Household Water Lead and Hematologic Toxic Effects in Chronic Kidney Disease.

Importance: The consequences of low levels of environmental lead exposure, as found commonly in US household water, have not been established. Objective: To examine whether commonly encountered levels of lead in household water are associated with hematologic toxicity among individuals with advanced kidney disease, a group known to have disproportionate susceptibility to environmental toxicants. Design, Setting, and Participants: Cross-sectional analysis of household water lead concentrations and hematologic outcomes was performed among patients beginning dialysis at a Fresenius Medical Care outpatient facility between January 1, 2017, and December 20, 2021. Data analysis was performed from April 1 to August 15, 2023. Exposure: Concentrations of lead in household water were examined in categorical proportions of the Environmental Protection Agency's allowable threshold (15 µg/L) and continuously. Main Outcomes and Measures: Hematologic toxic effects were defined by monthly erythropoiesis-stimulating agent (ESA) dosing during the first 90 days of incident kidney failure care and examined as 3 primary outcomes: a proportion receiving maximum or higher dosing, continuously, and by a resistance index that normalized to body weight and hemoglobin concentrations. Secondarily, hemoglobin concentrations for patients with data prior to kidney failure onset were examined, overall and among those with concurrent iron deficiency, thought to increase gastrointestinal absorption of ingested lead. Results: Among 6404 patients with incident kidney failure (male, 4182 [65%]; mean [SD] age, 57 [14] years) followed up for the first 90 days of dialysis therapy, 12% (n = 742) had measurable lead in household drinking water. A higher category of household lead contamination was associated with 15% (odds ratio [OR], 1.15 [95% CI, 1.04-1.27]) higher risk of maximum monthly ESA dosing, 4.5 (95% CI, 0.8-8.2) µg higher monthly ESA dose, and a 0.48% (95% CI, 0.002%-0.96%) higher monthly resistance index. Among patients with pre-kidney failure hemoglobin measures (n = 2648), a higher household lead categorization was associated with a 0.12 (95% CI, -0.23 to -0.002) g/dL lower hemoglobin concentration, particularly among those with concurrent iron deficiency (multiplicative interaction, P = .07), among whom hemoglobin concentrations were 0.25 (95% CI, -0.47 to -0.04) g/dL lower. Conclusion: The findings of this study suggest that levels of lead found commonly in US drinking water may be associated with lead poisoning among susceptible individuals.

Predicting SARS-CoV-2 infection among hemodialysis patients using multimodal data.

Background: The 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. Methods: We 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. Result: From 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. Conclusion: As 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.

Real-time prediction of intradialytic hypotension using machine learning and cloud computing infrastructure.

BACKGROUND: In 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. METHODS: We 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. RESULTS: We 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. CONCLUSIONS: Real-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.

Machine Learning Case Study: Patterns of Kidney Function Decline and Their Association With Clinical Outcomes Within 90 Days After the Initiation of Renal Dialysis.

A case study explores patterns of kidney function decline using unsupervised learning methods first and then associating patterns with clinical outcomes using supervised learning methods. Predicting short-term risk of hospitalization and death prior to renal dialysis initiation may help target high-risk patients for more aggressive management. This study combined clinical data from patients presenting for renal dialysis at Fresenius Medical Care with laboratory data from Quest Diagnostics to identify disease trajectory patterns associated with the 90-day risk of hospitalization and death after beginning renal dialysis. Patients were clustered into 4 groups with varying rates of estimated glomerular filtration rate (eGFR) decline during the 2-year period prior to dialysis. Overall rates of hospitalization and death were 24.9% (582/2341) and 4.6% (108/2341), respectively. Groups with the steepest declines had the highest rates of hospitalization and death within 90 days of dialysis initiation. The rate of eGFR decline is a valuable and readily available tool to stratify short-term (90 days) risk of hospitalization and death after the initiation of renal dialysis. More intense approaches are needed that apply models that identify high risks to potentially avert or reduce short-term hospitalization and death of patients with a severe and rapidly progressive chronic kidney disease.

Predicting mortality risk in dialysis: Assessment of risk factors using traditional and advanced modeling techniques within the Monitoring Dialysis Outcomes initiative.

