Burden of Surgical Infections in a Tertiary-Care Pediatric Surgery Service in Uganda

Abstract

Background:

Delayed presentation of surgical disease often leads to infection in low- and middle-income countries (LMICs). In addition, many primary infections require surgical intervention. The burden of infection in children's surgery in LMICs is poorly defined and may tax the limited availability of surgical resources.

Methods:

A prospective surgical database was reviewed for all children presenting to a Ugandan tertiary referral hospital from January 2012 to August 2016. All patients presenting with infection were included and analyzed by operative intervention and survival.

Results:

Of the 3,494 children admitted over the time period, 712 (20.4%) presented with infection. A total of 455 patients (64%) with an infection underwent an operation, with an in-hospital mortality rate of 12.5%. Operations involving infections represented 20% of the volume of the children's surgery department. Common conditions were abscesses (n = 308; 43.4%), typhoid intestinal perforations (n = 85; 12.0%), appendicitis (n = 78; 11.0%) and perforated bowel caused by ileocolic intussusception (n = 37; 5.2%). Patients with esophageal atresia presenting with aspiration pneumonia had an in-hospital mortality rate of 78.6%, those with abdominal sepsis a 67% mortality rate, and neonatal infants with necrotizing enterocolitis a 50% mortality rate.

Conclusion:

There is a high volume of infection in children requiring surgery, contributing to a high mortality rate. Resource allocation for children's surgical care in LMIC should be directed toward timely diagnosis and surgical intervention of these conditions.

The provision of surgical services historically has been minimized during the formation of national health plans in low- and middle-income countries (LMICs). The Lancet Commission on Global Surgery and the Disease Control Priorities-3 project highlighted the significant burden of surgical disease as well as the economic impact caused by a lack of surgical access, which disproportionally affects LMICs [1,2]. These investigations led, in part, to the World Health Assembly adopting resolution A68.15: Strengthening emergency and essential surgical and anesthesia care as a component of universal health coverage in 2015 and a new global focus on developing National Surgical, Obstetric and Anesthesia Plans in LMICs [3 -5].

Recent estimates suggest that 1.7 billion children around the world are without access to safe surgical care [6]. Despite the recent focus on surgery in national health plans, children's surgery continues to be considered by policy stakeholders to be both non-essential and too expensive to include in the development of health programs [7,8]. This has led to significant resource limitations, including a lack of dedicated equipment and supplies for children's surgery in LMICs [9] and an insufficient number of pediatric surgeons [10,11].

In addition to limited surgical capacity, a persistent gap remains in the epidemiology of children's surgical disease in LMICs as well as associated operative volumes and mortality rates [12]. Furthermore, evidence suggests that emergency operations including for trauma/burns and infections utilize the majority of the operative volume, leaving less available for complex elective cases [13 -15].

To better quantify the unmet need of children's surgery in LMICs, we assessed the impact of infection on operative volume and death in a children's surgery division in Kampala, Uganda. Children at that hospital often present with primary infections that require operative intervention, such as abscesses, appendicitis, or intestinal perforations caused by typhoid fever. In addition, infections can complicate surgical diseases that go undiagnosed or untreated, such as intestinal perforation as a consequence of ileocolic intussusception, aspiration pneumonia resulting from improperly diagnosed or treated esophageal atresia, or abdominal sepsis linked to gastroschisis.

Infection represents an important subset of surgical epidemiology, as infections often necessitate emergency procedures, which can tax a small surgical workforce with limited resources. We hypothesized that a large fraction of the total operative volume performed by pediatric surgeons involves infection. Furthermore, we suspect that mortality rates for certain conditions such as abdominal and neonatal sepsis are significant, given the limitations on intensive care unit (ICU) resources in Uganda.

