Background: Attaining the optimal balance between achieving adequate volume removal while preserving organ perfusion is a challenge for patients receiving maintenance hemodialysis. Current strategies to guide ultrafiltration are inadequate. Methods: We developed an approach to calculate plasma refill rate throughout hemodialysis using hematocrit and ultrafiltration data in a retrospective cohort of patients receiving maintenance hemodialysis at 17 dialysis units from January 2017-October 2019. We studied whether (1) plasma refill rate is associated with traditional risk factors for hemodynamic instability using logistic regression, (2) low starting plasma refill rate is associated with intradialytic hypotension using Cox proportional hazard regression, and (3) time-varying plasma refill rate throughout hemodialysis is associated with hypotension using marginal structural modeling. Results: During 180,319 hemodialysis sessions among 2554 patients, plasma refill rate had high within- and between-patient variability. Female sex and hypoalbuminemia were associated with low plasma refill rate at multiple time points during the first hour of hemodialysis. Low starting plasma refill rate had higher hazards of intradialytic hypotension while high starting plasma refill rate was protective (HR 1.26, 95% CI 1.18, 1.35 versus HR 0.79, 95% CI 0.73, 0.85, respectively). However, when accounting for time-varying plasma refill rate and time-varying confounders, compared to a moderate plasma refill rate, while a consistently low plasma refill rate was associated with increased risk of hypotension (OR 1.09, 95% CI 1.02, 1.16), a consistently high plasma refill rate had a stronger association with hypotension within the next 15 minutes (OR 1.38, 95% CI 1.30, 1.45). Conclusions: We present a straightforward technique to quantify plasma refill that could easily integrate with devices that monitor hematocrit during hemodialysis. Our study highlights how examining patterns of plasma refill may enhance our understanding of circulatory changes during hemodialysis, an important step to understanding how current technology might be utilized to improve hemodynamic instability.
Background: Most hemodialysis patients without residual kidney function accumulate fluid between dialysis session that needs to be removed by ultrafiltration. Ultrafiltration usually results in a decline in relative blood volume (RBV). Recent epidemiological research has identified RBV ranges that were associated with significantly better survival. The objective of this work was to develop an ultrafiltration controller to steer a patient’s RBV trajectory into these favorable RBV ranges. Methods: We designed a proportional-integral feedback ultrafiltration controller that utilizes signals from a device that reports RBV. The control goal is to attain the RBV trajectory associated with improved patient survival. Additional constraints such as upper and lower bounds of ultrafiltration volume and rate were realized. The controller was evaluated in in silico and ex vivo bench experiments, and in a clinical proof-of-concept study in two maintenance dialysis patients. Results: In all tests, the ultrafiltration controller performed as expected. In the in silico and ex vivo bench experiments, the controller showed robust reaction toward deliberate disruptive interventions (e.g. signal noise; extreme plasma refill rates). No adverse events were observed in the clinical study. Conclusions: The ultrafiltration controller can steer RBV trajectories toward desired RBV ranges while obeying to a set of constraints. Prospective studies in hemodialysis patients with diverse clinical characteristics are warranted to further explore the controllers impact on intradialytic hemodynamic stability, quality of life, and long-term outcomes.
Background and Aims Anemia is a common complication in hemodialysis (HD) patients. Its treatment with erythropoiesis-stimulating agents (ESAs) is challenging due to a nonlinear dose-response relationship and time delays between ESA administration and hemoglobin (Hgb) response. Anemia treatment protocols are frequently used in clinical settings. However, high variability of patient-specific disease characteristics complicate attainment and maintenance of desired Hgb levels. We developed a novel fully personalized ESA dose recommendation tool and present clinical results of a multi-center, randomized controlled trial (RCT) using this software. Method We conducted an RCT in adult HD patients in six dialysis facilities in the US. Patients were randomized 1:1 and treated with our personalized ESA dose recommendation tool for twenty-six weeks (intervention group) or continued to be treated using an anemia protocol that was used as part of standard of care in those clinics (control group). The recommendation tool utilized a physiology-based model of anemia to estimate patient-specific physiological key characteristics, such as red blood cell lifespan, to create a patient-individual model from recent routine clinical data (gender, height, weight, Hgb measurements, and ESA treatment). These key characteristics and model-based outcome predictions were used to generate patient-specific ESA dose recommendations to stabilize Hgb levels in a target window of 10–11 g/dL. Dose recommendations were disseminated to the anemia managers of patients enrolled in the study for evaluation and further decision-making (Figure 1). This procedure was repeated biweekly with updated clinical data. Results Ninety-six patients were enrolled in the RCT. Patients were included in the statistical analysis when they remained in the clinical study for at least 30 days (n = 45 control group, n = 46 intervention group). We evaluated outcome measures showing efficacy and efficiency of the treatment in achieving target Hgb levels. Hgb-related outcomes were significantly different between the two study groups, with an improved Hgb control in the intervention arm, manifesting in a reduction of the Hgb distance to target by more than 30%. Epoetin-beta utilization in the intervention group was decreased by over 20%, while iron-related parameters showed no difference between the two arms (Table 1). Acceptance rate of dose recommendations was high; roughly 95% of the recommendations were accepted and implemented by the clinical staff. Conclusion A therapy software for personalized anemia management was developed for use with epoetin beta. The model-based ESA dose recommendation tool was evaluated in a clinical RCT in HD patients. Hgb control improved significantly in the group using the novel software tool, while ESA usage decreased, thus providing more efficient anemia management for the individual patient while reducing epoetin-beta drug exposure.
This project is based on a mathematical model of erythropoiesis for anemia [1,2], which consists of five hyperbolic population equations describing the production of red blood cells under treatment with epoetin-alfa (EPO). Extended dynamic mode decomposition (EDMD) is utilized to approximate the non-linear dynamical systems by linear ones. This allows for efficient and reliable strategies based on a combination of EDMD and model predictive control (MPC), which produces results comparable with the one obtained in [7] for the original model. IntroductionAlmost all hemodialysis patients suffer from chronic anemia, due to the reduced functionality of the kidneys and the resulting low production of erythropoietin, a kidney-derived hormone that increases red blood cell output by the bone marrow. Therefore, physicians use erythropoietin stimulating agents, such as epoetin-alfa (EPO), to partially correct the anemia. The challenge in designing efficient therapies is due to the patients' differences in long-term response to EPO. In [2], the authors introduce a mathematical model for predicting such a response. As in [7], our aim is to design a feedback control strategy, based on Model Predictive Control (MPC) [3], to optimize the injections of EPO doses in order to reach a target hemoglobin level. In contrast to [7], we do not imply the EPO model from [2] during the optimization, but we utilize it to generate a data-driven approximation of such
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.