Missouri harnesses big data to track and manage behavioral health patients

data infographic
Missouri harnessed data analytics to monitor patients with behavioral health issues, which helped drive down hospitalization rates. (Pixabay)

NATIONAL HARBOR, Md.—Health officials in Missouri have invested heavily in data analytics to monitor patients with behavioral health needs—an investment that's already started paying dividends. 

The state was able to reduce hospitalization rates from 34% to 25% between 2012 and 2015, and it cut down emergency department visit rates from 45% to 31% in that same window. It also saved about $98 million in healthcare costs. 

Embracing big data and leveraging population health tools has helped providers in the state identify the needs of a notoriously vulnerable patient group and better monitor other chronic comorbidities, said Natalie Fornelli-Cook, manager of integrated care for the Missouri Department of Health.

"Giving our providers this data helps them so that they can see where they are," Fornelli-Cook said. 

RELATED: How Providence St. Joseph Health, Kaiser Permanente are embracing mental health first aid 

Behavioral health leaders in Missouri were among the dozens of speakers at the National Council for Behavioral Health's annual conference last week. 

The state's data-gathering process has been incremental but began by harnessing Medicaid claims data circa 2008, Fornelli-Cook said. However, that claims data came with a number of gaps, leading behavioral health providers in Missouri to conduct routine metabolic syndrome screenings for patients. 

Beginning in 2010, the state transitioned to a disease management outreach program that enrolled in 3,700 patients with mental illness to get a stronger grasp of their needs. That program also began to enroll patients with substance abuse disorders in 2014. 

RELATED: HHS adviser—Real-world challenges plant the seeds for effective data analytics 

What brought it all together was the 2012 launch of the Behavioral Healthcare Home program, the first of its kind in the U.S. Through the program, which auto-enrolled 18,000 people at launch, state officials were able to monitor hospitalizations and ER visits for behavioral health patients based on prior authorization alerts. 

Planning for population health and data collection requires a culture change, which is likely the largest hurdle for groups looking to emulate Missouri's approach, said Tara Crawford, clinical quality manager for the Missouri Coalition for Community Behavioral Healthcare. 

Collecting that data into a form that was useful for providers was also a challenge. Data were spread out across electronic health records, various spreadsheets and other databases outside the EHR, she said. In addition, the team wanted to ensure that clinicians had access to real-time data at both the individual patient level and in aggregate. 

That required several updates to the web-based tool utilized by clinicians, Crawford said. Healthcare organizations looking to try something similar need to find an effective vendor partner that can build a usable platform. She offered some advice on finding the best one: 

  • Spend time with potential vendors and learn as much as possible about what they can offer
  • Be honest and transparent about needs and wants for the program
  • Look past sweet talk or buzzwords and search for concrete information