Red blood cell distribution width and mortality and hospitalizations in peritoneal dialysis patients

Interpretable machine learning models for predicting 90-day death in patients in the intensive care unit with epilepsy

PURPOSE

This study aims to develop a machine learning-based model for predicting mortality risk in patients with epilepsy admitted to the intensive care unit (ICU), providing clinicians with an accurate prognostic tool to guide individualized treatment.

METHODS

We collected clinical data from clinical databases (MIMIC IV and eICU-CRD) of epilepsy patients 24 h after ICU admission. The clinical characteristics of ICU patients with epilepsy were carefully feature selected and processed. MIMIC IV as the training set and eICU-CRD database as the test set. Six models were developed and validated, and the best LightGBM model was selected by performance comparison and analysed for interpretability.

RESULTS

The final cohort comprised 429 patients for training and 1217 for testing. The training set exhibited a 90-day mortality rate of 9.32 %, and the test set had an in-hospital 90-day mortality rate of 4.10 %. Utilizing the LightGBM model, we achieved an AUC of 0.956 in the training set. External validation demonstrated promising results with accuracy of 0.898, precision of 0.975, AUC of 0.781, F1 score of 0.945, highlighting the model's potential for guiding clinical decision-making. Significant factors influencing model performance included the severity of illness, as measured by the OASIS score, and clinical parameters like heart rate and body temperature.

CONCLUSION

This study introduces a machine learning-based approach to predict mortality risk in ICU epilepsy patients, offering a valuable tool for clinicians to identify high-risk individuals and devise personalized treatment strategies, thus improving patient prognosis and treatment outcomes.

Association of Red Blood Cell Distribution Width and Neutrophil-to-Lymphocyte Ratio with Calcification and Cardiovascular Markers in Chronic Kidney Disease

We aimed to investigate the association between Red Blood Cell Distribution Width (RDW) and Neutrophil-to-Lymphocyte Ratio (NLR), simple, rapidly assessed markers from the complete blood count with vascular calcification (VC)/stiffness and cardiovascular disease (CVD) in chronic kidney disease (CKD). Dephosphorylated, uncarboxylated matrix Gla-protein (dp-ucMGP), and central/peripheral hemodynamics' parameters were measured in 158 CKD patients, including Hemodialysis and Peritoneal Dialysis. Spearman's rho analysis showed that RDW correlated with C-reactive protein (CRP) (r = 0.29, p < 0.001), dp-ucMGP (r = 0.43, p = < 0.0001), central diastolic blood pressure (DBP) (r = -0.19, p = 0.02), and albuminuria (r = -0.17, p = 0.03). NLR correlated with the duration of CVD (r = 0.32, p < 0.001), CRP (r = 0.27, p = 0.01), dp-ucMGP (r = 0.43, p < 0.0001), central DBP (r = -0.32, p < 0.0001) and eGFR (r = -0.25, p = 0.04). In multiple regression models, circulating dp-ucMGP was an independent predictor of RDW (β = 0.001, p = 0.001) and NLR (β = 0.002, p = 0.002). In CKD patients, RDW and NLR are associated with traditional and novel markers of VC and CVD.

Red blood cell distribution width at admission predicts outcome in critically ill patients with kidney failure: a retrospective cohort study based on the MIMIC-IV database

PURPOSE

We aimed to explore whether red blood cell distribution width (RDW) could serve as a biomarker to predict outcomes in critically ill patients with kidney failure in this study.

MATERIALS AND METHODS

This retrospective study was conducted with the Medical Information Mart for Intensive Care IV (MIMIC-IV).A total of 674 patients were divided into three groups based on tertiles of RDW. We used the generalized additive model, Kaplan-Meier curve, and Cox proportional hazards models to evaluate the association between RDW and clinical outcomes. We then performed subgroup analyses to investigate the stability of the associations between RDW and all-cause mortality.

RESULTS

Nonlinear and J-shaped curves were observed in the generalized additive model. Kaplan-Meier analysis showed that patients with elevated RDW had a lower survival rate. The Cox regression model indicated that high levels of RDW were most closely associated with ICU mortality and 30-day mortality (HR = 4.71, 95% CI: 1.69-11.64 and HR = 6.62, 95% CI: 2.84-15.41). Subgroup analyses indicated that the associations between RDW and all-cause mortality were stable.

CONCLUSIONS

Elevated levels of RDW were associated with an increased risk of all-cause mortality, and RDW could be an independent prognostic factor for kidney failure.

