Kansas pilot says AI helped older adults stay home longer

A Kansas pilot from Zemplee and Windsor Place At-Home Care found that Medicaid-eligible older adults who qualified for nursing home care spent less time in facilities and reported less depression and loneliness. The findings could point state Medicaid programs and health plans toward a lower-cost alternative to institutional care.

Why it matters: - A year in a nursing home now costs a national median of $115,000 to $130,000. - Medicaid covers 44% of long-term nursing home care nationally. - Kansas’s average daily Medicaid nursing facility rate rose about 32% between 2022 and 2025. - About 90% of older adults say they want to age at home. - Roughly 63 million Americans are caring for a family member, often while working.

What happened: - Zemplee and Windsor Place At-Home Care released preliminary results from a Kansas pilot in a white paper titled Nursing Homes Without Walls. - The pilot followed 121 Kansas Medicaid members who qualified for nursing home care but lived at home. - Participants averaged 73 years old. - Ninety-one percent had three or more chronic conditions. - Windsor Place ran the pilot with the University of Kansas Medical Center and UnitedHealthcare of Kansas. - The Kansas Department for Aging and Disability Services funded the program. - The pilot won two Gold awards at the 2026 McKnight’s Tech Awards.

The details: - Participants spent 54% fewer total days in skilled nursing facilities compared with the period before the pilot. - Depression scores fell 45%. - Loneliness scores fell 29%. - The depression and loneliness changes were statistically significant. - The technology cost less than $600 per person per month. - Of the 109 participants included in the analysis, people stayed in the program and at home for an average of 27.4 months. - Seven participants, or about 6%, had to be transitioned into nursing homes because of their conditions. - Hospitalizations fell 10% among the 19% highest-risk participants after an earlier increase tied to early intervention from case finding. - Every participant said the technology was helpful and easy to use. - Ninety percent said they felt more independent. - The in-home system used small sensors placed around the home, with no cameras or microphones. - Zemplee’s Attentive AI learns daily routines and flags changes such as disrupted sleep, reduced activity, or altered routines. - Participants also received daily vital-sign monitoring. - Amazon Echo Show devices helped participants stay connected with family and caregivers.

Between the lines: - The pilot points to a lower-cost model for people who qualify for nursing home care but may be able to remain safely at home with monitoring and support. - The results also suggest that passive sensing and early alerts may help caregivers intervene before a crisis and reduce isolation. - The findings are preliminary, so broader rollout would need to prove the model at larger scale and across different care settings.

What’s next: - Zemplee is inviting state Medicaid programs, health plans, and other care organizations to expand the model. - Organizations can contact the team at nhwwstudy@zemplee.com. - The full white paper is available here.

The bottom line: - The Kansas pilot suggests AI-powered home monitoring could help older adults stay out of nursing facilities longer while lowering costs and improving well-being.