How Data Transforms Behavioral Health Outcomes

Is the future of US mental healthcare at risk? As an author of a book on data and AI's role in mental health, I believe this moment, while alarming, also demands a strategic pivot. When federal funding is under threat, it becomes even more imperative to optimize every dollar spent and ensure interventions are effective. This is where the intelligent application of data and AI becomes critical. We must evolve from a reactive "more services" model to a data-driven, precision mental health strategy. This is especially true within the workplace, where individuals spend a significant portion of their lives and where mental health directly impacts productivity and retention. Here's how: 1. Precision Targeting/Needs Assessment: Instead of broad programs, AI and advanced analytics can pinpoint specific workforce segments experiencing elevated mental health risks or particular types of distress. My experience in F100 companies has consistently shown that understanding the specific 'why' behind workplace mental health challenges through data is crucial for effective program design. 2. Evaluating Program Effectiveness: Many mental health programs, while well-intentioned, often lack rigorous, data-driven evaluation. Leveraging AI and analytics allows us to measure the actual impact of various interventions (including manager training) on employee well-being and productivity. 3. Proactive Workforce Support: Data from HR and other systems, aggregated and anonymized, can identify early indicators of burnout or stress. This empowers companies to implement proactive interventions before employees reach a critical state, fostering a culture of preventative well-being and resilience rather than merely reacting to problems. These proposed funding changes are undoubtedly significant. However, they also serve as a stark reminder that innovation is paramount. By strategically embracing data and AI, we can move beyond simply reacting to budget constraints and instead build more resilient, efficient, and equitable mental health support systems, ensuring our workforce remains healthy, productive, and equipped to navigate the future. Ps. Check out "The Inclusion Equation: Leveraging Data & AI For Organizational Diversity and Well-being" for more details: https://lnkd.in/eXJyjMhT Data With Serena™️

CMS just signaled where Medicare Advantage quality measurement is headed—and it's squarely focused on behavioral health outcomes. The proposed 2027 MA rule, released last week, adds a Depression Screening and Follow-Up measure to Star Ratings. This isn't just another screening metric. It's an outcomes measure that tracks whether plans actually deliver effective treatment after a positive screen. The timeline: Data collection starts in 2027. Star Ratings impact hits in 2029. Why this matters: For MA plans: Depression screening and follow-up will directly affect Quality Bonus Payments. Plans that can demonstrate closed-loop care coordination between screening and treatment will have a significant competitive advantage. For primary care: PCPs are already screening. The gap is in the follow-up infrastructure. Long referral wait times, poor care coordination between behavioral health specialists and PCPs, and lack of systematic tracking all work against performance on this measure. For behavioral health integration: This measure essentially validates the integrated behavioral health model—embedded behavioral health in primary care settings so that patients are screened and identified early and ensure they get the follow-up with specialists. CMS is saying: screening without infrastructure for follow-up isn't enough. Context matters here: CMS is simultaneously removing 12 administrative process measures (appeals, customer service, etc.) to refocus Star Ratings on clinical outcomes and patient experience. The message is clear—less focus on operational metrics, more focus on whether plans are actually improving health outcomes. https://lnkd.in/g6gJz3eF

