We all know that data is important. Over the past two decades, business intelligence has become ubiquitous across industries, enabling organizations to collect, analyze, and visualize vast amounts of information to make informed decisions. Today, those same tools—bolstered with advanced artificial intelligence, machine learning, and powerful analytics—have emerged as an important asset in healthcare, a sector that remains an exploration of technological extremes.
On the one hand, clinical environments are increasingly marked by some of the most cutting-edge technologies, from sophisticated innovations in imaging and connected medical devices to breakthroughs in genomics and personalized medicine that are testing the limits of supercomputing. On the other hand, core business processes like approvals continue to rely on outdated methods from fax machines to paperwork.
Most striking, however, at a time when data is considered the new ‘oil’, is the continued inability of doctors to obtain, consolidate and use the ‘whole health’ data they need to determine the best course of treatment , achieve successful patient outcomes and maximize the financial success of their practices. The resulting data gap is significant.
Crossing one gap only to find another
Significant progress has been made in recent years to bridge the data gap between payers and providers. Despite the vast amount of clinical data at their disposal, physicians have historically struggled to access the population-level intelligence held by payers. In contrast, payers had insight into the trends shaping patient cohorts, but had limited intelligence about the key nuances of care that dramatically affect individual patient outcomes.
The shift from a fee-for-service to a value-based approach to care quickly illuminated the importance of addressing this fundamental data gap and forced payers and practices to more effectively share information to meet performance and payment incentives. Substantial gains followed, although the shift to a value-based approach revealed the importance and absence of an equally fundamental gap, viz. the lack of “overall health” data.
What is ‘whole health’ data and why is it important?
“Whole health” data is the information that, combined with traditional clinical data, enables practices to fully impact the single metric that matters most—patient outcomes. For this reason, and because value-based approaches depend on outcomes, “whole health” data is also what allows practices to excel in moving to incentive payment models, including HEDIS and star ratings that impact profitability .
From a data science perspective, “whole health” data should therefore include information related to each patient’s outcome. This includes SDOH, mental and behavioral health insights, and diagnoses and insights gained through the use of mobile devices – all consolidated on a platform that allows non-technical clinical users to discover actionable insights with the powerful algorithms and models provided by modern machine learning applications . A quick examination of each of these data categories reveals why it is imperative that practices incorporate them into patient assessments and care plans.
Research shows that 80% of an individual’s health is determined by factors other than access to quality care. The Physicians Foundation recently noted that eight in ten physicians “believe that the United States cannot improve health outcomes or reduce health care costs without addressing the social drivers of health.”
Not surprisingly, patients who cannot afford medications or travel to appointments face additional health challenges. The results can be tragic. For example, we now know that black Americans in very rural areas died at a 34% higher rate during the Omicron epidemic than white Americans in the same communities due to SDOH.
Fortunately, SDOH data, when accessible and used at the practice level, makes a difference. In fact, an accountable care organization recently analyzed high-cost emergency room visits and found that many patients live in the same poor neighborhood. Additional analysis with machine learning applications found a dramatic increase in emergency room visits on hot days, a metric that led clinicians to realize that patients’ homes lacked air conditioning, a fact that worsened existing health conditions. The non-clinical and counterintuitive step of purchasing air conditioners for the cohort dramatically reduced emergency room visits and made the ACO more profitable.
Mental health data
Behavioral health and physical health are, of course, linked to mental health issues, often associated with co-morbidities, including substance abuse, eating disorders, anxiety and depression. As the National Institute on Drug Abuse notes, “about half of people who suffer from a mental illness will also experience a substance use disorder at some point in their lives, and vice versa.”
Nonadherence to treatment plans and medications is also a significant problem, and substance abuse can be a factor in suicide, the leading cause of death for many age groups. Clearly, practices benefit from being aware of patients’ mental and behavioral health challenges before completing patient assessments and as they work to ensure healthy outcomes.
Data from a mobile device
The pandemic has dramatically accelerated the use of mobile smartphones and made them the primary communication channel for many patients and providers, from scheduling appointments to attending telemedicine consultations. Multiple mobile apps are also now available to track patient progress. Examples include those that track activity levels, blood pressure and blood sugar – all important factors for conditions such as hypertension and diabetes.
Therefore, practices must be able to incorporate cell phone data into their patient assessments and be able to update the electronic health record and treatment plans with real-time data.
Clearly, it is important for physicians and practices to incorporate “whole health” data into their health equity and value-based care efforts; however, this is a process that should not be undertaken alone. Any steps to incorporate external or new sources of patient-related data into the practice should be undertaken with a partner who understands the regulatory issues surrounding personally identifiable information and has a deep understanding of how the data will be stored. protected and analyzed. Finally, the partner must be well-versed in how practices can leverage the intelligence they gain with the many intuitive AI, machine learning, and business intelligence tools now available to business users.
Jeffrey Springer is SVP of Product Management at CitiusTech.