The emerging clinical consensus positions SUD alongside conditions like diabetes, hypertension, and heart disease: a chronic illness that requires ongoing medical management, evidence-based intervention, and coordinated care across multiple specialties. For health plans and TPAs, this reframing demands a parallel evolution in how UM programs evaluate appropriate levels of care. The gap between where SUD treatment is heading and how most UM programs are currently designed represents both a clinical risk and a financial one. 

The Gap Between SUD Care Reality and UM Practice 

The American Society of Addiction Medicine (ASAM) Criteria, 4th Edition (published December 2023), organizes SUD treatment into four broad levels of care: Level 1 (Outpatient), Level 2 (Intensive Outpatient), Level 3 (Residential), and Level 4 (Medically Managed Inpatient). Within each level, decimal gradations reflect increasing clinical intensity, with the medically managed sub-levels — 2.7, 3.7, and 4 — representing the points on the continuum where active withdrawal management and biomedical monitoring are required. Yet despite this well-established framework, UM programs often default to binary thinking: inpatient withdrawal management or not, residential or outpatient. 

The data illustrates both the scale and the shifting complexity of the challenge. According to the 2024 National Survey on Drug Use and Health (NSDUH), published by SAMHSA in July 2025, an estimated 48.4 million people in the U.S. — roughly 16.8% of the population aged 12 or older — met criteria for a substance use disorder in 2024. Yet 80% of those who needed treatment did not receive it. Among those who do enter the treatment system, the stakes of appropriate level of care placement are high. This is especially true as SUD presentations grow more medically complex: polysubstance use, alcohol-related hepatic disease, opioid use disorder with cardiac and pulmonary complications, and stimulant use with psychiatric comorbidities are increasingly common clinical presentations that demand more from UM review, not less. 

Health plans and TPAs face a dual risk when UM programs are not calibrated to this complexity: 

• Overutilization: Patients placed at higher levels of care than clinical evidence supports, driving avoidable inpatient costs and occupying limited beds needed by patients who genuinely require that level of care. 

• Underutilization: Patients stepped down prematurely without clinical basis, increasing the likelihood of relapse, readmission, and higher total cost of care over time. 

The regulatory landscape reinforces the need for rigor. While all U.S. states have enacted statutes addressing mental health and SUD insurance parity, the strength and scope of these laws vary widely. State insurance commissioners are increasingly scrutinizing how health plans apply UM criteria to behavioral health relative to medical and surgical benefits, and the data in the Mental Health and SUD Insurance Parity Summary of State Laws demonstrates that most states have enacted or are actively writing legislation to strengthen parity protections. For health plans operating across multiple states, the compliance picture is not uniform, and that makes evidence-based, consistently applied clinical review all the more critical. 

 Evidence-Based Reviews Across the SUD Continuum 

Accurate SUD level-of-care determination requires more than criteria matching. Clinical nuance is high: withdrawal severity, substance type, prior treatment history, co-occurring medical and psychiatric conditions, and social determinants of health all inform appropriate level of care (LOC) in ways that algorithmic review cannot reliably capture. Physician-led review, conducted by specialists with direct clinical experience in addiction medicine, is essential. 

MRIoA’s clinical review capabilities span the full SUD continuum, ranging from Level 1 (Outpatient) through Level 4 (Medically Managed Inpatient), including the clinical intensity within those levels. Appropriate level of care placement is not a one-time determination; the ASAM framework calls for ongoing reassessment as a patient’s clinical status evolves, with the goal of stepping down to the least intensive setting that safely supports recovery. 

The integrated treatment dimension matters here as well. MRIoA’s reviewer network includes addiction psychiatrists alongside internists, hepatologists, cardiologists, and other medical specialists. As SUD presentations increasingly intersect with complex medical conditions, the ability to bring multi-specialty clinical input to a single review becomes a meaningful differentiator. The growing clinical evidence base for medication-assisted treatment (MAT) also means that pharmacy-related reviews, MAT appropriateness, and concurrent medical management are becoming standard components of UM for SUD. 

MRIoA’s Capabilities for SUD and Detox Review 

MRIoA brings specialized depth to SUD clinical review that general UM programs cannot replicate: 

• Addiction psychiatry expertise at the intersection of SUD and medical management, drawing from a panel of 40+ behavioral health specialists who bring direct clinical experience to complex, comorbid presentations. 

• Defensible, criteria-grounded determinations using ASAM, InterQual, MCG, or client-specified criteria, applied by physicians with hands-on addiction medicine experience — not algorithmic matching. 

• Same-day review capability that meets the clinical urgency inherent in detox presentations, where level-of-care decisions must be made quickly and accurately. 

• State-matched reviewer licensing across all 50 states, which is particularly important for Medicaid SUD programs operating under state-specific utilization review requirements. 

 Matching Clinical Review to the Complexity of SUD 

As SUD treatment evolves toward integrated, chronic disease management, health plans and TPAs need a clinical review partner whose expertise matches the complexity of the cases they are seeing. Getting level of care right in SUD is not primarily a cost management exercise. It is a clinical quality imperative that directly affects members’ recovery outcomes, long-term total cost of care, and compliance with evolving regulations. 

MRIoA’s detox and SUD level of care review capabilities are purpose-built for this moment, grounded in 40+ years of clinical review experience and backed by physician specialists who bring real addiction medicine expertise to every case. 

Schedule a consultation to learn how MRIoA can help your organization ensure appropriate, evidence-based care decisions for every member.  

The post Getting Level of Care Right in SUD: Clinical Review Strategies for Withdrawal Management and Level of Care Placement appeared first on MRIoA.

Utilization management (including prior authorization) sits squarely at the center of this tension. When done well, UM ensures patients receive clinically appropriate, evidence-based care while protecting payers and patients from avoidable and unsustainable cost exposure. When done poorly, in opaque and inconsistent ways, it becomes a barrier to care that harms patients, frustrates physicians, and exposes organizations to regulatory and reputational risk.

The stakes are high and rising. Understanding how rigorous, independent UM actually works for specialty drugs is strategically and clinically imperative.

What Makes Specialty Drugs Different

High-cost specialty drugs are biologically complex therapies, including biologics, gene therapies, immunologics, and targeted oncology agents designed to treat serious, often rare or complex conditions. Costs can range from $5,000 per year for some biologics to well over $1 million for a single gene therapy treatment.

That price range reflects genuine clinical complexity. These therapies often target small patient populations, require specialized administration, and demand ongoing monitoring to assess response and safety. The clinical criteria governing their use are sophisticated, evolving rapidly, and highly specific to each subspecialty.

This is precisely why standard UM approaches fall short. A generalist reviewer applying broad clinical criteria to a gene therapy for a rare metabolic disorder is not equipped to make a sound determination. The science moves too fast, the populations too narrow, and the clinical nuance too deep for one-size-fits-all review. This complexity carries directly into how utilization management must be structured.

