Latest Github package now available

With the 1.1.1 release, the CV19 Index is now capable of making accurate predictions for any adult 18 or older. Previous releases of the CV19 Index were trained only on a Medicare data set and were appropriate only for Medicare members. This updated model used a training data set that was representative of the entire US population by combining the original Medicare data with additional data provided by HealthFirst. In addition to the release of a new model, the packaging of the CV19 index has been updated to make it easier to install and use.

Get the latest open source model on Github.

The package available to install now, does it use XGboost plus all disease conditions as input features? what else? age, gender, number of IP admits, ER visits? There are three models in the PDF paper so just want to make sure.

Hello! The model does use XGBoost. The model uses a primary diagnosis of pneumonia, influenza, acute bronchitis, or other respiratory infections specified using ICD-10-CM and AHRQ’s Clinical Classifications Software (CCS) CCSRs = RSP002, RSP003, RSP005, and RSP006. Features do include age, gender, and admissions for the above diagnoses. The open sourced package has an example of how to make the input data.