Veterans flagged by STORM for case review die less often in a VA trial

In VA's trial across 140 medical centers, patients whose STORM risk scores triggered a mandated case review had 22% lower odds of death within four months. Serious adverse events did not change.

  • 5 for society

    VA researchers report 22% lower odds of death within four months for patients under the review mandate across all 140 medical centers, as an exploratory outcome.

  • 2 for an organization

    Patients under the mandate were five times more likely to receive a case review and received on average 0.5 more risk mitigation strategies.

Depth runs from 1, a convenience or time saved, to 5, a life or a livelihood preserved.

Status
Reported benefit.
Evidence

foundstudiedreportedclaimed

Only the deployer supports the benefit.

Compared with

Patients with similar STORM risk at medical centers that had not yet raised the very high risk cutoff, so that no review was mandated. Measured.

The system’s part
The trial randomized when each medical center raised STORM's very high risk cutoff from the top 1% to the top 5%. Patients of similar predicted risk were thus compared with and without the flag and the review that it triggered. The trial does not compare STORM with another way to pick patients. The paper calls STORM predictive analytics and says that its algorithm was previously validated, but it does not say how the model was built or whether it learned from data.
Not yet checked
LimitsCosts and harmsLearns from dataLater outcome

Download data Cite Suggest a correction

Benefit type
Improved Physical Health
System
STORM (Stratification Tool for Opioid Risk Mitigation)
Made by
Veterans Health Administration
Used by
Veterans Health Administration, all 140 medical centers
Where
United States (all 140 VHA medical centers)
Stage
In routine use
Date
, when the source was published
Kind
Aggregate result
Broad benefit
Analyzing Data or Information, Organizing, Planning, and Prioritizing Work
Last revised
1 November 2026

Sources

  1. Peer-reviewed study, by the organization that used it.

    Strombotne KL, Legler A, Minegishi T, Trafton JA, Oliva EM, Lewis ET, Sohoni P, Garrido MM, Pizer SD, Frakt AB. Effect of a Predictive Analytics-Targeted Program in Patients on Opioids: a Stepped-Wedge Cluster Randomized Controlled Trial. Journal of General Internal Medicine 38(2):375-381, published online 2 May 2022 (PMC9060407). The authors work at VA's Partnered Evidence-based Policy Resource Center and at the Program Evaluation and Resource Center of VHA's Office of Mental Health and Suicide Prevention, and VHA funded the work.

    Mandated review inclusion was associated with a significant decrease in all-cause mortality within 4 months of inclusion (OR: 0.78; 95% CI: 0.65-0.94).

Details

AI system

Workflow

STORM computed each patient's risk of an overdose- or suicide-related event in the next year from VHA records. VHA Notice 2018-08 required interdisciplinary teams to review every patient whom STORM rated very high risk. Each review covered the patient's risks, prescriptions and risk mitigation strategies and was documented in the medical record.

Follow-up

Conditions for replication

The patient is prescribed opioid analgesics and has a predicted risk between the top 1% and 5% nationally. Clinicians have the time to complete case reviews for an expanded high-risk cohort. The risk scores keep their performance over time and across populations.

Entry details

Sector

Healthcare (opioid safety).

Purpose

The Veterans Health Administration built STORM to stratify patients prescribed opioids by their risk of an overdose- or suicide-related event and to support decisions on lowering that risk.

References

Other identifiers

Isrctn: ISRCTN16012111.

29 more questions not yet checked

model, release, provider, role, configuration, inputs and outputs, autonomy, human oversight, beneficiaries, reach, distribution and access, duration, every measurement question, every money question, ownership and control, dependencies, cross-border connections, geopolitical implications, replication attempts, availability, institutional decision, grounds, action and result, success criterion, every evidence question, observed or intended, alternative assessments, related entries, category mappings.

Cite this record

Peter H. Vartanian. (2026). Veterans flagged by STORM for case review die less often in a VA trial (AIBD-1008). Artificial Intelligence Benefits Database, edition of 1 November 2026. https://petervartanian.xyz/aibd/records/AIBD-1008.html (archived at https://doi.org/10.7910/DVN/AADSII, version 7.0)

BibTeX Data Field definitions