A predictive stroke analysis tool - MedCity News

Almost twenty years ago, the Institute of Medicine (now the National Academy of Medicine) published Unequal Treatment: Addressing Racial and Ethnic Disparities in Health Care. In this landmark report, the NAM stated that “a large body of published research reveals that racial and ethnic minorities experience lower quality health care and are less likely to receive even routine medical procedures than white Americans.”

Almost two decades later, social determinants of health (SDoH) – also called health-related social needs (HRSN) – remain a major cause of disparity and account for up to 50% of health outcomes. SDoH are environmental conditions, such as where people are born, live, learn, work, play, worship, and grow old. In mental health and substance use disorder (SUD), these disparities result in disproportionate overdose mortality among minority populations, limit access to care and impact where and how patients are treated, including over-representation in the criminal justice system.

The sad reality is that zip code, rather than genetic code, may still matter more in terms of our health care outcomes. For example, in many studies just ten miles can be the difference of 30 years of life expectancy. So outcomes-based care is here, but it’s just not openly available to everyone. We need a new approach.

Data is the answer, but what data?

Many experts point to data as the most effective tool to address these long-standing inequalities. Data is essential for identifying discrepancies, targeting efforts and resources to address those discrepancies, measuring progress and achieving accountability. One of the most significant data hurdles, however, is the need for more standardized and holistic data models. With so many types of data available to help track and guide healthcare decisions, what kind of data should behavioral health include in its outcome models?

For example, the federal government has minimum standards to report race/ethnicity data. But these standards were last revised in 1997 and may only partially reflect the diversity of today’s population. The Office of Minority Health (OMH) and the Affordable Care Act of 2010 (ACA) proposed more detailed categories. However, OMH, ACA and many other types are not universally accepted. This categorical complexity is amplified when we talk about country-level data with countries adopting different reporting standards and requirements.

Data insight, unlocked and empowered by technology, is the most effective way forward. To address gaps in standardization and quality, mental health providers must use technology to explore how facilities can harness the power of linked data from purpose-built electronic medical records (EMRs). With this approach, providers can identify high-cost and high-risk behavioral health patients through complete SDoH classification and analysis.

Data: Beyond clinical and postcode

Many mental health EMR systems have used clinical data to understand patient outcomes. The problem is that clinical data tell us only part of the story. For example, weekly teletherapy will only work if the patient has reliable access to the Internet, a computer, or a smartphone. The correct SSRI prescription will not help the patient’s depression if they do not understand the dosage, the label, or the delayed onset of action. And factors such as distance and available transportation to care can hinder a patient’s ability to make appointments.

SDoH should be the basis of a new data structure for behavioral health in the future. First, providers must identify patients facing adverse SDoH and include them in behavioral health management and population strategies. And predictive insights, along with artificial intelligence and automation, can help prioritize barriers to care for each patient, helping providers overcome the mental health crisis plaguing Americans today.

We can lay the foundation for predictive insights by adding smart features to EMRs, proactive population-level assessment data, biometric and wearable integration, and existing social vulnerability indices. Additionally, by linking demographics to treatment protocols and outcomes, we can trend and benchmark outcomes and begin applying and acting on risk models through intelligent referral and follow-up workflows. And global data models can identify risk patterns, drive earlier and more proactive outreach to at-risk patients, and reveal new insights into patient experience and risk factors.

These strategies can assist clinicians with targeted interventions that help patients manage their health more effectively and maximize clinical resources, leading to better patient outcomes and long-term, sustainable substance abuse management. By linking clinical and SDoH data into care analytics, behavioral health practices can support adjustments to approved treatment protocols and lengths of stay. And by looking beyond just the physiological, we can create personalized treatment pathways based on the individual and their barriers to care. These holistic models will help close gaps where clinical data only tell part of the story.

Related data, trips and results

Connecting the dots of the journey for those most vulnerable to mental challenges and risks can be a game-changer in the fight to curb substance abuse. For decades, SUD and behavioral health data have existed in silos. And payment incentives aren’t aligned with results—yet. It is a disjointed system that contributes to disjointed and contrasting results.

In general, behavioral health and SUDs lag behind other specialties in collecting health equity data to reduce disparities in outcomes. Data standardization, reporting requirements and interoperability are critical to rapidly addressing gaps. We can begin to lay the groundwork for an open approach to sharing results and data standardization through investments in data and technology that can serve as a powerful springboard to much-needed changes in addiction treatment and coverage.

Photo: tonefotografia, Getty Images

Source link

Leave a Reply

Your email address will not be published. Required fields are marked *