ÌÇÐÄlogoÈë¿ÚÃâ·Ñ¹Û¿´

RESULTSWe utilized data from 693 patients who contributed 42 656 hemodialysis sessions and 355 693 intradialytic SBP measurements. IDH occurred in 16.2% of hemodialysis treatments. Our model predicted IDH 15-75 min in advance with an AUROC of 0.89. Top IDH predictors were the most recent intradialytic SBP and IDH rate, as well as mean nadir SBP of the previous 10 dialysis sessions.CONCLUSIONSReal-time prediction of IDH during an ongoing hemodialysis session is feasible and has a clinically actionable predictive performance. If and to what degree this predictive information facilitates the timely deployment of preventive interventions and translates into lower IDH rates and improved patient outcomes warrants prospective studies.BACKGROUNDIn maintenance hemodialysis patients, intradialytic hypotension (IDH) is a frequent complication that has been associated with poor clinical outcomes. Prediction of IDH may facilitate timely interventions and eventually reduce IDH rates.METHODSWe developed a machine learning model to predict IDH in in-center hemodialysis patients 15-75 min in advance. IDH was defined as systolic blood pressure (SBP) <90 mmHg. Demographic, clinical, treatment-related and laboratory data were retrieved from electronic health records and merged with intradialytic machine data that were sent in real-time to the cloud. For model development, dialysis sessions were randomly split into training (80%) and testing (20%) sets. The area under the receiver operating characteristic curve (AUROC) was used as a measure of the model's predictive performance.

RESULTSWe analyzed 4,075 consecutive 5-minute segments from 89 hemodialysis treatments in 22 hemodialysis patients. While 891 (21.9%) segments showed saw-tooth pattern, 3,184 (78.1%) did not. In the test data set, the rate of correct SaO2 pattern classification was 96% with an area under the receiver operating curve of 0.995 (95% CI: 0.993 to 0.998).CONCLUSIONOur 1D-CNN algorithm accurately classifies SaO2 saw-tooth pattern. The SaO2 pattern classification can be performed in real time during an ongoing hemodialysis treatment, provide timely alert in the event of respiratory instability or sleep apnea, and trigger further diagnostic and therapeutic interventions.BACKGROUNDMaintenance hemodialysis patients experience high morbidity and mortality, primarily from cardiovascular and infectious diseases. It was discovered recently that low arterial oxygen saturation (SaO2) is associated with a pro-inflammatory phenotype and poor patient outcomes. Sleep apnea is highly prevalent in maintenance hemodialysis patients and may contribute to intradialytic hypoxemia. In sleep apnea, normal respiration patterns are disrupted by episodes of apnea because of either disturbed respiratory control (i.e., central sleep apnea) or upper airway obstruction (i.e., obstructive sleep apnea). Intermittent SaO2 saw-tooth patterns are a hallmark of sleep apnea. Continuous intradialytic measurements of SaO2 provide an opportunity to follow the temporal evolution of SaO2 during hemodialysis. Using artificial intelligence, we aimed to automatically identify patients with repetitive episodes of intermittent SaO2 saw-tooth patterns.METHODSThe analysis utilized intradialytic SaO2 measurements by the Crit-Line device (Fresenius Medical Care, Waltham, MA). In patients with an arterio-venous fistula as vascular access, this FDA approved device records 150 SaO2 measurements per second in the extracorporeal blood circuit of the hemodialysis system. The average SaO2 of a 10-second segment is computed and streamed to the cloud. Periods comprising thirty 10-second segments (i.e., 300 s or five minutes) were independently adjudicated by two researchers for the presence or absence of SaO2 saw-tooth pattern. We built one-dimensional convolutional neural networks (1D-CNN), a state-of-the-art deep learning method, for SaO2 pattern classification and randomly assigned SaO2 time series segments to either a training (80%) or a test (20%) set.

