From the *Department of Surgery, King Faisal Hospital; †Department of Surgery, University Teaching Hospital of Kigali; ‡Kacyiru Police Hospital; §Department of Pediatrics, University Teaching Hospital of Kigali; ¶Department of Surgery, University Teaching Hospital of Butare, Kigali, Rwanda; ‖World Health Organization, Regional Office for Africa (WHO/AFRO), Brazzaville, Republic of Congo; **Centers for Disease Control and Prevention, Atlanta, GA; and ††Maternal and Child Health Unit, Ministry of Health, Kigali, Rwanda.
Accepted for publication August 7, 2013.
The authors have no conflicts of interest to disclose.
Address for Correspondence: Edouard Ngendahayo, Department of Surgery, King Faisal Hospital, Kigali, Rwanda. E-mail: email@example.com.
In some settings, rotavirus vaccines have been associated with a small risk of intussusception. This study describes the demographics, clinical characteristics and outcomes of children with intussusception in Rwanda before vaccine introduction in May 2012.
We retrospectively reviewed data on pediatric patients treated for intussusception at University Teaching Hospital of Kigali from January 2009 to June 2012. Hospital registries were reviewed to identify intussusception cases. Archived patient files were abstracted to collect data on demographics, clinical presentation, surgery and outcome.
During the study period, 70 children ≤19 years of age were treated for intussusception and patient files were retrieved for 60 (86%) cases. Over half of patients (58%) were <1 year of age and 90% were transferred from a district hospital. Only 30% of patients presented within 3 days of symptom onset and 35% experienced a delay of ≥6 hours between surgical review and surgery. Surgical complications were experienced by 35% of children. Over 1 quarter (28%) of children died. Compared with children who survived, children who died were significantly more likely to have experienced complications (21% vs. 71%; P < 0.001) and to be female (28% vs. 65%; P = 0.009).
Mortality due to intussusception was high among Rwanda children. Delays in presentation and treatment were common. Assessing trends in the number of cases to monitor for vaccine-associated intussusception will be difficult. Additional work is needed to further understand risk factors for mortality, to calculate incidence rates and to monitor the safety of the rotavirus vaccination program.
Intussusception, 1 of the most common causes of acute intestinal obstruction in infants and young children, occurs when 1 segment of bowel (intussusceptum) invaginates into the distal bowel (intussuscipiens), resulting in venous congestion and bowel wall edema. Failure to diagnose and treat this condition may lead to bowel infarction and perforation and could be lethal.1 Two classic symptoms (abdominal pain and vomiting) and 2 classic signs (abdominal mass and rectal bleeding) help in making the diagnosis with 3 of the 4 symptoms and signs present in approximately 40% of cases.2 Clinical diagnosis is confirmed by ultrasonography, radiology or surgery. Mortality caused by intussusception has become rare in developed countries through timely diagnosis and treatment with enema or surgery. In developing countries, patients often present late or with more serious illness, leading to high rates of bowel resection, morbidity and mortality.1,3–5
The etiology of most intussusception cases is unknown. A viral illness precedes or contributes to pediatric intussusception in about 20% of cases.2 Adenovirus infection has been reported associated with increased risk of intussusception, but no link with natural rotavirus infection has been reported.6,7 However, due to an unexpected association between a previously available oral rotavirus vaccine (RRV-TV; RotaShield, Wyeth Laboratories, PA) and intussusception, there has been recent interest in the epidemiology, etiology and pathophysiology of intussusception in infants. RRV-TV was withdrawn from the market in the United States, the only country where it was introduced, after it was associated with intussusception and resulted in 1 excess case of intussusception for every 10,000 infants vaccinated.8,9 Two new rotavirus vaccines (RV5; RotaTeq, Merck and Co, Inc, Pennsylvania, PA and RV1; Rotarix, GSK Biologicals, Rixensart, Belgium) are currently available and recommended for use in all countries globally.10 No increased risk of intussusception was observed in clinical trials of RV5 or RV1 but postmarketing studies have identified a small risk of 1–2 excess cases per 100,000 infants vaccinated in some settings but not in others.11–13
Rwanda recently introduced RV5 into its national immunization program in May 2012. Understanding the epidemiology of intussusception before the introduction of rotavirus vaccine in Rwanda will help in postintroduction safety monitoring, but no such data are available. Furthermore, emergency laparotomy following pediatric intussusception is not rare in settings like Rwanda but information regarding surgical outcomes is limited. Knowing the outcome of surgery for intussusception may help both departments of pediatrics and surgery to improve patient care. The objectives of this study are to describe the demographics, clinical characteristics and outcomes of children ≤19 years of age that were treated for intussusception at the University Teaching Hospital (UTH) of Kigali from January 2009 through June 2012.
