Predictive healthcare scheduling to adjust patient outcomes
A method of modifying a healthcare employee schedule includes receiving an indication that a pre-scheduled healthcare employee belonging to a first class of healthcare employees will be absent from a scheduled shift at a first healthcare facility, receiving a first plurality of patient health profiles for a first plurality of patients expected to visit the first healthcare facility during the scheduled shift, receiving a first plurality of employee profiles for a plurality of healthcare employees, simulating a first predicted outcome score for each employee of the plurality of healthcare employees using a simulator, selecting a preferred replacement employee from the plurality of healthcare employees based on the simulated first predicted outcome scores, and scheduling the preferred replacement employee during the scheduled shift. Each first predicted outcome score is simulated using a computer-implemented machine learning model configured to generate treatment outcome positivity values based on patient health profiles and employee profiles.
1 . A method of modifying a healthcare employee schedule, the method comprising:
receiving an indication that a pre-scheduled healthcare employee will be absent from a scheduled shift at a first healthcare facility, the pre-scheduled healthcare employee belonging to a first class of healthcare employees;
receiving a first plurality of patient health profiles for a first plurality of patients, each patient health profile of the first plurality of patient profiles corresponding to one patient of the first plurality of patients, the first plurality of patients expected to visit the first healthcare facility during the scheduled shift;
receiving a first plurality of employee profiles for a plurality of healthcare employees scheduled at a second healthcare facility during the scheduled shift, each employee profile for the plurality of healthcare employees corresponding to one healthcare employee of the plurality of healthcare employees, wherein:
the healthcare employees of the plurality of healthcare employees belong to the first class of healthcare employees; and
each employee profile of the first plurality of employee profiles comprises at least one employee attribute;
receiving a second plurality of patient health profiles for a second plurality of patients, each patient health profile of the second plurality of patient profiles corresponding to one patient of the second plurality of patients, the second plurality of patients expected to visit the second healthcare facility during the scheduled shift, wherein each healthcare employee of the plurality of healthcare employees is scheduled at the second healthcare facility during the scheduled shift;
simulating, for each employee of the plurality of healthcare employees, a first predicted outcome score for the first plurality of patients using a simulator, a computer-implemented machine learning model, the first plurality of patient health profiles, and a respective employee profile of the first plurality of employee profiles, the computer-implemented machine learning model configured to generate treatment outcome positivity values and to accept as inputs pluralities of patient health profiles and at least one employee profile;
simulating, for each employee of the plurality of healthcare employees, a second predicted outcome score for the second plurality of patients using the simulator, the computer-implemented machine learning model, the second plurality of patient health profiles, and the respective employee profile of the first plurality of employee profiles;
selecting a preferred replacement employee from the plurality of healthcare employees based on the first predicted outcome scores and the second predicted outcome scores; and
responsive to selecting the preferred replacement employee, modifying an electronic scheduling system to schedule the preferred replacement employee at the first healthcare facility instead of the second healthcare facility during the scheduled shift.
2 . The method of claim 1 , wherein at least one healthcare employee of the plurality of healthcare employees is scheduled, in the electronic scheduling system, to work at a second healthcare facility during the scheduled shift.
3 . The method of claim 1 , wherein selecting the preferred replacement employee comprises using an optimization algorithm.
4 . The method of claim 3 , wherein the first plurality of patient profiles comprises electronic health records for the first plurality of patients.
5 . The method of claim 4 , and further comprising receiving a plurality of patient identifying information for the first plurality of patients, and wherein receiving the first plurality of patient health profiles comprises querying a patient health profile database with the plurality of patient identifying information.
6 . The method of claim 5 , and further comprising scheduling the first plurality of patients for a plurality of appointments at the first healthcare facility during the scheduled shift.
7 . The method of claim 6 , wherein scheduling the first plurality of patients comprises:
receiving identifying information of a pool of incoming patients, the first plurality of patients belonging to the pool of incoming patients;
querying the patient health profile database with the identifying information of the pool of incoming patients to receive patient health profiles for the pool of incoming patients; and
selecting the first plurality of patients from the pool of incoming patients based on the scheduled shift and the patient health profiles for the pool of incoming patients.
8 . The method of claim 7 , and further comprising selecting the plurality of healthcare employees from a pool of healthcare employees of the first class of healthcare employees based on employee availability during the scheduled shift.
9 . The method of claim 8 , wherein selecting the plurality of healthcare employees comprises automatically selecting the plurality of healthcare employees in response to receiving the indication that the pre-scheduled healthcare employee will be absent.
10 . The method of claim 1 , and further comprising:
receiving a second plurality of employee profiles for a second plurality of healthcare employees, each employee profile for the second plurality of healthcare employees corresponding to one healthcare employee of the second plurality of healthcare employees, wherein:
the employees of the second plurality of healthcare employees are scheduled at the first healthcare facility during the scheduled shift; and
each employee profile of the second plurality of employee profiles comprises at least one employee attribute;
simulating, for each employee of the first plurality of healthcare employees, the first predicted outcome score comprises:
creating a plurality of employee combinations, each employee combinations of the plurality of employee combinations including the second plurality of healthcare employees and one healthcare employee of the first plurality of healthcare employees; and
simulating the first predicted outcome score for each employee combination using the simulator, the computer-implemented machine learning model, the first plurality of patient health profiles, the second plurality of employee profiles, and the respective employee profile of the first plurality of employee profiles; and
selecting a preferred replacement employee from the first plurality of healthcare employees comprises selecting a preferred employee combination based on the simulated first predicted outcome scores, wherein the preferred replacement employee is a healthcare employee of the first plurality of healthcare employees belonging to the preferred employee combination.
