Prediction of the risk of developing end-stage renal diseases in newly diagnosed type 2 diabetes mellitus using artificial intelligence algorithms.

OBJECTIVES: Type 2 diabetes mellitus (T2DM) imposes a great burden on healthcare systems, and these patients experience higher long-term risks for developing end-stage renal disease (ESRD). Managing diabetic nephropathy becomes more challenging when kidney function starts declining. Therefore, developing predictive models for the risk of developing ESRD in newly diagnosed T2DM patients may be helpful in clinical settings. METHODS: We established machine learning models constructed from a subset of clinical features collected from 53,477 newly diagnosed T2DM patients from January 2008 to December 2018 and then selected the best model. The cohort was divided, with 70% and 30% of patients randomly assigned to the training and testing sets, respectively. RESULTS: The discriminative ability of our machine learning models, including logistic regression, extra tree classifier, random forest, gradient boosting decision tree (GBDT), extreme gradient boosting (XGBoost), and light gradient boosting machine were evaluated across the cohort. XGBoost yielded the highest area under the receiver operating characteristic curve (AUC) of 0.953, followed by extra tree and GBDT, with AUC values of 0.952 and 0.938 on the testing dataset. The SHapley Additive explanation summary plot in the XGBoost model illustrated that the top five important features included baseline serum creatinine, mean serum creatine within 1 year before the diagnosis of T2DM, high-sensitivity C-reactive protein, spot urine protein-to-creatinine ratio and female gender. CONCLUSIONS: Because our machine learning prediction models were based on routinely collected clinical features, they can be used as risk assessment tools for developing ESRD. By identifying high-risk patients, intervention strategies may be provided at an early stage.

Associations of the Serum Total Carbon Dioxide Level with Long-Term Clinical Outcomes in Sepsis Survivors.

INTRODUCTION: Sepsis is characterized by a dysregulated host response to infection that leads to multiple organ dysfunction and often complicated with metabolic acidosis. However, the associations between serum total carbon dioxide level (TCO2) and long-term clinical outcomes in sepsis survivors remains unknown. METHODS: A total of 7212 sepsis survivors aged ≥ 20 years who were discharged from January  1, 2008 to December 31, 2018 were included in our analyses. The sepsis survivors were further divided into high TCO2 (≥ 18 mmol/L) and low TCO2 (< 18 mmol/L) groups, comprising 5023 and 2189 patients, respectively. The following outcomes of interest were included: all-cause mortality, myocardial infarction, ischemic stroke, hospitalization for heart failure, ventricular arrhythmia, and end-stage renal disease (ESRD). RESULTS: After propensity score matching, the low TCO2 group was at higher risks of all-cause mortality (hazard ratio [HR] 1.28, 95% confidence interval [95% CI] 1.18-1.39), myocardial infarction (HR 1.83, 95% CI 1.39-2.43), and ESRD (HR 1.38, 95% CI 1.16-1.64) than the high TCO2 group. The results remained similar after considering death as a competing risk. CONCLUSION: Patients discharged from hospitalization for sepsis have higher risks of worse long-term clinical outcomes. Physicians may need to pay more attention to sepsis survivors whose TCO2 was low.

Echocardiographic features of left ventricular dysfunction and outcomes in chronic kidney disease.

OBJECTIVE: Heart failure (HF) imposes a substantial burden and the prevalence of HF is high in patients with chronic kidney disease (CKD). HF results in multiple hospital admissions, but whether HF subtypes worsen long-term outcomes and renal function in patients with CKD remains inconclusive. METHODS: The study comprised 10 904 patients with CKD aged ≥20 years who underwent echocardiography between 1 January 2011 and 31 December 2018. The patients were stratified into four groups: non-HF, HF with reduced ejection fraction (HFrEF), HF with mildly reduced ejection fraction (HFmrEF) and HF with preserved ejection fraction (HFpEF). The primary end points were all-cause mortality, major adverse cardiovascular events (MACEs) and adverse renal outcomes. RESULTS: In inverse probability of treatment weighting-adjusted method, the risk of all-cause mortality and MACEs relative to the non-HF group was greatest in the HFrEF group (HR 3.18 (95% CI 2.57 to 3.93) and HR 3.83 (95% CI 3.20 to 4.59)), followed by the HFmrEF (HR 2.75 (95% CI 2.22 to 3.42) and HR 3.08 (95% CI 2.57 to 3.69)) and HFpEF (HR 1.85 (95% CI 1.59 to 2.15) and HR 2.43 (95% CI 2.16 to 2.73) groups. In addition, the HFrEF group had the greatest risks of end-stage renal disease (HR 2.58 (95% CI 1.94 to 3.44)) compared with other groups. CONCLUSIONS: HF is associated with subsequent worse clinical outcomes, which may be more pronounced in patients with HFrEF, followed by those with HFmrEF and those with HFpEF relative to non-HF group.