Large language models (LLMs) such as ChatGPT are increasingly positioned to be integrated into various aspects of daily life, with promising applications in healthcare, including personalized nutritional guidance for patients with chronic kidney disease (CKD). However, for LLM-powered nutrition support tools to reach their full potential, active collaboration of healthcare professionals, patients, caregivers and LLM experts is crucial. We conducted a comprehensive review of the literature on the use of LLMs as tools to enhance nutrition recommendations for patients with CKD, curated by our expertise in the field. Additionally, we considered relevant findings from adjacent fields, including diabetes and obesity management. Currently, the application of LLMs for CKD-specific nutrition support remains limited and has room for improvement. Although LLMs can generate recipe ideas, their nutritional analyses often underestimate critical food components such as electrolytes and calories. Anticipated advancements in LLMs and other generative artificial intelligence (AI) technologies are expected to enhance these capabilities, potentially enabling accurate nutritional analysis, the generation of visual aids for cooking and identification of kidney-healthy options in restaurants. While LLM-based nutritional support for patients with CKD is still in its early stages, rapid advancements are expected in the near future. Engagement from the CKD community, including healthcare professionals, patients and caregivers, will be essential to harness AI-driven improvements in nutritional care with a balanced perspective that is both critical and optimistic.

No abstract available

COVID-19 has a higher rate of morbidity and mortality among dialysis patients than the general population. Identifying infected patients early with the support of predictive models helps dialysis centers implement concerted procedures (e.g., temperature screenings, universal masking, isolation treatments) to control the spread of SARS-CoV-2 and mitigate outbreaks. We collect data from multiple sources, including demographics, clinical, treatment, laboratory, vaccination, socioeconomic status, and COVID-19 surveillance. Previous early prediction models, such as logistic regression, SVM, and XGBoost, require sophisticated feature engineering and need improved prediction performance. We create deep learning models, including Recurrent Neural Networks (RNN) and Convolutional Neural Networks (CNN), to predict SARS-CoV-2 infections during incubation. Our study shows deep learning models with minimal feature engineering can identify those infected patients more accurately than previously built models. Our Long Short-Term Memory (LSTM) model consistently performed well, with an AUC exceeding 0.80, peaking at 0.91 in August 2021. The CNN model also demonstrated strong results with an AUC above 0.75. Both models outperformed previous best XGBoost models by over 0.10 in AUC. Prediction accuracy declined as the pandemic evolved, dropping to approximately 0.75 between September 2021 and January 2022. Maintaining a 20% false positive rate, our LSTM and CNN models identified 66% and 64% of positive cases among patients, significantly outperforming XGBoost models at 42%. We also identify key features for dialysis patients by calculating the gradient of the output with respect to the input features. By closely monitoring these factors, dialysis patients can receive earlier diagnoses and care, leading to less severe outcomes. Our research highlights the effectiveness of deep neural networks in analyzing longitudinal data, especially in predicting COVID-19 infections during the crucial incubation period. These deep network approaches surpass traditional methods relying on aggregated variable means, significantly improving the accurate identification of SARS-CoV-2 infections.

RESULTSThe QIP group comprised 44 patients (59 ± 17 years), the concurrent control group 48 patients (59 ± 16 years), the historic control group 57 patients (58 ± 15 years). Six-month post-AVF creation, the fraction of non-censored patients with catheter in place was 21% in the QIP cohort, 67% in the concurrent control group, and 68% in the historic control group. In unadjusted and adjusted analysis, catheter residence time post-fistula creation was shorter in QIP patients compared to either control groups (p < 0.001).CONCLUSIONScvO2-based assessment of fistula maturation is associated with shorter catheter residence post-AVF creation.INTRODUCTIONArteriovenous fistula (AVF) maturation assessment is essential to reduce venous catheter residence. We introduced central venous oxygen saturation (ScvO2) and estimated upper body blood flow (eUBBF) to monitor newly created fistula maturation and recorded catheter time in patients with and without ScvO2-based fistula maturation.METHODSFrom 2017 to 2019, we conducted a multicenter quality improvement project (QIP) in hemodialysis patients with the explicit goal to shorten catheter residence time post-AVF creation through ScvO2-based maturation monitoring. In patients with a catheter as vascular access, we tracked ScvO2 and eUBBF pre- and post-AVF creation. The primary outcome was catheter residence time post-AVF creation. We compared catheter residence time post-AVF creation between QIP patients and controls. One control group comprised concurrent patients; a second control group comprised historic controls (2014-2016). We conducted Kaplan-Meier analysis and constructed a Cox proportional hazards model with variables adjustment to assess time-to-catheter removal.