11 . The method of claim 1 , wherein the preferred replacement employee has a greatest difference between a corresponding first predicted outcome score and a corresponding second predicted outcome score.
12 . The method of claim 1 , wherein the pre-scheduled healthcare employee is a first pre-scheduled healthcare employee, the scheduled shift is a first scheduled shift, the plurality of healthcare employees is a first plurality of healthcare employees and the preferred replacement employee is a first preferred replacement employee, and further comprising:
receiving an indication that a second pre-scheduled healthcare employee will be absent from a second scheduled shift at the first healthcare facility, the pre-scheduled healthcare employee belonging to a first class of healthcare employees and the first scheduled shift at least partially overlapping with the second scheduled shift;
receiving a second plurality of employee profiles for a second plurality of healthcare employees scheduled at the second healthcare facility during the second scheduled shift, each employee profile of the second plurality of healthcare employees corresponding to one healthcare employee of the second plurality of healthcare employees, wherein:
the healthcare employees of the plurality of healthcare employees belong to the second class of healthcare employees; and
each employee profile of the second plurality of employee profiles comprises at least one employee attribute;
simulating, for each employee of the second plurality of healthcare employees, a third predicted outcome score for the second plurality of patients using the simulator, the computer-implemented machine learning model, the first plurality of patient health profiles, and a respective employee profile of the second plurality of employee profiles;
simulating, for each employee of the second plurality of healthcare employees, a fourth predicted outcome score for the second plurality of patients using the simulator, the computer-implemented machine learning model, the second plurality of patient health profiles, and the respective employee profile of the second plurality of employee profiles;
selecting a second preferred replacement employee from the plurality of healthcare employees based on the third predicted outcome scores and the fourth predicted outcome scores; and
responsive to selecting the preferred replacement employee, modifying the electronic scheduling system to schedule the second preferred replacement employee at the first healthcare facility instead of the second healthcare facility during the second scheduled shift.
13 . A system comprising:
a patient database;
an employee database;
an electronic employee scheduling system;
a processor; and
computer-readable memory encoded with instructions that, when executed, cause the processor to:
receive an indication that a pre-scheduled healthcare employee will be absent from a scheduled shift at a first healthcare facility, the pre-scheduled healthcare employee belonging to a first class of healthcare employees;
receive, from the patient database, a first plurality of patient health profiles for a first plurality of patients, each patient health profile of the first plurality of patient profiles corresponding to one patient of the first plurality of patients, the first plurality of patients expected to visit the first healthcare facility during the scheduled shift;
receive, from the employee database, a first plurality of employee profiles for a plurality of healthcare employees scheduled at a second healthcare facility during the scheduled shift, each employee profile for the plurality of healthcare employees corresponding to one healthcare employee of the plurality of healthcare employees, wherein:
the healthcare employees of the plurality of healthcare employees belong to the first class of healthcare employees; and
each employee profile of the first plurality of employee profiles comprises at least one employee attribute;
receive a second plurality of patient health profiles for a second plurality of patients, each patient health profile of the second plurality of patient profiles corresponding to one patient of the second plurality of patients, the second plurality of patients expected to visit the second healthcare facility during the scheduled shift, wherein each healthcare employee of the plurality of healthcare employees is scheduled at the second healthcare facility during the scheduled shift;
simulate, for each employee of the plurality of healthcare employees, a first predicted outcome score for the first plurality of patients using a simulator, a computer-implemented machine learning model, the first plurality of patient health profiles, and a respective employee profile of the first plurality of employee profiles, the computer-implemented machine learning model configured to generate treatment outcome positivity values and to accept as inputs pluralities of patient health profiles and at least one employee profile;
simulate, for each employee of the plurality of healthcare employees, a second predicted outcome score for the second plurality of patients using the simulator, the computer-implemented machine learning model, the second plurality of patient health profiles, and the respective employee profile of the first plurality of employee profiles;
select a preferred replacement employee from the plurality of healthcare employees based on the first predicted outcome scores and the second predicted outcome scores; and
responsive to selecting the preferred replacement employee, modify the electronic employee scheduling system to schedule the preferred replacement employee at the first healthcare facility instead of the second healthcare facility during the scheduled shift.
14 . The system of claim 13 , wherein the system further comprises a patient scheduling system and the instructions, when executed, further cause the processor to:
receive, from the patient scheduling system, a plurality of patient identifying information for the first plurality of patients; and
receive the first plurality of patient health profiles by querying the patient health profile database with the plurality of patient identifying information.
15 . The system of claim 14 , wherein the instructions, when executed, cause the processor to modify the patient scheduling system to schedule the first plurality of patients for a plurality of appointments at the first healthcare facility during the scheduled shift.
16 . The system of claim 15 , wherein the instructions, when executed, cause the processor to:
receive, from the employee database, a pool of healthcare employees of the first class of healthcare employees; and
select the plurality of healthcare employees from the pool of healthcare employees based on employee availability during the scheduled shift.
17 . The system of claim 16 , wherein the instructions, when executed, cause the processor to:
receive, from the patient database, identifying information of a pool of incoming patients, the first plurality of patients belonging to the pool of incoming patients;
query the patient database with the identifying information of the pool of incoming patients to receive patient health profiles for the pool of incoming patients; and
select the first plurality of patients from the pool of incoming patients based on the scheduled shift and the patient health profiles for the pool of incoming patients.