Systems and methods for transitioning patient care from signal-based monitoring to risk-based monitoring
A risk-based patient monitoring system for critical care patients combines data from multiple sources to assess the current and the future risks to the patient, thereby enabling providers to review a current patient risk profile and to continuously track a clinical trajectory. A physiology observer module in the system utilizes multiple measurements to estimate Probability Density Functions (PDF) of a number of Internal State Variables (ISVs) that describe components of the physiology relevant to the patient treatment and condition. A clinical trajectory interpreter module in the system utilizes the estimated PDFs of ISVs to identify under which probable patient states the patient can be currently categorized and assign a probability value that the patient will be in each of the identified states. The combination of patient states and their probabilities is defined as the clinical risk to the patient.
1 . A computer-based method of risk-based monitoring of a patient, the method comprising:
generating, by the computer, predicted probability density functions of internal state variables for a subsequent time step (t k+1 ), wherein the predicted probability density functions are calculated using posterior estimated probability density functions from a preceding time step (t k );
acquiring, by the computer at subsequent time step (t k+1 ), physiological data from a set of sensors connected with the patient;
generating a conditional likelihood kernel for the subsequent time step (t k+1 ), the conditional likelihood kernel comprising conditional probability density functions of the physiological data acquired at subsequent time step (t k+1 ) given the predicted probability density functions of internal state variables for subsequent time step (t k+1 );
continuously estimating a risk that a particular bio-marker of the patient is abnormal, wherein the particular bio-marker comprises a hidden internal state variable, because it exceeds, by being either above or below, a corresponding pre-defined clinically significant value for that bio-marker, by:
generating, using Bayes theorem operating on (a) the conditional likelihood kernel and (b) predicted probability density functions of internal state variables for subsequent time step (t k+1 ), posterior probability density functions for the plurality of the internal state variables for the subsequent time step (t k+1 ); and
generating, for the particular bio-marker and based on the posterior probability density functions for the subsequent time step (t k+1 ), a probability that the particular bio-marker exceeds the corresponding pre-defined threshold for that particular bio-marker; and
generating, for display on a display device, a graphical depiction of the risk that the bio-marker is abnormal.
2 . The method of claim 1 , wherein the hidden internal state variable is not measured directly.
3 . The method of claim 1 wherein:
the set of sensors includes (i) a heart rate sensor, and (ii) a pulse oximetry sensor;
the physiological data comprises heart rate from the heart rate sensor, and oxygen level from the pulse oximetry sensor;
the particular bio-marker comprises mixed venous oxygen saturation; and
the corresponding threshold is mixed venous oxygen saturation at or below a level indicating a patient state of inadequate oxygen delivery.
4 . The method of claim 1 wherein:
the set of sensors includes (i) a heart rate sensor, and (ii) a pulse oximetry sensor, and (iii) a respiratory rate sensor;
the physiological data comprises heart rate from the heart rate sensor, oxygen level from the pulse oximetry sensor, and respiratory rate from the respiratory rate sensor;
the particular bio-marker comprises arterial partial pressure of carbon dioxide blood; and
the corresponding threshold is arterial partial pressure of carbon dioxide at or above a level indicating a patient state of inadequate ventilation of carbon dioxide.
5 . The method of claim 1 wherein:
the set of sensors includes (i) a heart rate sensor, and (ii) a pulse oximetry sensor, and (iii) a respiratory rate sensor;
the physiological data comprises heart rate from the heart rate sensor, oxygen level from the pulse oximetry sensor, and respiratory rate from the respiratory rate sensor;
the particular bio-marker comprises blood pH; and
the corresponding threshold is a blood pH below a level indicating a patient state of acidosis.
6 . The method of claim 1 wherein:
the set of sensors includes (i) a heart rate sensor, and (ii) a pulse oximetry sensor;
the physiological data comprises heart rate from the heart rate sensor, and oxygen level from the pulse oximetry sensor;
the particular bio-marker comprises arterial lactate level; and
the corresponding threshold is an arterial lactate blood level greater than a level indicating a patient state of hyperlactatemia.
7 . A system for risk-based monitoring of a patient, the system comprising:
a data reception module configured to receive measurements of internal state variables from a set of sensors operably coupled to the patient;
an observation model configured to produce a conditional likelihood kernel comprising conditional probability density functions of the physiological data acquired at subsequent time step (t k+1 ) given the predicted probability density functions of internal state variables for subsequent time step (t k+1 ), the predicted probability density functions calculated using posterior estimated probability density functions from a preceding time step (t k );
an inference engine configured to generate, using Bayes theorem operating on (a) the conditional likelihood kernel and (b) predicted probability density functions of internal state variables for subsequent time step (t k+1 ), posterior probability density functions for the plurality of the internal state variables for the subsequent time step (t k+1 );
a clinical trajectory interpreter module configured to generate, for a particular bio-marker that is a hidden internal state variable, and based on the posterior probability density functions for the subsequent time step (t k+1 ), a probability that the particular bio-marker is abnormal by exceeding a corresponding pre-defined threshold comprising a clinically significant value for that bio-marker, wherein said probability may exceed said threshold being either above or below said corresponding pre-defined clinically significant value for that bio-marker; and
a user interaction module configured to generate, for display on a display device, the risk that the bio-marker is abnormal.