The Utilization Management Framework

Effective specialty drug UM operates across three stages, each serving a distinct clinical purpose.

Prior authorization establishes clinical necessity before treatment begins. For high-cost specialty drugs, this is a meaningful clinical checkpoint that protects patients from therapies that may not match their clinical profile and protects payers from expenditures with no corresponding benefit.

Concurrent review monitors ongoing treatment appropriateness. For specialty therapies, treatment trajectories can shift. A therapy that was clinically appropriate at initiation may need reassessment as patient response, disease progression, or emerging evidence changes the picture.

Retrospective review evaluates outcomes and cost-effectiveness following treatment. For value-based contracting and outcomes-linked agreements (a growing trend in specialty drug management), retrospective data is foundational.

Step therapy and evidence-based clinical criteria round out the framework, ensuring that treatment pathways reflect current clinical evidence rather than financial defaults.

What Good Utilization Management Actually Looks Like

Evidence-based, specialty-matched clinical criteria are the foundation for good utilization management. Every review should be grounded in current clinical evidence, not outdated criteria, financial targets, or administrative convenience. For specialty drugs, that means criteria developed and applied by reviewers with genuine subspecialty expertise in the condition and therapy being reviewed.

Specialty-matched physician reviewers are critical to a quality case review. When a health plan or PBM routes an oncology review to a generalist, or a rare disease case to a reviewer unfamiliar with the therapeutic landscape, the result is neither clinically sound nor defensible on appeal. Physician reviewers must be matched to the specialty of the case.

Turnaround times matter for both compliance and for patient outcomes. Meeting or exceeding CMS standards (72 hours for urgent reviews, seven days for standard) is not just a regulatory requirement. It is a clinical one. Delays in specialty drug authorization have direct consequences for patients managing serious conditions.

Transparent and written denial documentation is essential. Meaningful appeals require meaningful explanations. When payers cannot provide clear, evidence-based rationale for an adverse determination, they undermine the entire process and expose themselves to regulatory and legal risk.

Monitoring for disparate outcomes across patient populations should be a standard operational practice. Organizations serious about health equity cannot rely on external enforcement to identify disparities. They need internal visibility and analytical infrastructure to act on what they find.

The Case for Independent Review

Each of these best practices points toward the same conclusion: independent, physician-led review is the most defensible approach to specialty drug UM because it removes the financial conflicts that make in-house payer review vulnerable to criticism.

When a reviewer has no stake in the outcome of a coverage determination, the decision rests entirely on clinical evidence. That objectivity is not just ethically important. It is operationally valuable. Independent review produces decisions that hold up under appeal, withstand regulatory scrutiny, and can be documented and defended with transparency.

Organizations like MRIoA bring a combination of clinical depth, subspecialty breadth, and independence that in-house UM programs structurally cannot replicate. With more than 700 state-matched specialists spanning 150+ specialties, and a data infrastructure built on over 10 million clinical reviews, MRIoA provides the clinical rigor and analytical foundation that specialty drug UM demands.

Looking Ahead

The specialty drug pipeline shows no signs of slowing. Gene and cell therapies, value-based contracting models, AI-assisted UM tools, and continued CMS regulatory evolution will all reshape how payers manage high-cost therapies over the next several years. Each of these trends raises new questions about clinical criteria, outcomes measurement, and equitable access.

What will not change is the standard by which utilization management will be judged: does it ensure the right patient receives the right treatment at the right time or does it introduce barriers that serve financial interests at the expense of clinical ones?

Done right, UM is not the enemy of access to high-cost specialty drugs. It is one of its most important safeguards. The goal is getting it right every time, for every patient.

To learn more about how MRIoA supports health plans, PBMs, and TPAs in building clinically rigorous, defensible utilization management programs for specialty drugs, schedule a consultation with our team.

The post Safeguarding Care by Rethinking Utilization Management for High-Cost Specialty Drugs appeared first on MRIoA.

Keese brought an insider’s perspective on the noteworthy policies and political factors reshaping the business of healthcare, and what they mean for the members they serve.

The ACA Is Being Reshaped Through Regulation

The Trump administration is using regulatory tools rather than legislation to reshape how the Affordable Care Act (ACA) delivers health benefits. Two significant provisions in the U.S. Department of Health and Human Services (HHS) 2027 Notice of Benefit and Payment Parameters signal where policy is heading.

1. Expands Access to Catastrophic Coverage: Under the current framework, individuals over 30 cannot enroll in catastrophic plans without a hardship exemption. The proposed rule would extend eligibility up to 10 years for longer-term catastrophic coverage.
2. Non-Network Plans as Qualified Health Plans: Allows plans with narrow or no network requirements (also known as “skinny plans”) to be certified as Qualified Health Plans (QHP) on the exchange. This is a significant structural departure from the baseline ACA definition.

Keese’s assessment: the administration lacks the votes to pursue this legislatively, so it’s moving through every available regulatory channel instead. Healthcare organizations should treat the 2027 rule as a signal of structural change to come, not a technicality.

CMS Innovation Models and AI in Decision-Making

This administration has also doubled down on the use of Artificial Intelligence (AI) and technology to try to reduce fraud, waste, and abuse (FWA), while also pushing CMS innovation programs to be more efficient, cost-effective, and transparent. Several new models carry direct implications for payers and healthcare organizations:

• WISeR (Wasteful and Inappropriate Service Reduction): The first-ever prior authorization program in fee-for-service Medicare that leverages AI to expedite review for services vulnerable to FWA. Keese was skeptical WISeR survives a change in administration, but its existence sets a structural precedent.
• ACCESS (Advancing Chronic Care with Effective, Scalable Solutions): Tests new payment models using wearable devices and technology to improve chronic disease management. Applications were due in 30 states by April 2026 for a January 2027 start.
• MAHA ELEVATE (Make America Healthy Again – Enhancing Lifestyle and Evaluating Value-based Approaches Through Evidence): Tests 30 new chronic disease management models focused on MAHA principles.
• GLOBE (Global Benchmark for Efficient Drug Pricing): Applicable to Medicare Part B Drugs; triggers manufacturer rebates when a drug’s price exceeds an international benchmark.
• GUARD (Guarding U.S. Medicare Against Rising Drug Costs): The same international reference pricing model as GLOBE, applied to Medicare Part D drugs.

Regarding AI in coverage decisions, Keese was direct: using algorithms to drive claim review and coverage determinations is becoming politically untenable on both sides of the aisle.

Bipartisan legislation from Senators Warren (D-VA) and Marshall (R-KS) would require a human “learned intermediary” to review any AI-generated coverage denial before it’s issued. Organizations currently using AI in Utilization Management (UM) should treat this as near-certain regulatory direction, not a distant possibility.