INTRODUCTION: Several factors affect the survival of End Stage Kidney Disease (ESKD) patients on dialysis. Machine learning (ML) models may help tackle multivariable and complex, often non-linear predictors of adverse clinical events in ESKD patients. In this study, we used advanced ML method as well as a traditional statistical method to develop and compare the risk factors for mortality prediction model in hemodialysis (HD) patients. MATERIALS AND METHODS: We included data HD patients who had data across a baseline period of at least 1 year and 1 day in the internationally representative Monitoring Dialysis Outcomes (MONDO) Initiative dataset. Twenty-three input parameters considered in the model were chosen in an a priori manner. The prediction model used 1 year baseline data to predict death in the following 3 years. The dataset was randomly split into 80% training data and 20% testing data for model development. Two different modeling techniques were used to build the mortality prediction model. FINDINGS: A total of 95,142 patients were included in the analysis sample. The area under the receiver operating curve (AUROC) of the model on the test data with XGBoost ML model was 0.84 on the training data and 0.80 on the test data. AUROC of the logistic regression model was 0.73 on training data and 0.75 on test data. Four out of the top five predictors were common to both modeling strategies. DISCUSSION: In the internationally representative MONDO data for HD patients, we describe the development of a ML model and a traditional statistical model that was suitable for classification of a prevalent HD patient's 3-year risk of death. While both models had a reasonably high AUROC, the ML model was able to identify levels of hematocrit (HCT) as an important risk factor in mortality. If implemented in clinical practice, such proof-of-concept models could be used to provide pre-emptive care for HD patients.

Trajectories of clinical and laboratory characteristics associated with COVID-19 in hemodialysis patients by survival.

INTRODUCTION: The clinical impact of COVID-19 has not been established in the dialysis population. We evaluated the trajectories of clinical and laboratory parameters in hemodialysis (HD) patients. METHODS: We used data from adult HD patients treated at an integrated kidney disease company who received a reverse transcription polymerase chain reaction (RT-PCR) test to investigate suspicion of a severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) infection between May 1 and September 1, 2020. Nonparametric smoothing splines were used to fit data for individual trajectories and estimate the mean change over time in patients testing positive or negative for SARS-CoV-2 and those who survived or died within 30 days of first suspicion or positive test date. For each clinical parameter of interest, the difference in average daily changes between COVID-19 positive versus negative group and COVID-19 survivor versus nonsurvivor group was estimated by fitting a linear mixed effects model based on measurements in the 14 days before (i.e., Day -14 to Day 0) Day 0. RESULTS: There were 12,836 HD patients with a suspicion of COVID-19 who received RT-PCR testing (8895 SARS-CoV-2 positive). We observed significantly different trends (p < 0.05) in pre-HD systolic blood pressure (SBP), pre-HD pulse rate, body temperature, ferritin, neutrophils, lymphocytes, albumin, and interdialytic weight gain (IDWG) between COVID-19 positive and negative patients. For COVID-19 positive group, we observed significantly different clinical trends (p < 0.05) in pre-HD pulse rate, lymphocytes, neutrophils, and albumin between survivors and nonsurvivors. We also observed that, in the group of survivors, most clinical parameters returned to pre-COVID-19 levels within 60-90 days. CONCLUSION: We observed unique temporal trends in various clinical and laboratory parameters among HD patients who tested positive versus negative for SARS-CoV-2 infection and those who survived the infection versus those who died. These trends can help to define the physiological disturbances that characterize the onset and course of COVID-19 in HD patients.

Real-time prediction of intradialytic relative blood volume: a proof-of-concept for integrated cloud computing infrastructure.

BACKGROUND: Inadequate refilling from extravascular compartments during hemodialysis can lead to intradialytic symptoms, such as hypotension, nausea, vomiting, and cramping/myalgia. Relative blood volume (RBV) plays an important role in adapting the ultrafiltration rate which in turn has a positive effect on intradialytic symptoms. It has been clinically challenging to identify changes RBV in real time to proactively intervene and reduce potential negative consequences of volume depletion. Leveraging advanced technologies to process large volumes of dialysis and machine data in real time and developing prediction models using machine learning (ML) is critical in identifying these signals. METHOD: We conducted a proof-of-concept analysis to retrospectively assess near real-time dialysis treatment data from in-center patients in six clinics using Optical Sensing Device (OSD), during December 2018 to August 2019. The goal of this analysis was to use real-time OSD data to predict if a patient's relative blood volume (RBV) decreases at a rate of at least - 6.5 % per hour within the next 15 min during a dialysis treatment, based on 10-second windows of data in the previous 15 min. A dashboard application was constructed to demonstrate how reporting structures may be developed to alert clinicians in real time of at-risk cases. Data was derived from three sources: (1) OSDs, (2) hemodialysis machines, and (3) patient electronic health records. RESULTS: Treatment data from 616 in-center dialysis patients in the six clinics was curated into a big data store and fed into a Machine Learning (ML) model developed and deployed within the cloud. The threshold for classifying observations as positive or negative was set at 0.08. Precision for the model at this threshold was 0.33 and recall was 0.94. The area under the receiver operating curve (AUROC) for the ML model was 0.89 using test data. CONCLUSIONS: The findings from our proof-of concept analysis demonstrate the design of a cloud-based framework that can be used for making real-time predictions of events during dialysis treatments. Making real-time predictions has the potential to assist clinicians at the point of care during hemodialysis.