OBJECTIVESThe rising diversity of food preferences and the desire to provide better personalized care provide challenges to renal dietitians working in dialysis clinics. To address this situation, we explored the use of a large language model, specifically, ChatGPT using the GPT-4 model (openai.com), to support nutritional advice given to dialysis patients.RESULTSChatGPT generated a daily menu with five recipes. The renal dietitian rated the recipes at 3 (3, 3) [median (Q1, Q3)], the cooking instructions at 5 (5,5), and the nutritional analysis at 2 (2, 2) on the five-point Likert scale. ChatGPT's nutritional analysis underestimated calories by 36% (95% CI: 44-88%), protein by 28% (25-167%), fat 48% (29-81%), phosphorus 54% (15-102%), potassium 49% (40-68%), and sodium 53% (14-139%). The nutritional analysis of online available recipes differed only by 0 to 35%. The translations were rated as reliable by native speakers (4 on the five-point Likert scale).CONCLUSIONWhile ChatGPT-4 shows promise in providing personalized nutritional guidance for diverse dialysis patients, improvements are necessary. This study highlights the importance of thorough qualitative and quantitative evaluation of artificial intelligence-generated content, especially regarding medical use cases.METHODSWe tasked ChatGPT-4 with generating a personalized daily meal plan, including nutritional information. Virtual "patients" were generated through Monte Carlo simulation; data from a randomly selected virtual patient were presented to ChatGPT. We provided to ChatGPT patient demographics, food preferences, laboratory data, clinical characteristics, and available budget, to generate a one-day sample menu with recipes and nutritional analyses. The resulting daily recipe recommendations, cooking instructions, and nutritional analyses were reviewed and rated on a five-point Likert scale by an experienced renal dietitian. In addition, the generated content was rated by a renal dietitian and compared with a U. S. Department of Agriculture-approved nutrient analysis software. ChatGPT also analyzed nutrition information of two recipes published online. We also requested a translation of the output into Spanish, Mandarin, Hungarian, German, and Dutch.

RESULTSThe intervention group showed an improved median percentage of hemoglobin measurements within target at 47% (interquartile range, 39–58), with a 10% point median difference between the two groups (95% confidence interval, 3 to 16; P = 0.008). The odds ratio of being within the hemoglobin target in the standard-of-care group compared with the group receiving the personalized ESA recommendations was 0.68 (95% confidence interval, 0.51 to 0.92). The variability of hemoglobin levels decreased in the intervention group, with the percentage of patients experiencing fluctuating hemoglobin levels being 45% versus 82% in the standard-of-care group. ESA usage was reduced by approximately 25% in the intervention group.KEY POINTSWe conducted a randomized controlled pilot trial in patients on hemodialysis using a physiology-based individualized anemia therapy assistance software. Patients in the group receiving erythropoiesis-stimulating agent dose recommendations from the novel software showed improvement in hemoglobin stability and erythropoiesis-stimulating agent utilization.CONCLUSIONSOur results demonstrated an improved hemoglobin target attainment and variability by using personalized ESA recommendations using the physiology-based anemia therapy assistance software.CLINICAL TRIAL REGISTRATION NUMBER:NCT04360902.BACKGROUNDAnemia is common among patients on hemodialysis. Maintaining stable hemoglobin levels within predefined target levels can be challenging, particularly in patients with frequent hemoglobin fluctuations both above and below the desired targets. We conducted a multicenter, randomized controlled trial comparing our anemia therapy assistance software against a standard population-based anemia treatment protocol. We hypothesized that personalized dosing of erythropoiesis-stimulating agents (ESAs) improves hemoglobin target attainment.METHODSNinety-six patients undergoing hemodialysis and receiving methoxy polyethylene glycol-epoetin beta were randomized 1:1 to the intervention group (personalized ESA dose recommendations computed by the software) or the standard-of-care group for 26 weeks. The therapy assistance software combined a physiology-based mathematical model and a model predictive controller designed to stabilize hemoglobin levels within a tight target range (10–11 g/dl). The primary outcome measure was the percentage of hemoglobin measurements within the target. Secondary outcome measures included measures of hemoglobin variability and ESA utilization.