PBM Reform: 155 Bills in 40 States

Signed into law on February 3, 2026, the Consolidated Appropriations Act contains what Keese described as “the most sweeping Pharmacy Benefit Manager (PBM) reforms ever enacted,” including:

• De-Linking: Prohibits PBM compensation from being tied to a drug manufacturer’s list price; requires a flat fee model in Medicare and commercial markets beginning January 1, 2028.
• 100% Pass-through of Rebates: Requires PBMs to remit to clients 100% of rebates, fees, alternative discounts, and other remuneration received from manufacturers, GPOs, etc., with quarterly reporting to clients.
• Eliminates Spread Pricing: Requires CMS to define and enforce “reasonable and relevant” Medicare Part D contract terms, including reimbursement and dispensing fees, and establish an appeals process.
• Enforcement of Penalties: Grants CMS authority to impose monetary penalties, and funds CMS up to $188 million for enforcement.
• Increased Transparency: Allows CMS to track payment trends to pharmacies and pharmacy inclusion in PBM networks, including a designation of essential retail pharmacies.
• Audits: Requires audits once per plan year. The Secretary of Labor will establish reasonable confidentiality restrictions for audited Rebate contracts.

At the state level, Keese identified at least 155 bills across 40 states addressing PBM practices. Two stood out: Arkansas House Bill 1150, which sought to bar PBMs or insurers from owning pharmacies, is currently blocked by a federal judge on Commerce Clause grounds. Meanwhile, Tennessee SB 2040/ HB 1959 is a nearly identical bill moving through the legislature committees.

The driving force behind these bills is bipartisan criticism of the vertical integration of payers, PBMs, pharmacies, and drug distribution. Keese flagged the pressure for structural separation as a trend to monitor heading into the 2026 campaign cycle.

What to Watch in the Next 12 Months

Based on Keese’s analysis, these policy developments require near-term operational attention:

• PBM Contracts: Rebate pass-through, spread pricing prohibition, and compensation delinking are now federal law. Contracts and formulary structures built on the prior model need to be reviewed now.
• 2027 Exchange Plan Design: Skinny plan and catastrophic coverage changes could reshape the competitive landscape. Payers, employers, brokers, and consultants should track the final rule closely.
• AI in UM: AI-driven claim review or prior authorization workflows should align with the learned-intermediary standard now ahead of likely regulatory action.
• Midterm Election Outcomes: The potential loss of Senate health committee leadership could create uncertainty around FDA policy, public health programs, and the timing of the next major healthcare legislation.
• State PBM legislation: With 155 bills in 40 states, multistate organizations face a rapidly fragmented compliance landscape. Arkansas and Tennessee are the immediate signals but more states will likely follow.

Watch the full March 2026 Washington Insider Webinar recording for Jay Keese’s complete analysis, including Q&A on vertical integration, Senate succession, and prescription drug cash-pay models.

The post Spring 2026 Washington Insider Recap: What Healthcare Leaders Need to Know appeared first on MRIoA.

That has changed.

Today, UM outsourcing is a strategic way to extend capacity in day-to-day operations while maintaining consistency, defensibility, and operational control as UM programs become more complex and regulatory expectations continue to rise.

The following are four best practices that prevent those breakdowns and help UM outsourcing deliver on quality, compliance, and performance expectations.

Why UM Outsourcing is Increasing

The growth of UM outsourcing isn’t just about staffing shortage (though those are real). Across the healthcare ecosystem, health plans, TPAs, PBMs, and other organizations are navigating the challenges of:

• Rising costs
• Scrutiny of prior authorization and utilization management policies
• Workforce shortages
• Administrative costs and burden
• Member satisfaction expectations

Each of these challenges creates a demand for an external partner that can deliver clinical depth, operational precision, and regulatory reliability.

Meanwhile, delivering more data on decisions by clinicians with same-state, specialty-match credentials requires a deeper level of clinical depth, process rigor, and documentation. These requirements are increasingly difficult for internal programs to staff and support.

Most organizations pursue outsourcing in one of three ways:

• Full outsourcing of end-to-end UM operations
• Co-sourcing for overflow, after-hours needs, or specific lines of business
• Specialty carve-outs for high-cost services or high-complexity areas that require deep clinical expertise such as behavioral health, post-acute care, advanced imaging, or specialty pharmacy

While the models differ, success relies on the same three underlying fundamentals to ensure outsourcing achieves the desired goal.

Best Practice 1: Define Scope and Success Factors Before Contracting

Effective UM outsourcing starts with clarity before the contract is signed. That means there’s a shared understanding of what is being outsourced and what defines success.

This begins with a precise definition of which activities to contract for, including:

• Functions such as intake, clinical review, peer-to-peer support, determinations, notices, appeals support, and reporting
• Coverage across lines of business, geographic regions, products, networks, and benefit plans
• Volume assumptions are based on current demand, projected growth, and seasonal patterns like Open Enrollment.

Success metrics should be agreed upon early and tied to both speed and quality. Common measures include:

• Turnaround time (TAT) by request type
• Clinical accuracy
• Denial rationale quality
• Appeal outcomes
• Indicators tied to provider or member experience

When scope and success are defined early, performance becomes measurable and addressable. It also creates a foundation for a genuine partnership where both parties win.

Best Practice 2: Evaluate Clinical Depth, Compliance Readiness, and Scalability

Clinical credibility is vital for success because capacity alone does not hold up when decisions are questioned by providers, regulators, or auditors.

Clinical Depth and Specialty Alignment

Many UM decisions require specialty-level expertise. Strong partners can demonstrate access to board-certified clinicians across numerous specialties with same-state licensure, familiarity with the populations being served (Medicaid, Medicare Advantage, commercial, or ASO), and in-house clinical leadership when cases are complex.

Accreditations and Delegated Authority

Accreditation is not a checkbox. It is a meaningful quality signal that separates high-performing IROs from basic review vendors. Not all outsourcing partners are NCQA or URAC accredited, and many cannot meet the standards required for delegated authority.

When a partner holds delegated authority, they can issue final determinations on behalf of the health plan. That means:

• The client can offload risk and administrative burden
• Final decisions carry the clinical and regulatory credibility of an accredited organization
• The partner becomes embedded in operations, not just an overflow vendor

When evaluating outsourced partners, confirm their accreditation status and whether they qualify for delegation across determinations, quality management, and compliance oversight. This distinction dramatically affects the value and risk profile of the relationship.

Compliance Posture and Audit Readiness

In today’s environment, your partner needs to operate like they’re audit-ready every day. Look for:

• Clear and deep understanding of current federal and state regulatory UM requirements
• Transparent audit support and documentation standards
• Formal quality management program that identifies trends and drives corrective action
• Expertise to consult with your team on important compliance risks and policy/process improvements
• Accreditations from organizations like NCQA and URAC that evaluate, certify, and improve service organizations

Technology and Integration Capabilities

Scalability isn’t just about staffing, it’s also about infrastructure. Your partner should have clear integration paths (workflows, data exchange, reporting), insightful reporting that you can use, and defined surge capacity for expected and unexpected spikes. The goal is stability under pressure.