Use of the systemic inflammation response index (SIRI) as a novel prognostic marker for patients on peritoneal dialysis

BACKGROUND

The systemic inflammatory response index (SIRI), a novel inflammation maker, has proven to be associated with prognostic outcomes in various diseases. However, few studies have been conducted assessing how SIRI may influence outcomes of patients on peritoneal dialysis (PD). Herein, we assessed the predictive value of SIRI on mortality all-cause mortality, including cardiovascular disease (CVD) in PD patients.

METHODS

A total of 646 PD patients were enrolled in this study. PD patients received regular PD treatments at the Zhujiang Hospital from 1 January 2011 to 31 December 2018. SIRI values could be computed as follows: neutrophil count × monocyte count/lymphocyte count. Patients were divided into two groups according to the median level of SIRI. Cox regression analysis and Kaplan-Meier methods were applied to analyze the relationship between SIRI and mortality outcomes in PD patients.

RESULTS

During the median 31-month follow-up period, 97 (15.0%) PD patients died from all-causes, and 47 (49.0%) died of CVD. Kaplan-Meier analyses revealed that a high SIRI corresponded to the high mortality of all-cause deaths, including CVD (both p < 0.001) in patients on PD. After adjusting for potential confounders, the higher SIRI level was significantly associated with an increased all-cause mortality (HR: 2.007, 95% CI: 1.304-3.088, p = 0.002) and cardiovascular mortality (HR: 2.847, 95% CI: 1.445-5.608, p = 0.002).

CONCLUSIONS

SIRI was a promising predictor of mortality in PD patients, with a higher SIRI corresponding to increased risk of mortality.

Established risk prediction models for the incidence of a low lean tissue index in patients with peritoneal dialysis

Objective

The objective of this study is to investigate the incidence of low lean tissue index (LTI) and the risk factors for low LTI in peritoneal dialysis (PD) patients, including to establish risk prediction models.

Methods

A total of 104 PD patients were enrolled from October 2019 to 2021. LTI was measured by bioimpedance spectroscopy. Multivariate logistic regression and machine learning were used to analyze the risk factors for low LTI in PD patients. Kaplan–Meier analysis was used to analyze the survival rate of patients with low LTI.

Results

The interleukin-6 (IL-6) level, red cell distribution width (RDW), overhydration, body mass index (BMI), and the subjective global assessment (SGA) rating significantly differed between the low LTI and normal LTI groups (all p < 0.05). Multivariate logistic regression showed that IL-6 (1.10 [95% CI: 1.02–1.18]), RDW (1.87 [95% CI: 1.18–2.97]), BMI (0.97 [95% CI: 0.68–0.91]), and the SGA rating (6.33 [95% CI: 1.59–25.30]) were independent risk factors for LTI. Cox regression analysis showed that low LTI (HR 3.14, [95% CI: 1.12–8.80]) was the only significant risk factor for all-cause death in peritoneal dialysis patients. The decision process to predict the incidence of low LTI in PD patients was established by machine learning, and the area under the curve of internal validation was 0.6349.

Conclusions

Low LTI is closely related to mortality in PD patients. Microinflammatory status, high RDW, low BMI and low SGA rating are risk factors for low LTI in PD patients. The developed prediction model may serve as a useful tool for assessing low LTI in PD patients.

Red blood cell distribution width: A severity indicator in patients with COVID-19

Red blood cell distribution width (RDW) was frequently assessed in COVID-19 infection and reported to be associated with adverse outcomes. However, there was no consensus regarding the optimal cutoff value for RDW. Records of 98 patients with COVID-19 from the First People's Hospital of Jingzhou were reviewed. They were divided into two groups according to the cutoff value for RDW on admission by receiver operator characteristic curve analysis: ≤11.5% (n = 50) and >11.5% (n = 48). The association of RDW with the severity and outcomes of COVID-19 was analyzed. The receiver operating characteristic curve indicated that the RDW was a good discrimination factor for identifying COVID-19 severity (area under the curve = 0.728, 95% CI: 0.626-0.830, p < 0.001). Patients with RDW > 11.5% more frequently suffered from critical COVID-19 than those with RDW ≤ 11.5% (62.5% vs. 26.0%, p < 0.001). Multivariate logistic regression analysis showed RDW to be an independent predictor for critical illness due to COVID-19 (OR = 2.40, 95% CI: 1.27-4.55, p = 0.007). A similar result was obtained when we included RDW > 11.5% into another model instead of RDW as a continuous variable (OR = 5.41, 95% CI: 1.53-19.10, p = 0.009). RDW, as an inexpensive and routinely measured parameter, showed promise as a predictor for critical illness in patients with COVID-19 infection. RDW > 11.5% could be the optimal cutoff to discriminate critical COVID-19 infection.

