A care technology pilot may demonstrate fewer missed visits, improved medication support, greater independence or better communication between people receiving services and their support teams. Yet even when those improvements are credible, the technology may disappear when the grant ends. The problem is not necessarily that the innovation failed. It may be that the evidence was never translated into a payment decision.
Across the United States, this distinction is becoming increasingly important for Medicaid-funded Home- and Community-Based Services (HCBS), Long-Term Services and Supports (LTSS), intellectual and developmental disability (IDD) services, aging programs and community behavioral health. The wider challenge is explored through the Impact Insights Innovation, Pilots & Emerging Models Knowledge Hub, which connects emerging approaches with the operational conditions needed for lasting system change.
Evidence-led funding requires more than demonstrating that a device works or that a digital platform has been adopted. It requires understanding what changes in people's lives, how reliably those changes occur, which organizations benefit and whether the improvement justifies a sustainable funding commitment. This places pilot evaluation and structured learning at the center of payment reform.
The question is also closely connected to value-based payment and outcomes-led design. However, the transition from evidence to reimbursement is not automatic. States, health plans and providers must establish what is being purchased, who is eligible, how services are authorized, how outcomes are measured and who remains accountable when technology becomes part of everyday support.
1. Why Successful Care Technology Pilots Do Not Automatically Become Funded Services
Pilots are often designed to answer a relatively narrow question: can a particular technology improve a defined aspect of service delivery? Funding decisions require a wider answer. A Medicaid agency or managed care organization must consider whether the intervention is covered under an available authority, whether the expenditure is allowable, how it fits existing benefits and rates, and whether the evidence supports use beyond the original participants.
A remote monitoring pilot might show that a group of older adults experienced faster responses to changes in daily routines. A digital communication platform might improve coordination for people with complex support needs. Assistive technology could help someone with IDD manage aspects of their day with greater independence.
Each finding is valuable. None, by itself, establishes a reimbursable service.
The transition from a pilot to sustained payment normally requires answers to several distinct questions:
- Eligibility: Which people can appropriately benefit, and under what program or benefit authority?
- Effectiveness: What outcomes improve compared with an appropriate alternative?
- Cost: What is the full implementation and operating cost, including workforce and oversight?
- Accountability: Who delivers, monitors, maintains and evaluates the technology-enabled support?
- Payment: Which purchaser has authority and a defensible mechanism to pay?
These questions explain why scaling successful innovations is fundamentally different from completing a pilot. Scaling requires a repeatable operating model, reliable evidence and a financing structure capable of supporting the intervention over time.
A further difficulty is that pilot funding can conceal costs. Grant-funded project managers, donated equipment, temporary training support and supplier discounts may make an intervention appear less expensive than it would be under routine operating conditions. Evidence intended to influence payment must therefore reflect the cost of sustained delivery rather than the artificially favorable economics of an initial demonstration.
2. The Federal–State Funding Architecture Determines What Payment Is Possible
Medicaid offers several potential pathways for technology-enabled support, but it is not a single national reimbursement system. Federal law and regulation establish the framework within which states operate. States determine covered services, payment methodologies and administrative processes within their approved authorities, subject to applicable federal requirements.
For HCBS, a state may have relevant services within its Medicaid state plan, a Section 1915(c) waiver or another authorized arrangement. Depending on the approved program and individual circumstances, assistive technology, environmental modifications, remote supports or related services may be covered. Coverage descriptions, limitations, provider qualifications and authorization rules vary substantially.
Section 1115 demonstrations can support certain state-specific innovations where approved by CMS, but a demonstration is not a general permission to reimburse any emerging technology. Its approved terms, expenditure authorities, evaluation requirements and time limits matter.
Some states administer relevant services through fee-for-service Medicaid. Others use managed care organizations (MCOs) for some or all covered populations and benefits. Managed LTSS arrangements can create opportunities for plans to integrate technology into care coordination, network development or covered service delivery. However, the state contract, benefit design and plan responsibilities determine what is permissible.
Medicare presents a different financing structure. Certain technologies may be incorporated into covered clinical services, equipment benefits or eligible remote monitoring arrangements when the applicable requirements are met. That does not mean Medicare broadly pays for nonmedical community support technology or substitutes for Medicaid HCBS funding.
Commercial insurance, state general funds, aging-service grants, county resources, philanthropy and private payment may also contribute. The important distinction is between funding an innovation temporarily and establishing an enduring, legally supportable purchasing mechanism.
For state agencies, this is part of the wider challenge of designing funding and payment models that align service objectives, authorized expenditures and measurable public value.
3. Evidence Must Move Beyond Adoption and Activity
Technology suppliers commonly report installations, active users, completed assessments, alerts generated or staff time spent using a platform. These measures can establish implementation activity, but they are insufficient to demonstrate that the intervention improved care.