RESULTSIn total, 38 151 patients (52.2%) had complete data and made up the main cohort. Sensitivity analyses were conducted in 67 421 patients (92.3%) after eliminating variables with a high proportion of missing data points. Selected features diverged between datasets and workflows. A previously failed arteriovenous access appeared to be the most stable predictor for subsequent failure. Prediction of re-conversion based on the demographic and clinical information resulted in an area under the receiver operating characteristic curve (ROCAUC) between 0.541 and 0.571, whereas models predicting all cause mortality performed considerably better (ROCAUC 0.662 - 0.683).OBJECTIVEThe decision to convert from catheter to arteriovenous access is difficult yet very important. The ability to accurately predict fistula survival prior to surgery would significantly improve the decision making process. Many previously investigated demographic and clinical features have been associated with fistula failure. However, it is not conclusively understood how reliable predictions based on these parameters are at an individual level. The aim of this study was to investigate the probability of arteriovenous fistula maturation and survival after conversion using machine learning workflows.CONCLUSIONWhile group level depiction of major adverse outcomes after catheter to arteriovenous fistula or graft conversion is possible using the included variables, patient level predictions are associated with limited performance. Factors during and after fistula creation as well as biomolecular and genetic biomarkers might be more relevant predictors of fistula survival than baseline clinical conditions.METHODSA retrospective cohort study on multicentre data from a large North American dialysis organisation was conducted. The study population comprised 73 031 chronic in centre haemodialysis patients. The dataset included 49 variables including demographic and clinical features. Two distinct feature selection and prediction pipelines were used: LASSO regression and Boruta followed by a random forest classifier. Predictions were facilitated for re-conversion to catheter within one year. Additionally, all cause mortality predictions were conducted to serve as a comparator.

RESULTSOut of 978 patients, 193 (19.7%) tested positive for COVID-19 and had contact with other patients during the COV-Pos infectious period. Network diagrams showed no evidence that more exposed patients would have had a higher chance of infection. This finding was corroborated by logistic mixed effect regression (donor-to-potential recipient exposure OR: 0.63; 95% CI 0.32 to 1.17, p = 0.163). Separate analyses according to vaccination led to materially identical results.CONCLUSIONSTransmission of SARS-CoV-2 between in-center hemodialysis patients is unlikely. This finding supports the effectiveness of non-pharmaceutical interventions, such as universal masking and other procedures to control spread of COVID-19.BACKGROUNDIn-center hemodialysis entails repeated interactions between patients and clinic staff, potentially facilitating the spread of COVID-19. We examined if in-center hemodialysis is associated with the spread of SARS-CoV-2 between patients.METHODSOur retrospective analysis comprised all patients receiving hemodialysis in four New York City clinics between March 12th, 2020, and August 31st, 2022. Treatment-level clinic ID, dialysis shift, dialysis machine station, and date of COVID-19 diagnosis by RT-PCR were documented. To estimate the donor-to-potential recipient exposure ("donor" being the COVID-19 positive patient denoted as "COV-Pos"; "potential recipient" being other susceptible patients in the same shift), we obtained the spatial coordinates of each dialysis station, calculated the Euclidean distances between stations and weighted the exposure by proximity between them. For each donor, we estimated the donor-to-potential recipient exposure of all potential recipients dialyzed in the same shift and accumulated the exposure over time within the 'COV-Pos infectious period' as cumulative exposures. The 'COV-Pos infectious period' started 5 days before COVID-19 diagnosis date. We deployed network analysis to assess these interactions and summarized the donor-to-potential recipient exposure in 193 network diagrams. We fitted mixed effects logistic regression models to test whether more donor-to-potential recipient exposure conferred a higher risk of SARS-CoV-2 infection.

