To the Editor Belsher et al1 reviewed the diagnostic accuracy of suicide predictive models (PMs) and also simulated the impact of implementing a PM. As members of the implementation team for the Veterans Health Administration’s Recovery Engagement and Coordination for Health–Veterans Enhanced Treatment (REACH VET) program, we are responding to several concerns noted in the Discussion section including risks associated with implementing PM-based programs and the scarcity of evidence-based treatments (EBTs) for high-risk patients.