Happy 2019! To celebrate the new year, let’s visit an old topic: hospital length of stay (LOS). It’s a universal and critical metric in U.S. healthcare: the amount of time in an episode of hospitalization from admission to discharge. Most U.S. hospitals have some sort of desire to reduce LOS in some way.
But despite being a metric with a simple definition, there’s a world of nuance regarding LOS once you dive in. Some hospitals count observation lengths of stay separately from inpatient. Discharges may not be captured in EMRs at the true moment of a patient physically leaving the hospital. Patient units with different care missions can have drastically different lengths of stay, so averaging doesn’t always make sense. Shorter stays aren’t always good; longer stays aren’t always bad. Rural hospitals are not urban hospitals are not academic hospitals. Some important sources only have data that’s months old, so using it to fuel improvement projects is difficult. I could go on.
Case Study: Proactive Staffing and Patient Prioritization to Decompress ED and Reduce Length of Stay
In short, no single system is the source of everything relevant to determining a patient’s “true” LOS, and understanding the best opportunities for improvement is difficult. It’s not as easy as adding up some numbers and comparing averages. Googling “length of stay” produces over 25 million results. There are thousands of research articles regarding LOS improvement. Where do you start?
One way is by thinking about the LOS journey in three parts: strategic, tactical and technical.
First, identify the largest clinical variations in the hospital or hospital system that you’re able to address, and focus initial efforts there. Find possible clinical variations by analyzing avoidable days by patient unit, physician and diagnosis-related group (DRG), and comparing to national norms. The Centers for Medicare and Medicaid Services’ yearly table of geometric mean length of stay (GMLOS) by DRG is the best U.S. source for understanding potential avoidable days. Also, identify patient units and physicians with variances in discharge order and actual discharge times of day. Hospital COOs, CMOs and CNOs can decide on initial improvement projects around one or two of these areas.
Dig into and identify micro-populations with high opportunity for movement or discharge today. Examples include patients with a high likelihood of discharge; observation patients with outstanding discharge orders to home; and certain clinical populations like flu patients. Identifying even one or two additional patients a day for discharge can have a profound effect on LOS and patient flow as a whole. Daily bed huddles, formed by a combination of bed management and nursing unit management, can drive this action.
The tasks in the first two steps are very hard to accomplish without technology that assembles data and presents it as clean analysis inside a time frame tight enough to spur action. Hospital leaders need to get to the data in minutes, not months, or risk losing engaged stakeholders like hospitalists. Good technology allows operations improvement staff to spend their time driving change rather than wrangling data.
Hospital IQ is unique in its ability to quickly assemble a true picture of LOS from many data sources; compare performance to history and national norms to show areas of LOS improvement opportunity; and easily identify micro-populations currently in house to focus daily discharge efforts. As hospitals navigate the shift to value-based care and delivering cost and quality outcomes, reducing LOS has become more important than ever. Hospital IQ can serve as the foundation for hospitals to accomplish this critical goal in the year ahead.
Check out this case study to see how University Hospitals – Cleveland was able to reduce their length of stay by 15% with Hospital IQ.