US 11,627,917 B1 · Predicting chronic kidney disease progression

Abstract

Systems, methods, and computer-readable media are provided for identification of patients having an elevated near-term risk of chronic kidney disease progression, including quantitatively predicting whether or not an elevated risk of progression of Stage 3 or Stage 4 chronic kidney disease is likely within a time interval of up to 36 months subsequent to computing the prediction. Based on the prediction, appropriate care providers may be notified so that the risk of CKD progression may be mitigated. In some embodiments, serial measurements are obtained of urine osmolality, and a challenge with an AVP V2 antagonist and serum sodium concentration is provided. From a time series based on the serial measurements, estimates of each variable's velocity and/or doubling-time may be determined. These values then may be combined via a multivariable mathematical model for providing a leading indicator of near-term future abnormalities in kidney function.

Claims20

Dependent claims

A method for predicting a progression of Chronic Kidney Disease (CKD) in a patient, the method comprising:

measuring one or more physiologic variables of the patient at a plurality of time stamped data points, the one or more physiologic variables associated with kidney function;
administering one or more agents to the patient;
measuring, after a predetermined time period has elapsed, the one or more physiologic variables at one or more post-administration of one or more agents time stamped data points;
compiling a serial measurement time series comprising the plurality of time stamped data points and the one or more post-agent administration time stamped data points;
calculating a predicted probability that the one or more agents will retard the progression of the CKD by transforming the serial measurement time series wherein transforming the plurality of time stamped data points comprises using at least one of the following on the serial measurement time series:
a logistic regression model;
a neural network;
and
a support vector machine model
utilizing the predicted probability, evoking an action corresponding to treatment of the patient.

7 more elements

One or more non-transitory computer-readable media having computer-executable instructions embodied thereon that, when executed, facilitate a method for predicting a progression of Chronic Kidney Disease (CKD) in a patient the method comprising:

receiving measurements of one or more physiologic variables for the patient at a plurality of time stamped data points, the one or more physiologic variables associated with kidney function;
receiving an indication that one or more agents have been administered to the patient;
receiving measurements, after a predetermined time period has elapsed, of the one or more physiologic variables at one or more post-agent administration time stamped data points;
compiling a serial measurement time series comprising the plurality of time stamped data points and the one or more post-agent administration time stamped data points;
calculating a predicted probability that the one or more agents will retard the progression of the CKD by transforming the serial measurement time series wherein transforming the plurality of time stamped data points comprises using at least one of the following on the serial measurement time series:
a logistic regression model;
a neural network;
and
a support vector machine model;
determining if the predicted probability exceeds a predetermined threshold;
and
upon determining that the predicted probability exceeds the predetermined threshold, evoking an action corresponding to treatment of the patient.

9 more elements

A system for predicting the progression of Chronic Kidney Disease (CKD) in a patient, the system comprising:

one or more processers;
one or more sensors configured to measure one or more physiologic variables for the patient, the one or more physiologic variables associated with kidney function;
and
computer storage memory having computer-executable instructions stored thereon that, when executed by the processor, implement a method comprising:
receiving measurements of the one or more physiologic variables for the patient at a plurality of time stamped data points;
receiving an indication that one or more agents have been administered to the patient;
receiving measurements, after a predetermined time period has elapsed, of the one or more physiologic variables at one or more post-agent administration time stamped data points;
compiling a serial measurement time series comprising the plurality of time stamped data points and the one or more post-agent administration time stamped data points;
censoring any of the plurality of time stamped data points affected by at least one of:
error;
and
measurement artifact;
calculating a predicted probability of progression of the CKD of the patient by transforming the serial measurement time series;
determining if the predicted probability exceeds a predetermined threshold;
and
upon determining that the predicted probability exceeds the predetermined threshold, evoking an action corresponding to treatment of the patient.

13 more elements

Dependencies are inferred from the claim text.

Predicting chronic kidney disease progression

Life of this patent

Predicting chronic kidney disease progression

Life of this patent

Abstract

Claims20