Relationship between blood glucose levels and length of hospital stay in patients with acute pancreatitis: An analysis of MIMIC-III database

Prediction Models of Severity in Acute Biliary Pancreatitis

Background: Acute pancreatitis is a common condition with a variable prognosis. While the overall mortality rate of acute pancreatitis is relatively low, ranging between 3 and 5% in most cases, severe forms can result in significantly higher morbidity and mortality. Therefore, early risk assessment is crucial for optimizing management and treatment. The aim of the present study wasto compare simple prognostic markers and identify the best predictors of severity in patients with acute pancreatitis. Material and Methods: A retrospective analysis was carried outon 108 patients admitted in our center during one year with acute biliary pancreatitis. Acute pancreatitis severity was stratified based on the revised Atlanta criteria. Results: 108 subjects (mean age of 60.1 ± 18.6, 65.7% females) diagnosed with acute biliary pancreatitis were included. Based on the Atlanta criteria, 59.3% (64/108) of the subjects were classified as having mild acute biliary pancreatitis, 35.2% (38/108) as having a moderate-severe pancreatitis, and 5.5% (6/108) were classified as having severe acute pancreatitis. In univariate analysis, the following parameterswere associatedwith at least a moderate-severe form of acute pancreatitis: Balthazar score, fasting blood glucose (mg/dL), modified CTSI score, CRP values at 48 h, BISAP score at admission, CTSI score, Ranson score, duration of hospitalization (days), and the presence of leukocytosis (×1000/µL) (all p < 0.05).BISAP score at admission (AUC-0.91), CRP levels at 48 h (AUC-0.92), mCTSI (AUC-0.94), and CTSI score (AUC-0.93) had the highest area under the curve (AUC) for predicting the severity of acute pancreatitis. In multivariate analysis, the model including the following independent parameters was predictive for the severity of acute pancreatitis: CTSI score (p < 0.0001), BISAP score (p = 0.0082), and CRP levels at 48 h (p = 0.0091), respectively. The model showed a slightly higher AUC compared to the independent predictors (AUC-0.96). Conclusions: The use of a multiparametric prediction model can increase the accuracy of predicting severity in patients with acute biliary pancreatitis.

Improved outcome prediction in acute pancreatitis with generated data and advanced machine learning algorithms

OBJECTIVES

Traditional scoring systems have been widely used to predict acute pancreatitis (AP) severity but have limitations in predictive accuracy. This study investigates the use of machine learning (ML) algorithms to improve predictive accuracy in AP.

METHODS

A retrospective study was conducted using data from 101 AP patients in a tertiary hospital in Türkiye. Data were preprocessed, and synthetic data were generated with Gaussian noise addition and balanced with the ADASYN algorithm, resulting in 250 cases. Supervised ML models, including random forest (RF) and XGBoost (XGB), were trained, tested, and validated against traditional clinical scores (Ranson's, modified Glasgow, and BISAP) using area under the curve (AUC), F1 score, and recall.

RESULTS

RF outperformed XGB with an AUC of 0.89, F1 score of 0.82, and recall of 0.82. BISAP showed balanced performance (AUC = 0.70, F1 = 0.44, and recall = 0.85), whereas the Glasgow criteria had the highest recall but lower precision (AUC = 0.70, F1 = 0.38, and recall = 0.95). Ranson's admission criteria were the least effective (AUC = 0.53, F1 = 0.42, and recall = 0.39), probable because it lacked the 48th h features.

CONCLUSION

ML models, especially RF, significantly outperform traditional clinical scores in predicting adverse outcomes in AP, suggesting that integrating ML into clinical practice could improve prognostic assessments.

Association Between High-Density Lipoprotein Cholesterol and Length of Hospital Stay in Acute Pancreatitis: A Retrospective Cohort Study

Background

Acute pancreatitis (AP) is a complex inflammatory disorder with varying degrees of severity, impacting patient recovery and healthcare resource utilization. The length of hospital stay (LOS) is a pivotal indicator of recovery, and identifying factors influencing LOS can offer insights into AP management. High-density lipoprotein cholesterol (HDL-C), known for its cardioprotective properties, has been posited to influence AP outcomes; however, its relationship with LOS remains unclear.

