The POMRs on within 24 hours and 30 days after the procedure were 1.94% (95% CI = 11.09–3.17) and 1.17% (95% CI = 0.91–2.91), respectively and compares with a similar study conducted in a single center in Southwest Nigeria by Ademola et al.[12] A Kenyan multicenter study reported 24-hour and 7-day mortality rates of 0.8% and 1.7%, respectively[13] in contrast to middle-income countries (MICs) and high-income countries (HICs) with 0.7% and 0.3%, respectively[14] and represents more than 2-fold and 6-fold increased mortality in LMICs in comparison with MICs and HICs respectively.
The highest perioperative mortalities occurred in patients with gastroschisis and esophageal atresia, 100% and 60.0% mortality rates, respectively. Four patients with gastroschisis presented in very poor clinical condition and died before any form of intervention and were excluded from this study. Three had improvised silo application and two had laparotomies and primary surgical closure. In Kano, Northern Nigeria, Anyanwu et al. (2020) reported 87.2% mortality rate for gastroschisis, and later, observed a significant improvement in outcome (p = 0.035) when they reverted to sutureless application of improvised silo. [15] As with the present study, most mortalities were due to starvation and late presentation with sepsis. A Global PedSurg prospective study (2021), reported 90.0% POMRs for gastroschisis, in LICs, 31.9% in MICs and 1.4% in HICs.[14] Lack of prenatal diagnosis and immediate postnatal interventions might be responsible for the dismal outcome in LMICs. [12, 15, 16] Two mortalities in EA-TOF had thoracotomy with primary repair, and one had laparotomy in addition, for associated duodenal atresia. Others had gastrostomies with gastric banding due to late presentation with aspiration pneumonitis. All required prolonged ventilatory support, like the report by Puri et al. (2019) [17] though this could not be sustained owing to the unavailability of NICU and requisite expertise, especially for the nursing care of such patients. The need for prolonged oxygen or ventilatory support has been linked to higher neonatal perioperative mortality. [18]
Neonates accounted for the majority (60.7% ) of deaths, mirroring Ademola et al.’s finding of 58.8% [12] Neonates encounter distinct challenges as they transit from intrauterine life to postnatal life, and going through surgery can disrupt this delicate balance, making them more vulnerable to sepsis, nutritional deficiencies, and respiratory problems. [19]The neonatal surgical mortality (NSM) rate was 19.8%, lower than reports from India (33.33%),[17] Bangladesh (14.6%),[20] and Tunisia[21] but higher than those from HICs. [14, 22–24] The lower neonatal mortality rates in HICs have been linked to advanced perioperative support systems, prenatal diagnosis, improved perinatal care, availability of NICUs, and TPN. [12, 14, 25–27]
Most (72%) mortalities were attributed to preoperative child conditions and overwhelming sepsis was the leading cause of death (53.6%), consistent with other reports on pediatric perioperative mortality rates. [12, 14, 17] Late presentation, often with complications and sepsis, is common in the subregion.[28,29] In the present study, all the patients who died from intussusception presented more than 24 hours of the onset of symptoms in poor clinical conditions, and some required multiple surgeries. Limitations in theatre space availability created further delays in intervention contributing to the observed mortality. Presentation after 24 hours is known to increase the risk of perioperative mortality in children with intussusception. [28]
Factors related to anesthesia were responsible for or confounded the pre-existing condition in 20% of deaths, which is a significant contrast to de Bruin L et al.'s report of less than 0.02%.[6] The absence of pediatric anesthesiologists and the lack of suitable equipment for pediatric anesthesia in the participating hospitals may account for this difference. Emergency surgeries had significantly more mortalities than elective surgical cases, while the surgical procedure accounted for 16% of mortalities and this may be due to technical errors and suboptimal theatre conditions for pediatric surgeries. Consultant surgeons performed most of the more delicate surgeries and this might explain why more mortalities were seen when the consultant and not the Resident was the surgeon.
We found no significant difference in perioperative mortalities in surgeries for congenital and acquired cases contrary to Ademola[12] and this might be due to the exclusion of cardiac cases in the present study.
Multivariable logistic regression showed that higher ASA status and associated comorbidities were significant predictors of perioperative mortality, supporting earlier findings on the subject. [12, 14, 15] Associated comorbidities identified were associated congenital malformations, low birth weight, and immunosuppression had increased odds of 30-day mortality. Patients with ASA status III or greater represented 89.3% of deaths. Other forms of anesthesia aside from general anesthesia with or without ETT were significantly related to increased mortality. These forms of anesthesia were mostly sedation (with midazolam or diazepam) or sedation in combination with local infiltration with lignocaine. Although sedation using these agents is commonly used in pediatric surgery,[30] we believe that the provision of adequate ventilatory support could have made a significant difference for the patients in this study. It is possible that the lack of proper equipment or expertise, or both, may have contributed to the fatalities observed with this type of anesthesia. Inadequate or inappropriate use of anesthetic medication, limited ventilatory support, unavailability of airway devices, and the required skills are linked to increased pediatric perioperative mortality.[31] We hope to explore this in a follow-up study further. More mortalities were recorded when the anesthesiologist was a trainee. This was not surprising as most emergencies were undertaken by the trainees. While the results were not statistically significant, it might be necessary to provide additional oversight for trainees in anesthesiology undertaking major pediatric surgeries.
Limitations of the study
Despite the reported high POMRs reported in this study, the actual figure could be higher because data collection was at tertiary centers where there is access to expert pediatric surgical services. Pediatric surgeries performed outside these centers were not captured. The exclusion of cardiac, orthopedic, and plastic surgeries in the pediatric age group may affect the overall POMR for pediatric surgeries and this needs to be considered when interpreting the results.