Patients and Methods

The study was approved by the Institutional Review Board at Mulago Hospital in Kampala, Uganda (Protocol No. MREC: 464) and at Yale University School of Medicine (Protocol No. 1605017844). Mulago Hospital is a tertiary-care facility and the national referral hospital in Uganda. At the time of this study, it was one of only two hospitals in the country that offered surgical care by specialty-trained pediatric surgeons. A prospective surgical database is maintained on all patients treated by the pediatric surgical team at the hospital, and this database was reviewed for admissions between January 2012 and August 2016. All children less than 14 years old who were admitted to the children's surgical ward or treated in consultation by the pediatric surgery team with a diagnosis including infection were included in the analysis. Diagnoses determined to represent infections were selected, a priori, to include primary surgical infections such as abscesses and appendicitis, as well as surgical pathology complicated by infections such as intussusception associated with intestinal perforation, esophageal atresia linked to pneumonia, and infections requiring surgery such as typhoid complicated by intestinal perforation. We specifically used these inclusive definitions of infection to demonstrate the operative and clinical burden of infections for the pediatric surgery team.

Conditions treated by other subspecialists at Mulago Hospital, including osteomyelitis (orthopedic surgery) and infectious complications following burns (plastic surgery), will be under-represented in this investigation, as these patients are seen on other units and will not be captured in the database. Patients treated for medical infections such as typhoid disease were included only if they were evaluated by the surgical team.

Patients were excluded from the analysis if there were insufficient clinical or operative details in the database. Individual patient charts and operative logs were not reviewed for this investigation. Demographic and clinical data represented in the database include age, gender, travel distance to the hospital, referral status from district hospitals, admission diagnosis, final diagnosis, surgical intervention, procedure details, operative complications, and in-hospital death. The primary outcomes were infections at admission and in-hospital mortality rate divided into neonatal and pediatric conditions. Neonatal conditions were defined as common surgical pathology diagnosed in infants under 28 days such as gastroschisis or esophageal atresia. Pediatric conditions were defined as surgical pathology that occurs in children of all ages, including appendicitis, abscess, and ileal perforation from typhoid disease. Subgroup analysis was conducted on patients presenting with abdominal sepsis.

For patients presenting with an infection, we conducted descriptive analysis on the basis of the patient's confirmed diagnosis and in-patient survival. We conducted a subgroup analysis of patients presenting with abdominal sepsis. In our database, sepsis was defined by the consultant pediatric surgeon at the time of initial evaluation and was listed in the clinical diagnosis. We performed a descriptive analysis of confirmed diagnosis and in-patient death for patients presenting with abdominal sepsis. We further carried out a univariable analysis to investigate factors influencing survival in patients presenting with abdominal sepsis. We evaluated demographic predictors, age and gender, as well as factors that may affect time to receiving definitive care, including distance to the hospital, geo-political region, and referral from a different hospital.

A p value <0.05 was considered significant for all analyses. Data collection was performed using Microsoft Access 2016 (Redmond, WA); data analysis was performed using SAS v. 9.4 (Cary, NC).

Results

There were 3,494 patients treated by the pediatric surgery team over the time period. Sixteen patients were excluded from analysis because of insufficient data regarding initial diagnosis (four patients), operative details (six patients), or clinical outcomes (six patients). Of the 3,478 patients included in the study, 712 (20.4%) were admitted or treated with an infection (Table 1 for pediatric conditions; Table 2 for neonatal conditions). The total in-hospital mortality rate of the patients presenting with an infection was 12.5% (n = 89).

Table 1.

Epidemiology, Operative Volume, and Mortality Rate of Surgical Infections in Pediatric Population

Surgical condition presenting with infection	N (%)	Operations (%)	Post-operative deaths (%)	In-hospital deaths (%)
Abscess (all locations) and cellulitis	308 (43.4)	197 (64.6)	2 (1.0)	3 (1.0)
Ileal perforation (typhoid disease)	85 (12.0)	76 (89.4)	11 (14.5)	13 (15.3)
Appendicitis	78 (11.0)	70 (89.7)	0	0
Intussusception (perforated)	37 (5.2)	33 (89.2)	11 (33.3)	14 (37.8)
Necrotizing fasciitis	29 (4.1)	12 (41.4)	1 (8.3)	3 (10.3)
Primary peritonitis	26 (3.7)	19 (73.1)	1 (5.3)	2 (7.7)
Surgical site infection	18 (2.5)	4 (22.2)	0	1 (5.5)
Pneumonia (presenting symptom in primary surgery patient)	10 (1.4)	2 (20)	2 (100)	2 (20)

Other infections not listed because of low incidence (<0.5%): Phycomycosis and other invasive fungal disease, thoracic empyema, cervical adenitis, parotitis, genital warts, cholangitis, orchitis, recurrent urinary tract infection from congenital urogenital anomalies, measles, enteric colitis; pneumonia category includes patients with intestinal obstruction leading to aspiration pneumonia as presenting symptom or recurrent bronchial pneumonia from thoracic surgical pathology. Post-operative mortality rate is rate among patients who underwent surgery for their condition.