8 . The system of claim 7 wherein:
the sensors include (i) a heart rate sensor, and (ii) a pulse oximetry sensor;
the physiological data comprises heart rate from the heart rate sensor, and oxygen level from the pulse oximetry sensor;
the particular bio-marker comprises mixed venous oxygen saturation, which is the hidden internal state variable; and
the corresponding threshold is mixed venous oxygen saturation at or below a given threshold, indicating a patient state of inadequate oxygen delivery.
9 . The system of claim 7 wherein:
the set of sensors includes (i) a heart rate sensor, and (ii) a pulse oximetry sensor, and (iii) a respiratory rate sensor;
the physiological data comprises heart rate from the heart rate sensor, oxygen level from the pulse oximetry sensor, and respiratory rate from the respiratory rate sensor;
the particular bio-marker comprises arterial partial pressure of carbon dioxide, which is the hidden internal state variable; and
the particular bio-marker comprises arterial partial pressure of carbon dioxide (PaCO2) at or above a level indicating a patient state of inadequate ventilation of carbon dioxide.
10 . The system of claim 7 wherein:
the set of sensors includes (i) a heart rate sensor, and (ii) a pulse oximetry sensor, and (iii) a respiratory rate sensor;
the physiological data comprises heart rate from the heart rate sensor, oxygen level from the pulse oximetry sensor, and respiratory rate from the respiratory rate sensor;
the particular bio-marker comprises blood pH, which is the hidden internal state variable; and
the corresponding threshold is a blood pH below a level indicating a patient state of acidosis.
11 . The system of claim 7 wherein:
the sensors include (i) a heart rate sensor, and (ii) a pulse oximetry sensor;
the physiological data comprises heart rate from the heart rate sensor, and oxygen level from the pulse oximetry sensor;
the particular bio-marker comprises arterial lactate level, which is the hidden internal state variable; and
the corresponding threshold is an arterial lactate blood level greater than a level indicating a patient state of hyperlactatemia.
12 . The system of claim 7 , wherein the hidden internal state variable is not measured directly.
13 . A non-transient computer program product comprising executable code, which executable code, when executed by a computer processor, causes the computer processor to implement a method of risk-based monitoring of a patient, the method comprising:
generating, by the computer, predicted probability density functions of internal state variables for a subsequent time step (t k+1 ), wherein the predicted probability density functions are calculated using posterior estimated probability density functions from a preceding time step (t k );
receiving, at the computer at subsequent time step (t k+1 ), physiological data from the set of sensors connected with the patient;
generating a conditional likelihood kernel for the subsequent time step (t k+1 ), the conditional likelihood kernel comprising conditional probability density functions of the physiological data acquired at subsequent time step (t k+1 ) given the predicted probability density functions of internal state variables for subsequent time step (t k+1 );
continuously estimating a risk that a particular bio-marker of the patient is abnormal by being either above or below a corresponding pre-defined clinically significant value for that bio-marker, indicating that the patient is suffering a specific adverse medical condition, by:
generating, using Bayes theorem operating on (a) the conditional likelihood kernel and (b) predicted probability density functions of internal state variables for subsequent time step (t k+1 ), posterior probability density functions for the plurality of the internal state variables for the subsequent time step (t k+1 ); and
generating, for the for a particular bio-marker internal state variable, which is a hidden internal state variable, and based on the posterior probability density functions for the subsequent time step (t k+1 ), a probability that the particular bio-marker internal state variable exceeds the corresponding pre-defined threshold for that particular internal state variable bio-marker; and
generating, for display, on a display device, a graphical depiction of the risk that the patient is suffering the specific adverse medical condition.
14 . The computer program product of claim 13 , wherein the particular bio-marker comprises a hidden internal state variable, which hidden internal state variable is not measured directly.
15 . The computer program product of claim 14 , wherein:
the set of sensors includes (i) a heart rate sensor, and (ii) a pulse oximetry sensor;
the physiological data comprises heart rate from the heart rate sensor and oxygen level from the pulse oximetry sensor;
the particular bio-marker comprises mixed venous oxygen saturation; and
the corresponding threshold is mixed venous oxygen saturation at or below a level indicating a patient state of inadequate oxygen delivery.
16 . The computer program product of claim 14 , wherein:
the set of sensors includes (i) a heart rate sensor, and (ii) a pulse oximetry sensor and (iii) a respiratory rate sensor;
the physiological data comprises heart rate from the heart rate sensor, oxygen level from the pulse oximetry sensor, and respiratory rate from the respiratory rate sensor;
the particular bio-marker comprises arterial partial pressure of carbon dioxide blood; and
the corresponding threshold is arterial partial pressure of carbon dioxide at or above a level indicating a patient state of inadequate ventilation of carbon dioxide.
17 . The computer program product of claim 14 , wherein:
the set of sensor includes (i) a heart rate sensor, and (ii) a pulse oximetry sensor and (iii) a respiratory rate sensor;
the physiological data comprises heart rate from the heart rate sensor, oxygen level from the pulse oximetry sensor, and respiratory rate from the respiratory rate sensor;
the particular bio-marker comprises blood pH; and
the corresponding threshold is a blood pH below a level indicating a patient state of acidosis.
18 . The computer program product of claim 15 , wherein:
the set of sensor includes (i) a heart rate sensor, and (ii) a pulse oximetry sensor;
the physiological data comprises heart rate from the heart rate sensor and oxygen level from the pulse oximetry sensor;
the particular bio-marker comprises arterial lactate level; and
the corresponding threshold is an arterial lactate blood level greater than a level indicating a patient state of hyperlactatemia.