Length of stay is the most scrutinized number in hospice, by regulators, by referrers, and by the organizations themselves. It is also one of the easiest numbers to misread, because nearly everyone reads it as an average, and the average hides everything that matters.
Why the average misleads
Hospice length of stay is not normally distributed. Every hospice serves a mix of very short stays, often patients referred in the final days of life, and long stays that can run past six months. The average of that mix can look stable while both tails are moving in opposite directions. Two agencies with identical average length of stay can have completely different compliance exposure and completely different referral problems.
The distribution is what tells the story: what share of patients are served for less than seven days, what share for more than one hundred eighty, and how the middle is shaped.
The two tails mean different things
A heavy short-stay tail is primarily a referral-timing problem. Patients arriving in the last days of life receive less benefit from hospice care, families consistently report wishing the referral had come sooner, and a rising short-stay share usually traces to specific referral sources whose patients arrive late. Because it traces to sources, it is actionable: the fix lives in earlier-referral conversations with identifiable hospitals and physicians.
A heavy long-stay tail is primarily a compliance and eligibility-documentation matter, and it is where regulatory attention concentrates. A long-stay share drifting above what is typical for your market and diagnosis mix is worth understanding before someone else asks you to explain it.
Benchmark against the right comparison
Raw comparisons between agencies mislead because diagnosis mix drives length of stay. A hospice with a high share of dementia patients will run longer stays than one dominated by cancer diagnoses, legitimately. Meaningful benchmarking holds diagnosis constant: your distribution for a given diagnosis group against the same distribution for your market, your state, and the nation.
The same discipline applies within an organization. Comparing sites to each other only becomes fair, and only becomes useful, when the comparison accounts for what each site’s referral sources send it.
What to watch, on a rhythm
Three views cover most of what leadership needs: the short-stay share by referral source, which turns referral-timing problems into a named list of conversations; the long-stay share by diagnosis against market norms, which quantifies compliance exposure; and site-to-site distributions on a like-for-like basis, which separate operational differences from mix differences.
Claims data makes all three views possible for every hospice in every market, not just your own, which is what turns length-of-stay review from introspection into benchmarking.
Ariv Hospice builds these views from verified CMS claims, refreshed weekly, with your agencies flagged as yours. To see your own distributions against your markets, contact us at enquiries@arivhealth.com.