A $40M Lesson in Healthcare Planning: What One Closed Facility Can Teach Us About Strategy
In 2021, a major U.S. health system—among the ten largest—opened a freestanding emergency department (FSED) in a growing metropolitan area. It was a state-of-the-art, 33,000-square-foot facility, designed to serve up to 50,000 patient visits annually. But just four years later, the facility was shuttered, and what was once envisioned as a cornerstone of regional care strategy became an empty building with a hefty price tag.
What Went Wrong?
Sources familiar with the project suggest the facility struggled to reach double-digit patient volumes on many days—staggeringly low utilization for a facility this size. While the exact figures remain internal, what’s clear is that the site significantly underperformed. Estimates place the total loss—accounting for construction, equipment, and years of operating costs—at upwards of $40 million.
How could one of the most sophisticated systems in the country make such a large mistake?
By doing the same thing as everyone else. Almost every health system across the country has experienced a project that dramatically underperforms expectations. In the vast majority of cases, the expectations were unrealistic – a by-product of traditional assumptions-based modeling. This approach relies on Excel sheets of weighted factors and weighted factors, but is susceptible to bias and sensitive to incorrect assumptions.
What Could Have Been Known?
We took a look at this FSED to see what went wrong. Our team applied our predictive utilization models to the site and the surrounding region.
Our models take a novel approach to projecting healthcare utilization. Using AI, they learn from hundreds of thousands of real-world sites of care – letting patterns in the data define relationships between contributing factors instead of making assumptions. This method significantly boosts projection accuracy over traditional approaches.
The results were sobering: the model projected just 11,847 patient visits per year at the selected site—barely 20% of its target volume, but consistent with eyewitness accounts. If the health system had access to Zite AI’s technology, they would have known patient volumes would be low before breaking ground. The disaster could have been entirely avoided.
Better yet, this story could have had a happy ending. We analyzed the same model of care across the metropolitan area. The selected site of care had the lowest expected patient volumes of anywhere in the region. If the system opened the FSED just a few miles away, projected volumes exceeded 50,000 patients annually.
In short: the concept wasn’t flawed. The location was.
The Broader Implication
To be clear, we aren’t criticizing the health system. They have an outstanding (and well-deserved) reputation for clinical excellence and operational sophistication. Rather, the decision-making process employed by organizations across the industry is inherently flawed and outdated.
Traditional assumptions-based models have been the gold standard for decades, but they’re based on heuristics, weighted scoring sheets, and expert judgment. When assumptions are wrong, utilization ends up looking very different from expectations. Consequently, almost every hospital system in the country has a similar story of a project that ended in abject failure – completely missing the mark on expected utilization.
A Different Path
For the first time, healthcare organizations don’t need to rely on heuristics and assumptions. Modern artificial intelligence technology empowers strategic leaders to predict utilization without making assumptions. Instead, the data does the talking. Had modern predictive tools been part of the planning process for this health system, this site’s limitations would have been clear from the outset. So too would the stronger alternatives nearby.
Moving Forward
The stakes are high. Decisions have far-reaching financial consequences. Health systems are beginning to incorporate modern predictive analytics tools into their planning process, significantly reducing their risk on each decision to the benefit of their bottom line and the communities they serve. This closed FSED is more than a cautionary tale. It’s a reminder that the smartest strategies start with the clearest vision of the future.
