Health & Human ServicesIssue 024 records
Health & Human Services
Four sources. Algorithms are reaching Medicaid home- and community-based services: a Nebraska court let a due-process challenge to interRAI-driven disability funding proceed (Oct 29), and a KDADS-funded Kansas pilot reports 54% fewer nursing facility days with sensor-based AI monitoring (vendor-reported, no control group). San Diego County Behavioral Health is training its provider network for systemwide Eleos ambient documentation, and Indiana's child welfare adviser is vetting AI for paperwork and kinship search but not determinations. No new public health or public hospital evidence this edition.
- Evidence records
- 4
- Cross-source patterns
- 2
- Evidence classes
- 2 independent reporting1 vendor claim1 government evaluation
- Outcomes
- 3 emerging1 cautionary
- Source freshness
- 3 new this fortnight1 recent
- Research completed
- 2026-10-31
Choose a role to see its takeaway beside every record in the ledger.
Patterns across the evidence
Synthesis · Lighthouse Advisory interpretation
Algorithms are reaching Medicaid home- and community-based services from two directions: funding and monitoring
In Nebraska, a court let a due-process challenge proceed against algorithm-driven interRAI funding tiers that allegedly rely on undisclosed decision rules. In Kansas, a state-funded pilot reports fewer nursing facility days with sensor-based AI monitoring, but only as vendor-reported pre/post results. Both leave the people affected, and the state, without independently verifiable evidence of how the tool shapes outcomes.
Operating questionBefore an algorithm sets or influences HCBS services, can the agency explain individual results and show independently evaluated outcomes?
Supporting evidenceNational Health Law Program (plaintiff co-counsel; Legal Aid of Nebraska filed, joined by Nebraska Appleseed, NHeLP and TechTonic Justice); case concerns Nebraska Department of Health and Human ServicesZemplee Inc. and Windsor Place At-Home Care, with University of Kansas Medical Center and UnitedHealthcare of Kansas; funded by the Kansas Department for Aging and Disability Services
Child welfare and behavioral health AI is framed as administrative relief, ahead of measured results
San Diego County BHS is training its contracted network for a systemwide ambient documentation rollout with consent coaching, and Indiana's child welfare adviser is vetting AI for paperwork, staffing forecasts and kinship search while ruling out AI determinations. Neither source reports measured outcomes yet, so baselines set now determine whether benefits can be shown later.
Operating questionWhat baseline (documentation time, time to kin identification, edit rates) is the agency capturing before go-live?
Supporting evidenceCounty of San Diego Behavioral Health ServicesIndiana Capital Chronicle; Gov. Mike Braun's special advisor on abuse and neglect, Adam Krupp (former Indiana DCS director)
Evidence ledger
Full record · every source keeps its link and limitations
Nebraska court lets due-process challenge to DHHS's interRAI algorithm-driven disability funding proceed
In an October 29 order, a Lancaster County District Court overruled Nebraska's motion to partially dismiss, letting six claims proceed and lifting a stay on the rest, in Bessey v. Nebraska DHHS, which challenges the use of the interRAI assessment to set Medicaid HCBS funding for people with intellectual and developmental disabilities. The plaintiff's annual budget was cut about 35 percent after an interRAI assessment although his condition had not improved. The suit alleges DHHS uses undisclosed criteria, scoring methods and decision rules to turn assessment answers into funding levels. The ruling does not decide the merits.
Why it matters, evidence and limitations
- Why it matters
- State disability and aging agencies using standardized assessments plus scoring algorithms to set Medicaid budgets now face live litigation over explainability and notice, not just policy criticism.
- Evidence and measured results
- Procedural ruling reported by plaintiff co-counsel one day after the order. The order itself was not inspected. Allegations are the plaintiffs'; the State's position is not described in the release.
- Limitations and uncertainty
- Advocacy source that is party counsel. Interim procedural ruling only. interRAI is a standardized assessment instrument with state-defined scoring; the release does not claim it uses machine learning. Single plaintiff.