Table 2.

Epidemiology, Operative Volume, and Mortality Rate of Surgical Infections in Neonatal Population

Surgical condition presenting with infection	N (%)	Operations (%)	Post-operative deaths (%)	In-hospital deaths (%)
Necrotizing enterocolitis	19 (2.7)	5 (26.3)	3 (60)	9 (47.4)
Gastroschisis	19 (2.7)	3 (15.8)	3 (100)	19 (100)
Tracheoesophageal fistula (aspiration pneumonia)	14 (1.9)	6 (42.9)	3 (50)	11 (78.6)
Omphalocele	7 (1.0)	0	0	3 (42.9)
Intestinal anomaly	6 (0.8)	3 (50)	0	1 (16.7)
Neonatal sepsis (NOS)	3 (0.4)	1 (33.3)	0	1 (33.3)

Neonatal sepsis not otherwise specified (NOS) causes one each of pneumonia with concern for tracheoesophageal fistula, cervical adenitis, and urogenital anomaly. Gastroschisis, omphalocele, and intestinal anomalies represent abdominal infections. Post-operative mortality rate is rate among patients who underwent surgery for their condition.

A total of 455 patients (64%) underwent a procedure. Operations for infections represented 20% of the total operative volume of the children's surgery department at Mulago Hospital over the study period. Slightly less than half (47.1%; n = 335) of the patients included in this analysis were female. The median age was 2 years (interquartile range [IQR] 0.24–6.9 years), likely reflecting the high referral rate of neonatal and older infants to Mulago Hospital. Patients presented from 51 districts across Uganda (43.3% from Kampala District) and a median distance to the hospital of 20 km (IQR 10–50 km). Eleven patients (1.5%) were human immunodeficiency virus (HIV) positive.

Pediatric conditions

There were 308 patients (43.4%) presenting with either cellulitis or an abscess: 41 with cellulitis, 208 with a superficial abscess (35 rectal or gluteal), and 62 with a deep abscess (31 pyomyositis, 21 psoas, 8 liver, 2 intra-abdominal not otherwise specified). A total of 85 patients (12.0%) presented with ileal perforation as a result of typhoid disease. A series of 78 patients (11.0%) presented with appendicitis: 52 with uncomplicated appendicitis, 14 with perforated appendicitis without abscess, and 12 with perforated appendicitis and abscess. Thirty-seven patients (5.2%) presented with intestinal perforation secondary to prolonged ileocolic intussusception. A total of 29 patients (4.1%) had necrotizing fasciitis, and 26 patients (3.7%) had primary peritonitis in which there was purulent material in the abdomen at the time of operation, but no perforation or infection source was located or identified. Twenty three of 26 patients (88.5%) with primary peritonitis were female, and two were HIV positive. Another 18 patients (2.5%) presented with a surgical site infection as a referral from an outside hospital. The original operations were laparotomy, herniorrhaphy, and circumcision. There also were 10 patients who presented with a primary diagnosis of pneumonia caused by or being evaluated for surgical disease. This includes eight patients with suspected aspiration pneumonia caused by intestinal obstruction, one patient with recurrent pneumonia associated with aspiration of a foreign object and formation of a trachea-esophageal fistula, and one patient with recurrent pneumonia who was found to have a congenital diaphragmatic hernia.