Best Practice 3: Establish Governance that Keeps Performance Visible and Supports Continuous Improvement

Governance is what prevents outsourcing from turning into a black box. It should include regular performance reviews and clear operating rhythms, including regular check-ins that surface trends in volume, TAT, quality, and compliance risk early enough to act.

Clear ownership and escalation paths prevent issues from lingering. There should be escalation triggers for:

• Missed TAT
• Repeated documentation gaps
• High-risk or high-profile cases
• Provider friction or dissatisfaction
• Signs of guideline ambiguity or inconsistent application

The strongest partnerships treat QA as joint work. Find partners who schedule joint audits, regular calibration, and shared root-cause analysis to help distinguish between training gaps, workflow issues, and guideline ambiguity. When a partner is committed to making corrective actions tied to those findings, they demonstrate a focus on ongoing improvement. Additionally, look for an organization that has achieved NCQA and URAC accreditation.

Best Practice 4: Assess Operational Continuity and After-Hours Coverage Capabilities

One of the most overlooked dimensions of outsourcing evaluation is coverage continuity. Clinical volume doesn’t stop at 5 PM, nor do compliance requirements in some states. When a health plan misses turnaround time requirements over a weekend or holiday, the downstream consequences — member delays, provider friction, compliance exposure — are just as significant as a weekday failure.

When evaluating a UM outsourcing partner, assess their capacity for:

• After-hours, weekend, and holiday clinical review coverage
• Surge capacity during seasonal spikes due to open enrollment, plan design changes, etc.
• Staffing stability that doesn’t create backlogs when internal client teams are short

Member-friendly communication is equally important. Notices and determination letters that members can understand reduce unnecessary appeals, decrease call center volume, and directly support CAHPS and STAR rating performance. Evaluate whether your partner’s communication standards are built for clarity or simply legal defensibility.

Partners with strong operational continuity also tend to have the staffing infrastructure to absorb market uncertainty, like the enrollment volatility driven by Medicaid redeterminations or ACA market shifts, without passing that instability back to the plan.

Making Outsourcing Work

UM outsourcing delivers the most value when it operates as a clinical and operational partnership, not a short-term staffing solution. Choose partners who can combine:

• Clear scope and measurable success metrics
• Clinical depth, accreditations, and delegated authority
• Compliance readiness and scalable infrastructure
• Governance that enforces accountability and continuous improvement
• Operational continuity to include after-hours coverage and surge capacity

With a partner who demonstrates these capabilities, organizations can expand UM capacity without sacrificing quality, turnaround time, or provider and member experience.

To learn about our outsourced UM services, contact us to start the conversation.

The post 4 Best Practices to Get Better Outcomes and Better Member Experiences appeared first on MRIoA.

Legislative and regulatory shifts are moving quickly from Capitol Hill and state capitols into day-to-day operations. For health plans, third-party administrators (TPAs), and pharmacy benefit managers (PBMs), changes that once allowed long-range planning now demand near-term decisions across eligibility, utilization management (UM), and clinical staffing. Compliance depends as much on execution as on policy.

How H.R.1 Reframes Medicaid & Operations

The One Big Beautiful Bill Act (H.R.1) resets Medicaid. After years of leveraging Medicaid to expand access to coverage and care, the program is repositioned as a narrower safety net with greater emphasis on program integrity.

For health plans and PBMs, the operational impact is immediate. Tighter verification and documentation increases coverage churn and gap periods. Clinical reviews may hinge on coverage timing as well as clinical appropriateness, driving more continuity-of-care determinations, retroactive eligibility reviews, and appeals tied to administrative documentation.

While H.R.1 is federal policy, implementation is state-driven. Timelines, requirements, and enforcement will vary by state, increasing operational complexity for organizations in multiple states. Eligibility can no longer be treated as a front-end administrative step; it must be integrated into UM and appeals.

Operational priorities include:

• Embed eligibility verification into UM and appeal workflows
• Refresh medical necessity criteria to reflect state-by-state changes
• Prepare for increased appeal volume tied to eligibility and documentation disputes

State-Level Regulation Is Driving Day-to-Day Change

Alongside federal shifts, states continue to regulate the mechanics of UM, especially prior authorization (PA). Reforms increasingly targeted the use of artificial intelligence (AI), turnaround times, standardized definitions, reporting and transparency, and clinical oversight.

States are also beginning to regulate the use of AI in utilization management. Plans using algorithmic tools must be able to explain decision logic, show clinician involvement, and document consistent application criteria.

Common state trends include: 

• Streamlined prior authorization models and service exemptions
• Shorter turnaround time requirements 
• Expanded expectations for specialty-matched clinical review
• Increased reporting and transparency
• Electronic prior authorization with guardrails for AI use

These reforms reduce administrative burden if intake, routing, review logic, and documentation are redesigned to meet the new standards.

Impacts on Health Plan Clinical Operations

PA reform compresses timelines while raising documentation expectations. Added scrutiny of AI-supported reviews introduces requirements for audit trails, clinical accountability, and defensible application of criteria. All of this occurs amid ongoing staffing constraints.

Clinical operations cannot respond by simply adding headcount. The sustainable approach is to standardize routine reviews while directing complex cases and appeals to experienced clinicians. Operational workflows are as important as clinical accuracy.

Impacts on Pharmacy Operations

Pharmacy operations face a distinct set of pressures, driven by drug pricing policy and coverage expansion.

Beginning in 2026, changes affecting GLP-1 pricing, Medicare negotiation, and reimbursement models are expected to reshape cost and utilization patterns. PA criteria, step therapy requirements, and exception processes will need to evolve accordingly.

At the same time, coverage signals for select weight-loss indications are expanding. Broader access paired with tighter cost controls increases operational strain. When criteria are unclear, inconsistently applied, or poorly communicated — appeal volume rises quickly, followed by provider confusion and member dissatisfaction.

State-level PA reform increasingly applies to pharmacy as well. Turnaround time standards, reporting requirements, and transparency expectations increasingly mirror those on the medical side. Pharmacy workflows must meet these standards consistently with clear documentation, appropriate clinical oversight, and auditable decision paths. Data is essential as an operating tool to identify where criteria are working, where appeals overturn decisions, and where operational friction occurs.

Adapting Review Operations Without Adding Headcount

Regulatory change is colliding with workforce constraints. Budget pressure, clinical staffing shortages, and rising utilization make it unrealistic to “add more reviewers” for every new requirement.

The practical response is deliberate allocation of clinical effort. Routine, low-variability cases are increasingly handled through standardized, tech-enabled workflows. Complex cases, exceptions, and appeals are directed to experienced clinicians who can apply judgment where it matters most. Clear escalation paths and consistent documentation reduce rework while supporting compliance.