CHD4 plays a critical role in arsenite-induced oxidative damage in human urothelial carcinoma.

Inorganic arsenic (iAs), a known human carcinogen, induces oxidative DNA damage and epigenetic silencing of tumor suppressor genes related to tumor progression. Chromodomain-helicase-DNA-binding protein 4 (CHD4) is a chromatin remodeling protein that acts on DNA repair and DNA methylation under oxidative damage in malignancies, but the role of CHD4 in arsenical urothelial carcinoma (UC) is unidentified. Our purpose was to observe CHD4-related repair effects on As-stimulated oxidative damage in human UC. The markers of oxidative DNA damage 8-hydroxy-2'-deoxyguanosine (8-OHdG) and CHD4 were investigated by immunohistochemistry in 45 UC tissues from non-blackfoot disease (BFD) areas and BFD areas respectively. The cellular mechanisms of CHD4 involved in the oxidative DNA repair and DNA methylation were evaluated by immunocytochemistry and western blot. The expressions of CHD4 and 8-OHdG were significantly increased in UC patients from the As-exposed areas. The underlying mechanism of CHD4-mediated DNA repair and DNA methylation involved the activation of zinc finger MYND-type containing 8 (ZMYND8) and DNA methyltransferase (DNMTs) in SV-HUC-1, T24 and BFTC-905 cells. These results highlight the potential clinical significance of CHD4 in UCs from BFD areas. The CHD4-mediated oxidative DNA repair and epigenetic DNA methylation in UC cells stimulated by arsenic was revealed. CHD4 might be used as a prognostic indicator in arsenical UC.

The Pattern of Hemoglobin A1C Trajectories and Risk of Herpes Zoster Infection: A Follow-Up Study.

To investigate the risks of herpes zoster (HZ) infection among heterogeneous HbA1C trajectories of patients with newly diagnosed type 2 diabetes, this cohort study used data from the Chang Gung Research Database (CGRD), from the 10-year period of 1 January 2007 to 31 December 2017. We applied group-based trajectory modeling (GBTM) to identify the patterns of HbA1C trajectories, and multiple Cox proportional hazards regressions were used to estimate the hazard ratio (HR) for the risk of HZ infection with adjustment of age, sex, and comorbidities. This study enrolled 121,999 subjects to perform the analysis. The GBTM identified four HbA1C trajectories: 'good control' (58.4%), 'high decreasing' (8.9%), 'moderate control' (25.1%), and 'poor control' (7.6%) with the mean HbA1C of 6.7% (50 mmol/mol), 7.9% (63 mmol/mol), 8.4% (68 mmol/mol), and 10.7% (93 mmol/mol) respectively. The risk of HZ was significantly higher in the poor control with an HR = 1.44 (95% CI 1.26-1.64) after adjustment for confounders and comorbidities. The risk of HZ infection for the high decreasing group (initially poor then rapidly reaching optimal control) was nonsignificant compared to the good control group. In conclusion, the patients with poor glycemic control (mean HbA1C = 10.7%) had the highest risk of HZ infection. The patients with initial hyperglycemia then reaching optimal control could have a lower risk of HZ infection.

Could prokinetic agents protect long-term nasogastric tube-dependent patients from being hospitalized for pneumonia? A nationwide population-based case-crossover study.

BACKGROUND: Some studies have indicated that the use of prokinetic agents may reduce pneumonia risk in some populations. Nasogastric tube insertion is known to increase the risk of pneumonia because it disrupts lower esophageal sphincter function. The aim of this study was to evaluate whether prokinetic agents could protect long-term nasogastric tube-dependent patients in Taiwan from being hospitalized for pneumonia. METHODS: A case-crossover study design was applied in this study. Long-term nasogastric tube-dependent patients who had a first-time admission to a hospital due to pneumonia from 1996 to 2013 that was recorded in the Taiwan National Health Insurance Research Database were included. The case period was set to be 30 days before admission, and two control periods were selected for analysis. Prokinetic agent use during those three periods was then assessed for the included patients. Conditional logistic regression was used to calculate the odds ratio (OR) for pneumonia admission with the use of prokinetic agents. RESULTS: A total of 639 first-time hospitalizations for pneumonia among patients with long-term nasogastric tube dependence were included. After adjusting the confounding factors for pneumonia, no negative association between prokinetic agent use and pneumonia hospitalization was found, and the adjusted OR was 1.342 (95% CI 0.967-1.86). In subgroup analysis, the adjusted ORs were 1.401 (0.982-1.997), 1.256 (0.87-1.814), 0.937 (0.607-1.447) and 2.222 (1.196-4.129) for elderly, stroke, diabetic and parkinsonism patients, respectively. CONCLUSION: Prokinetic agent use had no negative association with pneumonia admission among long-term nasogastric tube-dependent patients in Taiwan.