Care transitions require complex coordination. Suboptimal care coordination results in subpar quality and higher costs. With value-based care there’s an incentive for robust care coordination.
With a new tool that identifies at risk “on the border” patients, they’ve:
- reduced nursing home stays,
- decreased hospital stay length
Keys to success in discharge planning
- Human-centered design to build trust with users.
Current tools focus on raw data and calculated scores. But the explainable AI discharge tool focuses on insights, transparency, and actionable relevant information. That means instead of just sharing information, like “64% chance of hospital readmission” they share an action to take. “Set up PCP appt.”
2. Stakeholder alignment is a core design principle.
Stakeholders may directly or indirectly work with patients. They tend to work in silos, so getting them aligned is key. For example, the payer may be focused on cost, and the provider is focused on quality. We need to focus on both, which can happen more easily when we get the stakeholders talking to each other.
3. Design with extensible data architectures.
Build for scale – think about how your current competencies could address similar challenges. Try to use easily available data. And think about how things may change. Policy changes could mean updating the tool.
Optimal care coordination is an essential part of value-based care. But it’s difficult! AI and machine learning can help.