Iatrogenic genitourinary fistula: an 18-year retrospective review of 805 injuries

Obstetric fistula is a recognized consequence of health system failure and the inability of the world’s poorest and most marginalized women to access emergency obstetric care [2]. This study considers a sample of women—over 13 % of those presenting for fistula repair surgery—who suffered not directly because of prolonged, obstructed labor, but because of accidents caused by health providers. The prevalence of IF points to gaps in the quality of obstetric and gynecological surgery. The characteristics of iatrogenic injuries prompt us to consider opportunities to improve the quality of service provision.

A range of obstetric and gynecological procedures led to the development of IF, including CS (57.4 %), CS/hysterectomy (19.8 %), gynecological hysterectomy (19.6 %), ruptured uterus repair (3.1 %), and induced abortion (0.1 %). Other surgeries, such as destructive vaginal operations or symphysiotomy, also carry risks of accidental harm from the provider [7], but none was reported to be a causative procedure in this series.

Data on whether the causative CSs were elective or emergency were not collected, but data on the duration of labor suggest that 15 women had an elective CS (labor ≤1 h). The mean duration of labor reported by the remaining women who underwent CS or CS/hysterectomy was 39.7 h. Women experiencing obstetric complications frequently present to the hospital late, and providers may have had inadequate time to prepare their patients [8]. The availability of skilled professionals able to perform surgery is likely to be particularly low outside of normal working hours [8, 9].

The median age of women who developed IF during a gynecological procedure was 42 years, which is consistent with the patient population requiring gynecological procedures and is in line with published data [8, 10, 11]. The median age of women who developed IF during an obstetric procedure was 28 years, older than the age reported for most obstetric fistula patients [3, 5, 12, 14].

Several factors are suspected to place a woman at risk of IF. These include prior uterine operation, endometriosis, cervical myoma, and prior pelvic radiation [10, 11, 14]. Scar tissue and adhesions from prior laparotomies can create challenges for providers performing obstetric and gynecological surgery. It is therefore reasonable to hypothesize that obstetric or gynecological surgery might carry a greater risk of iatrogenic injury for women who have undergone a laparotomy in the past [10]. The frequency of previous laparotomy in the general population is unknown, but in this sample of women with IF, a full quarter had undergone one or more previous laparotomies. One quarter of the women who had had a previous laparotomy had undergone more than one. In this series, 98.0 % of previous laparotomies were cesarean sections.

The different types of IF were not equally associated with previous laparotomy. Under 12 % of ureteric injuries and 17.1 % of vault fistulas occurred in women who had undergone previous laparotomy, but 40.2 % of women with VCVF had undergone at least one previous laparotomy. Among VCVF patients, previous CS was more common in women who delivered a live baby than in those with a stillbirth (48.6 % vs 28.1 %). Providers are more prompt in providing CS to women with a previous CS, and more often as an elective procedure.

While appropriate cesarean sections improve maternal and perinatal outcomes, they do not confer similar advantages when performed in low-risk groups [15]. The World Health Organization has pointed out the intrinsic risk associated with CS [16]; yet, obstetric practice is shifting from vaginal to cesarean birth in many parts of the world, including in some of the countries included in this study [15‐18]. It would seem that one risk associated with CS, particularly repeated CS, is that providers might be more likely to accidentally cause iatrogenic injury during a subsequent surgery.

Early detection of IF can help patients avoid prolonged morbidity and its consequences. Early management of IFs should be feasible, so long as providers recognize the problem [8, 14, 19]. Providers can identify many IFs when removing the Foley catheter shortly after surgery. Excluding ureteric injuries, a substantial number of small IFs could be healed by re-introducing the catheter and leaving it for a period of 4–6 weeks, with a regimen of plenty of oral fluids and sitz baths [20].

The median time before patients began leaking among those with a VCVF or vault fistula was 2 and 3 days respectively, and 7 days after the causative surgery in those with ureteric injury. In this analysis, IFs following gynecological hysterectomy were treated earlier than those following obstetric surgeries. Differences in patient populations may explain this finding: gynecological patients are mainly self-referred, establishing a relationship with a provider who may recognize the problem and ensure appropriate care. If the leaking starts after discharge, the patient will go back to the operating provider and will be referred appropriately. Obstetric patients typically arrive as emergencies, and the operating provider may not see the patient after her CS. Moreover, leaking after an emergency CS could be due to pressure necrosis; providers may not immediately recognize the iatrogenic cause.