We are swimming in data. Wearables stream physiologic signals in real time. Accelerometers track movement and position. EMRs generate metadata from every click. Patients self-report symptoms, behaviors, and context. Add in social and environmental factors, and we’re practically drowning in information. And all of it is tied back to a unique individual in a way that would have been unimaginable twenty years ago. The appetite for data is insatiable and technology has made us extraordinarily skilled at collecting it, eagerly anticipating a world where we would know exactly what it all means. In theory, these streams should let us see a patient’s trajectory long before they walk into a clinic or emergency department. But that reality doesn't exist yet. We haven’t yet learned how to decipher the data's meaning or how to distinguish meaningful signal from the noise that surrounds it. The real opportunity is translating what we collect into timely, meaningful insights for clinicians, shortening the distance between signal and action to improve outcomes. Remote physiologic monitoring (RPM) is one of the clearest examples of this opportunity. What started as a collection of devices and data is evolving into an operational layer that can influence how teams anticipate, triage, and intervene across care settings. It’s becoming less about the technology itself and more about the systems that surround it: how information moves, how decisions are made, and how early signals are converted into timely action. We’ve known for years what this can look like when done well. In 2018, the TIM-HF2 randomized controlled trial embedded physiologic and behavioral data into a structured clinical response pathway. When signals changed, clinicians intervened, and as a result, patients spent fewer days lost to unplanned hospitalizations or even death than those receiving usual care. It was a pre-pandemic proof point that real insight doesn’t come from monitoring alone; it comes from building a system that recognizes meaningful patterns and acts early enough to change the trajectory. As the post-pandemic world evolves and care increasingly expands beyond hospital walls into other environments, the datapoints collected are compounding exponentially. The organizations that excel in this next era will be those that know what to do with it. They'll build the connective tissue that turns scattered inputs into coherent insight, and coherent insight into safer, more proactive care. They will shift from responding to deterioration after it happens to anticipating it beforehand, and taking proactive action accordingly. The value and quality benefits that shift will afford is why the next frontier for clinical excellence won’t be defined by how much we monitor, but by how well we understand the signals we receive and act accordingly. #RPM #remotemonitoring #telehealth https://lnkd.in/ehxAHbaQ

Mental Health Can’t Be an Afterthought in Oncology Anymore We’ve sequenced tumors. We’ve mapped biomarkers. We’ve built AI systems that can predict treatment response from a tissue slide. Precision oncology has transformed cancer care. And yet. 1 in 3 cancer patients has a diagnosable mental health condition. 60% never receive a referral for screening. 2 in 5 who ask for support don’t get it. We’ve applied extraordinary rigor to the molecular signature of disease. We haven’t come close to applying that same rigor to the psychological experience of the people living with it. That’s starting to change. I’m a decent oncologist. I can talk your ear off about minimal residual disease testing and cytokine release syndrome protocols. But data-driven mental health care? I’m learning alongside everyone else. Medical school didn’t teach me measurement-based psychiatric care. I can order an NGS panel in my sleep, but ask me to interpret a longitudinal PHQ-9 trajectory and I’m reaching for help. This isn’t a criticism. It’s an opportunity. We need our mental health professional colleagues to educate us. What does precision mental health care actually look like? How do we build it into cancer care the way we’ve built in tumor boards and molecular testing? The Emerging Landscape Tech-Enabled Collaborative Care. The Collaborative Care Model—backed by over 90 randomized trials—is getting a digital upgrade. Behavioral health care managers and psycho-oncologists are integrating directly into oncology practices, tracking symptoms in real time. Early adopters are seeing 58% improvements in depression scores within three months. Depression remission in 86 days versus 614 days in usual care. That’s nearly two years of suffering addressed through better process engineering. Early Warning Systems. AI tools are scanning clinical notes and portal messages for linguistic markers of distress—identifying patients who need support before crisis hits. We’re shifting from reactive callbacks to proactive outreach. Virtual Care Without Borders. Geography no longer dictates access. Digital therapeutics deliver evidence-based cognitive behavioral therapy directly to a patient’s phone. These aren’t meditation apps—they’re prescription-grade interventions calibrated to the cancer journey. 24/7 Support That Works. AI-enabled platforms offer round-the-clock symptom monitoring with escalation to oncology-trained clinicians. A patient waking up terrified at 2 AM can reach out and get a response. The common thread? Technology isn’t replacing human empathy. It’s scaling it. I’m organizing a Virtual Summit on the Future of Oncology Mental Health Care—bringing together national thought leaders on where we are and where we’re headed. I’m looking for collaborators. If you’re a digital health innovator, or clinician doing this work—I want to learn from you. If you represent an organization aligned with this mission and interested in sponsoring or partnering—let’s talk.

