- Email: [email protected]
Mortality risk associated with surgical treatment of female genital fistula
IJG-07951; No of Pages 6 International Journal of Gynecology and Obstetrics xxx (2014) xxx–xxx
Contents lists available at ScienceDirect
International Journal of Gynecology and Obstetrics journal homepage: www.elsevier.com/locate/ijgo
CLINICAL ARTICLE
Mortality risk associated with surgical treatment of female genital fistula☆ Joseph Ruminjo a,⁎, Evelyn Landry a, Karen Beattie a, Adamu Isah b, Abu Jamil Faisel c, Sita Millimono d a
Fistula Care, EngenderHealth, NY, USA Fistula Care, EngenderHealth, Sokoto, Nigeria Fistula Care, EngenderHealth, Dhaka, Bangladesh d Fistula Care, EngenderHealth, Conakry, Guinea b c
a r t i c l e
i n f o
Article history: Received 20 August 2013 Received in revised form 4 February 2014 Accepted 7 April 2014 Keywords: Confidential inquiry Female genital fistula Mortality risk Surgical treatment
a b s t r a c t Objective: To describe the mortality risk associated with surgical treatment of female genital fistula and the contributory and contextual factors. Methods: In a descriptive study, confidential inquiries and clinical audits were conducted at 14 fistula repair sites in seven resource-poor countries between January 2005 and March 2013. Data collection included interviews with key personnel involved in the clinical management of the deceased, and a review of hospital records and patient files following an audit protocol. Results: Overall, 26 060 fistula repairs were performed at 44 sites located in 13 countries; 30 deaths were reported in this period. Twenty-one deaths were attributable to surgery, yielding a case fatality of 0.08 per 100 procedures. The cause of death in nearly half of the cases was various manifestations of sepsis and inflammation. Conclusion: The case fatality rate for fistula repair surgery in resource-poor countries was in the same range as that for comparable gynecologic operations in high-resource settings. Clinical and systemic issues to be addressed to reduce the case fatality rate include improvement of perioperative care and follow-up, assuring prudent referral or deferral of difficult cases, and maintaining better records. © 2014 Published by Elsevier Ireland Ltd. on behalf of International Federation of Gynecology and Obstetrics.
1. Introduction Intensive efforts at both national and international levels reduced annual maternal mortality globally to approximately 287 000 maternal deaths in 2010, representing a decrease of 47% from the level in 1990. Sub-Saharan Africa (56%) and southern Asia (29%) accounted for 85% of the global burden of maternal deaths in 2010 [1]; obstructed labor is the direct cause of 8% of these deaths [2]. Women who survive obstructed labor may suffer severe and longterm morbidity, and genital fistula being one of the most devastating associated conditions. There are no reliable data on the global burden of fistula [3], but a recent systemic review and meta-analysis estimated that there are more than 1 million cases [4]. Fistula repair carries some risk—as does all major surgery—including that of death. Although evidence for the occurrence of immediate,
☆ A preliminary version of this work was presented at the 2013 Global Maternal Health Congress in Arusha, Tanzania, co-sponsored by Management and Development for Health, Dar es Salaam, Tanzania, and the Maternal Health Task Force at the Harvard School of Public Health, Boston, USA. ⁎ Corresponding author at: EngenderHealth, 440 Ninth Avenue, FL 13, New York, NY 10001, USA. Tel.: +1 212 561 8000; fax: +1 212 561 8067. E-mail address: [email protected] (J. Ruminjo).
medium-term, and longer-term complications from such surgery has been reviewed [5–7], death is rarely mentioned except anecdotally. If death from fistula treatment is infrequent, research will require large caseloads involving multiple sites to compile adequate numbers for analysis. Hancock and Browning [8] and Waaldijk [9] have reported on several actual and potential causes of mortality from fistula treatment. A case fatality rate of 0.7% was reported from a small series of women undergoing repair surgery in Nigeria [10]. By contrast, there are data on mortality for comparable gynecologic surgery, such as hysterectomy, in resource-rich settings. Mortality risk associated with these procedures varies by surgical approach and associated conditions, although crude and attributable rates are often unspecified (Table 1) [11–14]. Mortality rates from diverse surgeries for pelvic organ prolapse are generally lower in these settings, but range from 0.05% to 0.38% [15,16]. The aim of the present study was to describe the mortality risk of surgical treatment for female genital fistula and the contributing and contextual factors. 2. Materials and methods The data presented in this descriptive study were collected between January 1, 2005, and March 31, 2013, from 44 facilities supported by EngenderHealth that provide services for fistula repair in 13 resource-
http://dx.doi.org/10.1016/j.ijgo.2014.02.015 0020-7292/© 2014 Published by Elsevier Ireland Ltd. on behalf of International Federation of Gynecology and Obstetrics.
Please cite this article as: Ruminjo J, et al, Mortality risk associated with surgical treatment of female genital fistula, Int J Gynecol Obstet (2014), http://dx.doi.org/10.1016/j.ijgo.2014.02.015
2
J. Ruminjo et al. / International Journal of Gynecology and Obstetrics xxx (2014) xxx–xxx
Table 1 Mortality rates for gynecologic surgical procedures comparable to fistula repair in resource-rich settings. Surgical procedure
Mortality rate, %
Comments
Reference
Vaginal hysterectomy
0.06
Wingo et al. 1985 [11]
Abdominal hysterectomy
0.14
“Simple” hysterectomy for benign disease
0.16
Overall hysterectomy mortality for benign disease
0.15
Hysterectomy for benign disease
0.38
Operations for female stress urinary incontinence
0.05
Varies with associated conditions; 119 972 vaginal hysterectomies; United States Varies with associated conditions; 317 389 abdominal hysterectomies; United States Population based; 29 192 women; Denmark 220 vaginal, 1349 abdominal, and 223 laparoscopic hysterectomies; Canada VALUE study: Vaginal Abdominal Laparoscopy Uterine Excision; 37 295 cases England, N. Ireland, Wales Literature review, meta-analysis of 282 journal articles on slings, anterior repairs, retropubic, and transvaginal colposuspension
poor countries with funding from the US Agency for International Development (USAID). Facility administrators consented to the requirements and procedures for the confidential inquiry at the start of EngenderHealth support. Administrators, key service delivery staff, and other interviewees gave voluntary verbal consent if the need arose for confidential inquiry. Institutional review board approval was not required because clinical quality audit for complications and death is part of routine program monitoring and evaluation. In its Fistula Care project, EngenderHealth developed a death investigation protocol for fistula services as part of routine program monitoring and evaluation in south Asia and Sub-Saharan Africa. Using the protocol and reporting tools, confidential inquiries were conducted for any women who died either during surgery for urinary or fecal fistula of obstetric or traumatic origin or up to 42 days after such surgery (Fistula program stakeholders agreed to the 42-day post-repair mortality timeline, which is consistent with standard reporting for maternal mortality and physiologic recovery after clinical trauma). Each supported site was asked to notify EngenderHealth within 3 days of the death of a woman who had undergone fistula repair surgery. EngenderHealth clinical staff or consultants collaborated with clinical and administrative staff at the repair facilities to conduct the audits. Interviews were conducted with the surgical team, facility director, and medical and paramedical staff involved in patient screening and other direct clinical care. If a death occurred after discharge, family members, field staff, and providers who might have seen the patient were also interviewed. The audit team reviewed ward and theater registers, patient records, records of anesthesia and analgesia regimen, and diagnostic investigations, in addition to the chronology of complications and their evolution. The team also reviewed referral records if the patient was treated elsewhere, and autopsy findings if available. The data collection form for the investigation was not designed with pre-coded response options; most questions were open-ended. Lastly, the audit team conducted on-site observations to identify medical quality issues that might have contributed to death. The audit was designed to be facilitative and constructive, focusing on potential systemic flaws directly or indirectly related to the death, and the approach emphasized diplomacy, sensitivity, and confidentiality. The objectives were to determine the cause of death, identify contributing factors, and ascertain whether death was attributable to the procedure. The findings were also used to determine whether the death was preventable and to design remedial actions. All information gathered during the audits was confidential. The names of the deceased and of service providers were excluded from reports; electronic and hardcopy reports were securely filed. The denominator for the present analysis was the number of fistula repairs reported over a 9-year study period from the supported sites.
