The predictD risk algorithm predicts the onset of major depression in the next year. The validation study was done in seven Spanish provinces in six different communities. A total of 4,574 primary care consultants who did not suffer from depression at the beginning of the study were included and followed up for one year.

Forty different risk factors related to the onset of depression were studied. For the predictD risk algorithm, 11 of these risk factors and one interaction were selected that together predict the incidence of new cases of major depression with good validity - area under the ROC curve (AUROC) of 0.82 - and good calibration (the probabilities of risk that the predictive model predicts are very similar to those that people actually have)¹.