Clinical risk model

1. A model-assisted system for predicting health care services, the system comprising:

at least one processor programmed to:

access a database storing medical records associated with a plurality of patients;

analyze a medical record associated with a patient of the plurality of patients to identify a characteristic of the patient;

receive an indication of a selected type of health care service of interest, the selected type of health care service being selected by a user from a plurality of types of health care services;

determine, based on the patient characteristic and using a trained machine learning model, a patient risk level indicating a likelihood that the patient will require a health care service corresponding to the selected type of health care service within a predetermined time period relative to a date the medical record is accessed, the machine learning model being trained using a machine learning system configured to:

access training data comprising a plurality of training medical records, a simulated effective date for each of the plurality of training medical records, and a corresponding healthcare service date for each of the plurality of training medical records, the simulated effective date representing a simulated date each of the plurality of training medical records is accessed;

extract features from the plurality of training medical records; and

generate a model weight for each of the extracted features using a training algorithm;

compare the patient risk level to a predetermined risk threshold;

generate, based on the comparison, a report indicating a recommended intervention for the patient;

schedule the recommended intervention for the patient, the recommended intervention being scheduled based on a comparison of a priority level for the patient to a priority level of at least one additional patient of the plurality of patients;

determine a first calibration factor indicating a difference between an average patient risk level and an average actual healthcare service usage for a first group of the plurality of patients;

determine, based on a comparison of the first calibration factor to a second calibration factor associated with a second group of the plurality of patients, a bias associated with the first group relative to the second group of the plurality of patients; and

apply, based on the determined bias, a correction factor to at least one variable of the trained machine learning model.

2. The system of claim 1 , wherein the medical record comprises structured data associated with the patient and analyzing the medical record comprises analyzing the structured data.

3. The system of claim 1 , wherein the medical record comprises unstructured data associated with the patient and analyzing the medical record comprises analyzing the unstructured data.

4. The system of claim 1 , wherein the patient characteristic comprises prior use of medical services by the patient.

5. The system of claim 1 , wherein the patient characteristic comprises an indication of a medical diagnosis for the patient.

6. The system of claim 1 , wherein the patient characteristic comprises at least one of a laboratory or diagnostic test result for the patient.

7. The system of claim 1 , wherein the at least one processor is further configured to generate the priority level for the patient based on the patient risk level.

8. The system of claim 1 , wherein the at least one processor is further configured to:

generate reports indicating recommended interventions for the plurality of patients; and

schedule, based on the reports, the recommended interventions for the plurality of patients within the predetermined time period.

9. The system of claim 1 , wherein the at least one processor is further configured to generate a report indicating the bias.

10. The system of claim 1 , wherein the first group comprises patients having a first ethnicity and the second group comprises patients having a second ethnicity.

11. The system of claim 1 , wherein the at least one processor is further configured to determine a confidence interval representing a range of values for the first calibration factor associated with a particular degree of confidence.

12. A computer-assisted method for predicting health care services, the method comprising:

accessing a database storing a medical record associated with a patient;

analyzing the medical record to identify a characteristic associated with the patient;

receiving an indication of a selected type of health care service of interest, the selected type of health care service being selected by a user from a plurality of types of health care services;

determining, based on the patient characteristic and using a trained machine learning model, a patient risk level indicating a likelihood that the patient will require a health care service corresponding to the selected type of health care service within a predetermined time period relative to a date the medical record is accessed, the machine learning model being trained using a machine learning system configured to:

access training data comprising a plurality of training medical records, a simulated effective date for each of the plurality of training medical records, and a corresponding healthcare service date for each of the plurality of training medical records, the simulated effective date representing a simulated date each of the plurality of training medical records is accessed;

extract features from the plurality of training medical records; and

generate a model weight for each of the extracted features using a training algorithm;

comparing the patient risk level to a predetermined risk threshold;

generating, based on the comparison, a report indicating a recommended intervention for the patient;

scheduling the recommended intervention for the patient, the recommended intervention being scheduled based on a comparison of a priority level for the patient to a priority level of at least one additional patient of the plurality of patients;

determining a first calibration factor indicating a difference between an average patient risk level and an average actual healthcare service usage for a first group of the plurality of patients;

determining, based on a comparison of the first calibration factor to a second calibration factor associated with a second group of the plurality of patients, a bias associated with the first group relative to the second group of the plurality of patients; and

applying, based on the determined bias, a correction factor to at least one variable of the trained machine learning model.

13. The method of claim 12 , wherein the medical record comprises structured data associated with the patient and analyzing the medical record comprises analyzing the structured data.

14. The method of claim 12 , wherein the medical record comprises unstructured data associated with the patient and analyzing the medical record comprises analyzing the unstructured data.

15. The method of claim 12 , wherein the patient characteristic comprises prior use of medical services by the patient.

16. The method of claim 12 , wherein the patient characteristic comprises an indication of a medical diagnosis for the patient.

17. The method of claim 12 , wherein the patient characteristic comprises at least one of a laboratory or diagnostic test result for the patient.

18. The method of claim 12 , wherein the method further comprises generating the priority level for the patient based on the patient risk level.

19. The method of claim 12 , wherein the method further comprises:

generating reports indicating recommended interventions for the plurality of patients; and

scheduling, based on the reports, recommended interventions for the plurality of patients within the predetermined time period.