Wingo et al. 1985 [11]
Loft et al. 1991 [12]
Kafy et al. 2006 [13]
McPherson et al. 2004 [14]
American Urological Association, Clinical Guidelines Panel, 1997 [15]
The results are reported as descriptive statistics (total numbers and percentages). There was no statistical analysis. 3. Results The 44 supported fistula repair sites reported 26 060 repairs over the 9-year period. The supported sites did not report to EngenderHealth the number of women returning for postoperative follow-up. Twenty seven repair facilities were public institutions; the others were private (n = 5) or faith-based (n = 12). Repair sites were located in both urban and rural settings. The duration of EngenderHealth’s support to these sites ranged from 1 to 8 years. Thirty deaths were reported across the 44 study sites. After clinical investigation, 21 deaths were deemed likely to be attributable to the procedure. In seven cases, there was no history or examination or laboratory findings indicative of likely attribution (e.g., on the basis of the sequence of clinical events or the time between surgery and death); there was a definitive other cause of death; or there were contrary autopsy findings. There was insufficient information to make a determination in two cases. The attributable case fatality from fistula surgery was 0.08 per 100 procedures. The gross case fatality rate, inclusive of non-attributable cases, was 0.12 per 100 procedures. Deaths were reported in 14 repair sites located in seven of the 13 countries. Across the sites reporting death, yearly mortality ranged from 0.00 to 0.36 per 100 procedures (Table 2). The number of reported cases per country ranged from zero to seven, or 0.00–0.29 per 100 procedures (Table 3). Age, which was reported for 15 of the 21 women, ranged from 17 to 65 years (median 25 years). None of the reported deaths occurred during the postpartum period. The duration of post-procedure survival, noted for 17 women, ranged from 1 to 32 days (median 10 days). No pattern was discernible by country. The complexity of fistula repair surgery, as assessed by the surgeons, varied. Seven women underwent simple fistula repair via the vaginal route; two women underwent trans-vaginal repairs of medium/ intermediate complexity; and six repairs were complex. Among the complex repairs, four women required adjunct or complementary surgery for separate or concurrent rectovaginal fistula and/or colostomy or other procedures: three of these repairs were done via the abdominal route, and one was done via a combined route. For six women, there were insufficient data to determine repair complexity. Spinal anesthesia was used exclusively in 11 of the 21 cases. General anesthesia was used in three cases, and “other” or combined (i.e., sedation, ketamine, or caudal epidural) was used in three cases. The type of anesthesia was not recorded in the clinical records of four women. No death was reported as directly attributable to anesthesia, but in three
Please cite this article as: Ruminjo J, et al, Mortality risk associated with surgical treatment of female genital fistula, Int J Gynecol Obstet (2014), http://dx.doi.org/10.1016/j.ijgo.2014.02.015
J. Ruminjo et al. / International Journal of Gynecology and Obstetrics xxx (2014) xxx–xxx Table 2 Fistula repair attributable mortality by year. Project year
No. of deaths
No. of repairs
Case fatality per 100 procedures
January 2005–September 2005 October 2005–September 2006 October 2006–September 2007 October 2007–September 2008 October 2008–September 2009 October 2009–September 2010 October 2010–September 2011 October 2011–September 2012 October 2012–March 2013 Total
0 3 1 3 1 3 5 4 1 21
141 846 2967 3150 3278 3871 4225 4759 2823 26 060
0.00 0.36 0.03 0.10 0.03 0.08 0.12 0.08 0.04 0.08
cases the women had unexplained perioperative convulsions; it is unclear whether these convulsions might have been related to anesthesia, electrolyte imbalance, or some other cause. The fistula surgery skills of the 17 principal surgeons for the 21 cases varied. Seven surgeons had competency for complex fistula surgery, three for medium/intermediate complex fistula surgery, and one for simple fistula surgery. For six surgeons, the skill level could not be ascertained. The primary cause of death in nearly 50% of cases (10/21) was related to manifestations of sepsis and inflammation (septicemia and peritonitis) (Table 4). Secondary causes of death were also given for some women, the most frequent being shock and/or hypotension. The audit uncovered contributing or underlying factors, some of which were specific and others that were general or systemic. The most frequent systemic factors were a lack of personnel to attend emergencies, emergency equipment, specific medications, and consumables. Surgeons did not always act on cautionary red flags in the patient history, physical examination, or diagnostic investigation. Postoperative care and documentation of new symptoms, signs, and differential diagnoses by medical and nursing staff were sometimes inadequate. At times, there was poor “handover” and continuity of care when “visiting surgeons” performed the surgery but did not leave documentation or adequate instructions for postoperative management by the resident team. Strength and certainty of diagnosis varied according to investigative routines and rigor, which often depended on available resources. In 14 cases, diagnosis was classified as presumptive rather than definitive. This was due to the lack or imprecision of available diagnostic investigations or documentation, and thus a greater reliance on patient history and a series of physical examinations conducted as part of routine care. Routine laboratory investigations carried out in the pre- and postoperative periods ranged from basic to extensive for 15 women. Investigations were not done for two women, and were unknown for four women. Laboratory investigations usually included a hemoglobin estimation and hematocrit levels, and occasionally a typhoid and/or Widal test, blood sugar levels, and stool microscopy. Less frequently reported tests included urea or electrolytes, blood creatinine levels, and HIV tests. Only a few cases included urine tests or blood cultures (positive in four cases); contrast urinary tract radiology or ultrasound testing was rarely carried out. An autopsy and/or histologic examination was conducted in only two cases: neither death was attributable to the repair (one woman suffered hemorrhagic shock from portal hypertension and bleeding esophageal varices; the other had metastatic brain tumors from an intestinal primary). Records stated definitively that autopsies were not conducted for 14 of the attributable cases. The remaining seven cases had no definitive statement about autopsy and no available results. It is likely that autopsy was not undertaken owing to cultural and/or religious prohibitions. In 9 of the 21 cases, the death was subjectively assessed as likely to have been preventable with available resources. In five cases, the death was probably not preventable. Data were insufficient to make a
3