Today, I'm happy to share a study that I was a part of previously, which just got published; we explored the possibility of predicting relapse in people with depression using smartphone behavioral data. https://lnkd.in/g-_VT6dg This study was particularly interesting because it used passive data, meaning participants didn't need to actively input information, and we relied on the app's existing tracking mechanisms to capture metrics such as step count and device usage. Testing various models, we were able to learn that a combination of activity, activity type, and location was the best way to predict depressive severity ratings and relapse. Ultimately, the value this study demonstrates is the utility of passive behavioral data, which is often unintrusive, constantly being collected, and not particularly sensitive (although the argument could be made for the app tracking your whereabouts). Traditionally, self-report measures can be inaccurate, or people get so busy that they forget to enter their data into mood-tracking apps. But having models that can use automatically collected data without user input allows us to gain more insight into naturalistic patterns of behavior. Moreover, if we're able to predict when an episode of depression occurs through such data, interventions can be introduced before things escalate. I'd argue that even a slightly inaccurate predictive model that overestimates relapse risk can be useful, though the counterargument is that it's not the optimal way to use a tool designed to reduce clinicians' workload while still ensuring the people they support receive quality care. Nevertheless, it's always exciting to see technology used in mental health, and I hope to be part of more projects like this in the future. #mentalhealth #mentalillness #psychology #psychiatry #wellness

From Access to Measurable Impact: The Next Competitive Edge in Behavioral Health In behavioral health today, payers, providers, and patients are all asking the same question: “Is this working?” Measurement‑Based Care (MBC) offers a clear, data‑driven answer. The American Psychological Association’s feature on MBC highlights a simple truth: when we collect, share, and act on patient‑reported data consistently, care doesn’t just become measurable — it becomes measurably better. Why this matters now 📈 Adoption gap + momentum: Fewer than 20% of clinicians use MBC — but adoption is accelerating as insurers, regulators, and large systems push for objective, repeatable outcome measures. 🔄 From episodic to continuous: The “Collect → Share → Act” model moves beyond static tools like the PHQ‑9 toward ongoing, patient‑driven data streams — a shift already reshaping digital health. 🤝 Payer alignment: MBC creates a direct bridge between documented progress and reimbursement in value‑based models. ⚡ Tech tailwinds: EHR integrations, mobile capture, and patient‑facing apps are lowering the biggest historical adoption barriers. The takeaway: In an era of value‑based care, the ability to prove improvement — not just enable it — is becoming table stakes. MBC isn’t a nice‑to‑have; it’s a differentiator for organizations that want to win contracts, build trust, and scale measurable impact. https://lnkd.in/g9XsHGNG

We’ve made real progress expanding access to mental health care. More kids are getting in the door, which matters. But I worry that in focusing so much on access, we’ve avoided a harder conversation about what happens next. Are kids actually getting better? What I see in practice is that families are left to figure that out on their own. There’s no clear way to track progress, no shared understanding of goals, and no consistent adjustment when something isn’t working. In other areas of healthcare, this would be a red flag. We measure. We adjust. We take action when something isn’t working. Mental health should be no different, especially for kids. A few things I believe need to change: • We have to use measurement-based care to track outcomes, not just visits • We need to adjust care based on what the data shows • We have to hold ourselves accountable for whether care is actually helping Access matters. But access without accountability can create the illusion of progress—and kids deserve better. I shared more in my latest Forbes piece. 👉 Read more here: https://lnkd.in/gFXZexYn #mentalhealth #youthmentalhealth #behavioralhealth #measurementbasedcare #MBC #healthcarequality #valuebasedcare #healthoutcomes #pediatrics #mentalhealthcare 

