System Dynamics vs Quality Improvement for Mental Health Care

Age: 18+

Sex: Any

Trial Phase: Academic

Sponsor: VA Office of Research and Development

Disqualifiers: Less than 12 months data, others

No Placebo Group All trial participants will receive the active study treatment (no placebo)

Compensation: Varies

Number of Visits: Unknown

What You Need to Know Before You Apply

What is the purpose of this trial?

This trial tests two methods to enhance mental health care for Veterans, focusing on depression, PTSD, and opioid use disorder. One method, Modeling to Learn (MTL), uses simulations to guide staff in improving care. The other method employs usual quality improvement (QI) techniques. The trial aims to determine which method better helps Veterans start and complete evidence-based treatments. Clinics not meeting certain VA quality standards and serving a sufficient number of unique patients with a team of mental health providers may be suitable for this trial. As an unphased trial, it offers Veterans the chance to contribute to innovative approaches in mental health care improvement.

Will I have to stop taking my current medications?

The trial information does not specify whether participants need to stop taking their current medications.

What prior data suggests that this protocol is safe for improving mental health care?

Research shows that both "Modeling to Learn" (MTL) and usual quality improvement (QI) methods are safe for people.

For the MTL approach, studies on machine learning and AI in mental health have demonstrated that these tools analyze data to predict and improve mental health without causing harm. No reports of serious side effects or negative events have emerged from these studies.

For usual quality improvement methods, past projects successfully improved access to and quality of mental health care in clinics. These improvements occurred safely, with no negative effects on patients.

Both approaches are well-tolerated and aim to enhance mental health care delivery without posing risks to participants. ^{1 2 3 4 5}

Why are researchers excited about this trial?

Researchers are excited about this trial because it explores two different approaches to improving mental health care. Modeling to Learn (MTL) is unique because it uses system dynamics modeling to help clinics understand complex interactions in mental health treatment, aiming to optimize decision-making and care strategies. Unlike typical quality improvement efforts that focus on incremental changes, MTL potentially allows for a more holistic understanding of clinic operations and patient outcomes. This could lead to more effective and efficient mental health care delivery, which is why there's a buzz around what these methods might reveal.

What evidence suggests that this trial's treatments could be effective for improving mental health care?

This trial will compare two approaches: Modeling to Learn (MTL) and Usual Quality Improvement (QI) in mental health care. Studies have shown that computer programs analyzing data, similar to the MTL approach, can effectively predict mental health conditions. These programs examine information to identify risk factors and help detect mental disorders early. For instance, some models analyze speech and behavior to spot signs of mental health issues before they become serious. By predicting future problems, this approach can improve care by optimizing scheduling and staffing decisions. Overall, research suggests that using these computer programs in mental health care can lead to better outcomes and more efficient resource use. ^{1 2 6 7 8}

Who Is on the Research Team?

Lindsey E. Zimmerman, PhD

Principal Investigator

VA Palo Alto Health Care System, Palo Alto, CA

Are You a Good Fit for This Trial?

This trial is for VA divisions and outpatient clinics that are performing below the median in mental health care quality, specifically for depression, PTSD, and opioid use disorder. Clinics must have a multidisciplinary team but can't be part of certain other QI programs or planning to implement a new EHR system during the study.

Inclusion Criteria

VA health care systems are below the median for 24 of 30 VA mental health recommendations for Strategic Analytics for Improvement & Learning (SAIL) and below the median for 3 of 8 VA evidence-based treatment approaches.

Must be below the overall VA quality median (as assessed by the Strategic Analytics for Improvement and Learning or SAIL), which includes 3 of 8 SAIL measures associated with four evidence-based psychotherapies and three evidence-based pharmacotherapies for depression, PTSD, and opioid use disorder.

VA divisions and community-based outpatient clinics (CBOCs) or 'clinics' from regional VA health systems

Exclusion Criteria

Clinics involved in Office of Veterans Access to Care (OVACS) quality improvement program at baseline

Clinics without an onsite multidisciplinary team of mental health or addiction service providers (minimum required: 1 psychiatrist, 1 psychologist, 1 social worker onsite)

Health care systems functioning above median VA mental health recommendations for Strategic Analytics for Improvement & Learning (SAIL) and below the median for 3 of 8 SAIL evidence-based treatment approaches. Only one health care system can be included per arm - MTL vs QI.

Timeline for a Trial Participant

Screening

Participants are screened for eligibility to participate in the trial

2-4 weeks

Implementation

Participatory System Dynamics (PSD) and usual Quality Improvement (QI) are implemented in clinics to improve mental health care access

12 months

Regular sessions with multidisciplinary teams

Follow-up

Participants are monitored for effectiveness of PSD and QI interventions

12 months

Evaluation

Evaluation of the effectiveness, scalability, and affordability of PSD compared to usual QI

6 months

What Are the Treatments Tested in This Trial?

Interventions

• Modeling to Learn (MTL)
• Usual quality improvement (QI)

Trial Overview The trial compares Participatory System Dynamics (PSD), which uses simulations to help staff make better decisions about resource allocation and scheduling, with usual quality improvement methods. It aims to see if PSD can increase patient access to evidence-based mental health care more effectively.

How Is the Trial Designed?

Group I: Usual quality improvement (QI) Experimental Treatment 1 Intervention

12 clinics randomly assigned to usual QI

Group II: Modeling to Learn (MTL) Experimental Treatment 1 Intervention

12 clinics randomly assigned to MTL

Who Is Running the Clinical Trial?

VA Office of Research and Development

Lead Sponsor

Trials

1,691

Recruited

3,759,000+

Citations

1. pmc.ncbi.nlm.nih.gov pmc.ncbi.nlm.nih.gov/articles/PMC12062285/

Machine learning-based predictive modelling of mental ... The study aims to apply machine learning techniques to predict mental health and identify its associated risk factors among Rwandan youth.

2. sciencedirect.com sciencedirect.com/science/article/pii/S2772442524000522

A comprehensive review of predictive analytics models for ... This study reviews the machine learning models, algorithms, and applications for the early detection of mental disease, particularly emphasizing the data ...

3. nature.com nature.com/articles/s41598-025-00386-8

Early detection of mental health disorders using machine ... This study presents a novel framework for the early detection of mental illness disorders using a multi-modal approach combining speech and behavioral data.

4. pmc.ncbi.nlm.nih.gov pmc.ncbi.nlm.nih.gov/articles/PMC11443461/

Machine Learning Techniques to Predict Mental Health ... This study aims to investigate the potential of machine learning in predicting mental health conditions among college students.

5. medinform.jmir.org medinform.jmir.org/2023/1/e44322

Predicting Undesired Treatment Outcomes With Machine ... By using commonly available data from routine care, the technical implementation of the model in clinical practice would be straightforward.

6. mdpi.com mdpi.com/2079-3197/13/8/186

Employing Machine Learning and Deep Learning Models ... This study employed a range of machine learning (ML) and deep learning (DL) models to analyze and classify mental health statuses based on textual data. Each ...

7. sciencedirect.com sciencedirect.com/science/article/pii/S2949916X24000525

Enhancing mental health with Artificial Intelligence Machine learning algorithms can sift through vast patient data, including medical histories, diagnostic tests, and clinical notes, to identify patterns ...

8. dl.acm.org dl.acm.org/doi/10.1145/3729605.3729680

Student Mental Health Risk Assessment Based on ... This study develops a federated learning framework, distributes modeling to local data points of different institutions, and uses differential ...