determination in seven cases. Preventable circumstances included performing an essentially elective surgery on a woman whose condition was already compromised; not paying due diligence to medical history, including previous investigations and medications; and, arguably, not ending an operation that was becoming progressively more complicated and hazardous. 4. Discussion The present study is one of the largest documented prospective, confidential inquiries into mortality risk for surgical treatment of female genital fistula. It provides a detailed assessment of the clinical and contextual risk from multiple country programs and several fistula repair sites over a wide geographic area and under varied economic and organizational conditions. The data indicate that mortality risk from surgical treatment of female genital fistula is within the same range as comparable gynecologic operations in high-resource settings, despite the severe constraints under which fistula surgical teams work in low-resource environments [11–16]. Detailed data for the mortality risk of fistula repair are scarce. In a series of more than 800 fistula repairs, Hancock and Browning [8] identified the likely causes of death as embolism and hyponatremia, but they did not cite specific case fatality rates. Similarly, Waaldijk [9], who reviewed more than 5000 fistula procedures under spinal anesthesia, did not cite a specific case fatality rate. His series included two deaths (from total spinal block and meningitis), but the case of meningitis might have been unrelated to the procedure. In a subsequent personal communication, Waaldijk was quoted as having found a 2% mortality rate, declining to 0.4% over the years [5]. More detailed data have come from a “surgical camp” or outreach service activity termed “Fistula Fortnight” for 569 women at four hospitals in northern Nigeria [10]. This program was deemed successful in that it strengthened clinical skills, provided medical equipment, and raised public awareness about treatment options. Four women, aged 17–42 years, died postoperatively on days 6, 13, 19, and 32, yielding a case fatality rate of 0.7%. Autopsies were not performed owing to cultural restrictions. Although this rate is almost nine times that found in the current series, the deaths were reportedly from malaria, hypoglycemic coma, hypertension, and leukemia, and might not have been directly attributable to the procedure. Nevertheless, that particular service model raised issues about the quality of perioperative screening, care, and follow-up, especially if no fistula surgeon is resident on-site during the woman’s postoperative recovery period. In the present audit, no post-repair deaths were deemed attributable to anesthesia or electrolyte imbalance. However, the three cases of unexplained perioperative convulsions would require diagnostics to rule out anesthetic complications and/or fluid or electrolyte imbalance; these tests are not available at most sites in the resource-poor settings. The primary cause of death for nearly half of the cases was infection and/or inflammation. Arrowsmith et al. [17] reported concern about antibiotic regimens that do not address emerging multi-resistant bacteria, as did Gordon Williams in a presentation at the International Society of Obstetric Fistula Surgeons (ISOFS 3rd Annual Meeting, Dakar, Senegal, October 7-9, 2010). Only a few women in the present series received blood culture and sensitivity testing for septicemia but, consistent with these concerns, their samples contained microbes that are resistant to many common antibiotics. The audits identified several systemic issues where recommendations for improvement are necessary. Although infection or inflammation sequelae were noted as the primary cause of death, contributory causes were also reported including anemia, malaria, dehydration, malnutrition, and other chronic morbidities. Preoperative screening, selection, and assessment of when the woman would be in the best condition to undergo surgery needs improvement and should be resolved before such elective surgery is performed. Routine use of a standard
Please cite this article as: Ruminjo J, et al, Mortality risk associated with surgical treatment of female genital fistula, Int J Gynecol Obstet (2014), http://dx.doi.org/10.1016/j.ijgo.2014.02.015
4
J. Ruminjo et al. / International Journal of Gynecology and Obstetrics xxx (2014) xxx–xxx
Table 3 Attributable case fatalities per 100 procedures by country and project year (October–September). Country Bangladesh No. of sites reporting No. of repairs No. of deaths Case fatality rate, % Beninc No. of sites reporting No. of repairs No. of deaths Case fatality rate, % DRCd No. of sites reporting No. of repairs No. of deaths Case fatality rate, % Ghanac No. of sites reporting No. of repairs No. of deaths Case fatality rate, % Guinea No. of sites reporting No. of repairs No. of deaths Case fatality rate, % Liberiac No. of sites reporting No. of repairs No. of deaths Case fatality rate, % Mali No. of sites reporting No. of repairs No. of deaths Case fatality rate, % Niger No. of sites reporting No. of repairs No. of deaths Case fatality rate, % Nigeria No. of sites reporting No. of repairs No. of deaths Case fatality rate, % Rwanda No. of sites reporting No. of repairs No. of deaths Case fatality rate, % Sierra Leone No. of sites reporting No. of repairs No. of deaths Case fatality rate, % Togoc No. of sites reporting No. of repairs No. of deaths Case fatality rate, % Uganda No. of sites reporting No. of repairs No. of deaths Case fatality rate, % Total No. of sites reporting Total no. of repairs Total no. of deaths Case fatality rate, %
2004–2005 a
2005–2006
2006–2007
2007–2008
2008–2009
2009–2010
2010–2011
2011–2012
2012–2013 b
Total
3 20 0 0.00
3 93 1 1.08
3 119 0 0.00
3 122 0 0.00
4 131 0 0.00
4 143 0 0.00
4 150 0 0.00
4 184 2 1.10
4 80 0 0.00
1042 3 0.29
– NS NS NS
1 NS NS NS
1 NS NS NS
– NS NS NS
– 110 0 0.00
1 21 0 0.00
1 20 0 0
– NS NS NS
– NS NS NS
151 0 0.00
0 NS NS NS
1 53 0 0.00
2 586 1 0.17
2 334 0 0.00
2 482 0 0.00
2 472 0 0.00
6 565 1 0.18
6 1221 1 0.08
7 829 0 0.00
4542 3 0.07
0 NS NS NS
1 21 0 0.00
1 42 0 0.00
– NS NS NS
– NS NS NS
– NS NS NS
– NS NS NS
– NS NS NS
– NS NS NS
63 0 0.00
0 NS NS NS
2 199 0 0.00
2 292 0 0.00
3 229 0 0.00
4 316 0 0.00
3 392 0 0.00
3 459 0 0.00
3 497 0 0.00
3 228 0 0.00
2612 0 0.00
– NS NS NS
– NS NS NS
– NS NS NS
1 59 0 0.00
– NS NS NS
– NS NS NS
– NS NS NS
– NS NS NS
– NS NS NS
59 0 0.00
– NS NS NS
– NS NS NS
– NS NS NS
– NS NS NS
1 46 0 0.00
1 40 0 0.00
1 91 0 0.00
1 53 0 0.00
3 125 0 0.00
355 0 0.00
– NS NS NS
– NS NS NS
1 27 0 0.00
4 213 0 0.00
3 158 0 0.00
4 220 2 0.91
4 333 0 0.00
4 209 0 0.00
5 213 0 0.00
1373 2 0.15
– NS NS NS
– NS NS NS
5 1081 2 0.19
5 1437 1 0.07
6 1347 1 0.07
6 1612 1 0.06
7 1507 0 0.00
9 1720 1 0.06
10 917 1 0.11
9621 7 0.07
– NS NS NS
2 145 0 0.00
2 147 0 0.00
2 83 0 0.00
3 167 0 0.00
3 259 0 0.00
3 278 0 0.00
4 114 1 0.88
3 19 0 0.00
1212 1 0.08
– NS NS NS
– NS NS NS
1 272 0 0.00
1 363 1 0.28
1 253 0 0.00
1 166 0 0.00
1 211 0 0.00
1 244 0 0.00
1 73 0 0.00
1582 1 0.06
– NS NS NS
– NS NS NS
– NS NS NS
– NS NS NS
– NS NS NS
1 97 0 0.00
– NS NS NS
– NS NS NS
– NS NS NS
97 0 0.00
1 121 0 0.00
2 335 0 0.00
2 401 0 0.00
2 310 1 0.32
2 268 0 0.00
2 449 0 0.00
2 611 2 0.33
3 517 1 0.19
3 339 0 0.00
3351 4 0.12
4 141 0 0.00
12 846 1 0.12
20 2967 3 0.10
23 3150 3 0.11
26 3278 1 0.03
27 3871 3 0.08
31 4225 5 0.12
35 4759 3 0.06
35 2823 1 0.04
26 060 21 0.08
Abbreviations: DRC, Democratic Republic of Congo; NS, not supported. a January–September 2005. b October 2012–March 2013 c Fistula repair services provided aboard a hospital ship which traveled to four countries (Benin, Ghana, Liberia and Togo). Counted as one site. d Includes repairs reported by the US Agency for International Development (USAID)-funded bilateral project in fiscal years 2005–2006, 2006–2007, and 2007–2008.
Please cite this article as: Ruminjo J, et al, Mortality risk associated with surgical treatment of female genital fistula, Int J Gynecol Obstet (2014), http://dx.doi.org/10.1016/j.ijgo.2014.02.015
J. Ruminjo et al. / International Journal of Gynecology and Obstetrics xxx (2014) xxx–xxx Table 4 Primary and secondary causes of death and contributory factors among the attributable deaths (n = 21).
