West Tennessee Healthcare piloted Xsolis’ Navigate, an AI-powered case management platform, to tackle avoidable delays in patient discharge. With predictive discharge modeling, smart task prioritization, and real-time EMR integration, the health system’s care management team gained the root-cause data they’d long lacked—leading to faster morning huddle prep, easier onboarding, and more timely discharges. Early targets include 60 minutes of daily leadership time savings, 105 minutes saved for frontline staff, and a 20% reduction in LOS variance and excess days. This blog highlights how West Tennessee Healthcare turned anecdotal insight into measurable, actionable improvement.
Length of stay (LOS) management has long been one of the most complex challenges facing hospital care management teams. During a recent session at Xsolis’ user conference, XCHANGE, The Bold Future of Length of Stay Management, Cassie Fleming, Senior Director of Product Management at Xsolis and Debbie Ashworth, Executive Director of Care Management at West Tennessee Healthcare, made the case for why the old ways of managing LOS are no longer enough and what a smarter, data-driven future looks like.
Before diving into West Tennessee’s journey, the speakers grounded attendees in the scale of the problem. According to an Advisory Board analysis of CMS fee-for-service claims data, avoidable delays account for roughly 26% of the average length of stay-approximately 1.2 of every 4.2 inpatient days. Across the healthcare system, that translates to 10.8 million avoidable inpatient days annually, the equivalent of 29,590 full hospital beds occupied for an entire year.
These aren’t just abstract numbers. Every avoidable delay represents a patient who isn’t getting home sooner, a bed that isn’t available for the next admission, and a cost that the health system absorbs unreimbursed. For care management leaders like Debbie, who oversees a team of 37 case managers, 21 social workers, 19 utilization management nurses, and 9 administrative assistants across the West Tennessee Healthcare system, the pressure to address these inefficiencies is relentless.
“Prior to implementing Xsolis’ Navigate solution, our biggest pain points were understanding what our discharge barriers were and what was causing avoidable delays,” Debbie shared. “We anecdotally understood the barriers, but we didn’t have the data to back it up.”
The speakers organized the LOS challenge around three core breakdowns that most hospitals will recognize immediately:
These three breakdowns compound one another. Without reliable data, you can’t fix your processes. Without fixed processes, patient flow suffers. And so, the cycle continues.
The heart of the session focused on how Navigate, Xsolis’ AI-powered case management workflow solution, was purpose-built to address each of these challenges, not as a technology overlay, but as a tool designed from the ground up around how case managers and social workers actually work.
Predictive models drive the Navigate workflow by generating discharge predictions early in an encounter and updating automatically as new clinical data becomes available. This gives care teams a forward-looking foundation for discharge planning rather than a reactive or backward-looking one.
From there, AI-powered workflows automatically prioritize tasks and flag cases of interest, helping case managers know exactly where to start their day. “Xsolis’ smart filters enable our case managers and social workers to sort their patient lists by expected discharge date, so they know where to start their day and which patients need to be prioritized,” Debbie noted.
Critically, Navigate integrates bi-directionally with the EMR, keeping discharge planning data synchronized across systems so that neither clinical staff nor care management teams are working from outdated information, and neither group is required to duplicate documentation.
West Tennessee Healthcare served as the pilot system for Navigate, partnering closely with the Xsolis product team across three structured phases covering workflow scope, technical integration, and go-live support.
West Tennessee’s clinical subject matter experts and leaders provided critical feedback that helped shape and improve the workflow and reporting, ensuring the final product could help solve for actual operational needs rather than theoretical ones.
“Unlike the EMR, Navigate was built specifically for case managers and social workers,” Debbie said. “Xsolis worked closely with our team to deliver the data and reporting we needed that we couldn’t get from our EMR to better manage length of stay.”
The early results from the pilot illustrate what’s possible when the right data is finally in the right hands. Leadership daily preparation time, particularly for morning huddles, was significantly transformed by having access to reporting at their fingertips instead of having to hunt down their data in multiple places.
Teams gained clear visibility into discharge planning progress and barriers through AI predictions directly in the workflow. Onboarding new team members became easier thanks to standardized workflows. And perhaps most importantly, the why behind avoidable delays finally became visible and actionable.
“The performance insights are a game changer,” Debbie reflected. “Gaining visibility into avoidable delay days alongside excess days allows us to more accurately identify opportunities for improvement. With root cause data to guide us, we know exactly where to focus our efforts to drive meaningful change.”
The improvement targets set for the initial implementation are ambitious, but grounded:
“Since implementing Navigate, the impact on our patients has been substantial,” Debbie added. “Eliminating discharge barriers has allowed us to discharge patients more timely, ensuring they receive the care they need sooner.”
The session made one thing unmistakably clear: the future of length of stay management isn’t about working harder, it’s about working with better tools and information.
When care management teams have AI-driven predictions, prioritized workflows, and meaningful root cause data at their fingertips, avoidable delays stop being an unavoidable reality and start becoming a solvable problem.
For health systems still relying on manual tracking, siloed EMR data, and anecdotal insight to manage LOS, the gap between where they are and where they could be is significant- and growing. The bold future described in this session is already here. The question is whether organizations are ready to step into it.
Predictive AI, generative AI, and agentic AI each serve different purposes in healthcare