Objective

This study aimed to investigate the potential correlation between HDL-C levels and LOS in patients with AP, considering the effects of demographic factors, comorbidities, and other clinical parameters.

Methods

A retrospective cohort study was conducted. Data collection adhered to the STROBE guidelines, and baseline clinical and laboratory variables were analyzed. Statistical analysis comprised univariate and multivariate regression models, Generalized Additive Models (GAM), and stratified linear regression models to assess the relationship between HDL-C and LOS, while accounting for confounding factors.

Results

After adjusting for key confounders, including age, sex, BMI, WBC, HB, PLT, CRP, ALT, AMY, TB, GLU, LDL-C, SCR, BUN, ALB, Ca2+, and the presence of comorbidities such as hypertension, gallstones, diabetes mellitus, liver dysfunction, renal insufficiency, smoking and alcohol consumption, the study revealed a nonlinear relationship between HDL-C levels and LOS, with an inflection point at 1.5 mmol/L. Below this threshold, HDL-C was significantly and inversely correlated with LOS, whereas above this threshold, HDL-C was positively correlated with LOS. Subgroup analyses emphasized that in non-diabetic, non-alcoholic and non-hyperlipidemic pancreatitis patients, there is a negative correlation between HDL-C levels and LOS.

Conclusion

HDL-C exhibits a U-shaped relationship with LOS in patients with AP, suggesting that both low and high levels of HDL-C may influence hospital stay duration. These findings underscore the importance of considering HDL-C levels in the clinical management of AP. Especially in patients who are non-diabetic, non-hyperlipidemic, and non-alcoholic, the management of HDL-C may significantly reduce hospital stay.

The relationship between intraoperative glucose levels and length of hospital stay in patients with a femoral neck fracture: a retrospective study based on the MIMIC-IV database

Objective

This retrospective study aimed to explore the relationship between intraoperative glucose (IG) and the length of hospital stay (LOS) in patients with femoral neck fractures via the Medical Information Mart for Intensive Care-IV (MIMIC-IV) database.

Methods

A generalized additive model was performed to explore the relationship between IG levels and LOS. Restricted cubic spline curves were used to analyze the dose-response relationship between IG levels and prolonged LOS (or 7-day LOS). Threshold effect analysis was conducted to assess the key points influencing their association. Receiver operating characteristic (ROC) curve and decision curve analysis (DCA) were performed to evaluate the predictive performance of IG levels for LOS.

Results

A total of 743 patients with femoral neck fractures were enrolled from the MIMIC-IV database. We found that there was a non-linear relationship between IG and the LOS (or prolonged LOS/>7 days LOS). Moreover, their relationship was still significant even after adjusting for potential confounders. The threshold effect showed that IG was significantly related to a prolonged LOS when it was >137 mg/dl, and IG was significantly related to a 7-day LOS when it was >163 mg/dl. ROC showed that IG had a better function in predicting a 7-day LOS in participants with IG >163 mg/dl than in predicting a prolonged LOS among participants with IG >137 mg/dl. Moreover, the DCA results showed that IG can obtain a favorable net benefit in clinical settings in predicting a 7-day LOS among participants with IG >163 mg/dl.

Conclusions

In summary, there was a non-linear relationship between IG levels and LOS. In patients with IG levels >163 mg/dl, using IG content to predict an LOS >7 days had a good function.

Computed tomography-based body composition indicative of diabetes after hypertriglyceridemic acute pancreatitis

BACKGROUND

Post‑acute pancreatitis prediabetes/diabetes mellitus (PPDM‑A) is one of the common sequelae of acute pancreatitis (AP). The aim of our study was to build a machine learning (ML)-based prediction model for PPDM-A in hypertriglyceridemic acute pancreatitis (HTGP).

