A prediction model for predicting the risk of acute respiratory distress syndrome in sepsis patients: a retrospective cohort study

Abstract

Background The risk of death in sepsis patients with acute respiratory distress syndrome (ARDS) was as high as 20–50%. Few studies focused on the risk identification of ARDS among sepsis patients. This study aimed to develop and validate a nomogram to predict the ARDS risk in sepsis patients based on the Medical Information Mart for Intensive Care IV database. Methods A total of 16,523 sepsis patients were included and randomly divided into the training and testing sets with a ratio of 7:3 in this retrospective cohort study. The outcomes were defined as the occurrence of ARDS for ICU patients with sepsis. Univariate and multivariate logistic regression analyses were used in the training set to identify the factors that were associated with ARDS risk, which were adopted to establish the nomogram. The receiver operating characteristic and calibration curves were used to assess the predictive performance of nomogram. Results Totally 2422 (20.66%) sepsis patients occurred ARDS, with the median follow-up time of 8.47 (5.20, 16.20) days. The results found that body mass index, respiratory rate, urine output, partial pressure of carbon dioxide, blood urea nitrogen, vasopressin, continuous renal replacement therapy, ventilation status, chronic pulmonary disease, malignant cancer, liver disease, septic shock and pancreatitis might be predictors. The area under the curve of developed model were 0.811 (95% CI 0.802–0.820) in the training set and 0.812 (95% CI 0.798–0.826) in the testing set. The calibration curve showed a good concordance between the predicted and observed ARDS among sepsis patients. Conclusion We developed a model incorporating thirteen clinical features to predict the ARDS risk in patients with sepsis. The model showed a good predictive ability by internal validation.

Critical evaluation of established risk prediction models for acute respiratory distress syndrome in adult patients: A systematic review and meta‐analysis

“…The prediction models were developed for ARDS prediction in the following patients: patients after surgery, 23,28,30 patients in emergency department, 25 patients in intensive care unit, 24,[31][32][33]36,37,45,47 patients with severe burns, 39 patients with severe injury, 27 patients with COVID-19, 38,40,42 patients with acute pulmonary edema, 25,26 patients with moderate hypoxia, 35 patients with severe acute pancreatitis 34,41,43 (Table 1).…”

“…ARDS in 16 studies, [31][32][33][34][35][36][37][38][39][40][41][42][43][45][46][47] ALI in 4 23,25,26,44 studies, ALI or ARDS in 5 studies. 23,24,26,28,30 The diagnosis of ARDS or ALI relied mainly on the Berlin or AECC criteria in the included studies, and Berlin in 15 studies, [32][33][34][35][36][37][38][39][40][41][43][44][45][46][47] and AECC in 9 studies. [23][24][25][26][27][28][29][30][31] In addition, 17 studies defined the follow-up duration (Table 1) [23]<...…”

“…27,29,31,36,39 For the models' development, predictors were mainly selected a priori, based on clinical knowledge or literature. Most studies used data-driven statistical methods to select predictors for final prediction models, [23][24][25][26][27][28][29][30][31][32]34,36,38,39,41,[43][44][45]47 including univariate or bivariate selection of statistically significant predictors, 24,[27][28][29]31,32,34,36,44,45 "stepwise" elimination, [26][27][28][29]38,44 and bootstrapping, 28 and the least absolute shrinkage and selection operator(LASSO) regression method. 43 Predictor selection was not described in five studies.…”

“…In total, 21 models were established by logistic regression methods, [23][24][25][26][27][28][29][30][31][32][34][35][36]38,39,41,[43][44][45]47 and seven models applied the XGBoost model (Table 2). It should be noted that Blum et al estab-lished two prediction models based on logistic regression, including the preoperative prediction model and the intraoperative prediction model.…”

“…25 Nine studies reported model calibration graphically using calibration plots. 27,35,38,39,[41][42][43]45,47 In total, 16 studies reported both discrimination and calibration measures.…”