Inflammation and/or infection Sepsis Septicemia Peritonitis Hematologic Pulmonary embolism Acute anemia Anemia Leukemia Splenic infarction Splenic sequestration Urinary Acute exacerbation of chronic renal failure Renal tubular necrosis/ureteric damage Renal failure Bladder lithiasis Chronic renal failure Uremia Previous urinary diversion Cardiovascular First-degree heart block Hypertension Congestive cardiac failure Hypotension and/or shock Cardiac arrhythmia Hypotension Metabolic Hypoglycemic coma Diabetes Hypokalemia Dehydration Central nervous system Cerebrovascular accident Perioperative convulsions/seizures Epilepsy Respiratory system Status asthmaticus Aspiration and/or pneumonitis Asthma Gastrointestinal system Gastroenteritis/bloody diarrhea Intestinal obstruction Acute abdominal distension Diarrhea Prolapsed colostomy Concurrent rectovaginal fistula Parasitemia Malaria Other Old age/frailty/general weakness Malnutrition Previous multiple surgeries Treated cervical cancer Nonclinical system factors
Primary cause
Secondary cause
Contributing factor
4 4 2
2 – –
– – –
2 1 – – – –
1 – 2 1 – –
– – 1 – 1 1
1 1 – – – – –
– – 1 1 – – –
– – – – 1 1 1
1 1 1 – – –
– – – 4 1 –
– – – – – 1
1 – – –
– 1 1 –
– 1 – 1
1 – –
– 3 –
– – 1
1 – – – – – – – – – – –
– 3 – – 3 1 1 – – – – 2
– – 1 – – – – 2 1 1 – 2
– – – – –
– – – – –
3 1 2 1 13
5
Fistula Care’s program in Nigeria, regular “Provider Network” retreats involve fistula surgeons, nurses, anesthetists, and managers from different facilities (I. Efem, personal communication, April 15, 2013), and fistula mentorship initiatives are taking place in the fistula program in Uganda (R. Mukisa, personal communication, April 15, 2013). The present analysis has some limitations. Clinical records were sometimes incomplete or contained poor quality data, including inconsistent documentation of perioperative observations and follow-up care. Lack of diagnostic resources, including those for autopsy and/or histology, meant that differential diagnoses and, consequently, attribution and preventability were often presumptive. In addition, given the understandably charged emotions surrounding these deaths, interview and recorded data might be inaccurate. Supported sites might have underreported deaths, because investigation requires extra work and (despite assurances to the contrary) providers and administrators may feel defensive or apprehensive about an audit. Related reviews have shown that, although maternal death audit is becoming more common in some low-income countries, facility-level data may underestimate the number of deaths [19,20]. In summary, the case fatality rate for fistula repair surgery in lowincome countries was in the same range as that for comparable gynecologic operations in high-income settings. Clinical and systemic issues must be addressed to reduce the case fatality rate. These include improving perioperative care and follow-up, assuring timely referral or deferral of difficult cases, and maintaining better records. Acknowledgments
checklist has been shown to improve safety in general surgery, and this can be adapted for fistula repair [18]. In addition, the recently launched global fistula training manual from the International Federation of Gynecology and Obstetrics includes guidance on developing competency to conduct audits, which might be useful for future analysis of systemic and clinical issues leading to case fatality [6]. A minimum package of resources is needed for diagnostics, including a quality laboratory, requisite infrastructure, trained technicians, equipment, reagents, and expendable supplies. Even when resources are available, diagnostic investigations might be too expensive. Personnel and physical resources should always be available for early identification and management of emergencies, in addition to better referral systems and more accountable handover between teams. Teamwork is crucial, for both clinical and ethical reasons, especially among visiting surgical and outreach teams and the resident staff. In
The Fistula Care project at EngenderHealth was funded by USAID under associate cooperative agreement GHS-A-00-07-00021-00. Views expressed here do not necessarily reflect those of USAID, the US Government or EngenderHealth. Conflict of interest The authors have no conflicts of interest. References [1] World Health Organization, UNICEF, UNFPA, The World Bank. Trends in maternal mortality: 1990 to 2010. http://www.who.int/reproductivehealth/publications/ monitoring/9789241503631/en/. Published 2012. [2] World Health Organization. The World Health Report 2005 – make every mother and child count. http://www.who.int/whr/2005/en/. Published 2005. [3] Stanton C, Holtz SA, Ahmed S. Challenges in measuring obstetric fistula. Int J Gynecol Obstet 2007;99(Suppl. 1):S4–9. [4] Adler AJ, Ronsmans C, Calvert C, Fillippi V. Estimating the prevalence of obstetric fistula: a systemic review and meta-analysis. BMC Pregnancy Childbirth 2013;13:246. [5] Abrams P, de Ridder D, de Vries C, Elneil S, Esegbona G, Mourad S, et al. Obstetric Fistula in the Developing World. Presented at An International Consultation on Vesicovaginal Fistula. Marrakech, Morocco, October 13–16, 2010. Montreal: Societe Internationale d’Urololgie; 2012. [6] Elneil S. Global competency-based fistula surgery training manual. http:/www.figo. org/publications/miscellaneous_publications/Global_Competency_Based_Fistula_ Surgery_Training_Manual. Published June 2011. [7] Gutman RE, Dodson JL, Mostwin JL. Complications of treatment of obstetric fistula in the developing world: gynatresia, urinary incontinence, and urinary diversion. Int J Gynecol Obstet 2007;99(Supplement 1):S57–64. [8] Hancock B, Browning A. Practical Obstetric Fistula Surgery. London: Royal Society of Medicine Press Limited; 2008. [9] Waaldijk K. Step-by-step Surgery of Vesicovaginal Fistulas: A Full-color Atlas. Cheltenham: Nelson Thornes Limited; 1994. [10] Ramsey K, liyasu Z, Idoko L. Fistula Fortnight: Innovative partnership brings mass treatment and public awareness towards ending obstetric fistula. Int J Gynecol Obstet 2007;99(Supplement 1):S130–6. [11] Wingo PA, Huezo CM, Rubin GL, Ory HW, Peterson HB. The mortality risk associated with hysterectomy. Am J Obstet Gynecol 1985;152(7):803–8. [12] Loft A, Anderson TF, Brønnum-Hansen H, Roepstorff C, Madsen M. Early postoperative mortality following hysterectomy. A Danish population based study, 1977–1981. Br J Obstet Gynaecol 1991;98(2):147–54. [13] Kafy S, Huang JY, Al-Sunaidi M, Wiener D, Tulandi T. Audit of morbidity and mortality rates of 1782 hysterectomies. J Minim Invasive Gynecol 2006;13(1):55–9. [14] McPherson K, Metcalfe MA, Herbert A, Maresh M, Casbard A, Hargreaves J, et al. Severe complications of hysterectomy: the VALUE study. BJOG 2004;111(7):688–94.
Please cite this article as: Ruminjo J, et al, Mortality risk associated with surgical treatment of female genital fistula, Int J Gynecol Obstet (2014), http://dx.doi.org/10.1016/j.ijgo.2014.02.015
6
J. Ruminjo et al. / International Journal of Gynecology and Obstetrics xxx (2014) xxx–xxx
[15] The American Urological Association, Female Stress Urinary Incontinence Clinical Guidelines Panel. Report on The Surgical Management of Female Stress Urinary Incontinence. www.auanet.org/common/pdf/education/Arc-SUI.pdf. Published 2007. [16] Drutz HP, Herschorn S, Diamant NE. Female Pelvic Medicine and Reconstructive Pelvic Surgery. Berlin: Springer Verlag Gmbh; 2003. [17] Arrowsmith SD, Ruminjo J, Landry EG. Current practices in treatment of female genital fistula: A cross-sectional study. BMC Pregnancy Childbirth 2010;10:73. [18] World Health Organization. Reproducing and Adapting the WHO Surgical Safety Checklist. http://www.who.int/patientsafety/safesurgery/using_checklist/en/. Accessed December 30, 2013.
[19] Graham WJ, Ahmed S, Stanton C, Abou-Zahr CL, Campbell OMR. Measuring maternal mortality: an overview of opportunities and options for developing countries. BMC Med 2008;6:12. [20] Immpact, University of Aberdeen. Immpact Toolkit: a guide and tools for maternal mortality programme assessment: Rapid Ascertainment Process for Institutional Deaths (RAPID). Module 4, Tool 2. Version 2.0 http://www.abdn.ac.uk/iahs/contentimages/rapid.pdf. Published 2007.