A service may achieve high digital adoption while producing little change in participant outcomes. Conversely, a modestly adopted intervention might deliver meaningful benefits for a narrowly defined group with substantial support needs.
The stronger evidence question is therefore not simply whether technology was used. It is whether its use changed decisions, practice, experience or outcomes in a way that matters.
For example, a digital medication support system could generate reminders and record acknowledgments. The meaningful outcomes might include improved medication routines, fewer preventable errors, greater confidence or reduced need for intrusive prompts. Those outcomes need to be assessed alongside incidents, participant feedback and the continuing role of staff and clinicians.
Evidence should distinguish:
- Activity evidence: whether the intervention was delivered and used.
- Implementation evidence: whether staff, participants and systems used it as intended.
- Outcome evidence: whether independence, safety, continuity or quality of life changed.
- Economic evidence: whether benefits justify the full cost under realistic operating conditions.
- Assurance evidence: whether results are reliable, equitable, sustained and subject to effective oversight.
Providers developing this approach can use the Quality Dashboard Builder to structure relevant indicators across quality, access, workforce and outcomes. The resource supports measurement design; it does not independently validate a technology's effectiveness or establish eligibility for reimbursement.
Crucially, evidence should include what did not improve. A technology that works for digitally confident participants but creates barriers for people with communication disabilities may require redesign rather than immediate expansion.
4. What Counts as Credible Evidence for a Payment Decision?
A payment decision does not always require the same evidence standard as a clinical trial. However, it does require evidence proportionate to the intervention's cost, risks, intended claims and potential consequences.
For a low-risk digital coordination tool, a well-designed operational evaluation may provide useful evidence about missed contacts, response times and service continuity. For technology that influences clinical escalation, safety-critical monitoring or high-stakes decisions, stronger validation and more intensive oversight may be necessary.
Evaluators should begin with a clear theory of change. If the intervention is intended to reduce avoidable emergency department use, what is the plausible pathway? Does it detect deterioration earlier, improve access to timely support or help prevent a breakdown in home care? Each proposed mechanism should be tested rather than assumed.
Comparison is particularly important. Participants may improve because staffing increased, a care coordinator became more involved or their underlying circumstances changed. Without an appropriate comparison, the technology may receive credit for improvements produced by other factors.
Randomized evaluation may be appropriate in some settings, but matched comparison groups, phased implementation, interrupted time-series analysis and carefully structured mixed-method evaluations can also provide useful evidence. The method should fit the intervention and decision being made.
Measurement also needs sufficient duration. A three-month pilot may establish feasibility while telling purchasers little about long-term maintenance, staff turnover, device replacement, sustained participant engagement or outcomes over a full year.
Through outcomes frameworks and indicators, organizations can define the difference between a promising operational signal and evidence strong enough to influence investment, contracting or rate-setting decisions.
5. Scenario: Remote Support Technology for a Person with IDD
Consider an illustrative provider supporting an adult with IDD who wants greater privacy and independence in their apartment. The person currently receives scheduled in-person support during parts of the evening. A remote support system could allow them to request assistance when needed, use agreed prompts for routine activities and maintain contact with a trained responder.
The person and their support team explore the option through person-centered planning. They agree which activities the technology may support, when in-person assistance remains necessary and how the person can withdraw consent or request a different arrangement. A guardian, where legally relevant, participates within the limits of their authority without displacing the person's own preferences.
During a six-month pilot, the provider records response times, missed support, participant experience, staff interventions and incidents. The person reports greater privacy and confidence. The provider also finds that remote responders require more training and backup coverage than originally anticipated.
The funding challenge emerges when the pilot ends. The technology is beneficial, but the relevant state waiver's service definitions and authorization rules determine whether the equipment, monitoring service and related staff time can be reimbursed. The provider cannot simply substitute a technology charge for authorized direct support hours.
A stronger proposal presents the person's outcomes, full operating costs, workforce safeguards and the specific payment authority under consideration. It also explains how the arrangement preserves choice and how in-person support will be restored if the technology becomes unsuitable.
The transferable lesson is that evidence should establish a viable support model, not merely justify purchasing a device. This is especially important for supported decision-making and autonomy, where financial efficiency must never become the sole basis for changing someone's support.
6. From Pilot Economics to Sustainable Rate Design
Once effectiveness has been demonstrated, the next challenge is determining how the intervention should be paid for. The answer depends partly on what the technology actually provides.
A device purchased once has a different cost profile from a continuously staffed remote support service. A care coordination platform may be embedded within an existing administrative rate, purchased through a contract or supported through an authorized payment arrangement. A technology-enabled service involving clinical staff may raise different reimbursement and professional requirements.