Neonatal conditions

Nineteen neonatal patients (2.7%) were treated for necrotizing enterocolitis. Another 19 (2.7%) presented with gastroschisis complicated by abdominal sepsis. A total of 154 patients were admitted with gastroschisis over the study period, meaning 12% of all gastroschisis admissions were complicated by abdominal sepsis. Similarly, 14 patients (1.9%) were admitted with esophageal atresia with or without a tracheoesophageal fistula that was complicated by aspiration pneumonia. A total of 36 patients presented with esophageal atresia (38.9% of the patients presented with pneumonia).

Operative and Mortality Rates

The overall operative mortality rate for patients presenting with an infection was 8.5%, whereas the overall operative mortality rate for all patients at Mulago Hospital was 10.1%. The highest operative rates were for infections arising from ileal perforations associated with typhoid disease (89.4%), intestinal perforation from intussusception (89.2%), and appendicitis (89.7%). The mortality rate for perforated intussusception was 37.8% and that for non-perforated intussusception was 19.9%. The highest mortality rates were caused by the neonatal conditions necrotizing enterocolitis (47.4%) and esophageal atresia (78.6%). All patients presenting with abdominal sepsis from untreated gastroschisis died. However, gastroschisis is a highly mortal disease in East Africa regardless of presenting symptoms (>90% in Uganda) because of a lack of ICU services and total parenteral nutrition) [16].

Forty-nine patients presented to the hospital with clinically recorded abdominal sepsis (Table 3). This includes the previously reported 19 patients with gastroschisis, eight patients who had bowel perforation from intussusception, seven patients with abdominal sepsis from omphalocele, six patients with abdominal sepsis without an additional primary diagnosis, and five patients with ileal perforation from typhoid. The all-cause mortality rate for patients presenting with abdominal sepsis was 67.3%. As gastroschisis is highly lethal regardless of presenting symptoms, we also calculated the overall mortality rate for abdominal sepsis excluding gastroschisis, which was 46.7% (14/30). Results from univariable analysis demonstrated that younger age was significantly correlated with in-hospital death (p = 0.0002). The median age of the children who died was three days (IQR 2–118 days) while the children who lived had a median age of 1 year (IQR 0.38–6.98 years) at the time of diagnosis. Distance to the hospital (p = 0.105), referral from an outside hospital (p = 0.622), and geopolitical region (p = 0.424) were not statistically significant factors.

Table 3.

Primary Diagnosis, Operative Volume, and Mortality Rate Associated with Abdominal Sepsis in Neonatal and Pediatric Population

Surgical condition presenting with infection	N (%)	Operations (%)	Post-operative deaths (%)	In-hospital deaths (%)
Gastroschisis	19 (39)	3 (15.8)	3 (100)	19 (100)
Intussusception (perforation)	8 (16)	8 (100)	5 (62.5)	5 (62.5)
Omphalocele	7 (14)	0	–	3 (42.9)
Abdominal sepsis (NOS)	6 (12)	2 (33.3)	0	1 (16.7)
Ileal perforation from typhoid disease	5 (10)	4 (80)	2 (50)	3 (60)
Umbilical hernia (perforation)	1 (2)	1 (100)	0	0
Ileal atresia (perforation)	1 (2)	1 (100)	1 (100)	1 (100)
Appendicitis	1 (2)	1 (100)	0	0
Primary peritonitis	1 (2)	1 (100)	1 (100)	1 (100)
Total	49	21 (42.9)	12 (57.1)	33 (67.3)

Post-operative mortality is rate among patients who underwent surgery for their condition.

NOS = not otherwise specified.

Discussion

High operative volume dedicated to infections

More than 20% of the operative volume in the children's surgery department over the five-year period of this study was dedicated to infections. This is concordant with our hypothesis that a large fraction of the case volume involves emergency cases. There were clear patterns explaining why infections may be more prevalent in a pediatric surgery practice in Uganda than in high-income countries. One cause is the higher incidence of endemic infectious disease. Over the study period, there were more typhoid-induced intestinal perforations (n = 85) than cases of appendicitis (n = 78). Another cause is the late presentation of surgical disease resulting in infection, as discussed below. Our findings are similar to those of previous studies describing pediatric operative volumes in LMICs and the burden of infections on surgical capacity. Over a two-year period in a tertiary hospital in Gambia, Bickler et al. found that 14.5% of the operative volume was dedicated to pediatric surgical infections [13]. In a Nigerian teaching hospital over a five-year period, 22.6% of the total admitting diagnoses were pediatric surgical infections [14].

