Systems and methods for healthcare provider dashboards

Inventors

Nida, Dean • Brown-Hayes, Michael • Chung, James • Espinoza, Carolyn • Emerson, Michael • Larson, John • Tschirpke, Jennifer • Driver, Stephani • Yoo, Daniel Jaesup • Curry, Jenifer

Assignees

Cambia Health Solutions Inc

Abstract

Systems and methods for generating dashboards for healthcare providers are provided. In one embodiment, a method comprises receiving, from a user of a client device, a selection of a patient, determining an order of a plurality of display modules, each display module displaying information relating to the patient, generating a dashboard including the plurality of display modules displayed in the order, and transmitting the dashboard to the client device for display to the user. In this way, members may be identified for intervention in a timely manner, and the most relevant information for a healthcare provider may be prioritized for display to a healthcare provider.

Core Innovation

A method receives, from a user of a client device, a selection of a patient, retrieves patient data of the patient from one or more databases, and predicts an actionability for each display module of a plurality of display modules based on the patient data and a user profile of the user using a machine learning model. The actionability for each display module is used to determine an order of the plurality of display modules, where each display module displays information relating to the patient.

The method generates a dashboard including the plurality of display modules displayed in the order and transmits the dashboard to the client device for display to the user. The method receives, from the client device, feedback regarding the order of the plurality of display modules and trains the machine learning model based on the feedback. The ordering therefore adapts over time based on user feedback about the dashboard module order.

A dashboard platform additionally builds patient-specific dashboards composed of a plurality of display modules and orders and selectively emphasizes modules based on predicted actionability and user profiles/preferences using an actionability ML model. The platform can incorporate care predictions and update the actionability model and/or user profile based on user reordering and feedback. Example display modules include contact information, eligibility, provider notes, pharmaceutical timelines, future clinical and cost predictions, power-of-attorney contact, and medical claims history with text and timeline views, along with smart search within claims history.

In parallel, the system manages clinical queues by generating clinical queues using multiple predictive/risk models and sorted/combined scores. A server transmits the queues to client devices, receives provider feedback indicating whether cases are opened or reviewed, and continuously retrains the models. This clinical-queue workflow prioritizes individuals for intervention using risk scores generated by multiple predictive models and supports patient selection from the queue for downstream dashboard generation.

Claims Coverage

The partial document contains three independent claims (method, computer-readable medium, and system) that share a common inventive theme: predicting actionability for multiple patient-related dashboard modules, ordering those modules, generating and transmitting an ordered dashboard, receiving feedback about the order, and training/updating the machine learning model based on the feedback. Across the independent claims, the approach is implemented either as a method, as a computer-readable storage medium with executable instructions, or as a client-server system.

Across the independent claims, the core inventive coverage is the prediction of actionability for multiple patient-related display modules using a machine learning model, ordering the modules based on the predicted actionability, generating and transmitting an ordered dashboard, receiving feedback about the ordering, and training/updating the machine learning model based on that feedback. Additional claim refinements tied to healthcare predictions and a clinical-queue workflow with risk scores are indicated in the dependent claim coverage summarized with the independent claims.

Stated Advantages

Documented Applications