Please cite this article as: Ruminjo J, et al, Mortality risk associated with surgical treatment of female genital fistula, Int J Gynecol Obstet (2014), http://dx.doi.org/10.1016/j.ijgo.2014.02.015
Contents lists available at ScienceDirect
International Journal of Gynecology and Obstetrics journal homepage: www.elsevier.com/locate/ijgo
CLINICAL ARTICLE
Mortality risk associated with surgical treatment of female genital fistula☆ Joseph Ruminjo a,⁎, Evelyn Landry a, Karen Beattie a, Adamu Isah b, Abu Jamil Faisel c, Sita Millimono d a
Fistula Care, EngenderHealth, NY, USA Fistula Care, EngenderHealth, Sokoto, Nigeria Fistula Care, EngenderHealth, Dhaka, Bangladesh d Fistula Care, EngenderHealth, Conakry, Guinea b c
a r t i c l e
i n f o
Article history: Received 20 August 2013 Received in revised form 4 February 2014 Accepted 7 April 2014 Keywords: Confidential inquiry Female genital fistula Mortality risk Surgical treatment
a b s t r a c t Objective: To describe the mortality risk associated with surgical treatment of female genital fistula and the contributory and contextual factors. Methods: In a descriptive study, confidential inquiries and clinical audits were conducted at 14 fistula repair sites in seven resource-poor countries between January 2005 and March 2013. Data collection included interviews with key personnel involved in the clinical management of the deceased, and a review of hospital records and patient files following an audit protocol. Results: Overall, 26 060 fistula repairs were performed at 44 sites located in 13 countries; 30 deaths were reported in this period. Twenty-one deaths were attributable to surgery, yielding a case fatality of 0.08 per 100 procedures. The cause of death in nearly half of the cases was various manifestations of sepsis and inflammation. Conclusion: The case fatality rate for fistula repair surgery in resource-poor countries was in the same range as that for comparable gynecologic operations in high-resource settings. Clinical and systemic issues to be addressed to reduce the case fatality rate include improvement of perioperative care and follow-up, assuring prudent referral or deferral of difficult cases, and maintaining better records. © 2014 Published by Elsevier Ireland Ltd. on behalf of International Federation of Gynecology and Obstetrics.
1. Introduction Intensive efforts at both national and international levels reduced annual maternal mortality globally to approximately 287 000 maternal deaths in 2010, representing a decrease of 47% from the level in 1990. Sub-Saharan Africa (56%) and southern Asia (29%) accounted for 85% of the global burden of maternal deaths in 2010 [1]; obstructed labor is the direct cause of 8% of these deaths [2]. Women who survive obstructed labor may suffer severe and longterm morbidity, and genital fistula being one of the most devastating associated conditions. There are no reliable data on the global burden of fistula [3], but a recent systemic review and meta-analysis estimated that there are more than 1 million cases [4]. Fistula repair carries some risk—as does all major surgery—including that of death. Although evidence for the occurrence of immediate,
☆ A preliminary version of this work was presented at the 2013 Global Maternal Health Congress in Arusha, Tanzania, co-sponsored by Management and Development for Health, Dar es Salaam, Tanzania, and the Maternal Health Task Force at the Harvard School of Public Health, Boston, USA. ⁎ Corresponding author at: EngenderHealth, 440 Ninth Avenue, FL 13, New York, NY 10001, USA. Tel.: +1 212 561 8000; fax: +1 212 561 8067. E-mail address: [email protected] (J. Ruminjo).
medium-term, and longer-term complications from such surgery has been reviewed [5–7], death is rarely mentioned except anecdotally. If death from fistula treatment is infrequent, research will require large caseloads involving multiple sites to compile adequate numbers for analysis. Hancock and Browning [8] and Waaldijk [9] have reported on several actual and potential causes of mortality from fistula treatment. A case fatality rate of 0.7% was reported from a small series of women undergoing repair surgery in Nigeria [10]. By contrast, there are data on mortality for comparable gynecologic surgery, such as hysterectomy, in resource-rich settings. Mortality risk associated with these procedures varies by surgical approach and associated conditions, although crude and attributable rates are often unspecified (Table 1) [11–14]. Mortality rates from diverse surgeries for pelvic organ prolapse are generally lower in these settings, but range from 0.05% to 0.38% [15,16]. The aim of the present study was to describe the mortality risk of surgical treatment for female genital fistula and the contributing and contextual factors. 2. Materials and methods The data presented in this descriptive study were collected between January 1, 2005, and March 31, 2013, from 44 facilities supported by EngenderHealth that provide services for fistula repair in 13 resource-
http://dx.doi.org/10.1016/j.ijgo.2014.02.015 0020-7292/© 2014 Published by Elsevier Ireland Ltd. on behalf of International Federation of Gynecology and Obstetrics.
Please cite this article as: Ruminjo J, et al, Mortality risk associated with surgical treatment of female genital fistula, Int J Gynecol Obstet (2014), http://dx.doi.org/10.1016/j.ijgo.2014.02.015
2
J. Ruminjo et al. / International Journal of Gynecology and Obstetrics xxx (2014) xxx–xxx
Table 1 Mortality rates for gynecologic surgical procedures comparable to fistula repair in resource-rich settings. Surgical procedure
Mortality rate, %
Comments
Reference
Vaginal hysterectomy
0.06
Wingo et al. 1985 [11]
Abdominal hysterectomy
0.14
“Simple” hysterectomy for benign disease
0.16
Overall hysterectomy mortality for benign disease
0.15
Hysterectomy for benign disease
0.38
Operations for female stress urinary incontinence
0.05
Varies with associated conditions; 119 972 vaginal hysterectomies; United States Varies with associated conditions; 317 389 abdominal hysterectomies; United States Population based; 29 192 women; Denmark 220 vaginal, 1349 abdominal, and 223 laparoscopic hysterectomies; Canada VALUE study: Vaginal Abdominal Laparoscopy Uterine Excision; 37 295 cases England, N. Ireland, Wales Literature review, meta-analysis of 282 journal articles on slings, anterior repairs, retropubic, and transvaginal colposuspension
poor countries with funding from the US Agency for International Development (USAID). Facility administrators consented to the requirements and procedures for the confidential inquiry at the start of EngenderHealth support. Administrators, key service delivery staff, and other interviewees gave voluntary verbal consent if the need arose for confidential inquiry. Institutional review board approval was not required because clinical quality audit for complications and death is part of routine program monitoring and evaluation. In its Fistula Care project, EngenderHealth developed a death investigation protocol for fistula services as part of routine program monitoring and evaluation in south Asia and Sub-Saharan Africa. Using the protocol and reporting tools, confidential inquiries were conducted for any women who died either during surgery for urinary or fecal fistula of obstetric or traumatic origin or up to 42 days after such surgery (Fistula program stakeholders agreed to the 42-day post-repair mortality timeline, which is consistent with standard reporting for maternal mortality and physiologic recovery after clinical trauma). Each supported site was asked to notify EngenderHealth within 3 days of the death of a woman who had undergone fistula repair surgery. EngenderHealth clinical staff or consultants collaborated with clinical and administrative staff at the repair facilities to conduct the audits. Interviews were conducted with the surgical team, facility director, and medical and paramedical staff involved in patient screening and other direct clinical care. If a death occurred after discharge, family members, field staff, and providers who might have seen the patient were also interviewed. The audit team reviewed ward and theater registers, patient records, records of anesthesia and analgesia regimen, and diagnostic investigations, in addition to the chronology of complications and their evolution. The team also reviewed referral records if the patient was treated elsewhere, and autopsy findings if available. The data collection form for the investigation was not designed with pre-coded response options; most questions were open-ended. Lastly, the audit team conducted on-site observations to identify medical quality issues that might have contributed to death. The audit was designed to be facilitative and constructive, focusing on potential systemic flaws directly or indirectly related to the death, and the approach emphasized diplomacy, sensitivity, and confidentiality. The objectives were to determine the cause of death, identify contributing factors, and ascertain whether death was attributable to the procedure. The findings were also used to determine whether the death was preventable and to design remedial actions. All information gathered during the audits was confidential. The names of the deceased and of service providers were excluded from reports; electronic and hardcopy reports were securely filed. The denominator for the present analysis was the number of fistula repairs reported over a 9-year study period from the supported sites.
