Machine Learning Prediction of Prediabetes in a Young Male Chinese Cohort with 5.8-Year Follow-Up.

The identification of risk factors for future prediabetes in young men remains largely unexamined. This study enrolled 6247 young ethnic Chinese men with normal fasting plasma glucose at the baseline (FPGbase), and used machine learning (Mach-L) methods to predict prediabetes after 5.8 years. The study seeks to achieve the following: 1. Evaluate whether Mach-L outperformed traditional multiple linear regression (MLR). 2. Identify the most important risk factors. The baseline data included demographic, biochemistry, and lifestyle information. Two models were built, where Model 1 included all variables and Model 2 excluded FPGbase, since it had the most profound effect on prediction. Random forest, stochastic gradient boosting, eXtreme gradient boosting, and elastic net were used, and the model performance was compared using different error metrics. All the Mach-L errors were smaller than those for MLR, thus Mach-L provided the most accurate results. In descending order of importance, the key factors for Model 1 were FPGbase, body fat (BF), creatinine (Cr), thyroid stimulating hormone (TSH), WBC, and age, while those for Model 2 were BF, white blood cell, age, TSH, TG, and LDL-C. We concluded that FPGbase was the most important factor to predict future prediabetes. However, after removing FPGbase, WBC, TSH, BF, HDL-C, and age were the key factors after 5.8 years.

Impact of prior pulmonary tuberculosis in treatment outcomes of HCAP and CAP patients in intensive care units.

BACKGROUND/PURPOSE: It is controversial whether healthcare-associated pneumonia (HCAP) belongs to a unique clinical entity or it shares common characteristics with community-acquired pneumonia (CAP). The impact of prior pulmonary tuberculosis (PTB) in clinical presentation and treatment outcome of ICU-admitted CAP and HCAP patients also remains unknown. METHODS: We report a nationwide, multi-center, retrospective study. ICU-admitted CAP and HCAP patients from six medical centers in Taiwan were enrolled for analysis. Patients were defined as either CAP or HCAP cases, and with and without prior PTB, according to the database of Taiwan CDC. The disease severity, microbiologic characteristics, and treatment outcomes between CAP and HCAP patients with or without prior PTB were compared and analyzed. RESULTS: A total of 414 ICU-admitted patients, including 176 CAP cases and 238 HCAP cases were included for analysis during the study period. In both CAP and HCAP subgroups, the pneumonia severities, proportions of organ dysfunction, and microbiologic characteristics were similar between patients with and without prior PTB. In survival analysis, patients with prior PTB had higher 30-day mortality than those without prior PTB (38.9% vs. 16.5%, p = 0.021) in the CAP population. Multivariate analysis revealed that a history of prior PTB was an independent clinical factor associated with higher 30-day mortality rate in CAP patients (HR = 4.45, 95% CI: 1.81-10.98, P = 0.001). CONCLUSION: History of prior PTB is an independent clinical factor for increased 30-day mortality rate in ICU-admitted CAP patients, but not in ICU-admitted HCAP patients.

Predictors of Early Onset Multiple Organ Dysfunction in Major Burn Patients with Ventilator Support: Experience from A Mass Casualty Explosion.

Organ dysfunction is common in patients with major burns and associated with poor outcomes. The risk factors for early onset multiple organ dysfunction syndrome (MODS) in major burn patients with invasive ventilator support has rarely been evaluated before. In this study, major burn patients with invasive ventilator support from 499 victims suffered in a mass casualty color dust explosion were retrospectively enrolled. The development of early MODS that occurred within 5 days after burn injury was determined and the risk factors associated with early MODS were analyzed. A total of 88 patients from five medical centers were included. Their mean total body surface area (TBSA) was 60.9 ± 15.8%, and 45 (51.1%) patients had early MODS. Hematologic failure was the most common organ failure (68.6%), followed by respiratory failure (48.9%). Independent clinical factors associated with early MODS included TBSA ≥55% (OR: 3.83; 95% CI: 1.29-11.37) and serum albumin level <2.1 g/dL upon admission (OR: 3.43; 95% CI: 1.01-11.57). Patients with early MODS had prolonged ventilator dependence and longer ICU admission than those without early MODS. Our results showed that early MODS in major burn patients with invasive ventilator support is very common and can be predicted early on admission.