Prioritizing Compliance for 2026, 2027, and Beyond

As regulatory volume increases, prioritization is critical. Near-term focus should center on:

• Turnaround time performance where state standards drive complaint or audit risk
• Member notices and appeals so timelines and documentation align with updated requirements
• Eligibility-related reviews where mid-care coverage changes create avoidable appeals and continuity-of-care risk

Medicare Advantage may see regulatory stability in 2027, creating a narrow window in 2026 to harden eligibility processes, tighten UM workflows, and reduce operational friction before the next wave of reform. Beyond 2027, pressure is unlikely to ease as Medicaid enrollment is projected to decline while utilization remains elevated. At the same time, CMS has signaled further reforms tied to risk adjustment, quality measurement, and technology-enabled oversight.

From Policy Awareness to Operational Readiness

Policy change is inevitable, but operational readiness is a choice. The work now is building clinical and pharmacy operations that adapt quickly, remain compliant, and support sound clinical judgment without disrupting care, consistency, or clinical decision-making.

The post From Policy to Practice: What Today’s Regulatory Shifts Mean for Clinical and Pharmacy Operations appeared first on MRIoA.

If your organization isn’t prepared, the risks are clear: fines, compliance penalties, reputational harm, and provider dissatisfaction. 

At MRIoA, our mission is to give health plans clarity and confidence in navigating these complex changes. Here’s a closer look at four states where new regulations are going into effect and what they mean for your business. 

Georgia – H.B. 197 (Effective Jan 1, 2026) 

What’s changing: Georgia requires insurers to make a good-faith effort for a pre-denial peer-to-peer review. That effort must include either a callback system, a telecommunications platform, or a public website to schedule callbacks. In addition, insurers must implement and maintain a program to reduce prior authorization requirements, with the first filing due July 1, 2026.  

• Georgia General Assembly – HB 197: https://www.legis.ga.gov/legislation/64277 

Why it matters: These requirements add new operational and reporting burdens. Failure to comply risks fines, member delays, and strained provider relationships. 

Takeaway: Don’t wait until the deadline. Begin building callback workflows now and assess how your PA program can be reduced without compromising compliance. 

Illinois – H.B. 4055 (Effective Jan 1, 2026) 

What’s changing: Illinois will restrict plans from requiring prior authorization for FDA-approved therapies for hereditary bleeding disorders more frequently than every six months, or for the length of the prescription, whichever is shorter.  

• Illinois General Assembly – HB4055: https://www.ilga.gov/legislation/billstatus.asp?DocNum=4055&GAID=19&GA=103&DocTypeID=HB&LegID=152792&SessionID=112 

Why it matters: This is not a full PA ban, but it requires significant adjustments to authorization logic and systems. Failure to comply risks penalties, reputational harm, and potential patient impact. 

Takeaway: Update your PA rules and test your systems now to avoid costly mistakes. 

Texas – H.B. 3812 & S.B. 815 (Effective Sept 1, 2025) 

What’s changing: H.B. 3812 requires that certain determinations be made by a Texas-licensed physician, who may not hold an administrative medicine license. 

• Texas Legislature – HB 3812: https://capitol.texas.gov/BillLookup/History.aspx?LegSess=88R&Bill=HB3812 

S.B. 815 prohibits utilization review agents from using automated decision systems to make adverse determinations. The law also grants the commissioner authority to audit UR agents’ practices.  

• Texas Legislature – SB 815: https://capitol.texas.gov/BillLookup/History.aspx?LegSess=88R&Bill=SB815 

Why it matters: Health plans must align physician oversight models with licensing requirements and ensure compliance processes can withstand audits. 

Takeaway: Evaluate your current utilization review oversight and ensure your staffing and governance models meet Texas’ new requirements. 

The Bottom Line 

These are not minor updates. They are business-impacting regulatory changes with real financial, operational, and reputational stakes. 

MRIoA provides the clinical expertise, regulatory guidance, and actionable insights that health plans, TPAs, and PBMs need to adapt quickly and with confidence. 

Join us for the Regulatory & Compliance Roundtable on October 2, 2025 to hear from experts, understand what these laws mean for your organization, and walk away with strategies to reduce risk and ensure compliance. 

View the Roundtable Recording

The post Is your Organization Ready for Fall 2025 State Regulatory Changes?  appeared first on MRIoA.

According to the American Medical Association, ten states passed significant prior authorization reforms in 2024, tightening turnaround times, mandating greater transparency, and easing administrative burdens. At the same time, nine states and the District of Columbia passed reforms in the prior year, and nearly 30 states are considering similar legislation now.  

This wave of legislation is fundamentally changing how health plans, TPAs, and PBMs will be evaluated and held accountable. Non-compliance is no longer measured only in fines, but in lost revenue, reputational damage, and competitive disadvantage.  

Why These Changes Matter to the C-Suite 

Revenue at Risk 
Fines, lawsuits, and contract losses can dwarf the cost of proactive compliance. Illinois’ new restrictions on prior authorization for hereditary bleeding disorder drugs require precise operational updates. A missed adjustment is not just a workflow error. It is a potential revenue and reputation crisis. 

Operational Disruption 
Georgia’s H.B. 197 mandates callback systems and PA-reduction programs. That means new infrastructure, new processes, and new reporting obligations. For executives, this translates into resourcing decisions and system investments that must be made quickly. 

Regulatory Scrutiny 
Texas’ new laws (S.B. 815 and H.B. 3812) create strict requirements for utilization review oversight and give regulators the authority to audit payer processes at any time. That puts executive accountability squarely in the spotlight. 

Competitive Differentiation 
New Jersey’s A.B. 1825 establishes step therapy exception requirements that will create winners and losers. Those who adapt quickly will maintain provider trust and member satisfaction. Those who lag risk being viewed as slow, opaque, and non-compliant. That is a costly brand position in a competitive market. 

What Executives Should Be Asking 

• Where is our greatest compliance risk exposure: financial, operational, or reputational? 
• Do we have the infrastructure to meet callback, exception, and reporting mandates across multiple states? 
• How are we monitoring and auditing utilization review processes to ensure readiness for regulator scrutiny? 
• Are we positioned to turn compliance into an advantage, not just an obligation? 

MRIoA’s Role: From Compliance Burden to Strategic Readiness 

For leadership, the question is not just “are we compliant?” It is “how do we make compliance an advantage?” 

MRIoA helps executives answer that question by delivering: 

• Regulatory Foresight – We track and interpret state-by-state changes before they become pain points, giving leadership the ability to plan strategically instead of reactively. 
• Operational Assurance – With the largest state-matched reviewer network in the nation, we ensure compliance requirements do not stall operations, disrupt provider relations, or inflate costs. 
• Business Intelligence – Our analytics surface where regulatory risk lives in your workflows, allowing you to redirect resources, reduce waste, and protect revenue. 

