METHOD FOR CONTINUOUS FORECASTING OF DISEASE SEVERITY, DEATH AND HOSPITALIZATION DURATION Russian patent published in 2017 - IPC G06F19/00 

FIELD: medicine.

SUBSTANCE: device for patient information contains a controller that takes the physiological and/or laboratory data on the current patient who receives treatment in a medical institution, from patient monitoring devices; a unit that determines the duration of stay, showing how much time was required for the current patient treatment; and a forecasting unit, which predicts the outcome variable by application of the time-dependent outcome variable estimation algorithm to physiological and/or laboratory data on the current patient. The algorithm for outcome variables estimation includes time-dependent weight values or scales, using a specific residence time during which the current patient received treatment. The patient monitoring system includes a patient information device and a patient monitoring device that collects physiological data from the patient. A method for outcome variable prediction for a patient comprises the following steps: access to at least one of a plurality of data fields for previous patients including physiological and/or laboratory data and residence time indicating how much time did it take for each patient to receive treatment during formation of physiological and/or laboratory data, and the corresponding outcome variables, in the clinical database; formation of a time-dependent algorithm for outcome variables estimation based on a set of physiological and/or laboratory data for previous patients corresponding to the residence time and the corresponding outcome variables. The algorithm for outcome variables estimation includes time-dependent weighting values or scales. Machine readable media containing the software that, when loaded by the processor, programs the processor to execute variable outcome prediction for the patient. Method for outcome probability forecasting contains the following stages: measurement of physiological and/or laboratory data for the current patient receiving treatment in a medical institution; determination of the residence time that current patient spent receiving treatment in a medical institution; application of the time-dependent algorithm to predict an outcome, using the measured physiological and/or laboratory data and the specified residence time that the current patient spent receiving treatment in a medical institution, to forecast the outcome. The algorithm for outcome estimation includes time-dependent weighting values or scales.

EFFECT: higher prediction accuracy.

15 cl, 4 dwg, 1 tbl