Machine learning directed interventions associate with decreased hospitalization rates in hemodialysis patients.

BACKGROUND: An integrated kidney disease company uses machine learning (ML) models that predict the 12-month risk of an outpatient hemodialysis (HD) patient having multiple hospitalizations to assist with directing personalized interdisciplinary interventions in a Dialysis Hospitalization Reduction Program (DHRP). We investigated the impact of risk directed interventions in the DHRP on clinic-wide hospitalization rates. METHODS: We compared the hospital admission and day rates per-patient-year (ppy) from all hemodialysis patients in 54 DHRP and 54 control clinics identified by propensity score matching at baseline in 2015 and at the end of the pilot in 2018. We also used paired T test to compare the between group difference of annual hospitalization rate and hospitalization days rates at baseline and end of the pilot. RESULTS: The between group difference in annual hospital admission and day rates was similar at baseline (2015) with a mean difference between DHRP versus control clinics of -0.008 ± 0.09 ppy and -0.05 ± 0.96 ppy respectively. The between group difference in hospital admission and day rates became more distinct at the end of follow up (2018) favoring DHRP clinics with the mean difference being -0.155 ± 0.38 ppy and -0.97 ± 2.78 ppy respectively. A paired t-test showed the change in the between group difference in hospital admission and day rates from baseline to the end of the follow up was statistically significant (t-value = 2.73, p-value < 0.01) and (t-value = 2.29, p-value = 0.02) respectively. CONCLUSIONS: These findings suggest ML model-based risk-directed interdisciplinary team interventions associate with lower hospitalization rates and hospital day rate in HD patients, compared to controls.

Machine Learning for Prediction of Patients on Hemodialysis with an Undetected SARS-CoV-2 Infection.

Background: We developed a machine learning (ML) model that predicts the risk of a patient on hemodialysis (HD) having an undetected SARS-CoV-2 infection that is identified after the following ≥3 days. Methods: As part of a healthcare operations effort, we used patient data from a national network of dialysis clinics (February-September 2020) to develop an ML model (XGBoost) that uses 81 variables to predict the likelihood of an adult patient on HD having an undetected SARS-CoV-2 infection that is identified in the subsequent ≥3 days. We used a 60%:20%:20% randomized split of COVID-19-positive samples for the training, validation, and testing datasets. Results: We used a select cohort of 40,490 patients on HD to build the ML model (11,166 patients who were COVID-19 positive and 29,324 patients who were unaffected controls). The prevalence of COVID-19 in the cohort (28% COVID-19 positive) was by design higher than the HD population. The prevalence of COVID-19 was set to 10% in the testing dataset to estimate the prevalence observed in the national HD population. The threshold for classifying observations as positive or negative was set at 0.80 to minimize false positives. Precision for the model was 0.52, the recall was 0.07, and the lift was 5.3 in the testing dataset. Area under the receiver operating characteristic curve (AUROC) and area under the precision-recall curve (AUPRC) for the model was 0.68 and 0.24 in the testing dataset, respectively. Top predictors of a patient on HD having a SARS-CoV-2 infection were the change in interdialytic weight gain from the previous month, mean pre-HD body temperature in the prior week, and the change in post-HD heart rate from the previous month. Conclusions: The developed ML model appears suitable for predicting patients on HD at risk of having COVID-19 at least 3 days before there would be a clinical suspicion of the disease.

Artificial intelligence enabled applications in kidney disease.

