Mortality Prediction in Pediatric Patients with Sepsis of Abdominal Origin: A Retrospective Study

Aims: The aim of this study was to assess the prediction of mortality in pediatric patients with sepsis of abdominal origin. Methods: We performed a retrospective study of patients with sepsis of abdominal origin admitted to the pediatric intensive care unit (PICU) in Shanghai Children’s Hospital between May 2014 and April 2018. Results: A total of 143 patients were enrolled in this study. The mortality rate in pediatric patients with sepsis of abdominal origin was 11% (17/143). PRISMIII score, the grade of acute gastrointestinal injury (AGI), hematologic malignancy, acute kidney injury, the initial number of organ dysfunction, alanine aminotransferase (ALT), and blood platelet (PLT) were potential risk factors for increased mortality. Multivariate logistic regression analysis indicated that higher Pediatric risk of mortality III (PRISM III) score and the grade of AGI were independent risk factors of mortality in patients with sepsis of abdominal origin. Furthermore, the area under receiver-operating characteristic (ROC) curve (AUC) for PRISM III score and the grade of AGI were 0.91 (95% confidence interval (CI) 0.86–0.96, P<0.001) and 0.83 (95% CI 0.74–0.92, P<0.001), in which the optimal cut-off value was 10.5 and 2.5, respectively. In addition, the Kaplan-Meier curve indicated that the 28-day survival rate was significantly lower in patients with higher PRISM III score (>10.5) and the grade of AGI (>2.5). Conclusions: Higher PRISM III score and the grade of AGI were associated with poor outcomes in pediatric patients with sepsis of abdominal origin.


Introduction
Sepsis, defined as life-threatening organ dysfunction caused by a dysregulated host response to infection, remains a high mortality rate in pediatric [1,2]. The abdomen is the common site of sepsis in 8.3%-19.6% of the patients in the pediatric intensive care unit (PICU) [3][4][5][6].

Study Design
We performed a retrospective analysis of patients with sepsis admitted to the PICU in Shanghai Children's Hospital between May 2014 and April 2018. Sepsis was defined according to the criteria of the Surviving Sepsis Campaign International Guidelines in 2012 [8]. All the patients enrolled in the study were confirmed to have the original infection of the abdomen. This study was approved by the Ethics Committee of Shanghai Children's Hospital (No. 2016R010-E02). Informed content was waived because of its retrospective design.

Patients
Patients aged 29 days to 18 years old who were diagnosed with sepsis of abdominal origin between May 2014 and April 2018 were screened for inclusion. Patients with PICU stay less than 48 hours were excluded.
Sepsis and organ dysfunction were diagnosed according to the criteria of Surviving Sepsis Campaign International Guidelines in 2012 [8]. Patients were treated in accordance to current guidelines for treatment of sepsis. The patients with acute kidney injury, fluid overload or hemodynamic instability were treated with continuous renal replacement therapy (CRRT). If the patients presented respiratory failure, mechanical ventilation was used as an adjuvant treatment.

Data Collection and Definitions
All the data were retrieved from electronic medical records. Data included demographic data (such as, age, gender, weight, etc.), comorbid diseases, the length of PICU stay, the initial number of organ dysfunction, and Pediatric Risk of Mortality score (PRISM III) [9] within 24 hours following admission, the need for mechanical ventilation, CRRT, positive bacterial cultures, the grade of acute gastrointestinal injury (AGI) [10] and the outcome on discharge. Additional data about laboratory indexes included total bilirubin level (TBIL), creatinine (CREA), prothrombin time (PT), lactate (LAC), blood platelet (PLT), C-reactive protein (CRP) and procalcitonin (PCT), etc within 24 hours following admission.

Statistical Analyses
Data were analyzed using SPSS (v.22.0). Continuous variables were summarized as means ± standard derivations (SD) for normal distribution data and as median (Inter-Quartile Range) for abnormal distribution data. T-test was used to compare the continuous variables with normally distributed data; otherwise, the Mann-Whitney U test was used.
The chi-square test was used to compare the categorical data. To identify factors associated with mortality, a multivariable logistic regression model was constructed using variables for which the P-value was ≤0.001 in univariable analysis. In order to appreciate the accuracy of independent risk factors as a prognostic marker, a receiver operating characteristic (ROC) curve was generated. A value of P < 0.05 was considered statistically significant.  Table   1. Patients with hematologic malignancy, in this study, had higher PICU mortality (P = 0.001). Non-survivors had a higher PRISM III score (P < 0.001) and were more likely to have acute kidney injury on admission (P = 0.04). The rate of mechanical ventilation and CRRT was significantly higher in the non-survivor group compared with the survivor group (all P < 0.05).