Study Design and Setting
This study is a retrospective review of infants and children operated on for intussusception at UTH of Kigali from January 2009 to June 2012. UTH of Kigali is the main national referral hospital located in Kigali City, the capital of Rwanda. It is a 513-bed consultant and teaching hospital and receives patients from all the 4 provinces of the country and Kigali City. The department of surgery has 6 operating rooms, with 1 fully dedicated to all surgical emergencies.
Data Collection and Analysis
Treatment of intussusception with enema is not practiced at UTH, and all patients are surgically treated. The pediatric emergency, operating room and ward registries were independently reviewed by 3 different data collectors to identify all pediatric patients ≤19 years of age admitted and surgically operated on for intussusception. The 3 independent lists from each registry were then merged and deduplicated. To retrieve a patient’s file from the archive, the patient’s hospital ID was entered into the electronic “open clinic” record system to obtain the patient’s code that linked to the archived file. Patients whose file could not be located in the archive were excluded from the analysis. Demographics, clinical presentation, surgery related data and outcome were abstracted from patient files using a standardized data collection sheet and were summarized using descriptive statistics. Characteristics of children who died and who survived were compared using χ2 statistics. Variables that were identified in bivariate analyses as significant predictors of mortality were included in multivariate analyses.
This study was approved by the Ethics Committee of the UTH of Kigali.
From January 2009 to June 2012, 70 children ≤19 years of age were identified as having been operated on at UTH of Kigali for intussusception. Patient files could be retrieved for 60 (86%) of these cases and 10 cases (7 from 2009, 2 from 2010 and 1 from 2011) were excluded from further analysis because the files could not be found. Over half of patients (58%) were <1 year of age with the peak of cases occurring in 5–7 month of age and two-thirds (68%) were under 2 years of age, with male predominance (62%) (Fig. 1; Table 1). Many patients (90%) were referred to UTH of Kigali from a district hospital. The number of cases identified increased from 8 in 2009 to 15 in 2010 and to 26 in 2011. No clear seasonal pattern was identified.
At least 3 of the 4 classic symptoms (abdominal pain and vomiting) and signs (rectal bleeding and abdominal mass) were found in 80% of patients (Table 2). An abdominal ultrasound scan was positive in 43% of cases. Three-quarters of the intussusceptions were of the ileocolic type.
Less than one-third of patients (30%) presented for care within 3 days of symptom onset and over one-third of patients (37%) presented at least 1 week after symptom onset (Table 3). Delays between surgical review and surgery also occurred with over one-third of children (35%) experiencing a delay of at least 6 hours between review and surgery. Most surgeries were conducted by a senior (55%) or junior (30%) resident. Bowel resection was required in 70% of cases. Many surgeries (78%) were completed in 2 or fewer hours. Complications were experienced by 35% of children with surgical site infections and chest infections as the most common complications. Over 1 quarter (28%) of children died.