Artificial intelligence (AI) is considered as the next natural progression of traditional statistical techniques. Advances in analytical methods and infrastructure enable AI to be applied in health care. While AI applications are relatively common in fields like ophthalmology and cardiology, its use is scarcely reported in nephrology. We present the current status of AI in research toward kidney disease and discuss future pathways for AI. The clinical applications of AI in progression to end-stage kidney disease and dialysis can be broadly subdivided into three main topics: (a) predicting events in the future such as mortality and hospitalization; (b) providing treatment and decision aids such as automating drug prescription; and (c) identifying patterns such as phenotypical clusters and arteriovenous fistula aneurysm. At present, the use of prediction models in treating patients with kidney disease is still in its infancy and further evidence is needed to identify its relative value. Policies and regulations need to be addressed before implementing AI solutions at the point of care in clinics. AI is not anticipated to replace the nephrologists' medical decision-making, but instead assist them in providing optimal personalized care for their patients.

Wearable health devices and personal area networks: can they improve outcomes in haemodialysis patients?

Digitization of healthcare will be a major innovation driver in the coming decade. Also, enabled by technological advancements and electronics miniaturization, wearable health device (WHD) applications are expected to grow exponentially. This, in turn, may make 4P medicine (predictive, precise, preventive and personalized) a more attainable goal within dialysis patient care. This article discusses different use cases where WHD could be of relevance for dialysis patient care, i.e. measurement of heart rate, arrhythmia detection, blood pressure, hyperkalaemia, fluid overload and physical activity. After adequate validation of the different WHD in this specific population, data obtained from WHD could form part of a body area network (BAN), which could serve different purposes such as feedback on actionable parameters like physical inactivity, fluid overload, danger signalling or event prediction. For a BAN to become clinical reality, not only must technical issues, cybersecurity and data privacy be addressed, but also adequate models based on artificial intelligence and mathematical analysis need to be developed for signal optimization, data representation, data reliability labelling and interpretation. Moreover, the potential of WHD and BAN can only be fulfilled if they are part of a transformative healthcare system with a shared responsibility between patients, healthcare providers and the payors, using a step-up approach that may include digital assistants and dedicated 'digital clinics'. The coming decade will be critical in observing how these developments will impact and transform dialysis patient care and will undoubtedly ask for an increased 'digital literacy' for all those implicated in their care.

Remote Treatment Monitoring on Hospitalization and Technique Failure Rates in Peritoneal Dialysis Patients.

Background: An integrated kidney disease healthcare company implemented a peritoneal dialysis (PD) remote treatment monitoring (RTM) application in 2016. We assessed if RTM utilization associates with hospitalization and technique failure rates. Methods: We used data from adult (age ≥18 years) patients on PD treated from October 2016 through May 2019 who registered online for the RTM. Patients were classified by RTM use during a 30-day baseline after registration. Groups were: nonusers (never entered data), moderate users (entered one to 15 treatments), and frequent users (entered >15 treatments). We compared hospital admission/day and sustained technique failure (required >6 consecutive weeks of hemodialysis) rates over 3, 6, 9, and 12 months of follow-up using Poisson and Cox models adjusted for patient/clinical characteristics. Results: Among 6343 patients, 65% were nonusers, 11% were moderate users, and 25% were frequent users. Incidence rate of hospital admission was 22% (incidence rate ratio [IRR]=0.78; P=0.002), 24% (IRR=0.76; P<0.001), 23% (IRR=0.77; P≤0.001), and 26% (IRR=0.74; P≤0.001) lower in frequent users after 3, 6, 9, and 12 months, respectively, versus nonusers. Incidence rate of hospital days was 38% (IRR=0.62; P=0.013), 35% (IRR=0.65; P=0.001), 34% (IRR=0.66; P≤0.001), and 32% (IRR=0.68; P<0.001) lower in frequent users after 3, 6, 9, and 12 months, respectively, versus nonusers. Sustained technique failure risk at 3, 6, 9, and 12 months was 33% (hazard ratio [HR]=0.67; P=0.020), 31% (HR=0.69; P=0.003), 31% (HR=0.69; P=0.001), and 27% (HR=0.73; P=0.001) lower, respectively, in frequent users versus nonusers. Among a subgroup of survivors of the 12-month follow-up, sustained technique failure risk was 26% (HR=0.74; P=0.023) and 21% (HR=0.79; P=0.054) lower after 9 and 12 months, respectively, in frequent users versus nonusers. Conclusions: Our findings suggest frequent use of an RTM application associates with less hospital admissions, shorter hospital length of stay, and lower technique failure rates. Adoption of RTM applications may have the potential to improve timely identification/intervention of complications.