Factors associated with mortality in congenital malformations of the gastrointestinal tract in a tertiary center in Senegal

Introduction

Congenital malformations (CMs) represent a large burden of pediatric surgical disease in African tertiary centers.1 2 CMs are the third leading cause of death among children aged less than 5 years worldwide.3 Malformations of the gastrointestinal tract (GIT) are the third most common CM.4 In high-income countries (HICs), these malformations are more frequently diagnosed prenatally or during the neonatal period with early management in dedicated centers and with improved survival during the last decades.5 6 However, prenatal diagnosis of GIT malformations is rare in low and middle-income countries (LMICs).5 6 Postnatal diagnosis is often delayed, and the resulting mortality is very high compared with HICs.6

The global mortality of CMs of the GIT in LMICs is 39.8%, while it is only 5.6% in HICs.6 However, some conditions have a very high contrast in mortality rate. This is the case for esophageal atresia (EA), for which mortality is 85.7% in LMICs versus 7.2% in HICs.6 Many reasons have been advocated to justify this difference. However, in sub-Saharan Africa, studies on risk factors for mortality from GIT malformations are scarce.7 In Senegal, the mortality of several GIT malformations was reported to be very high.8–12 We assume that the identification of independent factors of mortality in patients with CMs of the GIT will allow a better understanding of mortality in this group of children and therefore lead to the improvement of care for better survival. With this review, our aim was to identify risk factors for mortality in patients with CMs of the GIT admitted at our University Teaching Hospital (UTH), Albert Royer National Children’s Hospital Center in Dakar, the capital city of Senegal.

Methods

Results

From 2018 to 2021, among 3667 patients admitted to our department, 242 had CM of the GIT, implying a frequency of 6.6% at our institution. Among the 242 patients, 16 were excluded from the present study.

In this study, the patients’ ages ranged from 0 to 15 years; neonates represented 42.48%, and males represented 60.18%, with a sex ratio of 1.5:1. Concerning the origin of patients, 64.60% of them were from Dakar and its neighborhood. Parental consanguinity status was reported in 85 cases (37.61%), of whom 24.71% of patients were issued from a consanguineous marriage. Status of prenatal US was reported in 19.91% of cases, of which 13.33% were abnormal, 20% were not done and 66.67% were normal. The gestational age at birth was reported in 48 patients (20.24%), of whom 29.17% noted prematurity, 4.17% noted postmaturity, and 66.66% noted normal gestational age at birth. CMs of the GIT were isolated in 94.25% of cases. Malformations encountered included anorectal malformation in 30.53%, HD in 27.88%, EA in 16.37%, malrotation in 12.83%, duodenal atresia (DA) in 7.08%, intestinal atresia in 2.21%, congenital esophageal stenosis in 1.77%, and colonic atresia (CA) in 0.88%. The admission rate progressively increased from 2018 to 2020 with 53, 72, and 101 cases, respectively. A total of 85 patients (37.61%) were referred, 53.54% came from home, and 8.85% changed the hospital based on parental decisions, that is, self-referral.

Among our population study, 63 deaths (27.88%) were registered. Analysis of differences between the mortality and survival groups showed no significance for sex, parental consanguinity or year of admission. The variables between the two groups that were statistically significant included being a neonate (p < 0.001), geographical origin (p < 0.001), prenatal US status (p = 0.002), gestational age at birth (p = 0.002), presence of an associated malformation (p = 0.002), diagnosis (p < 0.001), and admission mode (p < 0.001). These results are presented in table 1.

Variables that showed a significant difference between the two groups were considered for logistic regression analysis. Unadjusted OR showed significance for being a neonate, originating out of Dakar, prematurity, presence of an associated malformation, and some diagnoses, such as DA, CA and EA, for two admission modes: referral and self-referral. Finally, based on adjusted OR, two variables showed statistically significant differences: the presence of associated malformation increased the risk of mortality by 13 (p = 0.026), while the diagnosis of EA increased the risk of mortality by 46 (p < 0.001), as represented in table 2.

Discussion

Our study retrospectively reviewed 226 cases of CMs of the GIT, of whom approximately a quarter died. This mortality rate is similar to those of other LMICs, as some authors reported mortality ranging from 19.23% to 50%.6 13–16 This is far less than the approximately 5% mortality rate reported in Western studies.6 In a recent review, originating from a low-income country was identified as an independent risk factor for mortality in CMs of the GIT, increasing the risk by approximately threefold compared with patients from HICs.6 Several reasons have been pointed out to explain this difference. These include lack of prenatal diagnosis, unavailability of pediatric anesthesiologists, lack of surgical safety checklists, unavailability of ventilation, and high cost of parenteral nutrition (PN), which makes it unavailable for many patients in need.6 Other possible factors are delayed referral to tertiary hospitals, inappropriate medical transfer of patients, and inadequate resuscitation during the transfer and on admission at the referral hospital.7

We found that EA was independently associated with increased mortality, increasing it by approximately 46 times. In previous studies in Senegal, the mortality of this condition was very high; for example, in previous decades, it was nearly 100%.17 In the last decade, mortality decreased to 73%–80%,9 11 17 which is still very high compared with HICs, where mortality reaches 7%.6 Reasons for this high mortality in our environment were identified by some earlier studies, including lack of prenatal diagnosis, which leads to birth out of tertiary centers, which in turn leads to delayed diagnosis and management, since patients should be referred.5 11 17 18 Delayed diagnosis exposes newborns to pulmonary compromising due to frequent episodes of choking, which leads to permanent respiratory distress, found in 40% of patients in a local study.9 The lack of surgical neonatal intensive care units (NICUs), with the hospitalization of newborns in medical NICUs, also increases mortality in our settings.17 19

We highlighted that patients with associated malformations had a 13-fold likelihood of death. This may be linked to the severity of associated anomalies, which lead to more complex management, considering the lack of well-equipped NICUs in our context, as well as elsewhere in sub-Saharan Africa.15 19 20

Additionally, some factors frequently associated with mortality in several studies were not shown to be independent risk factors for mortality in our review. The first is being neonate. In fact, several authors have reported a higher mortality rate in neonates than in older patients.14 19 However, another study did not find it to be an independent factor of mortality.21 A recent Somalian study showed improved neonatal perioperative survival compared with other African settings. However, the highest mortality rate was found in patients with EA.22 High mortality in neonates should therefore be linked to the diagnosis, delayed presentation, subsequent complications, and lack or underequipment of NICUs.19 23–25 Second, we did not find the geographical origin of our patient to be an independent predictor of mortality, unlike other authors.6 This can be explained by the fact that the distance from the tertiary hospital can interact with other factors, such as medical transportation during the referral, quality of resuscitation prior to referral and severity of the diagnosis.

We included 226 patients out of the 242 identified during the study period. First, missing data on prenatal US and gestational age at birth did not allow better analysis of the link between these elements and mortality. Second, data on the condition of referral (medical or not), place of birth (tertiary center or not), need for ventilation (yes vs no), need for PN (yes vs no), qualification of anesthesiologist (resident vs consultant), and qualification of the main surgeon (resident vs consultant) were not analyzed in our review.

In conclusion, our study found that mortality in patients with CMs of the GIT was very high, as in other sub-Saharan tertiary centers. Two factors were independently associated with mortality in Dakar in these conditions: presence of an associated malformation and a diagnosis of EA. Patients with these elements should benefit from specific care from diagnosis to postoperative care. Future multicenter studies on predictors of mortality in EA and polymalformation are encouraged to improve survival in these patients in our milieu.