Red blood cell distribution width and peritoneal dialysis-associated peritonitis prognosis

Objective

Red blood cell distribution width (RDW) is a parameter of the heterogeneity of circulating erythrocyte size. Recent researches have pointed out a link among RDW, chronic kidney disease, and inflammation. We sought to investigate the prognostic value of baseline RDW in patients with peritoneal dialysis-associated peritonitis (PDAP).

Methods

Our study included 337 peritonitis episodes experienced by 202 patients who were undergoing continuous ambulatory peritoneal dialysis (CAPD) at a single center from 2013 to 2018. Episodes were categorized according to the tertiles of baseline RDW levels (T1, <13.2%; T2, 13.2−14.3%; T3, >14.3%). Routine logistic regression and generalized estimating equation (GEE) were used to estimate the association between RDW and treatment failure, which was defined as relapse/recurrent episodes, catheter removal, or death during therapy.

Results

After adjusting for other potential predictors, RDW exhibited an incremental relationship with the risk of treatment failure. The baseline RDW of T3 indicated a 43% and 52% increased venture of treatment failure in logistic and GEE analyses, respectively, compared with T1. As a continuous variable, the fitting curve based on restricted cubic spiline showed that the relationship was nonlinearly but positively correlated. The multivariate model A (combined RDW with baseline age, albumin, serum ferritin, and duration on CAPD) showed an area under the curve of 0.671 (95% confidence interval, 0.5920.749) for the prediction of treatment failure.

Conclusions

A Higher baseline level of RDW was significantly associated with a greater rate of treatment failure among PDAP episodes independent of other potential predictors.

Red Blood Cell Distribution Width Is Associated with Deterioration of Renal Function and Cardiovascular Morbidity and Mortality in Patients with Diabetic Kidney Disease

We sought to investigate the possible association between Red Blood Cell Distribution Width (RDW), vascular calcification, oxidative stress and renal function and all-cause/cardiovascular (CV) mortality, CV events and progression of kidney disease in a cohort of patients with Diabetic Kidney Disease (DKD). Carotid intima media thickness (cIMT) and oxidized low-density cholesterol were measured in 104 Type 2 Diabetes Mellitus (T2DM) patients with established DKD, distributed in all five stages of kidney disease and 38 diabetics with normal renal function. All patients were followed for 7 years with end-points all-cause and CV mortality, CV events and progression to End-Stage Renal Disease (ESRD). RDW was positively correlated with diabetes duration (r = 0.19, p = 0.023) and albuminuria (r = 0.29, p = 0.002). Multivariate regression analysis revealed that RDW was a strong, independent predictor of cIMT value (β = 0.031, p = 0.012). Kaplan-Meier curves and Cox proportional hazard models revealed that after adjustment for several cofounders, RDW was a significant and independent predictor for all-cause mortality, CV mortality, CV event and progression to ESRD (HR 1.75, p = 0.001, HR 2.03, p = 0.001, HR = 1.66, p < 0.0001 and HR 2.14, p = 0.001 respectively). RDW predicts mortality, CV events and deterioration of renal function in DKD, probably reflecting atherosclerosis.

Association between red blood cell distribution width and long-term mortality among patients undergoing percutaneous coronary intervention with previous history of cancer

Background: The number of patients suffering from coronary heart disease with cancer is rising. There is scarce evidence concerning the biomarkers related to prognosis among patients undergoing percutaneous coronary intervention (PCI) with cancer. Thus, the aim of this study was to investigate the association between red blood cell distribution width (RDW) and prognosis in this population.Methods: A total of 172 patients undergoing PCI with previous history of cancer were enrolled in this retrospective study. The endpoint was long-term all-cause mortality. According to tertiles of RDW, the patients were classified into three groups: Tertile 1 (RDW <12.8%), Tertile 2 (RDW ≥12.8% and <13.5%) and Tertile 3 (RDW ≥13.5%).Results: During an average follow-up period of 33.3 months, 29 deaths occurred. Compared with Tertile 3, mortality of Tertile 1 and Tertile 2 was significantly lower in the Kaplan-Meier analysis. In multivariate Cox regression analysis, RDW remained an independent risk factor of mortality (HR: 1.938, 95% CI: 1.295-2.655, p < 0.001). The all-cause mortality in Tertile 3 was significantly higher than that in Tertile 1 (HR: 5.766; 95% CI: 1.426-23.310, p = 0.014).Conclusions: An elevated RDW level (≥13.5%) was associated with long-term all-cause mortality among patients undergoing PCI with previous history of cancer.