Compared with children who survived, children who died were significantly more likely to have experienced complications (21% vs. 71%, respectively; P < 0.001) and to be female (28% vs. 65%, respectively; P = 0.009; Table 4). These factors remained significant predictors of mortality in multivariate analyses. Children who died also had higher rates of bowel resection (82% vs. 65%) and longer delays between surgical review and surgery (median of 8 hours vs. 4 hours) than children who survived but these differences were not statistically significant.
Naturally occurring intussusception was documented in Rwandan infants and children before rotavirus vaccine introduction with 70 cases identified by retrospective record review at UTH of Kigali from January 2009 to June 2012. Infants <1 year of age and males experienced the greatest burden, the classic signs and symptoms were frequently reported with 80% of cases reporting 3 of the 4 classic signs and symptoms, and the ileocolic type was the most common accounting for 75% of cases. However, surgery was the only treatment option available at UTH of Kigali and delays between symptom onset and presentation for care and between surgical review and surgery were common. Mortality was high with 28% of intussusception cases dying following surgery. Children who died were significantly more likely to have experienced complications associated with surgery and to be female compared with children who survived.
Data on intussusception-associated mortality in Africa are limited and case-fatality ratios vary widely with a ratio as high as 53% reported in 1 Nigerian study 2 decades ago.14 In contrast, mortality rates in developed countries are <1%. The high rate of surgery in many developing countries like Rwanda, where lower risk radiologic treatment options such as air or hydrostatic reduction enema are not available, may account for some of the observed difference in mortality rates between developing and developed countries. Complications arising from surgery were identified as a risk factor for mortality in our study. In our study, 70% of participants underwent bowel resection compared with approximately 40% resection rate among surgically treated children in a retrospective study in South African children and in a prospective study in infants in the United Kingdom and Ireland.5,15 Both of these studies also identified children with intussusception who had nonsurgical interventions and the overall mortality rates were low (2% and 0.4%, respectively). Delays in presentation, diagnosis and treatment may also play a role in the increased mortality but our small sample size and high levels of delay precluded us from fully evaluating these factors. Education of parents, medical providers and surgeons to recognize the early classic signs and symptoms of intussusception may help facilitate early presentation, early transfer and timely surgery in Rwanda.
This study has limitations. First, this study was a retrospective review with case capture that relied on a diagnosis for intussusception to be recorded in a registry book. To minimize the number of cases missed, we independently reviewed registries from 3 different hospital departments. However, it is possible that cases were missed because diagnoses were inaccurately recorded or cases were misdiagnosed. Additionally, we were unable to locate the patient file for 15% of cases identified by the record review. The apparent increasing trend in the number of cases identified each year more likely reflects better recording and availability of records than a true increasing trend of the number of intussusception cases. The true number of cases that sought care for intussusception at UTH of Kigali during the study period may be greater than the 70 cases that we did identify. Second, given the high level of cases that were referred to UTH of Kigali from district hospitals throughout the country, we do not have a defined catchment area that can be used to calculate population-based incidence rates. Additionally, we did not capture cases that presented to other hospitals and health centers, that did not seek treatment or that were treated for an alternate diagnosis which further limited our ability to understand the full burden of intussusception in Rwandan children. Third, given our small sample size, we were only able to detect large differences between children who died and children who survived, which limited our ability to identify risk factors associated with increased mortality. Finally, our study was conducted at a single hospital that serves as a national referral hospital and cases referred to this hospital for treatment may not be representative of all intussusception cases in Rwanda.
This study has several key lessons that will be important for postvaccine introduction safety monitoring. First, year-to-year variability in the number of cases was large. Thus, simply assessing trends in the number of cases over time to monitor for risk of intussusception associated with the vaccine will be difficult. Second, a better understanding of the age distribution among children <1 year of age, ideally by week of age, is needed as these are the children who receive the rotavirus vaccine. Expanding intussusception surveillance to additional hospitals with the capacity to diagnose and treat intussusception in infants will help to increase the number of cases identified. Third, an epidemiologic assessment of the risk of intussusception following vaccination with RV5 using a case-series or a cohort approach will require a large number of cases of intussusception due to the expected low level of risk which will be difficult to achieve in Rwanda alone. Data from Rwanda can be combined with that from other countries for a regional assessment of vaccine safety. However, to date, Rwanda is the only country in Africa to have introduced RV5 into their national immunization program.
