Length of Stay (LOS)
Length of Stay is the number of days between hospital admission and discharge, serving as a primary operational efficiency metric and a key driver of hospital profitability under the DRG prospective payment system.
What is Length of Stay?
Length of Stay (LOS) is calculated as the number of days from the date of hospital admission to the date of discharge. For single-day stays (same-day admission and discharge), LOS is counted as 1 day. Under the Medicare DRG prospective payment system, hospitals receive a fixed payment per admission regardless of how many days the patient occupies a bed — making LOS a direct cost driver that does not generate additional revenue once the DRG threshold is crossed.
Geometric Mean vs Arithmetic Mean LOS
CMS publishes both geometric mean LOS and arithmetic mean LOS for each DRG, but uses geometric mean as the benchmark for DRG payments. Geometric mean LOS is less influenced by outlier long-stay cases (it calculates the nth root of the product of all values rather than the simple average), making it a better central tendency measure for right-skewed hospital LOS distributions. A hospital's performance is most meaningfully compared to the geometric mean LOS for each specific DRG — not a blended average across all DRGs.
LOS Outlier Thresholds
For Medicare, the day outlier threshold is defined as the geometric mean LOS + a fixed number of days (varies by DRG, typically around 1.94× geometric mean + a fixed amount). When a patient's LOS exceeds this threshold AND the estimated cost of care exceeds the cost outlier threshold, the hospital receives an additional per-diem payment for the marginal days. However, the incremental payment typically covers only 80 cents on the dollar of marginal cost — so outlier admissions are generally still financially suboptimal.
LOS Drivers and Analytics
Common drivers of prolonged LOS include: delayed specialist consultations, slow laboratory or imaging turnaround, weekend discharge barriers, discharge planning initiated too late, and social barriers (no safe discharge destination, lack of home support). Analytics should track LOS by DRG against geometric mean benchmarks, by physician and service, and by time-of-day/day-of-week discharge patterns to identify addressable operational bottlenecks.