Readmission risk assessment

A method comprising:

• identifying, by a first computing process, a patient condition for a patient admitted to a clinical facility;
• selecting, by a second computing process, from a plurality of readmission risk prediction models, a readmission risk prediction model based on the patient condition, each readmission risk prediction model having been generated for a given condition using clinically relevant data from patients diagnosed with that given condition;
• receiving, by a third computing process, a first set of patient data for the patient while the patient is admitted to the clinical facility;
• computing, by a fourth computing process, a first readmission risk score using the readmission risk prediction model and the first set of patient data, the first readmission risk score representing a probability of readmission for the patient;
• providing, by a fifth computing process, a first indication of readmission risk for the patient based on the readmission risk score for presentation to a clinician treating the patient;
• providing, by a sixth computing process, an inpatient treatment recommendation for the patient based on the first readmission risk score;
• receiving, by a seventh computing process, a second set of patient data for the patient after the patient has been discharged from the clinical facility;
• computing, by an eight computing process, a second readmission risk score using the readmission risk prediction model and the second set of patient data;
• providing, by a ninth computing process, a second indication of readmission risk for the patient based on the second readmission risk score;
• and
• providing, by a tenth computing process, an outpatient treatment recommendation for the patient based on the second readmission risk score;
• wherein each of the computing processes is performed by one or more computing devices.

A computerized system comprising:

• one or more processors;
• and
• one or more computer storage media storing instructions that, when used by the one or more processors, cause the one or more processors to:
• identify a patient condition for a patient admitted to a clinical facility;
• select, from a plurality of readmission risk prediction models, a readmission risk prediction model based on the patient condition, each readmission risk prediction model having been generated for a given condition using clinically relevant data from patients diagnosed with that given condition;
• receive a first set of patient data for the patient while the patient is admitted to the clinical facility;
• compute a first readmission risk score using the readmission risk prediction model and the first set of patient data, the first readmission risk score representing a probability of readmission for the patient;
• provide a first indication of readmission risk for the patient based on the readmission risk score for presentation to a clinician treating the patient;
• provide an inpatient treatment recommendation for the patient based on the first readmission risk score;
• receive a second set of patient data for the patient after the patient has been discharged from the clinical facility;
• compute a second readmission risk score using the readmission risk prediction model and the second set of patient data;
• providing a second indication of readmission risk for the patient based on the second readmission risk score;
• and
• providing an outpatient treatment recommendation for the patient based on the second readmission risk score.