In conclusion, to further describe the epidemiology of intussusception in Rwanda, this study should be expanded to include additional hospitals in Rwanda with the capacity to diagnose and treat intussusception in infants and children. Initiation of prospective surveillance to identify intussusception cases will also help to increase case capture. These baseline data will play an important role in the monitoring of rotavirus vaccine safety. These efforts should be coupled with active surveillance for rotavirus diarrhea so that the full benefits and potential risks of the vaccine can be evaluated. Additionally, efforts to address and reduce the high mortality rates due to intussusception such as reducing the delay to surgery should also be undertaken.
1. Ekenze SO, Mgbor SO, Okwesili OR. Routine surgical intervention for childhood intussusception in a developing country. Ann Afr Med. 2010; 9:27–30
2. Sigmund H, Daneman A. Fonkalsrun EW Jr, Coran AG. Intussusception. Pediatric Surgery. 2006; MO, USA Mosby & Elsevier 1314–1337
3. Ekenze SO, Mgbor SO. Childhood intussusception: the implications of delayed presentation. Afr J Paediatr Surg. 2011; 8:15–18
4. Abdur-Rahman LO, Adeniran JO, Taiwo JO, et al. Bowel resection in Nigerian children. Afr J Paediatr Surg. 2009; 6:85–87
5. Moore SW, Kirsten M, Müller EW, et al. Retrospective surveillance of intussusception in South Africa, 1998-2003. J Infect Dis. 2010; 202:(suppl)S156–S161
6. Bines JE, Liem NT, Justice FA, et al. Intussusception Study Group Risk factors for intussusception in infants in Vietnam and Australia: adenovirus implicated, but not rotavirus. J Pediatr. 2006; 149:452–460
7. Bahl R, Saxena M, Bhandari N, et al. Delhi Intussusception Study Hospital Group Population-based incidence of intussusception and a case-control study to examine the association of intussusception with natural rotavirus infection among Indian children. J Infect Dis. 2009; 200:(suppl 1)S277–S281
8. Centers for Disease Control and Prevention Withdrawal of rotavirus vaccine recommendation. MMWR. Morb Mortal Wkly Rep. 1999; 48:1007
9. Peter G, Myers MG. National Vaccine Advisory Committee; National Vaccine Program Office Intussusception, rotavirus, and oral vaccines: summary of a workshop. Pediatrics. 2002; 110:e67
10. Health Section of the Secretariat of the League of Nations Rotavirus vaccines: an update. Releve epidemiologique hebdomadaire / Section d’hygiene du Secretariat de la Societe des Nations. Wkly Epidemiol Rec. 2009; 84:533–540
11. Patel MM, López-Collada VR, Bulhões MM, et al. Intussusception risk and health benefits of rotavirus vaccination in Mexico and Brazil. N Engl J Med. 2011; 364:2283–2292
12. Shui IM, Baggs J, Patel M, et al. Risk of intussusception following administration of a pentavalent rotavirus vaccine in US infants. JAMA. 2012; 307:598–604
13. Buttery JP, Danchin MH, Lee KJ, et al. PAEDS/APSU Study Group Intussusception following rotavirus vaccine administration: post-marketing surveillance in the National Immunization Program in Australia. Vaccine. 2011; 29:3061–3066
14. E. BJ, Ivanoff B. Acute Intussusception in Infants and Children: Incidence, Clinical Presentation and Management: A Global Perspective. 2002; Geneva:World Health Organization
15. Samad L, Marven S, El Bashir H, et al. Prospective surveillance study of the management of intussusception in UK and Irish infants. Br J Surg. 2012; 99:411–415