No abstract available

DISCUSSIONConcurrent combined monitoring of intradialytic ScvO2 and RBV change may provide additional insights into a patient's circulatory status. Patients with low ScvO2 and small changes in RBV may represent a specifically vulnerable group of patients at particularly high risk for adverse outcomes, possibly related to poor cardiac reserve and fluid overload.INTRODUCTIONIn maintenance hemodialysis (HD) patients, low central venous oxygen saturation (ScvO2 ) and small decline in relative blood volume (RBV) have been associated with adverse outcomes. Here we explore the joint association between ScvO2 and RBV change in relation to all-cause mortality.FINDINGSBaseline comprised 5231 dialysis sessions in 216 patients. The median RBV change was -5.5% and median ScvO2 was 58.8%. During follow-up, 44 patients (20.4%) died. In the adjusted model, all-cause mortality was highest in patients with ScvO2 below median and RBV change above median (HR 6.32; 95% confidence interval [CI] 1.37-29.06), followed by patients with ScvO2 below median and RBV change below median (HR 5.04; 95% CI 1.14-22.35), and ScvO2 above median and RBV change above median (HR 4.52; 95% CI 0.95-21.36).METHODSWe conducted a retrospective study in maintenance HD patients with central venous catheters as vascular access. During a 6-month baseline period, Crit-Line (Fresenius Medical Care, Waltham, MA) was used to measure continuously intradialytic ScvO2 and hematocrit-based RBV. We defined four groups per median change of RBV and median ScvO2 . Patients with ScvO2 above median and RBV change below median were defined as reference. Follow-up period was 3 years. We constructed Cox proportional hazards model with adjustment for age, diabetes, and dialysis vintage to assess the association between ScvO2 and RBV and all-cause mortality during follow-up.

CONCLUSIONAs found in our study, the dynamics of the prediction model are frequently changing as the pandemic evolves. County-level infection information and vaccination information are crucial for the success of early COVID-19 prediction models. Our results show that the proposed model can effectively identify SARS-CoV-2 infections during the incubation period. Prospective studies are warranted to explore the application of such prediction models in daily clinical practice.BACKGROUNDThe coronavirus disease 2019 (COVID-19) pandemic has created more devastation among dialysis patients than among the general population. Patient-level prediction models for severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection are crucial for the early identification of patients to prevent and mitigate outbreaks within dialysis clinics. As the COVID-19 pandemic evolves, it is unclear whether or not previously built prediction models are still sufficiently effective.METHODSWe developed a machine learning (XGBoost) model to predict during the incubation period a SARS-CoV-2 infection that is subsequently diagnosed after 3 or more days. We used data from multiple sources, including demographic, clinical, treatment, laboratory, and vaccination information from a national network of hemodialysis clinics, socioeconomic information from the Census Bureau, and county-level COVID-19 infection and mortality information from state and local health agencies. We created prediction models and evaluated their performances on a rolling basis to investigate the evolution of prediction power and risk factors.RESULTFrom April 2020 to August 2020, our machine learning model achieved an area under the receiver operating characteristic curve (AUROC) of 0.75, an improvement of over 0.07 from a previously developed machine learning model published by Kidney360 in 2021. As the pandemic evolved, the prediction performance deteriorated and fluctuated more, with the lowest AUROC of 0.6 in December 2021 and January 2022. Over the whole study period, that is, from April 2020 to February 2022, fixing the false-positive rate at 20%, our model was able to detect 40% of the positive patients. We found that features derived from local infection information reported by the Centers for Disease Control and Prevention (CDC) were the most important predictors, and vaccination status was a useful predictor as well. Whether or not a patient lives in a nursing home was an effective predictor before vaccination, but became less predictive after vaccination.

Omics applications in nephrology may have relevance in the future to improve clinical care of kidney disease patients. In a short term, patients will benefit from specific measurement and computational analyses around biomarkers identified at various omics-levels. In mid term and long term, these approaches will need to be integrated into a holistic representation of the kidney and all its influencing factors for individualized patient care. Research demonstrates robust data to justify the application of omics for better understanding, risk stratification, and individualized treatment of kidney disease patients. Despite these advances in the research setting, there is still a lack of evidence showing the combination of omics technologies with artificial intelligence and its application in clinical diagnostics and care of patients with kidney disease.