MRIoA does not just reduce the burden of compliance. We give executives the clarity and confidence to make faster, better, and more defensible business decisions. 

The Strategic Imperative 

The regulatory landscape is not slowing down. With dozens of states advancing new prior authorization reforms in 2024 and 2025, compliance is now a C-suite responsibility. It is no longer about avoiding penalties. It is about protecting revenue, sustaining trust, and creating competitive advantage. 

Watch MRIoA’s Regulatory & Compliance Roundtable from October 2, 2025. Gain insights into what these state laws mean for your organization and learn how to turn regulatory change into strategic readiness. 

Vide the Roundtable Recording

The post Compliance Is a Business Risk You Can’t Ignore  appeared first on MRIoA.

Heather Grigo, MD

May 2025

The U.S. is currently experiencing a mental health crisis with rising levels of unmet behavioral health needs across all ages. It is estimated that more than one in five U.S. adults (23.1%) live with a mental illness¹. Additionally, 17.3% of people aged 12 and older had a substance use disorder in the past year². Depression is on the rise, with recent data indicating that its prevalence among U.S. adolescents and adults has increased by 60% over the past decade³. Among adolescents aged 12 to 17, 13.4% had serious thoughts of suicide, 6.5% made a suicidal plan, and 3.7% attempted suicide in the past year⁴. Spending has also trended upward: from 2019 to 2022 utilization and spending rates for mental health care services among commercially insured adults increased by 38.8% and 53.7%, respectively⁵. The ability to effectively address this crisis may partly depend on the development and deployment of innovative technologies capable of meeting these complex and widespread challenges. 

Artificial intelligence (AI) has the potential to transform mental health care by providing insights and solutions that were previously unattainable through conventional methods. The integration of AI into mental health care dates back to the mid-20th century, when scientists envisioned using robots to imitate cognitive processes. This was followed by the development in the 1960s of a rule-based chatbot (ELIZA) that simulated a Rogerian psychotherapist, the creation of rule-based AI systems designed to mimic human expertise in the 1980s, and then in the late 20th century with computerized cognitive-behavioral therapy programs⁶. The evolution of AI’s role in mental health care has increased exponentially since the early 21st century.  We have seen advances in the early identification of mental health conditions and development of treatment recommendations, as well as the use of specific tools to conduct virtual therapy⁶. By integrating artificial intelligence into mental health care, there is an opportunity to address some of the most pressing challenges in the field. Whether improving access to care with AI-driven therapy chatbots, reducing clinicians’ administrative burden by incorporating ambient AI listening tools into their practice, or enhancing the ability to diagnose patients using AI-enabled Clinical Decision Support Systems, the application of AI has the potential to profoundly impact patients, clinicians, and the mental health care system as a whole. 

Improving Access to Care with AI-Driven Chatbots 

The significant increase in demand for mental health services is placing a strain on already taxed resources. In addition, there is a nationwide shortage of providers. According to the National Center for Health Workforce Analysis, as of December 2023, over half of the U.S. population lives in a Mental Health Professional Shortage Area⁷. There is also a critical workforce shortage of child and adolescent psychiatrists. Data from the American Academy of Child and Adolescent Psychiatry⁸ shows that on average, there are only 14 child and adolescent psychiatrists for every 100,000 children in the U.S., with most states in severe shortage. The resulting extended wait times for appointments or long-distance travel for medically necessary care delays care and increases the burden on parents. Prompt access to mental health care is critical, particularly when demand is already high. Broadening the availability of mental health care to those without access – or who prefer alternative options due to perceived stigma or other reasons – could reduce symptom severity and improve functioning, thereby mitigating the risk of crises and hospitalizations. 

One way to increase access to services is through the use of AI-driven therapy chatbots. These interventions use natural language processing (NLP) to interpret the meaning and intent behind the user’s input. By classifying real-time, multimodal input according to the user’s emotional state, they can simulate human conversation and respond empathically, offering opportunities for therapeutic dialogue⁹. Due to their convenience, they also enable the delivery of personalized mental health interventions at scale. This scalability is crucial in addressing the current shortage of mental health professionals. 

Cognitive-behavioral therapy (CBT)-based digital therapeutics and rule-based AI chatbots have already shown effectiveness in treating depression and anxiety¹⁰. One study found a chatbot was able to establish a working alliance comparable to traditional, human-delivered services across different treatment modalities¹¹. AI-driven chatbots have also been developed to provide therapy for children with autism spectrum disorder¹². These tools can teach emotional recognition and social skills by analyzing a child’s facial expressions and adjusting their interactions accordingly. 

Recently, the first randomized controlled trial demonstrating the effectiveness of a fully generative-AI-powered therapy chatbot for treating clinical-level mental health symptoms showed that users experienced greater reductions in depression, anxiety and clinically high-risk feeding and eating disorder (CHR-FED) symptoms compared to the waitlist control group¹³. In addition, during the four-week trial, users demonstrated sustained engagement and reported that their alliance with the chatbot was comparable to that of human therapists. Notably, patients also appeared to have an overall positive perception of using chatbots for mental health conditions¹⁴. Since perception can influence the adoption of a technology, high levels of perceived usefulness highlights the ongoing potential impact of chatbots in mental health care.  

Reducing Administrative Burden with Ambient AI Listening Tools 

The complexity and time-consuming nature of administrative tasks produces a significant burden on providers, many of whom already have schedules at capacity. According to the American Psychiatric Association, most psychiatrists spend just 60% of their time with patients. As providers struggle to balance time spent seeing patients with administrative responsibilities-such as documenting patient care and completing prior authorization requests, this can lead to decreased job satisfaction and even burnout. A recent survey of practicing physicians by the American Medical Association found that physicians and their staff spend an average of 13 hours each week completing prior authorizations, and 89% of physicians surveyed reported that these requirements somewhat or significantly increase physician burnout¹⁵. At a time of already limited provider availability, physician burnout could further exacerbate existing workforce shortages. 

Automatic note generation using HIPAA-compliant ambient AI listening tools has the ability to record clinical interactions and is increasingly integrated directly into providers’ electronic health record systems. Adopting these tools enables mental health clinicians to improve efficiency by reducing documentation time, allowing them to focus on what matters most: personalized care. In addition, by saving time on administrative activities, providers may be able to see more patients, thereby improving access to mental health services through increased clinical productivity. According to a qualitative study evaluating physician perspectives on an ambient AI scribe pilot, those interviewed had predominantly positive views on the impact of ambient AI scribes on temporal demand, work-life integration, patient engagement, and overall workload. However, there were predominantly negative perspectives on note construction, specifically accuracy and style¹⁶. 

There are also applications for AI in automating other administrative tasks such as billing, scheduling, and basic patient communication¹⁷. When used responsibly and ethically, AI-enabled tools can optimize the prior authorization process, allowing providers to spend less time on insurance-related tasks. By streamlining many of the administrative burdens that weigh down the healthcare system, AI integration has the potential to reduce burnout in the existing mental health workforce—thereby addressing one of the major challenges currently facing the provision of mental health care in the United States.   