Wingo et al. 1985 [11]
Loft et al. 1991 [12]
Kafy et al. 2006 [13]
McPherson et al. 2004 [14]
American Urological Association, Clinical Guidelines Panel, 1997 [15]
The results are reported as descriptive statistics (total numbers and percentages). There was no statistical analysis. 3. Results The 44 supported fistula repair sites reported 26 060 repairs over the 9-year period. The supported sites did not report to EngenderHealth the number of women returning for postoperative follow-up. Twenty seven repair facilities were public institutions; the others were private (n = 5) or faith-based (n = 12). Repair sites were located in both urban and rural settings. The duration of EngenderHealth’s support to these sites ranged from 1 to 8 years. Thirty deaths were reported across the 44 study sites. After clinical investigation, 21 deaths were deemed likely to be attributable to the procedure. In seven cases, there was no history or examination or laboratory findings indicative of likely attribution (e.g., on the basis of the sequence of clinical events or the time between surgery and death); there was a definitive other cause of death; or there were contrary autopsy findings. There was insufficient information to make a determination in two cases. The attributable case fatality from fistula surgery was 0.08 per 100 procedures. The gross case fatality rate, inclusive of non-attributable cases, was 0.12 per 100 procedures. Deaths were reported in 14 repair sites located in seven of the 13 countries. Across the sites reporting death, yearly mortality ranged from 0.00 to 0.36 per 100 procedures (Table 2). The number of reported cases per country ranged from zero to seven, or 0.00–0.29 per 100 procedures (Table 3). Age, which was reported for 15 of the 21 women, ranged from 17 to 65 years (median 25 years). None of the reported deaths occurred during the postpartum period. The duration of post-procedure survival, noted for 17 women, ranged from 1 to 32 days (median 10 days). No pattern was discernible by country. The complexity of fistula repair surgery, as assessed by the surgeons, varied. Seven women underwent simple fistula repair via the vaginal route; two women underwent trans-vaginal repairs of medium/ intermediate complexity; and six repairs were complex. Among the complex repairs, four women required adjunct or complementary surgery for separate or concurrent rectovaginal fistula and/or colostomy or other procedures: three of these repairs were done via the abdominal route, and one was done via a combined route. For six women, there were insufficient data to determine repair complexity. Spinal anesthesia was used exclusively in 11 of the 21 cases. General anesthesia was used in three cases, and “other” or combined (i.e., sedation, ketamine, or caudal epidural) was used in three cases. The type of anesthesia was not recorded in the clinical records of four women. No death was reported as directly attributable to anesthesia, but in three
Please cite this article as: Ruminjo J, et al, Mortality risk associated with surgical treatment of female genital fistula, Int J Gynecol Obstet (2014), http://dx.doi.org/10.1016/j.ijgo.2014.02.015
J. Ruminjo et al. / International Journal of Gynecology and Obstetrics xxx (2014) xxx–xxx Table 2 Fistula repair attributable mortality by year. Project year
No. of deaths
No. of repairs
Case fatality per 100 procedures
January 2005–September 2005 October 2005–September 2006 October 2006–September 2007 October 2007–September 2008 October 2008–September 2009 October 2009–September 2010 October 2010–September 2011 October 2011–September 2012 October 2012–March 2013 Total
0 3 1 3 1 3 5 4 1 21
141 846 2967 3150 3278 3871 4225 4759 2823 26 060
0.00 0.36 0.03 0.10 0.03 0.08 0.12 0.08 0.04 0.08
cases the women had unexplained perioperative convulsions; it is unclear whether these convulsions might have been related to anesthesia, electrolyte imbalance, or some other cause. The fistula surgery skills of the 17 principal surgeons for the 21 cases varied. Seven surgeons had competency for complex fistula surgery, three for medium/intermediate complex fistula surgery, and one for simple fistula surgery. For six surgeons, the skill level could not be ascertained. The primary cause of death in nearly 50% of cases (10/21) was related to manifestations of sepsis and inflammation (septicemia and peritonitis) (Table 4). Secondary causes of death were also given for some women, the most frequent being shock and/or hypotension. The audit uncovered contributing or underlying factors, some of which were specific and others that were general or systemic. The most frequent systemic factors were a lack of personnel to attend emergencies, emergency equipment, specific medications, and consumables. Surgeons did not always act on cautionary red flags in the patient history, physical examination, or diagnostic investigation. Postoperative care and documentation of new symptoms, signs, and differential diagnoses by medical and nursing staff were sometimes inadequate. At times, there was poor “handover” and continuity of care when “visiting surgeons” performed the surgery but did not leave documentation or adequate instructions for postoperative management by the resident team. Strength and certainty of diagnosis varied according to investigative routines and rigor, which often depended on available resources. In 14 cases, diagnosis was classified as presumptive rather than definitive. This was due to the lack or imprecision of available diagnostic investigations or documentation, and thus a greater reliance on patient history and a series of physical examinations conducted as part of routine care. Routine laboratory investigations carried out in the pre- and postoperative periods ranged from basic to extensive for 15 women. Investigations were not done for two women, and were unknown for four women. Laboratory investigations usually included a hemoglobin estimation and hematocrit levels, and occasionally a typhoid and/or Widal test, blood sugar levels, and stool microscopy. Less frequently reported tests included urea or electrolytes, blood creatinine levels, and HIV tests. Only a few cases included urine tests or blood cultures (positive in four cases); contrast urinary tract radiology or ultrasound testing was rarely carried out. An autopsy and/or histologic examination was conducted in only two cases: neither death was attributable to the repair (one woman suffered hemorrhagic shock from portal hypertension and bleeding esophageal varices; the other had metastatic brain tumors from an intestinal primary). Records stated definitively that autopsies were not conducted for 14 of the attributable cases. The remaining seven cases had no definitive statement about autopsy and no available results. It is likely that autopsy was not undertaken owing to cultural and/or religious prohibitions. In 9 of the 21 cases, the death was subjectively assessed as likely to have been preventable with available resources. In five cases, the death was probably not preventable. Data were insufficient to make a
3