RESULTSDuring 180,319 HD sessions among 2554 patients, PRR had high within-patient and between-patient variability. Female sex and hypoalbuminemia were associated with low PRR at multiple time points during the first hour of HD. Low starting PRR has a higher hazard of IDH, whereas high starting PRR was protective (hazard ratio [HR], 1.26, 95% confidence interval [CI], 1.18 to 1.35 versus HR, 0.79, 95% CI, 0.73 to 0.85, respectively). However, when accounting for time-varying PRR and time-varying confounders, compared with a moderate PRR, while a consistently low PRR was associated with increased risk of hypotension (odds ratio [OR], 1.09, 95% CI, 1.02 to 1.16), a consistently high PRR had a stronger association with hypotension within the next 15 minutes (OR, 1.38, 95% CI, 1.30 to 1.45).KEY POINTSDirectly studying plasma refill rate (PRR) during hemodialysis (HD) can offer insight into physiologic mechanisms that change throughout HD. PRR at the start and during HD is associated with intradialytic hypotension, independent of ultrafiltration rate. A rising PRR during HD may be an early indicator of compensatory mechanisms for impending circulatory instability.CONCLUSIONSWe present a straightforward technique to quantify plasma refill that could easily integrate with devices that monitor hematocrit during HD. Our study highlights how examining patterns of plasma refill may enhance our understanding of circulatory changes during HD, an important step to understand how current technology might be used to improve hemodynamic instability.BACKGROUNDAttaining the optimal balance between achieving adequate volume removal while preserving organ perfusion is a challenge for patients receiving maintenance hemodialysis (HD). Current strategies to guide ultrafiltration are inadequate.METHODSWe developed an approach to calculate the plasma refill rate (PRR) throughout HD using hematocrit and ultrafiltration data in a retrospective cohort of patients receiving maintenance HD at 17 dialysis units from January 2017 to October 2019. We studied whether (1) PRR is associated with traditional risk factors for hemodynamic instability using logistic regression, (2) low starting PRR is associated with intradialytic hypotension (IDH) using Cox proportional hazard regression, and (3) time-varying PRR throughout HD is associated with hypotension using marginal structural modeling.

No abstract available

Analysis of medical images, such as radiological or tissue specimens, is an indispensable part of medical diagnostics. Conventionally done manually, the process may sometimes be time-consuming and prone to interobserver variability. Image classification and segmentation by deep learning strategies, predominantly convolutional neural networks, may provide a significant advance in the diagnostic process. In renal medicine, most evidence has been generated around the radiological assessment of renal abnormalities and histological analysis of renal biopsy specimens' segmentation. In this article, the basic principles of image analysis by convolutional neural networks, brief descriptions of convolutional neural networks, and their system architecture for image analysis are discussed, in combination with examples regarding their use in image analysis in nephrology.

RESULTSIn the effectiveness analysis concerning mRNA vaccines, we observed 850 SARS-CoV-2 infections and 201 COVID-19 related deaths among the 28110 patients during a mean follow up of 44 ± 40 days. In the effectiveness analysis concerning viral-carrier vaccines, we observed 297 SARS-CoV-2 infections and 64 COVID-19 related deaths among 12888 patients during a mean follow up of 48 ± 32 days. We observed 18.5/100-patient-year and 8.5/100-patient-year fewer infections and 5.4/100-patient-year and 5.2/100-patient-year fewer COVID-19 related deaths among patients vaccinated with mRNA and viral-carrier vaccines respectively, compared to matched unvaccinated controls. Estimated vaccine effectiveness at days 15, 30, 60 and 90 after the first dose of a mRNA vaccine was: for infection, 41.3%, 54.5%, 72.6% and 83.5% and, for death, 33.1%, 55.4%, 80.1% and 91.2%. Estimated vaccine effectiveness after the first dose of a viral-carrier vaccine was: for infection, 38.3% without increasing over time and, for death, 56.6%, 75.3%, 92.0% and 97.4%.CONCLUSIONIn this large, real-world cohort of hemodialyzed patients, mRNA and viral-carrier COVID-19 vaccines were associated with reduced COVID-19 related mortality. Additionally, we observed a strong reduction of SARS-CoV-2 infection in hemodialysis patients receiving mRNA vaccines.BACKGROUNDHemodialysis patients have high-risk of severe SARS-CoV-2 infection but were unrepresented in randomized controlled trials evaluating the safety and efficacy of COVID-19 vaccines. We estimated the real-world effectiveness of COVID-19 vaccines in a large international cohort of hemodialysis patients.METHODSIn this historical, 1:1 matched cohort study, we included adult hemodialysis patients receiving treatment from December 1, 2020, to May 31, 2021. For each vaccinated patient, an unvaccinated control was selected among patients registered in the same country and attending a dialysis session around the first vaccination date. Matching was based on demographics, clinical characteristics, past COVID-19 infections and a risk score representing the local background risk of infection at vaccination dates. We estimated the effectiveness of mRNA and viral-carrier COVID-19 vaccines in preventing infection and mortality rates from a time-dependent Cox regression stratified by country.

