waaldijk [6] . Although these techniques seem to produce interesting results, it is difficult to draw conclusions from this because of the difficult follow-up of patients over a long term [18] [19] . Moreover, the high cost of synthetic strips makes it impossible to afford them in developing countries [18] .