Late presentation of surgical disease results in infection

The clinical reality in Uganda is that patient-related and system-related resource constraints lead to late presentation of disease and delayed time to definitive care. Examples demonstrated in this investigation include the 12% of patients born with gastroschisis who presented with abdominal sepsis. With no abdominal silos or sterile bowel bags in rural hospitals, and a lack of emphasis on continuing education for rural providers, babies present to Mulago several days after birth with exposed bowel, severe dehydration, and sepsis [16]. Gastroschisis is a highly lethal disease in East Africa because of a lack of ICU services and total parenteral nutrition [16], although recent nursing-based interventions have improved the survival rate significantly to roughly 50% [18]. However, children with gastroschisis who present with abdominal sepsis have a 100% mortality rate. A similar pattern was demonstrated in esophageal atresia, where 39% of children presenting with the condition over the study period had it complicated by pneumonia.

Another representative example of late presentation of surgical disease complicated by infection is ileocolic intussusception. In our study, 16% of children (37/231) had perforated bowel at the time of surgery. Hospitals in Uganda do not have the capacity to perform pneumatic reduction; therefore, all ileocolic intussusception is treated surgically. Many children are misdiagnosed as having other causes of abdominal pain, such as malaria; and rural hospitals lack radiology infrastructure, including ultrasound scanners and fluoroscopy. Although other groups have established ultrasound-guided hydro-reduction in East Africa [19], the current resource gaps in Uganda contribute to delays in diagnosis and referral. Qualitatively, a significant portion of the children seen at Mulago Hospital who have intussusception will present with a history of more than a week of abdominal pain.

Over the study period, 49 patients presented to Mulago Hospital with a clinical diagnosis of abdominal sepsis. The underlying pathology also represented late presentation of surgical disease. The causes of abdominal sepsis included perforated bowel caused by typhoid disease, intussusception, anorectal malformation, umbilical hernia, ileal atresia, and appendicitis. Our a priori hypothesis was that surrogate markers for delays in receiving care (distance to hospital, transfer from an outside hospital, geopolitical region) would lead to high mortality rates in children with abdominal sepsis. However, these predictors did not have a statistically significant association with in-hospital survival. This result may be biased by the likelihood that many children are dying before reaching Mulago Hospital or are not being transferred at all. The number of cases of abdominal sepsis also is likely to be significantly underestimated in our study by relying on the clinical diagnosis recorded in the database. This reduced the power of our investigation. However, we chose to be conservative with our analysis, given the inherent limitations of the database.

Lack of ICU resources may contribute to high mortality rates associated with infections

There is no neonatal ICU in Mulago Hospital. There are four pediatric ICU beds, but without ventilators and other necessary critical care resources. Although we were unable to evaluate directly how a lack of intensive care resources affected the mortality rate from infections, several findings from this study correlate with under-resourced ICUs. The mortality rate for infections that require significant medical management was high. For example, the mortality rate from necrotizing enterocolitis was nearly 50%. Similarly, for patients presenting with perforated bowel from intussusception, the mortality rate was almost 40%. The in-hospital mortality rate for abdominal sepsis was greater than 60%. The inability to ventilate children in an ICU setting consistently contributed to a greater than 70% mortality rate in patients with esophageal atresia presenting with pneumonia. Other resource limitations such as inadequate operating room access and other well-described delays to care prevent us from correlating these data directly with a lack of ICU care. However, qualitatively, the lack of ICU resources dramatically affects clinical and surgical decision-making in patients with complex intra-abdominal infections every day at Mulago. Our high mortality rates from abdominal sepsis are supported in the literature. The SPROUT study was a global cross-sectional, point prevalence study across multiple pediatric ICUs to determine epidemiologic and clinical outcomes associated with severe sepsis. In the Africa region of that study, the overall mortality rate from all-cause sepsis in pediatric ICUs was 40% [20].

