Analysis of risk factors for prolonged mechanical ventilation in patients with sepsis after abdominal surgery

Abstract

To analyze the risk factors of prolonged mechanical ventilation (PMV) in patients with sepsis complicated by abdominal surgery, and to evaluate the predictive value of risk factors for PMV. A retrospective case-control study was conducted. The clinical data of patients with postoperative abdominal sepsis complicated with invasive mechanical ventilation who were admitted to the intensive care unit (ICU) of the Affiliated Hospital of Guizhou Medical University from January 1, 2018 to December 31, 2020 were collected. The patients were divided into PMV group (duration of mechanical ventilation longer than 48 hours) and non-PMV group (duration of mechanical ventilation shorter than 48 hours) according to the duration of mechanical ventilation in ICU. The patient's gender, age, body mass index (BMI), underlying diseases, mean arterial pressure (MAP), complete blood count, blood biochemistry, arterial blood gas, cardiac function indicators, procalcitonin (PCT) at admission to the ICU, the acute physiology and chronic health evaluation II (APACHE II) and the sequential organ failure assessment (SOFA) scores in the first 24 hours of admission to the ICU, and other clinical information were recorded. Univariate and multivariate Logistic regression models were used to analyze the risk factors for PMV. Receiver operator characteristic curve (ROC curve) was drawn to evaluate the predictive value of related indicators for PMV. A total of 195 patients with sepsis after abdominal surgery who received invasive mechanical ventilation were enrolled, including 127 males (65.1%) and 68 females (34.9%), with the median age of 65 (21, 93) years old. There were 91 patients (46.7%) in the non-PMV group and 104 patients (53.3%) in the PMV group. Univariate analysis showed that the APACHE II score, SOFA score, cardiac troponin T (cTnT), N-terminal pro-B type natriuretic peptide (NT-proBNP) in the PMV group were significantly higher than those in the non-PMV group. Oxygenation index (PaO2/FiO2), total protein (TP) and prealbumin (PA) in the PMV group were all lower than those in the non-PMV group when admitted to ICU. In the PMV group, serum creatinine (SCr), blood urea nitrogen (BUN), cystatin C (Cys C) were significantly increased, prothrombin time (PT) was significantly prolonged, the proportion of patients with septic shock and hypertension were significantly increased as compared with those in the non-PMV group. Multivariate analysis showed that low PaO2/FiO2 at ICU admission [odds ratio (OR) = 0.995, 95% confidence interval (95%CI) was 0.992-0.999, P = 0.010], high ln PCT (OR = 1.301, 95%CI was 1.088-1.555, P = 0.004), high ln cTnT (OR = 1.562, 95%CI was 1.079-2.261, P = 0.018) and septic shock (OR = 4.967, 95%CI was 2.461-10.026, P = 0.000) were the independent risk factors for PMV in patients with sepsis after abdominal surgery. ROC curve analysis showed that the PaO2/FiO2, ln cTnT, ln PCT and septic shock had certain predictive value for PMV, the area under the ROC curve (AUC) of the four variables were 0.607, 0.638, 0.690 and 0.711, the sensitivity was 50.0%, 62.5%, 86.5% and 74.0%, and the specificity was 71.4%, 62.6%, 48.3% and 68.1%, respectively. The AUC for the joint prediction of the four variables was 0.803, with a sensitivity of 76.0% and a specificity of 78.0%. It suggested that the multivariate joint prediction of PMV was more accurate. Decreased PaO2/FiO2, increased PCT, increased cTnT and the occurrence of septic shock are independent risk factors for PMV in patients with sepsis complicated by abdominal surgery. The combination of above four indices was more accurate than one single variable in predicting PMV and had higher diagnostic value.