Enhancing Clinical Decision-Making Using AI-Enabled Systems 

The ability of artificial intelligence to aid in detecting specific mental health conditions has the potential to improve diagnostic accuracy. Certain AI-enabled Clinical Decision Support Systems (AI-CDSS) can assist mental health practitioners in diagnosing or predicting the onset of conditions¹⁸. The Cognoa ASD Diagnosis Aid and EarliPoint System are FDA Class II devices designed for use by health care providers to assist in the diagnosis of autism spectrum disorder (ASD) in children. The Cognoa ASD Diagnosis Aid uses a machine learning algorithm to predict autism spectrum disorder based on clinical data and videos uploaded by caregivers through a mobile application. AI is used to analyze the combined data and produce a value that, when compared to pre-defined thresholds, determines if a diagnosis of ASD is present. The EarliPoint System assists healthcare providers in diagnosing ASD in children between 16 and 30 months of age who are at risk for developmental delays by tracking the child’s visual reactions to social information presented in videos. 

One AI-CDSS device, Limbic Access, is currently integrated within the National Health Service in the United Kingdom. It uses conversational data collected through a mobile application. A thorough symptom profile is created for each user, and the generated insights regarding the most likely diagnoses is presented to the treating clinician as a digital report to aid in diagnostic and treatment decision-making¹⁹. Researchers have also developed a combination of machine learning and deep learning algorithms to detect schizophrenia, depression and anxiety, bipolar disorder, posttraumatic stress disorder, anorexia nervosa, and attention-deficit/hyperactivity disorder²⁰. Although results can vary in terms of predictive accuracy and precision, all studies that implemented deep learning methods reported an accuracy of at least 63.62%. By improving diagnostic accuracy within mental health care, clinicians may be able to make more informed and timely decisions, which could ultimately improve patient outcomes by mitigating the progression of mental health conditions. 

In terms of therapeutic outcomes, several studies have evaluated the ability of machine learning models to develop personalized care recommendations based on predicted clinical response to treatment, with promising results across a range of psychiatric conditions^21,22. A recent meta-analysis of existing literature evaluated 14 studies investigating the impact of machine learning models, deep learning models, and hybrid models combining multiple AI approaches on both diagnostic accuracy and therapeutic efficacy²³. The results showed that AI models have a robust impact on both diagnostic accuracy and therapeutic outcomes (with pooled effect sizes of 0.85 for diagnostic accuracy and 0.84 for therapeutic efficacy). Among the AI methodologies reviewed, machine learning models demonstrated the highest diagnostic and therapeutic performance. 

Ethical Implications and Challenges 

Despite many promising applications, the integration of AI in mental health care presents clear challenges. There are ethical considerations surrounding data privacy as more clinicians incorporate AI tools into clinical care. It is an imperative to protect the privacy, security and consent of patients and health care professionals, particularly given the rapid advancement and widespread adoption of AI, especially as regulations struggle to keep pace.  

Some studies have also identified risks associated with conversational agent interventions (CAIs), including chatbots. Misunderstandings on the part of the CAI could result in ineffective or even harmful interventions, and there may be no built-in crisis warning system to respond in the event of an emergency²⁴. Additionally, there is a growing need for transparency in the development of AI technologies. Machine learning systems in health care may be subject to algorithmic bias²⁵. A balanced approach is necessary to ensure the safety, quality, and fairness of AI systems. This includes the refinement of AI models for broader and more diverse populations. The impact of AI on healthcare costs, access, and outcomes also warrants further research and evaluation to guide future development. Finally, in order to ensure responsible use, AI should serve as a method to augment and enhance, not replace, sound clinical decision-making and judgement. 

The American Psychiatric Association recognizes the potentially revolutionary role of AI in automating elements of medicine to improve both clinician and patient experience and generate better outcomes. However, it urges caution in the application of untested technologies within clinical medicine²⁶. Multiple efforts are underway to pass international laws and create guidelines for the responsible use of AI in healthcare, including initiatives by the World Health Organization²⁷. With the engagement of key stakeholders, the National Academy of Medicine is currently working to develop an AI Code of Conduct framework. Their goal is the intentional design of the future of AI-enabled health, health care and biomedical science that advances the vision of health and well-being for all²⁸.  

Conclusion and Future Outlook 

The mental health crisis in the United States continues to intensify with growing prevalence of mental health conditions, increasing demand for care, and a shortage of clinical providers. Artificial intelligence offers a powerful set of tools to reimagine how care is delivered, accessed, and optimized. We are already seeing the promise of AI-enabled technologies for increasing therapy access, improving clinical efficiency, and enhancing diagnostic and treatment capabilities.  

While the benefits are compelling, this transformative potential should be balanced with a clear understanding of the ethical and practical risks. Healthcare leaders need to consider issues such as data privacy, algorithmic bias, system transparency, and the absence of crisis safeguards. Above all, AI must remain a complement to, not a replacement for, human clinical judgment  in all areas that involve medical decision-making.  

To fully realize AI’s potential in mental health care, ongoing interdisciplinary collaboration is essential. This includes not only clinicians, patients, and AI developers, but also policymakers, ethicists, professional societies, and the broader public, who can advocate for and shape robust regulatory frameworks and standards. Continued research and impact assessments will also play a crucial role in refining these technologies for real-world use. 

Artificial intelligence represents a promising frontier in the evolution of mental health care. If deployed strategically and ethically, it may help bridge longstanding gaps in access to care, improve clinical outcomes, and strengthen the resilience of the mental health system for the future. 

References: 