‎A mother in the Bronx can find dozens of mental health programs within 5 miles. A veteran in rural upstate may have to drive hours for the same support. ‎That’s the reality my analysis of 4,959 programs, 712 agencies, and 715 facilities in New York revealed. ‎ ‎Using data, I built a set of dashboards that answer critical questions for policymakers, providers, and advocates: ‎ ‎• 𝘞𝘩𝘦𝘳𝘦 𝘪𝘴 𝘤𝘢𝘳𝘦 𝘤𝘰𝘯𝘤𝘦𝘯𝘵𝘳𝘢𝘵𝘦𝘥 𝘷𝘴. 𝘸𝘩𝘦𝘳𝘦 𝘪𝘵’𝘴 𝘴𝘤𝘢𝘳𝘤𝘦? ‎• 𝘞𝘩𝘪𝘤𝘩 𝘢𝘨𝘦𝘯𝘤𝘪𝘦𝘴 𝘵𝘳𝘶𝘭𝘺 𝘰𝘧𝘧𝘦𝘳 𝘤𝘰𝘮𝘱𝘳𝘦𝘩𝘦𝘯𝘴𝘪𝘷𝘦 𝘴𝘶𝘱𝘱𝘰𝘳𝘵 (3+ 𝘱𝘳𝘰𝘨𝘳𝘢𝘮 𝘤𝘢𝘵𝘦𝘨𝘰𝘳𝘪𝘦𝘴)? ‎• 𝘞𝘩𝘢𝘵 𝘱𝘰𝘱𝘶𝘭𝘢𝘵𝘪𝘰𝘯𝘴 𝘢𝘳𝘦 𝘶𝘯𝘥𝘦𝘳𝘴𝘦𝘳𝘷𝘦𝘥 𝘰𝘳 ����𝘵 𝘳𝘪𝘴𝘬? ‎• 𝘏𝘰𝘸 𝘥𝘰 𝘤𝘦𝘳𝘵𝘪𝘧𝘪𝘤𝘢𝘵𝘪𝘰𝘯 𝘢𝘯𝘥 𝘤𝘰𝘮𝘱𝘭𝘪𝘢𝘯𝘤𝘦 𝘷𝘢𝘳𝘺 𝘢𝘤𝘳𝘰𝘴𝘴 𝘵𝘩𝘦 𝘴𝘵𝘢𝘵𝘦? ‎ 𝗞𝗲𝘆 𝗜𝗻𝘀𝗶𝗴𝗵𝘁𝘀: ‎ ‎☑️ Only 25% of agencies provide broad, multi-category services. ‎☑️ Just 3 agencies have multi-location reach; most care remains hyper-local. ‎☑️ Nearly 1 in 3 programs target vulnerable populations, but coverage is patchy. ‎☑️ Over 60% of agencies offer only one category, limiting holistic care. ‎☑️ Agencies with stronger compliance also tend to offer broader programs, showing a direct link between structure and quality. ‎ ‎𝗥𝗲𝗰𝗼𝗺𝗺𝗲𝗻𝗱𝗮𝘁𝗶𝗼𝗻𝘀 𝗳𝗼𝗿 𝗦𝘁𝗮𝗸𝗲𝗵𝗼𝗹𝗱𝗲𝗿𝘀: ‎ ‎✅ 𝗕𝗿𝗼𝗮𝗱𝗲𝗻 𝗦𝗲𝗿𝘃𝗶𝗰𝗲 𝗦𝗰𝗼𝗽𝗲: Incentivize agencies to expand beyond single-category care. ‎✅ 𝗘𝘅𝗽𝗮𝗻𝗱 𝗚𝗲𝗼𝗴𝗿𝗮𝗽𝗵𝗶𝗰 𝗥𝗲𝗮𝗰𝗵: Strengthen referral systems and partnerships to cover rural & underserved areas. ‎✅ 𝗘𝗹𝗲𝘃𝗮𝘁𝗲 𝗖𝗼𝗺𝗽𝗹𝗶𝗮𝗻𝗰𝗲: Link certifications with funding and capacity-building to boost service quality. ‎✅ 𝗧𝗮𝗿𝗴𝗲𝘁 𝗩𝘂𝗹𝗻𝗲𝗿𝗮𝗯𝗹𝗲 𝗚𝗿𝗼𝘂𝗽𝘀: Prioritize investment in programs addressing high-risk populations. ‎ ‎Behind every bar chart is a person waiting for care. The data makes it clear: mental health access is uneven, fragmented, and in need of bold solutions. ‎ ‎I believe data like this can inform funding priorities, workforce planning, and equity-driven policies in mental health. ‎ ‎If you work in behavioral health, government, or nonprofit advocacy, I’d love your perspective on how insights like these can drive smarter decisions. ‎ ‎#MentalHealth #PublicHealth #HealthEquity #DataForGood #DataAnalytics #Healthcare ‎#HealthcareAnalytics #GlobalHealth #BehavioralHealth #Excel #Datafam