In patients with kidney failure treated by hemodialysis, intradialytic arterial oxygen saturation (SaO2) time series present intermittent high-frequency high-amplitude oximetry patterns (IHHOP), which correlate with observed sleep-associated breathing disturbances. A new method for identifying such intermittent patterns is proposed. The method is based on the analysis of recurrence in the time series through the quantification of an optimal recurrence threshold ([Formula: see text]). New time series for the value of [Formula: see text] were constructed using a rolling window scheme, which allowed for real-time identification of the occurrence of IHHOPs. The results for the optimal recurrence threshold were confronted with standard metrics used in studies of obstructive sleep apnea, namely the oxygen desaturation index (ODI) and oxygen desaturation density (ODD). A high correlation between [Formula: see text] and the ODD was observed. Using the value of the ODI as a surrogate to the apnea-hypopnea index (AHI), it was shown that the value of [Formula: see text] distinguishes occurrences of sleep apnea with great accuracy. When subjected to binary classifiers, this newly proposed metric has great power for predicting the occurrences of sleep apnea-related events, as can be seen by the larger than 0.90 AUC observed in the ROC curve. Therefore, the optimal threshold [Formula: see text] from recurrence analysis can be used as a metric to quantify the occurrence of abnormal behaviors in the arterial oxygen saturation time series.

RESULTSWe included 9,211 patients (age 65.4 ± 13.7 years, dialysis vintage 4.2 ± 3.7 years) eligible for the study. The 30-day mortality rate was 20.8%. In LR models, several potentially modifiable factors were associated with higher mortality: body mass index (BMI) 30-40 kg/m2 (OR: 1.28, CI: 1.10-1.50), single-pool Kt/V (OR off-target vs on-target: 1.19, CI: 1.02-1.38), overhydration (OR: 1.15, CI: 1.01-1.32), and both low (<2.5 mg/dl) and high (≥5.5 mg/dl) serum phosphate levels (OR: 1.52, CI: 1.07-2.16 and OR: 1.17, CI: 1.01-1.35). On-line hemodiafiltration was protective in the model using KPIs (OR: 0.86, CI: 0.76-0.97). SHapley Additive exPlanations analysis in XGBoost models shows a high influence on prediction for several modifiable factors as well, including inflammatory parameters, high BMI, and fluid overload. In both LR and XGBoost models, age, gender, and comorbidities were strongly associated with mortality.CONCLUSIONBoth conventional and machine learning techniques showed that KPIs and modifiable risk factors in different dimensions ascertained 6 months before the COVID-19 suspicion date were associated with 30-day COVID-19-related mortality. Our results suggest that adequate dialysis and achieving KPI targets remain of major importance during the COVID-19 pandemic as well.INTRODUCTIONPatients with end-stage kidney disease face a higher risk of severe outcomes from SARS-CoV-2 infection. Moreover, it is not well known to what extent potentially modifiable risk factors contribute to mortality risk. In this historical cohort study, we investigated the incidence and risk factors for 30-day mortality among hemodialysis patients with SARS-CoV-2 infection treated in the European Fresenius Medical Care NephroCare network using conventional and machine learning techniques.METHODSWe included adult hemodialysis patients with the first documented SARS-CoV-2 infection between February 1, 2020, and March 31, 2021, registered in the clinical database. The index date for the analysis was the first SARS-CoV-2 suspicion date. Patients were followed for up to 30 days until April 30, 2021. Demographics, comorbidities, and various modifiable risk factors, expressed as continuous parameters and as key performance indicators (KPIs), were considered to tap multiple dimensions including hemodynamic control, nutritional state, and mineral metabolism in the 6 months before the index date. We used logistic regression (LR) and XGBoost models to assess risk factors for 30-day mortality.

No abstract available