Health Service Data Patterns

by admin
6 years ago

Data Analytics: Using health service data to identify service patterns which can inform and improve the way we practice.

The Issue:

Frontline clinical staff may have a hunch that, for example, children receiving behavioral health services whose parents have PTSD or trauma may discontinue care before program completion (i.e., drop out) more often than other families. Traditionally, a staff member would need to recognize this pattern, and raise the issue through multiple levels of supervision before a data analyst could pull and analyze the data to confirm the pattern, a process that would often take months or years, if at all. Using pattern recognition techniques, such as machine learning, data patterns can be identified and machines can learn when those patterns are associated with positive or negative outcomes. New approaches can regularly feed alerts of service patterns associated with outcomes to executive level staff and data teams so that staff can respond in real time to adjust practice to meet the needs of every child, youth, family, adult and older adult in care. In our example, when this pattern is recognized, frontline staff can be trained to address caregiver trauma in children’s behavioral health service plans, improving retention of these families through to successful program completion.

Our Work:

  • Work with organizations to develop analytical models to evaluate service and outcomes data
  • Develop decision trees to support determination processes
  • Identify patterns and priorities of incoming service populations
  • Develop alerts for clients who are at risk of negative outcomes
  • Uncover clusters of similar clients, service patterns and outcomes
  • Implement machine learning techniques to identify strengths and weaknesses in health services programs
  • Move organizations closer to identifying what works for whom

The Solution:

MHData is able to develops custom analytical approaches for ‘big’ health data so that:

  • Organizations can learn more from the data they already capture to improve quality of care and respond to changing service needs.

  • Organizations can learn what works for whom.