Kansas Medicaid aging-in-place pilot with camera-free sensors and AI reports 54% fewer skilled nursing facility days (vendor white paper)
Preliminary findings released October 27 from a KDADS-funded pilot of 121 Kansans on Medicaid who qualified for nursing home care but lived at home: compared with the pre-pilot period, total skilled nursing facility days fell 54%, depression scores fell 45% and loneliness scores 29% (reported as statistically significant). Of 109 participants analyzed, people stayed in the program for an average of 27.4 months and 7 (about 6%) moved to nursing homes. Technology cost was reported at under $600 per person per month.
Why it matters, evidence and limitations
- Why it matters
- State aging and disability agencies and Medicaid HCBS programs weighing remote monitoring to delay nursing facility placement get a state-funded pilot with outcome measures, though only as vendor-reported results.
- Evidence and measured results
- Pre/post comparison without a control group, reported by the vendor in a press release summarizing a white paper. The full white paper sits behind a request form and was not inspected. 109 of 121 participants analyzed; methods for the significance tests are not described in the release.
- Limitations and uncertainty
- Vendor-authored claims; preliminary; no control group; bundled interventions (home care services, vitals monitoring, video devices); full methods not public without a form; results may reflect selection and regression to the mean.
San Diego County Behavioral Health Services trains contracted providers for systemwide Eleos SMARTscribe ambient documentation rollout (training Oct 26–30)
San Diego County BHS told its contracted providers it is preparing a systemwide launch of Eleos, including the SMARTscribe ambient documentation tool. Live virtual training runs October 27–30, 2026 in fourteen 3.5-hour sessions of about 100 providers each, with two optional sessions on talking to clients about SMARTscribe; self-paced training began October 26 and covers how to explain SMARTscribe and AI consent. The county states the goals are documentation quality, lower administrative burden and more provider capacity for direct care.
Why it matters, evidence and limitations
- Why it matters
- A large county behavioral health plan is moving ambient AI documentation from pilots to its contracted provider network, with client consent conversations built into training.
- Evidence and measured results
- Operational county notice; it states goals and training logistics only. No outcome data, go-live date, provider counts or evaluation plan are given.
- Limitations and uncertainty
- Government notice, not an evaluation; benefits are stated intentions; classification as government source reflects authorship only. No results yet.
Indiana child welfare adviser vets AI for paperwork, staffing forecasts and kinship searches, ruling out AI determinations
Adam Krupp, who left the Indiana DCS director post in April for a special advisor role to Gov. Mike Braun on abuse and neglect, is completing a comprehensive evaluation of the state's child welfare system and has been vetting AI's cost, implementation timelines and uses. He said AI should not make child welfare determinations but could reduce paperwork and data entry, forecast staffing from caseload trends, and mine large case files for possible kinship placements, a search he said can take a case manager about a week but AI 60 seconds.
Why it matters, evidence and limitations
- Why it matters
- A state child welfare system is likely to receive formal AI recommendations in an upcoming assessment, framed around administrative relief and kinship search rather than risk scoring.
- Evidence and measured results
- News report based on an interview and weekly status reports. The kinship-search time comparison is the adviser's claim, not a measured result. No tool, vendor or pilot is named.
- Limitations and uncertainty
- Plans and opinions, not deployments; the assessment report is not yet published; claims are unquantified apart from an anecdotal time comparison.
How to read this edition
Source findings, measured results and limitations come from the cited publications. Patterns, operating questions, role takeaways and implementation considerations are Lighthouse Advisory interpretation, stated as questions to validate locally rather than guaranteed outcomes. Vendor and operator claims are labeled as claims. Full research method.
- Independent reporting
- Independent reporting with attributable sources but without a formal evaluation design.
- Vendor claim
- A supplier-provided assertion that has not been upgraded to independent evidence.
- Government evaluation
- A public body’s measured evaluation or documented pilot.