determination in seven cases. Preventable circumstances included performing an essentially elective surgery on a woman whose condition was already compromised; not paying due diligence to medical history, including previous investigations and medications; and, arguably, not ending an operation that was becoming progressively more complicated and hazardous. 4. Discussion The present study is one of the largest documented prospective, confidential inquiries into mortality risk for surgical treatment of female genital fistula. It provides a detailed assessment of the clinical and contextual risk from multiple country programs and several fistula repair sites over a wide geographic area and under varied economic and organizational conditions. The data indicate that mortality risk from surgical treatment of female genital fistula is within the same range as comparable gynecologic operations in high-resource settings, despite the severe constraints under which fistula surgical teams work in low-resource environments [11–16]. Detailed data for the mortality risk of fistula repair are scarce. In a series of more than 800 fistula repairs, Hancock and Browning [8] identified the likely causes of death as embolism and hyponatremia, but they did not cite specific case fatality rates. Similarly, Waaldijk [9], who reviewed more than 5000 fistula procedures under spinal anesthesia, did not cite a specific case fatality rate. His series included two deaths (from total spinal block and meningitis), but the case of meningitis might have been unrelated to the procedure. In a subsequent personal communication, Waaldijk was quoted as having found a 2% mortality rate, declining to 0.4% over the years [5]. More detailed data have come from a “surgical camp” or outreach service activity termed “Fistula Fortnight” for 569 women at four hospitals in northern Nigeria [10]. This program was deemed successful in that it strengthened clinical skills, provided medical equipment, and raised public awareness about treatment options. Four women, aged 17–42 years, died postoperatively on days 6, 13, 19, and 32, yielding a case fatality rate of 0.7%. Autopsies were not performed owing to cultural restrictions. Although this rate is almost nine times that found in the current series, the deaths were reportedly from malaria, hypoglycemic coma, hypertension, and leukemia, and might not have been directly attributable to the procedure. Nevertheless, that particular service model raised issues about the quality of perioperative screening, care, and follow-up, especially if no fistula surgeon is resident on-site during the woman’s postoperative recovery period. In the present audit, no post-repair deaths were deemed attributable to anesthesia or electrolyte imbalance. However, the three cases of unexplained perioperative convulsions would require diagnostics to rule out anesthetic complications and/or fluid or electrolyte imbalance; these tests are not available at most sites in the resource-poor settings. The primary cause of death for nearly half of the cases was infection and/or inflammation. Arrowsmith et al. [17] reported concern about antibiotic regimens that do not address emerging multi-resistant bacteria, as did Gordon Williams in a presentation at the International Society of Obstetric Fistula Surgeons (ISOFS 3rd Annual Meeting, Dakar, Senegal, October 7-9, 2010). Only a few women in the present series received blood culture and sensitivity testing for septicemia but, consistent with these concerns, their samples contained microbes that are resistant to many common antibiotics. The audits identified several systemic issues where recommendations for improvement are necessary. Although infection or inflammation sequelae were noted as the primary cause of death, contributory causes were also reported including anemia, malaria, dehydration, malnutrition, and other chronic morbidities. Preoperative screening, selection, and assessment of when the woman would be in the best condition to undergo surgery needs improvement and should be resolved before such elective surgery is performed. Routine use of a standard
Please cite this article as: Ruminjo J, et al, Mortality risk associated with surgical treatment of female genital fistula, Int J Gynecol Obstet (2014), http://dx.doi.org/10.1016/j.ijgo.2014.02.015
4
J. Ruminjo et al. / International Journal of Gynecology and Obstetrics xxx (2014) xxx–xxx
Table 3 Attributable case fatalities per 100 procedures by country and project year (October–September). Country Bangladesh No. of sites reporting No. of repairs No. of deaths Case fatality rate, % Beninc No. of sites reporting No. of repairs No. of deaths Case fatality rate, % DRCd No. of sites reporting No. of repairs No. of deaths Case fatality rate, % Ghanac No. of sites reporting No. of repairs No. of deaths Case fatality rate, % Guinea No. of sites reporting No. of repairs No. of deaths Case fatality rate, % Liberiac No. of sites reporting No. of repairs No. of deaths Case fatality rate, % Mali No. of sites reporting No. of repairs No. of deaths Case fatality rate, % Niger No. of sites reporting No. of repairs No. of deaths Case fatality rate, % Nigeria No. of sites reporting No. of repairs No. of deaths Case fatality rate, % Rwanda No. of sites reporting No. of repairs No. of deaths Case fatality rate, % Sierra Leone No. of sites reporting No. of repairs No. of deaths Case fatality rate, % Togoc No. of sites reporting No. of repairs No. of deaths Case fatality rate, % Uganda No. of sites reporting No. of repairs No. of deaths Case fatality rate, % Total No. of sites reporting Total no. of repairs Total no. of deaths Case fatality rate, %
2004–2005 a
2005–2006
2006–2007
2007–2008
2008–2009
2009–2010
2010–2011
2011–2012
2012–2013 b
Total
3 20 0 0.00
3 93 1 1.08
3 119 0 0.00
3 122 0 0.00
4 131 0 0.00
4 143 0 0.00
4 150 0 0.00
4 184 2 1.10
4 80 0 0.00
1042 3 0.29
– NS NS NS
1 NS NS NS
1 NS NS NS
– NS NS NS
– 110 0 0.00
1 21 0 0.00
1 20 0 0
– NS NS NS
– NS NS NS
151 0 0.00
0 NS NS NS
1 53 0 0.00
2 586 1 0.17
2 334 0 0.00
2 482 0 0.00
2 472 0 0.00
6 565 1 0.18
6 1221 1 0.08
7 829 0 0.00
4542 3 0.07
0 NS NS NS
1 21 0 0.00
1 42 0 0.00
– NS NS NS
– NS NS NS
– NS NS NS
– NS NS NS
– NS NS NS
– NS NS NS
63 0 0.00
0 NS NS NS
2 199 0 0.00
2 292 0 0.00
3 229 0 0.00
4 316 0 0.00
3 392 0 0.00
3 459 0 0.00
3 497 0 0.00
3 228 0 0.00
2612 0 0.00
– NS NS NS
– NS NS NS
– NS NS NS
1 59 0 0.00
– NS NS NS
– NS NS NS
– NS NS NS
– NS NS NS
– NS NS NS
59 0 0.00
– NS NS NS
– NS NS NS
– NS NS NS
– NS NS NS
1 46 0 0.00
1 40 0 0.00
1 91 0 0.00
1 53 0 0.00
3 125 0 0.00
355 0 0.00
– NS NS NS
– NS NS NS
1 27 0 0.00
4 213 0 0.00
3 158 0 0.00
4 220 2 0.91
4 333 0 0.00
4 209 0 0.00
5 213 0 0.00
1373 2 0.15
– NS NS NS
– NS NS NS
5 1081 2 0.19
5 1437 1 0.07
6 1347 1 0.07
6 1612 1 0.06
7 1507 0 0.00
9 1720 1 0.06
10 917 1 0.11
9621 7 0.07
– NS NS NS
2 145 0 0.00
2 147 0 0.00
2 83 0 0.00
3 167 0 0.00
3 259 0 0.00
3 278 0 0.00
4 114 1 0.88
3 19 0 0.00
1212 1 0.08
– NS NS NS
– NS NS NS
1 272 0 0.00
1 363 1 0.28
1 253 0 0.00
1 166 0 0.00
1 211 0 0.00
1 244 0 0.00
1 73 0 0.00
1582 1 0.06
– NS NS NS
– NS NS NS
– NS NS NS
– NS NS NS
– NS NS NS
1 97 0 0.00
– NS NS NS
– NS NS NS
– NS NS NS
97 0 0.00
1 121 0 0.00
2 335 0 0.00
2 401 0 0.00
2 310 1 0.32
2 268 0 0.00
2 449 0 0.00
2 611 2 0.33
3 517 1 0.19
3 339 0 0.00
3351 4 0.12
4 141 0 0.00
12 846 1 0.12
20 2967 3 0.10
23 3150 3 0.11
26 3278 1 0.03
27 3871 3 0.08
31 4225 5 0.12
35 4759 3 0.06
35 2823 1 0.04
26 060 21 0.08
Abbreviations: DRC, Democratic Republic of Congo; NS, not supported. a January–September 2005. b October 2012–March 2013 c Fistula repair services provided aboard a hospital ship which traveled to four countries (Benin, Ghana, Liberia and Togo). Counted as one site. d Includes repairs reported by the US Agency for International Development (USAID)-funded bilateral project in fiscal years 2005–2006, 2006–2007, and 2007–2008.
Please cite this article as: Ruminjo J, et al, Mortality risk associated with surgical treatment of female genital fistula, Int J Gynecol Obstet (2014), http://dx.doi.org/10.1016/j.ijgo.2014.02.015
J. Ruminjo et al. / International Journal of Gynecology and Obstetrics xxx (2014) xxx–xxx Table 4 Primary and secondary causes of death and contributory factors among the attributable deaths (n = 21).
