Enter a participant's admission assessment to reproduce the predictions of the three random forest models described in Development and Preliminary Validation of Models to Predict Risk of Disengagement from Coordinated Specialty Care Programs for First Episode Psychosis. Leave any item blank if it was not recorded; the models were trained with missing values and use them directly.
Upload a CSV with one row per participant. Column names must match the variable names in the reference table below (an optional ID column is carried through). Blank cells, NA and columns that are absent are treated as missing. Values are used as entered, so apply the same coding as the development data.
Coding and ranges observed in the development data. Site variables describe the participant's program and were calculated from its participants; programs with fewer than 26 discharged participants had them set to missing.
| Variable | Description | Coding | Observed range | % recorded |
|---|