Result
The organ dysfunction indicators were compared between the survivors and non-survivors to evaluate the risk factor for the in-hospital mortality of patients with sepsis of abdominal origin. The grade of AGI, Glasgow coma scale score <8, the serum level of platelet, and alanine aminotransferase (ALT) were the risk factors for the PICU mortality (all P < 0.05).
Variables with a P ≤ 0.001 were selected for a multivariate analysis using a stepwise logistic regression model.  The ROC curves analysis was employed for the estimation of the predictive value of the PRISM III score, the initial number of organ dysfunction, and the grade of AGI (Table   3Figure 1  We investigated the association between PRISM III score, the grade of AGI, and the 28-day survival rate in the patients with sepsis of abdominal origin in PICU by using the Kaplan-Meier method. The patients with a higher PRISM III score or the grade of AGI showed a shorter 28-day survival rate (Figure 5, P < 0.001 and Figure 4, P < 0.05). Therefore, elevated PRISM III score and a higher grade of AGI is associated with poor outcome of the patients with sepsis of abdominal origin.

Discussion
Abdominal is the common site of sepsis, and the mortality rate of patients with sepsis of abdominal origin is high. Therefore, it is urgently needed to identify predictors of mortality in patients with sepsis of abdominal origin. In the present study, we found that PRISM III score, the initial number of organ dysfunction, with hematologic malignancy, the grade of AGI, ALT, PLT, and GCS score <8 were potential risk factors for the patients with sepsis of abdominal origin. Moreover, the higher PRISM III score and the grade of AGI were independent predictors for prognosis of pediatric patients with sepsis of abdominal origin. PRISM III score and the grade of AGI showed a good accuracy to predict the 28-day survival rate of pediatric patients with sepsis of abdominal origin.
In our study, the incidence of sepsis of abdominal origin in pediatric patients was 14% 10 (143/1012), and the mortality rate was 11% (17/143). Similarly, in a study of 1051 pediatric patients from Colombian, the sepsis was of abdominal origin in 18.3% of the episodes, and the mortality rate was 18% [5]. In another study of the 7087 sepsis patients from 1265 ICUs in 75 countries, the abdominal contributed to 19.6% of sepsis with the ICU mortality of 29.4% [6].
Our study found that hematologic malignancy, acute kidney injury, the initial number of organ dysfunction, ALT, and PLT were the potential risk factors for increased mortality.
Children with hematologic malignancy were at high risk of sepsis. Ruth et al. [11]reported that among patients with specific comorbidities, mortality was highest in patients with malignancies (22.4%), hematological or immunological disorders (20.3%). Higher intraabdominal pressures may create a state of inadequate renal perfusion pressure, contributing to the development of acute kidney injury (AKI). Harris et al. [12] found that critically ill patients with AKI had higher mortality. Julie et al. [13] reported that sepsis of abdominal origin remained associated with both mild and severe AKI. The mortality in the present study also reflected the high rate of multiorgan dysfunction. Scott et al. [4] found that 67% of patients had MODS at sepsis recognition and relevant to the prognosis. Our study found that Gram-negative isolates were most frequently present in patients with sepsis of abdominal origin. Studies have found that endotoxin can cause mitochondrial dysfunction and ultrastructural damage of hepatocytes, resulting in a significant increase in ALT [14]. Thrombocytopenia is a well-known biomarker for disease severity in patients with sepsis and has been thought to result from platelet consumption and activation [15,16].
The grade of AGI, suggesting poor gastrointestinal function and PRISM III score, might be considered an indicator of poor outcome. Our present study indicated that a higher grade of AGI (>2.5) and a higher PRISM III score (>10.5) were independently associated with an 11 increased risk of PICU mortality. Consistently, Hu et al. [17] reported that the grade of AGI was positively correlated with the 28-and 60-day mortality (P < 0.001) and was independently associated with the 60-day mortality (HR 1.65, 95% CI 1.28-2.12; P = 0.008) among the 550 patients at 14 general intensive care units (ICUs). All these results suggested that the grade of AGI was useful for identifying the severity of gastrointestinal function and could be used as a predictor of sepsis of abdominal origin. In our study, ROC analysis indicated a strong predictive power for PRISM III (area under the curve = 0.91).
Similarly, Bilan et al. [18] found that the PRISM III score had a strong predictive power (AUC = 0.898) and was well fit for the designed study (goodness-of-fit P = 0.161). In another study, PRISM III score (AUC=0.82, P < 0.001) and Lactate level (AUC = 0.79, P < 0.001) predicted mortality in 1109 critically ill children.
Our study has several limitations. First, This is a retrospective study with a limited number of patients from a single center. Second, long-term follow-up data was unavailable. If available, we would enable a more complete and robust analysis as to the risk of mortality in the future.

Conclusion
In summary, our study demonstrates that higher PRISM III score and the higher grade of   The relationship between the grade of AGI and the 28-day survival rate of pediatric patients with sepsis of abdominal origin.
21 Figure 5 The relationship between PRISM III and the 28-day survival rate of pediatric patients with sepsis of abdominal origin.

Supplementary Files
This is a list of supplementary files associated with the primary manuscript. Click to download.