The way forward

There currently is one dedicated operating room available to the pediatric surgeons at Mulago Hospital. When surgical emergencies occur, this by necessity bumps elective cases, including complex congenital repairs and oncologic resections. This is especially critical at Mulago, as it is the national referral hospital of the country. By recent international children's surgery guidelines, Mulago Hospital is one of two hospitals in the country capable of performing complex operations [21]. The fact that 20% of the operations performed at Mulago are dedicated to conditions involving infections has greater implications for the children's surgery health system in the country. These data should help inform pediatric surgical health care development. Pediatric operating rooms and appropriate clinical staff should be reserved for necessary emergency cases to prevent a backlog of elective cases.

There are current solutions in progress in Uganda. Recent work by the U.K.-based non-governmental organization Archie Foundation and now KidsOR has increased the number of pediatric operating rooms significantly and has demonstrated that a single dedicated pediatric operating room in Uganda can avert 6,447 disability-adjusted life years [22]. In addition, there has been a recent expansion of a pediatric surgery fellowship at Mulago National Referral Hospital, which has increased the number of pediatric surgeons in the country greatly. A recent pediatric surgical emergency course has been implemented for general surgeons operating in regional district-level hospitals in several parts of the country, which may ultimately decrease the referral of simple surgical infections such as abscess, appendicitis, and small bowel resections to Mulago and may improve the initial care at these hospital prior to transport [23,24]. Furthermore, Mulago Hospital is undergoing major renovations, and a new Women's and Children's Hospital has recently opened next to Mulago Hospital, both of which will increase operative space and capacity for children.

Limitations

Our study has the inherent limitations of any single-institution study. Furthermore, this prospective surgical database was designed to capture basic information regarding diagnosis, operation, and operative outcomes. It does not include clinical details such as vital signs, laboratory values, concomitant medical conditions, or non-surgical treatment received such as blood transfusions or duration of symptoms. We are unable to risk stratify patients from the database retrospectively and so cannot comment on why, in some cases, patients do not receive surgery for a common pathology. Local studies have demonstrated the inaccuracy of medical records as a source of such information [25]. These clinical data are not collected and recorded routinely in a meaningful way for any patients at Mulago Hospital, and our database and investigation reflect this reality. All definitions, including sepsis, must be stated clearly in the initial clinical encounter notes as defined by the admitting consultant pediatric surgeon. For example, although 85 patients presented with perforated bowel attributable to typhoid infection, only five were recorded as septic at the time of evaluation. There also is no hospital-level database at Mulago to estimate operative rates of medical disease. For example, it is not possible to obtain the total number of patients admitted with typhoid disease (usually treated medically) to generate an overall operative rate for the disease. Lastly, our study concentrates on the burden of surgical infections treated by a pediatric surgery team at a tertiary referral hospital. As such, our data do not include pediatric infections treated by surgeons who operate on adults in this area, and thus likely under-estimates the total burden of surgical infections in children in the region.

Despite these limitations, the prospective nature of this combined clinical and operative database remains informative about children's surgical epidemiology in Uganda and provides details on the workload of a pediatric surgery team at Mulago Hospital. Future data collection will attempt to improve the quality and especially to adjust risk accurately, given the limitations of the environment.

Conclusion

The burden of infections in pediatric surgery departments in LMICs is significant. At national referral hospitals, with large referral bases for complex cases including congenital anomalies and cancer operations, increasing surgical resources to account for the significant burden of infections should be considered seriously.

Footnotes

Financial Information

There is no financial support to disclose for this research.

Author Disclosure Statement

All authors associated with this work have no actual or potential competing financial interests to disclose.

References

Meara

, Leather

, Hagander

, et al. Global Surgery 2030: Evidence and solutions for achieving health, welfare, and economic development. Lancet, 2015; 386:569–624.

Mock

, Donkor

, Gawande

, et al. Essential surgery: Key messages from Disease Control Priorities, 3rd edition. Lancet, 2015; 385:2209–2219.

World Health Organization: Strengthening emergency and essential surgical care and anesthesia as a component of universal health coverage. Provisional Agenda Item 51 2015. Accessed June 1, 2019.