Rate design should reflect the complete cost structure. This may include licensing, connectivity, equipment, installation, training, technical support, cybersecurity, maintenance, replacements, supervision, evaluation and contingency arrangements. Workforce costs remain important even where the technology reduces certain manual tasks.
Purchasers should also distinguish between costs that are fixed and those that vary with the number of participants. A small rural provider may face relatively high implementation costs per person. A large organization may achieve economies of scale but incur additional integration and governance expenses.
The implications for rate-setting and cost modeling are significant. If payment is based on unrealistic assumptions about utilization or staff substitution, providers may be unable to sustain the service. If rates are too generous relative to demonstrated value, public funds may be diverted from other essential supports.
State Medicaid agencies and their actuaries, where relevant, must consider applicable rate and financing requirements. In managed care, payment arrangements also need to be understood within state contracts and the broader framework governing plan payments and covered benefits.
A useful funding model therefore separates the cost of implementing the intervention from the ongoing cost of delivering it. Temporary implementation funding may be justified even when the longer-term service is expected to operate within a stable rate.
7. How Value-Based Payment Could Change Technology Investment
Traditional fee-for-service arrangements generally reward defined units of covered activity. Technology that prevents avoidable activity or improves coordination may create benefits that are difficult to recognize within a payment system focused on individual encounters or hours.
Value-based payment offers a possible alternative by connecting some reimbursement to outcomes, quality or performance. It could support technology investments that improve continuity, reduce avoidable utilization or strengthen independence, provided the payment design recognizes the relevant intervention and the provider's contribution.
However, the financial incentives are not always aligned. A community provider may pay for technology while a health plan or hospital realizes savings. A state may invest in a service that reduces costs elsewhere in the public system. Participants may experience meaningful improvements in quality of life that do not translate into immediate healthcare savings.
These issues make cost-versus-outcome analysis more important than a narrow calculation of avoided claims.
Possible payment arrangements include quality incentives, prospective technology-enabled service payments, negotiated case rates, shared savings or broader provider contracts that support an agreed service model. Each depends on the applicable program, purchasing authority, contractual structure and evidence available.
Shared savings deserves particular caution. Savings must be measurable, appropriately attributed and calculated against a credible baseline. Providers should not be exposed to substantial financial risk for outcomes dominated by housing conditions, hospital decisions, local service shortages or other factors beyond their control.
Value-based arrangements also need safeguards against selecting only people most likely to achieve favorable outcomes. Risk adjustment, transparent eligibility criteria, minimum access standards and monitoring of disparities help prevent payment incentives from undermining equity.
Technology becomes more financially credible when payment recognizes both its contribution to measurable outcomes and the legitimate cost of maintaining safe, person-centered support.
8. Managed Care Organizations as Purchasers, Evaluators and System Partners
Where states use managed care for relevant Medicaid populations and services, MCOs may have an important role in evaluating technology-enabled care. Plans can examine utilization patterns, care coordination performance, member experience and network capacity across groups of participants.
This creates opportunities for identifying where technology could improve service delivery, but plans must distinguish between benefits they are responsible for providing and activities funded through other authorities. State contracts, federal requirements and applicable payment rules shape their options.
A plan considering technology investment should ask whether the intervention addresses a demonstrated need within its covered population. It should also consider whether the provider network has the capacity to deliver the model, whether data can be shared appropriately and whether the intervention creates additional administrative burden.
For providers, the central operational concern is often that each plan establishes a different process. Separate reporting specifications, portals, performance definitions and payment conditions can increase costs and undermine consistent implementation.
Stronger contract management and provider performance arrangements can reduce this problem by defining common responsibilities, clear evidence requirements, realistic reporting schedules and appropriate escalation routes.
Plans should also distinguish a provider's service performance from the technology supplier's performance. A failed alert caused by a supplier outage should trigger a coordinated review of the technical system, provider response arrangements and member safety. It should not disappear within a general provider quality score.
State oversight remains essential. Delegating operational functions to an MCO does not remove the state's responsibilities under its Medicaid program.
9. Scenario: A Managed LTSS Plan Tests Predictive Home Care Scheduling
Consider an illustrative managed LTSS plan experiencing recurring missed home care visits across part of its provider network. A technology supplier proposes predictive scheduling software that identifies potential coverage gaps using staff availability, travel patterns and historical cancellations.
The plan funds a limited pilot with several participating agencies. Early results suggest that supervisors can identify scheduling problems sooner and arrange alternative coverage before a visit is missed.
However, the initial data also show that rural members receive fewer improvements because replacement workers are less readily available. The software predicts a shortage accurately but cannot create workforce capacity.
