A retrospective review to identify criteria for incorporating the Singapore flap and gracilis muscle flap into obstetric fistula repair

Abstract Objective To identify criteria to guide surgeons regarding indications for use of the Singapore and gracilis muscle flaps in obstetric fistula repair. Methods This is a retrospective case series. Obstetric fistula surgeons in Lilongwe, Malawi, have been incorporating plastic surgery techniques with the Singapore and gracilis muscle flaps since collaborating with plastic surgeons in 2016. We describe the surgical outcomes of procedures utilizing each flap individually and those using both. Results Between February 2016 and June 2019, 69 patients received a flap at the time of obstetric fistula repair at the Fistula Care Center in Lilongwe, Malawi. A total of 32 (46.4%) received a Singapore flap, 20 (29.0%) received a gracilis flap, and 17 (24.6%) received both types of flap. Conclusion Based on our outcomes, we note the possible advantage of incorporating the gracilis flap even when it is thought that the Singapore flap is sufficient. However, more data are needed.


| INTRODUCTION
Since 2016, surgeons at the Fistula Care Centre (FCC) in Lilongwe, Malawi, have been incorporating Singapore fasciocutaneous flaps and gracilis muscle flaps for complex obstetric fistula repairs. This development arose from a collaboration with plastic surgeons out of the need for additional techniques for complex repairs. Since then, we have performed approaching 70 fistula repairs incorporating one or both flaps. We have been encouraged by the outcomes as these are the most complex fistulae to repair-quoted to have as low as a 52% success rate. 1 We have since held two workshops at the FCC along with a plastic surgeon (RB), during which we taught other fistula surgeons the flap techniques (Figs 1-4). For surgical techniques on flap creation, please see our papers on the Singapore flap for vaginal reconstruction and the gracilis muscle flap for complex repairs. 2,3 From the second larger workshop we found that indications for use of the flaps are subjective. Therefore, the aim of the present study was to review our outcomes to date and attempt to determine criteria to guide others in the use of these flaps.

| MATERIALS AND METHODS
This is a retrospective review of all obstetric fistula repairs that used a Singapore flap, gracilis muscle flap, or both. Cases using a flap for vaginal reconstruction without the presence of a fistula were not included in this review. This is part of a larger study on the outcomes of obstetric fistula repair at our center, which has been approved by the Malawian For patients with successful closure of the fistula, a 1-hour pad weight test is done to quantify the amount of urethral leakage; this is because many women with a complex fistula experience some degree of urethral leakage postoperatively despite a healed fistula.  Table 1 outlines the variables of interest in forming criteria to guide employment of each flap.

| Singapore flap
The mean age of the women who received a Singapore flap was 27.2 ± 9.0 years; these women had lived with a fistula for an aver- iii indicating significant scarring, previous repair, or were circumferential in nature. Ten (31%) women had undergone a previous repair. Average urethral length was 1.88 ± 0.76 cm, bladder length 7.20 ± 1.59 cm, and vaginal length 6.17 ± 1.90 cm. Scarring was considered moderate to severe in most cases. Dye tests were negative in 22 (68.8%) women. Nine (32.1%) women did not leak urine from the urethra on the cough test, indicating complete "dryness." This was measured objectively using a 1-hour pad weight test.
Mean pad weight was 26.5 ± 21.2 g. Thirteen (40.6%) women experienced a postoperative complication, which was primarily minor wound breakdown from the donor site.

| DISCUSSION
The literature reveals that, overall, outcomes for cases including flaps are better than for complex fistula repairs without flaps. Singapore flaps are employed in cases of moderate to severe scarring where there is a compromise of vaginal length, likely due to a large fistula. Gracilis flaps are employed when there has been a previous repair attempt, the fistula injury is severe, but vaginal reconstruction is not required. Values are given as number (percentage) and mean ± SD unless otherwise indicated.
Both flaps are employed when the injury is severe and vaginal reconstruction is required owing to a short length or severe scarring.
Incorporating a gracilis flap even in instances where the Singapore flap is felt to be sufficient is likely warranted given our lowest success rate for continence when using the Singapore flap alone. Although the Singapore flap carries a vascular supply, it is less robust than that of the gracilis muscle and is technically easier to compromise. Improving blood supply in compromised areas is paramount in any reconstructive procedure. Augmentation of the Singapore flap with the vascularity of the gracilis is beneficial to suture line healing and minimizing contracture. This is most poignantly demonstrated in two patients who had both a Singapore flap and a gracilis flap, who were discharged initially with a failed repair but returned to the center healed.
Our incontinence outcomes with the Singapore flap are lower than with the gracilis flap, suggesting again that perhaps the gracilis flap should be used more often in repairing large fistulae even when it seems that the Singapore flap is adequate. We have found that the Singapore flap The strengths of the present study are the large number of complex cases and the ability to follow outcomes because we were operating in a national referral center for women with obstetric fistulae. The major limitation is the retrospective nature of these data and the small sample size. However, it is exceedingly difficult to randomize cases with such heterogeneity of injury and these are relatively new techniques for use in obstetric-related fistula repairs. Future prospective studies, however, could use our suggested criteria in following outcomes.
In summary, we suggest the Singapore flap be used for vaginal reconstruction when large fistula injury has caused vaginal length shortening and/or severe tissue loss. We suggest the gracilis flap alone in cases of complex injury and previously failed repair attempts without vaginal tissue compromise. Finally, both flaps should be used when all of these variables are present, and perhaps more often in order to gain improved closure rates in all cases. We next plan to follow the quality of life and long-term outcomes including pad weights and urethral leakage to understand the changes over time. Ongoing work to decrease urethral leakage is still necessary to improve the impact of surgical repair on overall continence.

AUTHOR CONTRIBUTIONS
RP came up with the concept and initiated this study, drafted the manuscript and finalized edits. PC collected data and edited the manuscript. RP, RB, CC, LH, and JW contributed to the manuscript editing and surgical procedures performed in the study.