METHODS

We retrospectively enrolled 165 patients for our study. Demographic and laboratory data and body composition were collected. Multivariate logistic regression was applied to select features for ML. Support vector machine (SVM), linear discriminant analysis (LDA), and logistic regression (LR) were used to develop prediction models for PPDM-A.

RESULTS

65 patients were diagnosed with PPDM-A, and 100 patients were diagnosed with non-PPDM-A. Of the 84 body composition-related parameters, 15 were significant in discriminating between the PPDM-A and non-PPDM-A groups. Using clinical indicators and body composition parameters to develop ML models, we found that the SVM model presented the best predictive ability, obtaining the best AUC=0.796 in the training cohort, and the LDA and LR model showing an AUC of 0.783 and 0.745, respectively.

CONCLUSIONS

The association between body composition and PPDM-A provides insight into the potential pathogenesis of PPDM-A. Our model is feasible for reliably predicting PPDM-A in the early stages of AP and enables early intervention in patients with potential PPDM-A.

Eliminating COVID-19 as the immediate culprit for igniting pancreatitis

The impact of severe acute respiratory syndrome coronavirus 2 infection on the potential development of pancreatitis is a subject of ongoing debate within academic discourse. Establishing a causal link between COVID-19 and pancreatitis may not be fully supported by relying only on retrospective studies or case reports. This study examined the relationship between COVID-19 phenotypes and pancreatitis by Mendelian randomization (MR) method. The identification of instrumental variables (single nucleotide polymorphisms) that exhibit a robust association with the COVID-19 phenotypes was accomplished through a meticulous process of rigorous screening procedures. We included acute pancreatitis and chronic pancreatitis (CP) as the outcomes in the MR analysis, even though no definitive studies exist between COVID-19 and CP. A direct causal relationship between genetically predicted COVID-19 phenotypes and pancreatitis risk cannot be established. There is an ongoing debate over the designation of COVID-19 as a definitive cause of pancreatitis.

Development and Validation of a Dynamic Nomogram for Predicting in-Hospital Mortality in Patients with Acute Pancreatitis: A Retrospective Cohort Study in the Intensive Care Unit

Purpose

The aim of this study is to develop and validate a predictive model for the prediction of in-hospital mortality in patients with acute pancreatitis (AP) based on the intensive care database.

Patients and Methods

We analyzed the data of patients with AP in the Medical Information Mart for Intensive Care-IV (MIMIC-IV) database and Electronic Intensive Care Unit Collaborative Research Database (eICU-CRD). Then, patients from MIMIC-IV were divided into a development group and a validation group according to the ratio of 8:2, and eICU-CRD was assigned as an external validation group. Univariate logistic regression and least absolute shrinkage and selection operator regression were used for screening the best predictors, and multivariate logistic regression was used to establish a dynamic nomogram. We evaluated the discrimination, calibration, and clinical efficacy of the nomogram, and compared the performance of the nomogram with Acute Physiology and Chronic Health Evaluation II (APACHE-II) score and Bedside Index of Severity in AP (BISAP) score.

Results

A total of 1030 and 514 patients with AP in MIMIC-IV database and eICU-CRD were included in the study. After stepwise analysis, 8 out of a total of 37 variables were selected to construct the nomogram. The dynamic nomogram can be obtained by visiting https://model.sci-inn.com/KangZou/. The area under receiver operating characteristic curve (AUC) of the nomogram was 0.859, 0.871, and 0.847 in the development, internal, and external validation set respectively. The nomogram had a similar performance with APACHE-II (AUC = 0.841, p = 0.537) but performed better than BISAP (AUC = 0.690, p = 0.001) score in the validation group. Moreover, the calibration curve presented a satisfactory predictive accuracy, and the decision curve analysis suggested great clinical application value of the nomogram.

Conclusion

Based on the results of internal and external validation, the nomogram showed favorable discrimination, calibration, and clinical practicability in predicting the in-hospital mortality of patients with AP.