Alternate dose regimen identification system
Methods and systems for performing dose regimen modification are provided. The methods and systems perform operations comprising: receiving prescription related data for treating a patient with an expected level of efficacy, the prescription related data comprising medication regimen information including dose and interval; determining, using a model, a first amount of drug waste based on the prescription related data; comparing the first amount of drug waste to a threshold value; and in response to determining that the first amount of drug waste transgresses the threshold value, identifying an alternate medication regimen that is associated with a treatment having a given level of efficacy corresponding to the expected level of efficacy, the alternate medication regimen being associated with a second amount of drug waste that is lower than the first amount of drug waste; and triggering a notification associated with the alternate medication regimen.
1. A method comprising:
receiving, by one or more processors, prescription related data for treating a patient with an expected level of efficacy, the prescription related data comprising medication regimen information including dose and interval;
determining, using a model, a first amount of drug waste based on the prescription related data;
comparing the first amount of drug waste to a threshold value;
in response to determining that the first amount of drug waste transgresses the threshold value,
identifying an alternate medication regimen that is associated with a treatment having a given level of efficacy corresponding to the expected level of efficacy, the alternate medication regimen being associated with a second amount of drug waste that is lower than the first amount of drug waste; and
triggering a notification associated with the alternate medication regimen, wherein triggering the notification comprises generating a graphical user interface comprising a plurality of interactive regions representing different medication regimens, the graphical user interface comprising:
a prescription region representing the prescription related data and an indicator identifying the dose and interval,
a first plurality indicators representing a first set of alternate medication regimens having a first level of accuracy, and
a second plurality indicators representing a second set of alternate medication regimens having a second level of accuracy;
receiving input that selects a given indicator of the first plurality of indicators or the second plurality of indicators;
in response to receiving the input, updating the prescription region to represent a drug dose vial combination and interval associated with a given alternate medication regimen represented by the given indicator and to represent drug waste associated with the given alternate medication regimen; and
training the model by performing operations comprising:
obtaining a batch of training data comprising a first set of the plurality of training prescription related data features and alternate medication regimens that have been selected;
processing the first set of the plurality of training prescription related data features by the machine learning technique to generate an estimated alternate medication regimen;
computing a loss based on a deviation between the estimated alternate medication regimen and the corresponding alternate medication regimens of the first set of the plurality of training prescription related data features; and
updating parameters of the machine learning technique based on the computed loss.
2. The method of claim 1 , wherein the threshold value comprises 14 milligrams per dose or more.
3. The method of claim 1 , wherein identifying the alternate medication regimen further comprises:
generating a plurality of different medication regimens each being associated with the expected level of efficacy;
computing a plurality of amounts of drug waste associated with the plurality of different medication regimens; and
selecting the alternate medication regimen from the plurality of different medication regimens based on the plurality of amounts of drug waste.
4. The method of claim 3 , further comprising:
selecting the alternate medication regimen based on a criterion and the plurality of amounts of drug waste.
5. The method of claim 4 , wherein the criterion comprises a waste value that is below the threshold value.
6. The method of claim 4 , wherein the criterion comprises a zero waste value.
7. The method of claim 4 , wherein the criterion comprises a combination of drug dose vials that is associated with a specified level of adherence, a specified level of side effects, injection quantity, or a specified level of efficacy.
8. The method of claim 4 , wherein the criterion comprises an inventory parameter of drug dose vials.
9. The method of claim 8 , wherein the inventory parameter comprises a rate at which a specified type of drug dose vial is dispensed or an amount of the drug dose vial that is available.
10. The method of claim 4 , wherein the dose and interval of the prescription related data is a first dose and a first interval, wherein a first of the plurality of different medication regimens comprises a second dose and the first interval, the second dose being greater than the first dose, wherein a second of the plurality of different medication regimens comprises a third dose and the first interval, the third dose being less than the first dose, wherein a fourth of the plurality of different medication regimens comprises a second dose and a second interval.
11. The method of claim 1 , wherein the prescription related data is associated with a drug comprising Monoclonal antibodies with linear pharmacokinetics.
