Prescription assistance device and prescription assistance method
A prescription assistance device includes a control unit configured to acquire attribute data indicating an attribute of a patient, time series data on a state of the patient, and drug history data indicating a prescription history of a drug for the patient, configured to predict, based on the attribute data, the time series data, and the drug history data, an effect of the drug for each of a plurality of prescription candidates in which combinations of a type, an administration amount, and an administration timing of the drug are different from one another, and configured to present a prediction result to a user.
1. A prescription assistance device comprising:
a processor configured to:
acquire attribute data indicating an attribute of a patient, time series data on a state of the patient, and drug history data indicating a prescription history of a drug for the patient;
generate, receive, or read, a plurality of prescription candidates in which combinations of a type, an administration amount, and an administration timing of the drug are different from one another;
predict, based on the attribute data, the time series data, and the drug history data, an effect of the drug for each of the plurality of prescription candidates in which combinations of a type, an administration amount, and an administration timing of the drug are different from one another;
present a prediction result to a user of the effect of the drug for each of the plurality of prescription candidates via an output interface;
receive, from the user, an operation for selecting one of the plurality of prescription candidates outputted on the output interface via an input interface; and
control a pump so as to administer a drug of a type specified by the prescription candidate selected by the user to the patient at an administration amount and an administration timing specified by the prescription candidate selected by the user.
2. The prescription assistance device according to claim 1 , wherein
the time series data includes results of vital measurement on the patient at a plurality of time points; and
the processor is configured to:
predict, as the effect of the drug for each of the plurality of prescription candidates, results of vital measurement on the patient after drug administration.
3. The prescription assistance device according to claim 2 , wherein the vital measurement on the patient includes urine volume measurement on the patient.
4. The prescription assistance device according to claim 1 , wherein
the time series data includes results of a blood test on the patient at a plurality of time points; and
the processor is configured to:
predict, as the effect of the drug for each of the plurality of prescription candidates, results of blood tests on the patient after drug administration.
5. The prescription assistance device according to claim 4 , wherein the blood test on the patient includes at least one of a cardiac function test, a renal function test, and an electrolyte test on the patient.
6. The prescription assistance device according to claim 1 , wherein the processor is configured to:
predict, as the effect of the drug for each of the plurality of prescription candidates, the number of days until the patient leaves a bed after drug administration.
7. The prescription assistance device according to claim 1 , wherein the drug for each of the plurality of prescription candidates includes at least one of a diuretic, a vasodilator, a cardiotonic, a vasopressor, a heart rate controlling drug, an analgesic, and an infusion.
8. The prescription assistance device according to claim 1 , wherein the processor is configured to:
receive, from the user, an operation for adjusting at least one of the plurality of prescription candidates.
9. The prescription assistance device according to claim 1 , wherein the processor is configured to:
predict the effect of the drug for each of the plurality of prescription candidates using a trained model to which the attribute data, the time series data, and the drug history data are input; and
output the effect of the drug for each of the plurality of prescription candidates from the attribute data, the time series data, and the drug history data input into the trained model.
10. The prescription assistance device according to claim 9 , wherein when a drug selected from the plurality of prescription candidates is administered to the patient, the processor is configured to:
acquire data on the state of the patient at a timing determined for each type of the drug; and
feed the acquired data on the state of the patient at the timing determined for each type of the drug to the trained model.
11. The prescription assistance device according to claim 1 , wherein when a therapeutic intervention different from the drug administration is performed on the patient, the processor is configured to:
exclude at least a part of the time series data before the therapeutic intervention and predicts the effect of the drug for each of the plurality of prescription candidates.
12. The prescription assistance device according to claim 11 , wherein the therapeutic intervention includes at least one of respiratory support and blood circulation support.
13. The prescription assistance device according to claim 1 , wherein the processor is configured to adjust the plurality of prescription candidates with reference to at least one of the acquired attribute data, the acquired time series data, and the acquired drug history data.
14. A prescription assistance method comprising:
acquiring, by a prescription assistance device, attribute data indicating an attribute of a patient, time series data on a state of the patient, and drug history data indicating a prescription history of a drug for the patient;
generating, receiving, or reading, by the prescription assistance device, a plurality of prescription candidates in which combinations of a type, an administration amount, and an administration timing of the drug are different from one another;
predicting, by the prescription assistance device, an effect of the drug for each of the plurality of prescription candidates in which combinations of the type, the administration amount, and the administration timing of the drug are different from one another based on the attribute data, the time series data, and the drug history data;
presenting, by the prescription assistance device, a prediction result to a user of the effect of the drug for each of the plurality of prescription candidates via an output interface;
receiving, from the user, an operation for selecting one of the plurality of prescription candidates outputted on the output interface via an input interface; and
controlling, by the prescription assistance device, a pump so as to administer a drug of a type specified by the prescription candidate selected by the user to the patient at an administration amount and an administration timing specified by the prescription candidate selected by the user.
15. The prescription assistance method according to claim 14 , further comprising:
adjusting, by a prescription assistance device, the plurality of prescription candidates with reference to at least one of the acquired attribute data, the acquired time series data, and the acquired drug history data.
16. A prescription assistance method on a computer using a processor, the prescription assistance method comprising:
acquiring, by the processor, attribute data indicating an attribute of a patient, time series data on a state of the patient, and drug history data indicating a prescription history of a drug for the patient;
generating, by the processor, a plurality of prescription candidates in which combinations of a type, an administration amount, and an administration timing of the drug are different from one another with reference to at least one of the acquired attribute data, the acquired time series data, and the acquired drug history data;
predicting, by the processor, an effect of the drug for the plurality of prescription candidates in which combinations of the type, the administration amount, and the administration timing of the drug are different from one another based on the attribute data, the time series data, and the drug history data;
displaying a prediction result for the effect of the drug for the plurality of prescription candidates in which combinations of the type, the administration amount, and the administration timing of the drug are different from one another based on the attribute data, the time series data, and the drug history data via an output interface;
receiving, from the user, an operation for selecting one of the plurality of prescription candidates outputted on the output interface via an input interface; and
controlling, by the processor, a pump so as to administer a drug of a type specified by the prescription candidate selected by the user to the patient at an administration amount and an administration timing specified by the prescription candidate selected by the user.
17. The prescription assistance method according to claim 16 , wherein the time series data includes results of vital measurement on the patient at a plurality of time points, and further comprising:
predicting as the effect of the drug for the plurality of prescription candidates, results of vital measurement on the patient after drug administration.
18. The prescription assistance method according to claim 16 , wherein the time series data includes results of a blood test on the patient at a plurality of time points, and further comprising:
predicting, as the effect of the drug for the plurality of prescription candidates, results of blood tests on the patient after drug administration.
19. The prescription assistance method according to claim 16 , further comprising:
predicting, as the effect of the drug for the plurality of prescription candidates, the number of days until the patient leaves a bed after drug administration.
20. The prescription assistance method according to claim 16 , further comprising:
predicting the effect of the drug for the plurality of prescription candidates using a trained model to which the attribute data, the time series data, and the drug history data are input; and
outputting the effect of the drug for the plurality of prescription candidates from the attribute data, the time series data, and the drug history data input into the trained model.