It appears that the ureters are not at equal risk of being accidentally damaged during CS [8, 21]. The left ureter is more likely to be affected during CS for several reasons. First, it is half a centimeter nearer to the cervix than the right ureter [22]. Second, the large sigmoid colon in African women causes dextro-rotation of the gravid uterus, bringing the left ureter forward [23]. Finally, many right-handed operators stand on the right side of the patient when performing CS, making it more likely to inadvertently injure the left ureter.

All cadres of health providers in this series performed procedures that resulted in IF, from assistant medical officers to specialists. National data on the cadres of staff performing different types of procedures are typically unavailable, but the profile described here is not surprising based on the human resources in the countries involved. Medical officers are typically the most likely cadre to carry out emergency surgeries such as CS or CS/hysterectomies; thus, their role in 59 % of IFs likely reflects their high involvement in at-risk procedures.

In both Malawi and Tanzania, nonphysician clinicians perform the majority of obstetric surgery. In Malawi, 88 % of emergency obstetric operations in district hospitals are performed by clinical officers [24]. In one study in Tanzania, more than 85 % of obstetric and gynecological surgeries were performed by assistant medical officers [25]. In light of this, the 73.8 % of IFs caused by CO/AMOs in Malawi and Tanzania seems reasonable. In all countries, specialists would be more likely to conduct elective procedures than emergency ones. This explains why specialists performed only 9.6 % of the causative procedures, but 38.6 % of the gynecological hysterectomies that resulted in IF.

This analysis has several limitations. First, the women themselves provided much of the information recorded in the patient records. While women mostly know their obstetric histories, recall can be a challenge. In some cases, providers may not have fully educated patients about procedures performed and the reasons for performing them. The fistula surgeon’s assessment complemented each woman’s account, helping to determine the most likely obstetric history. The first author’s determination of the cadre causing IF was dependent on his and his colleagues’ knowledge of local facilities and their staffing, which could be subject to recall bias and may be difficult for others to reproduce. Finally, these data do not indicate the overall prevalence of IF, but instead point to the proportion of fistula of iatrogenic origin among women in need of fistula repair surgery. This series of IF repairs was drawn from a sample of nearly 6,000 fistula repairs that took place in 65 facilities across 11 countries, ranging from sub-district hospitals to tertiary referral facilities. Reliable population-level information about fistula prevalence is unavailable, as is information about the total number of obstetric and gynecological surgical procedures performed. This means that some of the denominators that would put the findings into context are unavailable.

Prevention of IF is an urgent matter that needs to be addressed in developing countries. Providers performing obstetric and gynecological surgery must have the appropriate competencies. Training, combined with mentoring and ongoing supervision, is essential. In addition, women with obstructed labor must be able to quickly access a health facility with the staffing and infrastructure to provide high-quality emergency obstetric care. As such, it is critical to strengthen referral systems (emergency communication and transport) and address financial barriers that result in delays in care-seeking behavior. Training and equitable deployment of skilled birth attendants at all levels of the health system will ensure that providers can recognize signs of abnormal labor progression and make appropriate decisions about referral. Such efforts should be complemented by community-level interventions that promote household preparation for birth, increase male partner involvement in maternal health, and empower women to take action to ensure their own health and well-being.

Whenever women arrive at facilities in need of care, it is critical that providers are able to make informed and timely decisions. Facilities and providers must consider the quality of emergency obstetric care, including the decision-making process leading to a CS. The partograph is an essential tool for monitoring progress in labor and for diagnosing obstructed labor [26]. Providers must manage women in labor according to best medical practice; some women wait hours or days in a hospital before receiving any intervention [27, 28]. Facilities must likewise treat providers fairly, compensating them for their service and providing adequate resources, training, and supervision.

Providers must have the knowledge and experience to be able to provide high-quality services. As shown elsewhere [6, 7], the high number of cesarean deliveries for dead infants emphasizes the importance of re-evaluating alternative vaginal delivery methods. Many providers may benefit from training on alternatives to cesarean section, including vacuum extraction, symphysiotomy, and craniotomy for dead babies. Provider training should likewise highlight optimal operative techniques [9]. In dissection of the bladder and the lower uterine segment, blunt dissection can damage blood supply to the bladder, especially if the patient has had a previous CS. Sharp dissection is always preferable [14, 22]. The training, mentoring, and supervision of providers will improve care and therefore the health outcomes of women.