12. The method of claim 1 , wherein triggering the notification comprises causing a message to be displayed by a client device associated with a pharmacy benefit manager, the message comprising the first amount of drug waste and the alternate medication regimen.
13. The method of claim 1 , wherein the graphical user interface includes a third plurality of indicators that represent medication regimens of the first and second sets of alternate medication regimens having a specified type of drug dose vial.
14. The method of claim 1 , wherein the model comprises a machine learning technique comprising a neural network, the machine learning technique trained to establish a relationship between a plurality of training prescription related data features and alternate medication regimens.
15. A system comprising:
one or more processors coupled to a memory comprising non-transitory computer instructions that when executed by the one or more processors perform operations comprising:
receiving prescription related data for treating a patient with an expected level of efficacy, the prescription related data comprising medication regimen information including dose and interval;
determining, using a model, a first amount of drug waste based on the prescription related data;
comparing the first amount of drug waste to a threshold value;
in response to determining that the first amount of drug waste transgresses the threshold value,
identifying an alternate medication regimen that is associated with a treatment having a given level of efficacy corresponding to the expected level of efficacy, the alternate medication regimen being associated with a second amount of drug waste that is lower than the first amount of drug waste; and
triggering a notification associated with the alternate medication regimen, wherein triggering the notification comprises generating a graphical user interface comprising a plurality of interactive regions representing different medication regimens, the graphical user interface comprising:
a prescription region representing the prescription related data and an indicator identifying the dose and interval,
a first plurality indicators representing a first set of alternate medication regimens having a first level of accuracy, and
a second plurality indicators representing a second set of alternate medication regimens having a second level of accuracy;
receiving input that selects a given indicator of the first plurality of indicators or the second plurality of indicators;
in response to receiving the input, updating the prescription region to represent a drug dose vial combination and interval associated with a given alternate medication regimen represented by the given indicator and to represent drug waste associated with the given alternate medication regimen; and
training the model by performing operations comprising:
obtaining a batch of training data comprising a first set of the plurality of training prescription related data features and alternate medication regimens that have been selected;
processing the first set of the plurality of training prescription related data features by the machine learning technique to generate an estimated alternate medication regimen;
computing a loss based on a deviation between the estimated alternate medication regimen and the corresponding alternate medication regimens of the first set of the plurality of training prescription related data features; and
updating parameters of the machine learning technique based on the computed loss.
16. The system of claim 15 , wherein the threshold value comprises 14 milligrams per dose or more.
17. A non-transitory computer readable medium comprising non-transitory computer-readable instructions for performing operations comprising:
receiving prescription related data for treating a patient with an expected level of efficacy, the prescription related data comprising medication regimen information including dose and interval;
determining, using a model, a first amount of drug waste based on the prescription related data;
comparing the first amount of drug waste to a threshold value;
in response to determining that the first amount of drug waste transgresses the threshold value,
identifying an alternate medication regimen that is associated with a treatment having a given level of efficacy corresponding to the expected level of efficacy, the alternate medication regimen being associated with a second amount of drug waste that is lower than the first amount of drug waste; and
triggering a notification associated with the alternate medication regimen, wherein triggering the notification comprises generating a graphical user interface comprising a plurality of interactive regions representing different medication regimens, the graphical user interface comprising:
a prescription region representing the prescription related data and an indicator identifying the dose and interval,
a first plurality indicators representing a first set of alternate medication regimens having a first level of accuracy, and
a second plurality indicators representing a second set of alternate medication regimens having a second level of accuracy;
receiving input that selects a given indicator of the first plurality of indicators or the second plurality of indicators;
in response to receiving the input, updating the prescription region to represent a drug dose vial combination and interval associated with a given alternate medication regimen represented by the given indicator and to represent drug waste associated with the given alternate medication regimen; and
training the model by performing operations comprising:
obtaining a batch of training data comprising a first set of the plurality of training prescription related data features and alternate medication regimens that have been selected;
processing the first set of the plurality of training prescription related data features by the machine learning technique to generate an estimated alternate medication regimen;
computing a loss based on a deviation between the estimated alternate medication regimen and the corresponding alternate medication regimens of the first set of the plurality of training prescription related data features; and
updating parameters of the machine learning technique based on the computed loss.