Inflammation and/or infection Sepsis Septicemia Peritonitis Hematologic Pulmonary embolism Acute anemia Anemia Leukemia Splenic infarction Splenic sequestration Urinary Acute exacerbation of chronic renal failure Renal tubular necrosis/ureteric damage Renal failure Bladder lithiasis Chronic renal failure Uremia Previous urinary diversion Cardiovascular First-degree heart block Hypertension Congestive cardiac failure Hypotension and/or shock Cardiac arrhythmia Hypotension Metabolic Hypoglycemic coma Diabetes Hypokalemia Dehydration Central nervous system Cerebrovascular accident Perioperative convulsions/seizures Epilepsy Respiratory system Status asthmaticus Aspiration and/or pneumonitis Asthma Gastrointestinal system Gastroenteritis/bloody diarrhea Intestinal obstruction Acute abdominal distension Diarrhea Prolapsed colostomy Concurrent rectovaginal fistula Parasitemia Malaria Other Old age/frailty/general weakness Malnutrition Previous multiple surgeries Treated cervical cancer Nonclinical system factors
Primary cause
Secondary cause
Contributing factor
4 4 2
2 – –
– – –
2 1 – – – –
1 – 2 1 – –
– – 1 – 1 1
1 1 – – – – –
– – 1 1 – – –
– – – – 1 1 1
1 1 1 – – –
– – – 4 1 –
– – – – – 1
1 – – –
– 1 1 –
– 1 – 1
1 – –
– 3 –
– – 1
1 – – – – – – – – – – –
– 3 – – 3 1 1 – – – – 2
– – 1 – – – – 2 1 1 – 2
– – – – –
– – – – –
3 1 2 1 13
5
Fistula Care’s program in Nigeria, regular “Provider Network” retreats involve fistula surgeons, nurses, anesthetists, and managers from different facilities (I. Efem, personal communication, April 15, 2013), and fistula mentorship initiatives are taking place in the fistula program in Uganda (R. Mukisa, personal communication, April 15, 2013). The present analysis has some limitations. Clinical records were sometimes incomplete or contained poor quality data, including inconsistent documentation of perioperative observations and follow-up care. Lack of diagnostic resources, including those for autopsy and/or histology, meant that differential diagnoses and, consequently, attribution and preventability were often presumptive. In addition, given the understandably charged emotions surrounding these deaths, interview and recorded data might be inaccurate. Supported sites might have underreported deaths, because investigation requires extra work and (despite assurances to the contrary) providers and administrators may feel defensive or apprehensive about an audit. Related reviews have shown that, although maternal death audit is becoming more common in some low-income countries, facility-level data may underestimate the number of deaths [19,20]. In summary, the case fatality rate for fistula repair surgery in lowincome countries was in the same range as that for comparable gynecologic operations in high-income settings. Clinical and systemic issues must be addressed to reduce the case fatality rate. These include improving perioperative care and follow-up, assuring timely referral or deferral of difficult cases, and maintaining better records. Acknowledgments
checklist has been shown to improve safety in general surgery, and this can be adapted for fistula repair [18]. In addition, the recently launched global fistula training manual from the International Federation of Gynecology and Obstetrics includes guidance on developing competency to conduct audits, which might be useful for future analysis of systemic and clinical issues leading to case fatality [6]. A minimum package of resources is needed for diagnostics, including a quality laboratory, requisite infrastructure, trained technicians, equipment, reagents, and expendable supplies. Even when resources are available, diagnostic investigations might be too expensive. Personnel and physical resources should always be available for early identification and management of emergencies, in addition to better referral systems and more accountable handover between teams. Teamwork is crucial, for both clinical and ethical reasons, especially among visiting surgical and outreach teams and the resident staff. In
The Fistula Care project at EngenderHealth was funded by USAID under associate cooperative agreement GHS-A-00-07-00021-00. Views expressed here do not necessarily reflect those of USAID, the US Government or EngenderHealth. Conflict of interest The authors have no conflicts of interest. References [1] World Health Organization, UNICEF, UNFPA, The World Bank. Trends in maternal mortality: 1990 to 2010. http://www.who.int/reproductivehealth/publications/ monitoring/9789241503631/en/. Published 2012. [2] World Health Organization. The World Health Report 2005 – make every mother and child count. http://www.who.int/whr/2005/en/. Published 2005. [3] Stanton C, Holtz SA, Ahmed S. Challenges in measuring obstetric fistula. Int J Gynecol Obstet 2007;99(Suppl. 1):S4–9. [4] Adler AJ, Ronsmans C, Calvert C, Fillippi V. Estimating the prevalence of obstetric fistula: a systemic review and meta-analysis. BMC Pregnancy Childbirth 2013;13:246. [5] Abrams P, de Ridder D, de Vries C, Elneil S, Esegbona G, Mourad S, et al. Obstetric Fistula in the Developing World. Presented at An International Consultation on Vesicovaginal Fistula. Marrakech, Morocco, October 13–16, 2010. Montreal: Societe Internationale d’Urololgie; 2012. [6] Elneil S. Global competency-based fistula surgery training manual. http:/www.figo. org/publications/miscellaneous_publications/Global_Competency_Based_Fistula_ Surgery_Training_Manual. Published June 2011. [7] Gutman RE, Dodson JL, Mostwin JL. Complications of treatment of obstetric fistula in the developing world: gynatresia, urinary incontinence, and urinary diversion. Int J Gynecol Obstet 2007;99(Supplement 1):S57–64. [8] Hancock B, Browning A. Practical Obstetric Fistula Surgery. London: Royal Society of Medicine Press Limited; 2008. [9] Waaldijk K. Step-by-step Surgery of Vesicovaginal Fistulas: A Full-color Atlas. Cheltenham: Nelson Thornes Limited; 1994. [10] Ramsey K, liyasu Z, Idoko L. Fistula Fortnight: Innovative partnership brings mass treatment and public awareness towards ending obstetric fistula. Int J Gynecol Obstet 2007;99(Supplement 1):S130–6. [11] Wingo PA, Huezo CM, Rubin GL, Ory HW, Peterson HB. The mortality risk associated with hysterectomy. Am J Obstet Gynecol 1985;152(7):803–8. [12] Loft A, Anderson TF, Brønnum-Hansen H, Roepstorff C, Madsen M. Early postoperative mortality following hysterectomy. A Danish population based study, 1977–1981. Br J Obstet Gynaecol 1991;98(2):147–54. [13] Kafy S, Huang JY, Al-Sunaidi M, Wiener D, Tulandi T. Audit of morbidity and mortality rates of 1782 hysterectomies. J Minim Invasive Gynecol 2006;13(1):55–9. [14] McPherson K, Metcalfe MA, Herbert A, Maresh M, Casbard A, Hargreaves J, et al. Severe complications of hysterectomy: the VALUE study. BJOG 2004;111(7):688–94.
Please cite this article as: Ruminjo J, et al, Mortality risk associated with surgical treatment of female genital fistula, Int J Gynecol Obstet (2014), http://dx.doi.org/10.1016/j.ijgo.2014.02.015
6
J. Ruminjo et al. / International Journal of Gynecology and Obstetrics xxx (2014) xxx–xxx
[15] The American Urological Association, Female Stress Urinary Incontinence Clinical Guidelines Panel. Report on The Surgical Management of Female Stress Urinary Incontinence. www.auanet.org/common/pdf/education/Arc-SUI.pdf. Published 2007. [16] Drutz HP, Herschorn S, Diamant NE. Female Pelvic Medicine and Reconstructive Pelvic Surgery. Berlin: Springer Verlag Gmbh; 2003. [17] Arrowsmith SD, Ruminjo J, Landry EG. Current practices in treatment of female genital fistula: A cross-sectional study. BMC Pregnancy Childbirth 2010;10:73. [18] World Health Organization. Reproducing and Adapting the WHO Surgical Safety Checklist. http://www.who.int/patientsafety/safesurgery/using_checklist/en/. Accessed December 30, 2013.
[19] Graham WJ, Ahmed S, Stanton C, Abou-Zahr CL, Campbell OMR. Measuring maternal mortality: an overview of opportunities and options for developing countries. BMC Med 2008;6:12. [20] Immpact, University of Aberdeen. Immpact Toolkit: a guide and tools for maternal mortality programme assessment: Rapid Ascertainment Process for Institutional Deaths (RAPID). Module 4, Tool 2. Version 2.0 http://www.abdn.ac.uk/iahs/contentimages/rapid.pdf. Published 2007.
Please cite this article as: Ruminjo J, et al, Mortality risk associated with surgical treatment of female genital fistula, Int J Gynecol Obstet (2014), http://dx.doi.org/10.1016/j.ijgo.2014.02.015