Intra-abdominal infection in acute pancreatitis in eastern China: microbiological features and a prediction model

Abstract

Background: This study aimed to investigate the microbiol distribution of intra-abdominal infection in patients with acute pancreatitis, and to develop a reliable prediction model to guide the use of antibiotics.

Methods: Inpatient with acute pancreatitis between January 2015 and June 2020 were enrolled in the study. Participants were divided into the intra-abdominal infection group and non-infection group. Isolated pathogens and antibiotic susceptibility were documented. Characteristics parameters, laboratory results, and outcomes were also compared. Least absolute shrinkage and selection operator (LASSO) regression model was used to select the risk factors associated with intra-abdominal infection in patients with acute pancreatitis. Logistic regression analysis, random forest model, and artificial neural network were also used to validate the performance of the selected predictors in intra-abdominal infection prediction. A novel nomogram based on selected predictors was established to provide individualized risk of developing intra-abdominal infection in patients with acute pancreatitis.

Results: A total amount of 711 participants were enrolled in the study, and of these, 182 (25.6%) had intra-abdominal infection. Of the 247 isolated pathogens, 45 (18.2%) were multidrug-resistant bacteria, and antibiotic susceptibility was lower than that of China Antimicrobial Surveillance Network 2020. The LASSO method identified 5 independent predictors [intra-abdominal pressure (IAP), acute physiology and chronic health evaluation II (APACHE II), computed tomography severity index (CTSI), the severity of pancreatitis, and intensive care unit (ICU) admission] of intra-abdominal infection, which were validated by three different models. The area under the curve was >0.95 for all 5 predictors. A clinically useful nomogram based on these predictors was successfully established.

Conclusions: Multidrug-resistant bacteria were quite common in intra-abdominal infection. IAP, APACHE II, CTSI, the severity of pancreatitis, and ICU admission were identified as risk factors and the new nomogram based on these could help clinicians estimate the risk of intra-abdominal infection and optimize antimicrobial prescription for acute pancreatitis patients.

Keywords: Acute pancreatitis; intra-abdominal infection; microbiology; prediction model.

Figure 1

Flow chart of the study.

Figure 2

Antimicrobial susceptibility of the four main gram-negative bacteria. (A) Escherichia coli; (B) Klebsiella pneumoniae; (C) Pseudomonas aeruginosa; (D) Acinetobacter baumannii.

Figure 3

Ninety-day survival curves of hospitalized patients with acute pancreatitis (intra-abdominal infection vs. non intra-abdominal infection).

Figure 4

Selection of risk factors of intra-abdominal infection using the LASSO logistic regression algorithm. (A) LASSO coefficient profiles of the 43 candidate variables. Vertical line was plotted at the given lambda, selected by 10-fold cross-validation with minimum classification error and minimum classification error plus 1 standard error, respectively. For the optimal lambda that gives minimum classification error plus 1 standard error, 5 features with a non-0 coefficient were selected. (B) Penalization coefficient lambda in the LASSO model was tuned using 10-fold cross-validation and the “lambda.1se” criterion. Area under the curve (AUC) metrics (y-axis) were plotted against log(lambda) (bottom x-axis). Top x-axis indicates the number of predictors for the given log(lambda). Red dots indicate average AUC for each model at the given lambda, and vertical bars through the red dots show the upper and lower values of the AUC according to the 10-fold cross-validation. Vertical black lines define the optimal lambda that gives the minimum classification error plus 1 standard error.

Figure 5

Performance of the logistic regression algorithm in intra-abdominal infection prediction. (A,B) Receiver-operating characteristic curves; (C,D) calibration curves.

Figure 6

Development and assessment of the random forest algorithm in intra-abdominal infection prediction. (A) Relationship between out-of-bag error and number of trees. In total, 114 trees are selected to establish a random forest model; (B) feature importance; (C,D) receiver-operating characteristic curves; (E,F) calibration curves.

Figure 7

Development and assessment of the artificial neural network algorithm in intra-abdominal infection prediction. (A) Structure of artificial neural network; (B) feature importance; (C,D) receiver-operating characteristic curves; (E,F) calibration curves.

Figure 8

Nomogram to predict intra-abdominal infection was developed using the LASSO model selected predictors (intra-abdominal pressure, APACHE II score, CTSI, ICU admission, and severity grade were identified as risk factors). ***, P<0.001. APACHE II, Acute Physiology and Chronic Health Evaluation II; CTSI, CT severity index; ICU, Intensive care unit.