Albutt

, Sonderman

, Citron

, et al. Healthcare leaders develop strategies for expanding national surgical, obstetric, and anaesthesia plans in WHO AFRO and EMRO regions. World J Surg, 2019; 43:360–367.

Sonderman

, Citron

, Meara

. National surgical, obstetric, and anesthesia planning in the context of global surgery: The way forward. JAMA Surg, 2018; 153:959–960.

Mullapudi

, Grabski

, Ameh

, et al. Estimates of number of children and adolescents without access to surgical care. Bull World Health Org, 2019; 97:254–258.

Bickler

, Rode

. Surgical services for children in developing countries. Bull World Health Org, 2002; 80:829–835.

Global Initiative for Children's Surgery: A model of global collaboration to advance the surgical care of children. World J Surg, 2019; 43:1416–1425.

Ozgediz

, Poenaru

. The burden of pediatric surgical conditions in low and middle income countries: A call to action. J Pediatr Surg, 2012; 47:2305–2311.

10.

Krishnaswami

, Nwomeh

, Ameh

. The pediatric surgery workforce in low- and middle-income countries: Problems and priorities. Semin Pediatr Surg, 2016; 25:32–42.

11.

Lalchandani

, Dunn

. Global comparison of pediatric surgery workforce and training. J Pediatr Surg, 2015; 50:1180–1183.

12.

Greenberg

, Ng-Kamstra

, Ameh

, et al. An investment in knowledge: Research in global pediatric surgery for the 21st Century. Semin Pediatr Surg, 2016; 25:51–60.

13.

Bickler

, Sanno-Duanda

. Epidemiology of paediatric surgical admissions to a government referral hospital in the Gambia. Bull World Health Org, 2000; 78:1330–1336.

14.

Thanni

, Shonubi

, Akiode

. A retrospective audit of paediatric surgical admission in a sub-urban tertiary hospital. W Afr J Med, 2005; 24:10–12.

15.

Butler

, Tran

, Nagarajan

, et al. Epidemiology of pediatric surgical needs in low-income countries. PloS One, 2017; 12:e0170968.

16.

Wesonga

, Fitzgerald

, Kabuye

, et al. Gastroschisis in Uganda: Opportunities for improved survival. J Pediatr Surg, 2016; 51:1772–1777.

17.

Wood

, Nthumba

, Stepita-Poenaru

, Poenaru

. Pediatric surgical site infection in the developing world: A Kenyan experience. Pediatr Surg Int, 2012; 28:523–527.

18.

Wesonga

A SM.

Reducing gastroschisis mortality: A quality improvement initiative at a Ugandan pediatric surgery unit [abstract]. Pan-Afr Pediatr Surg Assoc 2018.

19.

Wakjira

, Sisay

, Zember

, et al. Implementing ultrasound-guided hydrostatic reduction of intussusception in a low-resource country in sub-Saharan Africa: Our initial experience in Ethiopia. Emerg Radiol, 2018; 25:1–6.

20.

Weiss

, Fitzgerald

, Pappachan

, et al. Global epidemiology of pediatric severe sepsis: The sepsis prevalence, outcomes, and therapies study. Am J Respir Crit Care Med, 2015; 191:1147–1157.

21.

Optimal Resources for Children's Surgical Care: Executive Summary. World J Surg, 2019; 43:978–980.

22.

Yap

, Muzira

, Cheung

, et al. A cost-effectiveness analysis of a pediatric operating room in Uganda. Surgery, 2018; 164:953–959.

23.

Kisa

, Grabski

, Ozgediz

, et al. Unifying children's surgery and anesthesia stakeholders across institutions and clinical disciplines: Challenges and solutions from Uganda. World J Surg, 2019; 43:1435–1449.

24.

Nensi Ruzgar

, Ullrich

, Kisa

, Ozgediz

Needs assessment of rural providers for the development of a pediatric emergency surgical care course in Uganda [abstract]. Buthane Round Table 2019.

25.

Anderson

, Ilcisin

, Abesiga

, et al. Surgical volume and postoperative mortality rate at a referral hospital in Western Uganda: Measuring the Lancet Commission on Global Surgery indicators in low-resource settings. Surgery, 2017; 161:1710–1719.