Veterans trial targets the most dangerous window after psychiatric discharge
The weeks after psychiatric discharge carry a steep risk of suicide-related behaviors, and the Veterans Coordinated Community Care Study is built around that clinical reality. In this multicenter randomized controlled trial, investigators are testing whether a targeted post-discharge outreach program can reduce suicide-related behaviors among high-risk Veterans Health Administration psychiatric inpatients identified by a machine learning model.
The design is deliberate. Rather than offering the same preventive follow-up to all psychiatric inpatients, the trial concentrates on a subgroup with the greatest predicted danger after leaving the hospital. That approach matters because post-discharge outreach can be labor-intensive, and the health system needs evidence on whether limited preventive resources should be focused on those most likely to benefit.
Risk prediction is being used to direct prevention to a narrower cohort
The study enrolled 850 VHA psychiatric inpatients, all selected as high risk by a predictive model built to identify patients vulnerable to post-discharge suicide. Participants were randomized in equal numbers to either the Coping Long Term with Active Suicide Program intervention or treatment as usual. This is a pragmatic test of whether prediction-guided care can do more than stratify risk on paper; it asks whether an algorithm can help deliver a prevention strategy at the moment the system often loses contact with patients.
That sequence reflects a familiar problem in suicide prevention. Discharge from inpatient psychiatry often marks a drop in daily monitoring, medication oversight, and therapeutic contact, precisely when patients may still be struggling with acute despair, impulsivity, or unresolved psychosocial stressors. A targeted intervention in this window aims to blunt the transition from hospital containment to community vulnerability.
CLASP will be judged against standard care over 6 months
The intervention arm receives CLASP, a post-discharge clinical outreach program designed for active suicide prevention. The comparator is treatment as usual, giving the trial a direct look at whether structured follow-up adds meaningful protection beyond routine post-hospital care. The primary endpoint is suicide-related behaviors over 6 months, a clinically consequential outcome that captures more than death alone and better reflects the high-frequency harms that often precede lethal events.
The 6-month follow-up period is long enough to measure whether the effect of early outreach persists beyond the immediate post-discharge interval, when contact intensity often fades. By measuring average treatment effects and adjusting for loss to follow-up, the trial is set up to answer a practical systems question: does the intervention produce a measurable reduction in SRBs in this very specific high-risk group?
A one-size-fits-all approach has rarely been efficient in suicide prevention.
Trial design also allows for variation in response
Investigators plan to examine heterogeneity of treatment effects, which could reveal whether some patients benefit more than others from the same outreach strategy. In clinical terms, that is important because suicide risk is not monolithic. The factors that drive a veteran’s post-discharge danger may include prior attempts, psychiatric symptom burden, treatment engagement, social instability, substance use, or other vulnerabilities not captured by a single risk score alone.
If the intervention works best in certain subgroups, that information could shape how the Veterans Health Administration deploys CLASP in real-world care. If benefit is broad, the program could be used more confidently as a targeted standard for high-risk discharges. If benefit is limited or uneven, the model may still have value, but the prevention strategy would need refinement.
The trial design does not assume the predictive model is perfect; it assumes the model is useful enough to identify a higher-yield population for testing a preventive service.
Recruitment and analysis timeline point to a near-term evidence gap
Recruitment is underway and is scheduled to end in September 2024. Six-month follow-up will conclude and analysis will begin in summer 2025. Those dates matter because the trial is still in motion, and no outcome data have yet been reported. Clinicians and policymakers are therefore waiting on a study that may determine whether post-discharge outreach should be deployed more selectively across VHA psychiatric settings.
This is an especially important question for large health systems, where population-level suicide prevention must compete with finite staffing and uneven access. A targeted intervention that can lower SRBs in a model-defined high-risk cohort would support more efficient use of outreach personnel, care managers, and post-discharge support services. A neutral result would be just as consequential, because it would mean the predictive enrichment strategy did not translate into a detectable clinical advantage over routine care.
Why the study could influence VA suicide prevention policy
The Veterans Health Administration is the natural setting for a trial like this because it combines a large psychiatric inpatient population with the infrastructure needed to test coordinated community follow-up. The study also sits at the intersection of two growing clinical priorities: suicide prevention and precision targeting of services. Instead of asking only whether an intervention works in the abstract, the investigators are asking whether it works when directed by a risk model to the patients most likely to need it.
Biologically and behaviorally, that approach is plausible. The post-discharge period often features abrupt changes in supervision, routine, sleep, access to lethal means counseling, medication continuity, and social support. A structured outreach program may reduce SRBs by restoring continuity, identifying warning signs early, and reinforcing coping strategies before risk escalates.
The result will be especially informative if the intervention shows an effect on behaviors short of suicide death, because those events are more frequent and may provide an earlier signal of benefit. If no reduction emerges, the field will need to reassess whether the model identifies the right patients, whether CLASP is intensive enough, or whether different forms of follow-up are needed after psychiatric hospitalization.
For now, the study remains a carefully constructed test of a single question: can targeted outreach, delivered after discharge to Veterans at highest predicted risk, lower suicide-related behaviors during the months when the system is least able to watch them closely?