1. National Institutes of Mental Health: 2022 National Survey on Drug Use and Health, https://www.nimh.nih.gov/health/statistics/mental-illness
2. National Institutes of Mental Health: 2022 National Survey on Drug Use and Health, https://www.nimh.nih.gov/health/statistics/mental-illness
3. National Center for Health Statistics: Depression Prevalence in Adolescents and Adults: United States, August 2021-August 2023, https://www.cdc.gov/nchs/products/databriefs/db527.html 
4. Cantor JH, McBain RK, Ho P, Bravata DM, Whaley C. Telehealth and In-Person Mental Health Service Utilization and Spending, 2019 to 2022. JAMA Health Forum. 2023;4(8):e232645, doi:10.1001/jamahealthforum.2023.2645
5. National Institutes of Mental Health: 2022 National Survey on Drug Use and Health, https://www.nimh.nih.gov/health/statistics/mental-illness
6. Cantor JH, McBain RK, Ho P, Bravata DM, Whaley C. Telehealth and In-Person Mental Health Service Utilization and Spending, 2019 to 2022. JAMA Health Forum. 2023;4(8):e232645, doi:10.1001/jamahealthforum.2023.2645
7. Olawade DB, Wada OZ, Odetayo A, David-Olawade AC, Asaolu F, Eberhardt J. Enhancing mental health with Artificial Intelligence: Current trends and future prospects. Journal of Medicine, Surgery, and Public Health. 2024; Volume 3, Article 100099, https://doi.org/10.1016/j.glmedi.2024.100099
8. National Center for Health Workface Analysis: HRSA Behavioral Health Workforce, 2023, https://bhw.hrsa.gov/sites/default/files/bureau-health-workforce/Behavioral-Health-Workforce-Brief-2023.pdf
9. American Academy of Child & Adolescent Psychiatry: Severe Shortage of Child and Adolescent Psychiatrists Illustrated in AACAP Workforce Maps, 2022, https://www.aacap.org/AACAP/zLatest_News/Severe_Shortage_Child_Adolescent_Psychiatrists_Illustrated_AACAP_Workforce_Maps.aspx
10. He Y, Yang L, Qian C, Li T, Su Z, Zhang Q, Hou X. Conversational Agent Interventions for Mental Health Problems: Systematic Review and Meta-analysis of Randomized Controlled Trials. J Med Internet Res. 2023;25:e43862, doi: 10.2196/43862
11. Fitzpatrick KK, Darcy A, Vierhile Delivering Cognitive Behavior Therapy to Young Adults With Symptoms of Depression and Anxiety Using a Fully Automated Conversational Agent (Woebot): A Randomized Controlled Trial. JMIR Ment Health. 2017;4(2):e19, doi: 10.2196/mental.7785
12. Darcy A, Daniels J, Salinger D, Wicks P, Robinson A. Evidence of Human-Level Bonds Established With a Digital Conversational Agent: Cross-sectional, Retrospective Observational Study. JMIR Form Res. 2021 May 11;5(5):e doi: 10.2196/27868.
13. Olawade DB, Wada OZ, Odetayo A, David-Olawade AC, Asaolu F, Eberhardt J. Enhancing mental health with Artificial Intelligence: Current trends and future prospects. Journal of Medicine, Surgery, and Public Health. 2024; Volume 3, Article 100099, https://doi.org/10.1016/j.glmedi.2024.100099
14. Heinz MV, Mackin DM, Trudeau BM, Bhattacharya S, Wang Y, Banta HA, Jewett AD, Salzhauer AJ, Griffin TZ , Jacobson NC. Randomized Trial of a Generative AI Chatbot for Mental Health Treatment. NEJM AI 2025;2(4), doi: 1056/AIoa2400802
15. Abd-Alrazaq AA, Alajlani M, Ali N, Denecke K, Bewick BM, Househ Perceptions and Opinions of Patients About Mental Health Chatbots: Scoping Review. J Med Internet Res 2021;23(1):e17828, doi: 10.2196/17828
16. American Medical Association: 2024 AMA Prior Authorization Survey, htps://www.ama-assn.org/system/files/prior-authorization-survey.pdf
17. Shah SJ, Crowell T, Jeong Y, et al. Physician Perspectives on Ambient AI Scribes. JAMA Netw 2025;8(3):e251904. doi:10.1001/jamanetworkopen.2025.1904
18. American Psychiatric Association: The Basics of Augmented Intelligence: Some Factors Psychiatrists Need to Know Now (2023), https://www.psychiatry.org/News-room/APA-Blogs/The-Basics-of-Augmented-Intelligence
19. Kleine A-K, Kokje E, Hummelsberger P, Lermer E, Schaffernak I, Gaube AI-enabled clinical decision support tools for mental healthcare: A product review. Artificial Intelligence in Medicine. 2025; Volume 160, Article103052. https://doi.org/10.1016/j.artmed.2024.103052.
20. Kleine A-K, Kokje E, Hummelsberger P, Lermer E, Schaffernak I, Gaube AI-enabled clinical decision support tools for mental healthcare: A product review. Artificial Intelligence in Medicine. 2025; Volume 160, Article103052. https://doi.org/10.1016/j.artmed.2024.103052.
21. Iyortsuun NK, Kim S-H, Jhon M, Yang H-J, Pant S. A Review of Machine Learning and Deep Learning Approaches on Mental Health Diagnosis. Healthcare. 2023; 11(3):285. https://doi.org/10.3390/healthcare11030285
22. Meinke C, Lueken U, Walter H, Hilbert K. Predicting treatment outcome based on resting-state functional connectivity in internalizing mental disorders: A systematic review and meta-analysis. Neuroscience & Biobehavioral Reviews. 2024; Volume 160, Article 105640, https://doi.org/10.1016/j.neubiorev.2024.105640.
23. Vieira A, Liang X, Guiomar R, Mechelli Can we predict who will benefit from cognitive-behavioural therapy? A systematic review and meta-analysis of machine learning studies. Clinical Psychology Review. 2022; Volume 97, Article 102193, https://doi.org/10.1016/j.cpr.2022.102193. 
24. Rony MKK, Das DC, Khatun MostT, et al. Artificial intelligence in psychiatry: A systematic review and meta-analysis of diagnostic and therapeutic efficacy. DIGITAL HEALTH. 2025;11. doi:10.1177/20552076251330528
25. He Y, Yang L, Qian C, Li T, Su Z, Zhang Q, Hou X. Conversational Agent Interventions for Mental Health Problems: Systematic Review and Meta-analysis of Randomized Controlled Trials. J Med Internet Res. 2023;25:e43862, doi: 10.2196/43862
26. Davenport T, Kalakota The potential for artificial intelligence in healthcare. Future Healthc J. 2019 Jun;6(2):94-98. doi: 10.7861/futurehosp.6-2-94
27. American Psychiatric Association: The Basics of Augmented Intelligence: Some Factors Psychiatrists Need to Know Now (2023), https://www.psychiatry.org/News-room/APA-Blogs/The-Basics-of-Augmented-Intelligence
28. Goldberg CB, Adams L, Blumenthal D, Brennan PF, Brown N, Butte AJ, Cheatham M, deBronkart D, Dixon J, Drazen J, Evans BJ, Hoffman SM, Holmes C, Lee P, Manrai AK, Omenn GS, Perlin JB, Ramoni R, Sapiro G, Sarkar R, Sood H, Vayena E, Kohane IS, for the RAISE Consortium. To Do No Harm — and the Most Good — with AI in Health Care. NEJM AI. 2024;1(3) doi: 10.1056/AIp2400036
29. Adams L, Fontaine E, Lin S, Crowell T, Chung VCH, Gonzalez AA. Artificial Intelligence in Health, Health Care, and Biomedical Science: An AI Code of Conduct Principles and Commitments Discussion Draft. NAM Perspect. 2024 Apr 8;2024:10.31478/202403a. doi: 10.31478/202403a.

The post Integrating Artificial Intelligence into Mental Health Care: Promise, Progress, and Ethical Precautions appeared first on MRIoA.