Address Comorbidities with Tech-Supported Approaches to Integrated Care

Consider this common scenario: Trying to assess a high-need patient with diabetes, a care coordinator with access only to the patient’s physical health record is unaware of her history of depression. Conversely, a behavioral care coordinator reviewing her depression charts, may be unable to access her physical health records containing her diabetes history.

Lack of data sharing and communication between the physical and behavioral health worlds leads to suboptimal care including missed diagnoses and lack of treatment. Our fragmented system of care – behavioral and physical health each in their own independent silos – is a barrier to coordinating these interconnected aspects of health.

The scope of the challenge presented by comorbidities is significant:

  • 34 million people – 17 percent of American adults – had comorbid mental health and medical conditions in 2011.1
  • An estimated 8.1 million adults have both a mental illness and a substance use disorder (SUD), with less than half (48 percent) receiving either mental health care or SUD treatment at a specialty facility.2
  • Individuals with comorbid conditions are at heightened risk of returning to the hospital after discharge.3
  • Patients with both a chronic physical condition, such as diabetes, and a comorbid mental health disorder, such as depression, have a 200 percent higher mortality rate than individuals with only diabetes.4
  • Monthly costs for a patient with a chronic disease and depression are $560 more than for a patient with a chronic disease without depression.5

Caring for those with comorbid mental and physical health conditions – which may share common risk factors – calls for an integrated, holistic approach for identifying and managing these individuals in a population. The use of innovative technology tools can ease the path to getting there.

Of course, there is no one “correct” model for integrating care. At the most basic level of integration, providers periodically communicate about shared patients. Co-located care may involve slightly more integration – primary care and behavioral health providers share a common facility but maintain separate cultures and develop separate treatment plans for patients. At the next level, the two groups of providers share some information systems. At the most integrated level, behavioral, physical, social and pharmaceutical data is fully shared, and technology, work flows and care delivery are integrated and perhaps most importantly, providers approach patients’ symptoms in a holistic way.

Common elements of advanced integration models typically include:

  • Shared information systems that facilitate coordination and communication across providers and facilities
  • Screening for depression, anxiety, and other behavioral disorders using validated screening tools
  • Team-based care with non-physician staff to support primary care physicians and co-manage treatment
  • Standardized use of evidence-based guidelines
  • Individualized, person-centered care that incorporates family members and caregivers into the treatment plan
  • Systematic review and measurement of patient outcomes using registries and patient tracking tools

Addressing comorbid conditions through integrated care approaches is gaining momentum. Colorado Medicaid, for example, recently selected five provider organizations to manage integrated physical and behavioral health services throughout the state. But the industry still has a long way to go. According to preliminary findings from ODH survey research, while eight out of ten health care companies say that data integration is important, only one-half have highly integrated behavioral, physical and social data.

Technology can facilitate integrated care in complex patient populations in several ways:

  • Leveraging analytics to facilitate risk stratification
  • Improving care coordination, thereby enabling providers to address gaps in care
  • Expanding the reach of providers beyond the four walls of their offices
  • Delivering cost-effective, evidence-based care

Health care organizations are beginning to embrace developing technologies such as artificial intelligence, machine learning and predictive analytics that can help discern patterns among complex data sets and identify patients who are ascending the risk curve.

Consider a primary care physician who, upon noticing that the patient’s diabetes is worsening, added a new prescription to her medicine regimen, without realizing that the medication led to her hospitalization for a depressive episode. With its ability to recognize patterns, machine learning can help providers realize that the rate of refills or additional prescriptions is an indicator of severity in the patient’s depression or other comorbid conditions.

A recent research paper co-authored by ODH executives found that machine learning methods and advanced predictive models can help improve health care organizations’ ability to identify and predict future high-cost patients suffering from schizophrenia. The paper, “Predicting Future High-Cost Schizophrenia Patients Using High-Dimensional Administrative Data”, also found that the presence of co-morbid physical conditions, such as diabetes and kidney disease, contributed to higher costs.

Another critical role technology can play is serving as a data repository to allow the sharing of information between key stakeholders such as payers and providers. Information sharing, which is at the heart of integrated care, fosters improved care quality by enabling providers to have a broader perspective beyond the services they provide.

As we move increasingly to a value-based model of care, payers’ and providers’ interests will become more closely aligned. To make this happen, however, requires shared visibility into all the data that drives care decisions. Payers have some of this data in the form of claims, while providers have clinical data through their EMRs. The vision of data interoperability has yet to be realized, however, so health care organizations need to bridge this gap through portals and dedicated platforms.

Data sharing will facilitate the development of analytics and insights to segment populations, ensuring that they receive appropriate care, and to create the integrated, longitudinal patient view that is essential for care delivery for each individual.

Integrating mountains of data from multiple, disparate sources can be a daunting task. A good strategy, therefore, is to start slowly, perhaps by sharing just a small volume of data to develop insights about the patient, and then building from there.

Consider, again, the patient with comorbid diabetes and depression we introduced at the start of this paper. Simply by sharing diagnoses – so that her psychiatrist knows about her diabetes – would prompt the psychiatrist to start asking her about her sugar intake, diet, exercise regimen and how her mood affects her compliance with her diabetes medications. That is a meaningful first step and a far better approach than simply sitting on the sidelines and falling behind the curve.

References

  1. “Bringing Behavioral Health into the Care Continuum: Opportunities to Improve Quality, Costs and Outcomes,” American Hospital Association, January, 2012.
  2. https://www.samhsa.gov/data/sites/default/files/NSDUH-FFR1-2015/NSDUH-FFR1-2015/NSDUH-FFR1-2015.htm#cmhi
  3. “Bringing Behavioral Health into the Care Continuum: Opportunities to Improve Quality, Costs and Outcomes,” American Hospital Association, January, 2012.
  4. “Behavioral Population Health Management,” Tractica, 4Q 2016.

5.            “Bringing Behavioral Health into the Care Continuum: Opportunities to Improve Quality, Costs and Outcomes,” American Hospital Association, January, 2012.

Have a Comment?