The plan and providers therefore expand the evaluation beyond scheduling alerts. They examine completed visits, participant complaints, travel time, overtime, staff retention and the number of unresolved gaps. Members are asked whether they experience more reliable support rather than simply whether they like the technology.
Financial analysis shows that the intervention reduces some avoidable administrative work, but meaningful improvement requires additional workforce coordination and backup staffing. A payment arrangement based only on software licensing would understate the actual cost of service continuity.
The plan considers a time-limited provider incentive linked to independently checked continuity measures, alongside support for implementation. It also maintains minimum service expectations and investigates whether incentives could disadvantage agencies serving remote areas.
The example illustrates why a successful technology pilot may lead to a revised operating model rather than straightforward technology reimbursement. Evidence identifies the combination of software, staffing and purchasing arrangements needed to improve outcomes.
10. The Workforce Is Part of the Intervention, Not an External Cost
Care technology frequently changes how staff work rather than eliminating the need for staff. Remote monitoring requires responders. Digital care coordination requires accurate records and timely follow-up. Assistive technology requires assessment, installation, support and periodic review.
When workforce requirements are excluded from the original pilot, financial projections can become misleading. A platform may reduce duplicate documentation while increasing supervisory review or technical support. It may also create new responsibilities for DSPs, home care aides, case managers and clinical staff.
These changes should be evaluated through workforce innovation and role redesign, rather than assuming that technology creates automatic labor savings.
Provider executives need to understand whether workers have sufficient training, protected implementation time and access to support when systems fail. Competence should be demonstrated through practice observation, case review, appropriate escalation and participant feedback, not simply completion of an online module.
There is also a retention dimension. Technology that reduces unnecessary paperwork or improves scheduling may strengthen job satisfaction. Poorly implemented systems can instead increase surveillance concerns, documentation pressure and frustration.
For payment purposes, these workforce effects matter because sustained quality depends on the human infrastructure surrounding the technology. A lower-cost intervention that creates unstable staffing or delayed responses may not represent better value.
11. Governance, Privacy and Supplier Accountability
As care technology moves from pilot status into routine delivery, governance expectations should become more demanding rather than less. A temporary project may operate with intensive oversight from a small implementation team. A funded service must remain safe and accountable across ordinary staffing changes, supplier updates and changing participant needs.
Organizations should establish who owns clinical or operational decisions, who investigates incidents, who approves changes to the system and who is responsible when a supplier fails to meet agreed standards.
Privacy obligations depend on the organizations, information and services involved. HIPAA may apply to covered entities and business associates, while other federal and state privacy requirements may also be relevant. Behavioral health and substance use information can require additional safeguards, including where 42 CFR Part 2 applies.
Sound data governance and information accountability should address consent, access controls, data retention, security incidents, information quality and the ability to explain how information influences decisions.
The Digital Transformation, AI and Cybersecurity Readiness Assessment can support leadership teams in examining these operational and supplier risks before expanding a technology-enabled service.
Where predictive analytics or AI are involved, human review remains essential. A model that identifies elevated risk should support professional assessment, not independently determine eligibility, remove services or override the person's preferences.
Payment arrangements should make these responsibilities explicit. A supplier's performance claims cannot replace provider governance, and a provider contract cannot eliminate statutory or regulatory obligations.
12. Scenario: Behavioral Health Technology and the Limits of Avoided-Cost Claims
Consider an illustrative community behavioral health organization supporting adults with serious mental illness. The provider pilots a digital check-in platform that allows participants to report changes in wellbeing and request earlier contact from their support teams.
Participants value being able to communicate between scheduled appointments. Staff identify several emerging concerns sooner, and the organization records improvements in follow-up times. A technology supplier suggests that the platform could substantially reduce emergency department use and inpatient admissions.
The provider's quality team examines the evidence more carefully. Hospital utilization is influenced by many factors, including medication access, housing stability, crisis-service availability and the severity of individual circumstances. The pilot is too small to establish a reliable reduction in hospital spending.
Instead of claiming unproven savings, the organization develops a stronger evaluation around timely response, continuity of engagement, participant-reported recovery outcomes, crisis escalation and service accessibility. It also examines whether people without smartphones or reliable internet access are excluded.
The organization discusses future funding with relevant purchasers, recognizing that Medicaid behavioral health coverage, managed care arrangements, state funding and grant conditions vary. The technology may be supported as part of an existing service model or through a separately authorized arrangement, but no universal payment pathway can be assumed.
The lesson is that credible evidence sometimes supports a narrower claim than a supplier initially proposes. That restraint strengthens the case for investment because it separates demonstrated value from hoped-for financial returns.
13. Measuring Community Value Without Reducing Everything to Healthcare Savings
Some of the most important benefits of community-based technology are not easily captured through medical claims. A person may gain confidence using public transportation, communicate more independently, remain connected with friends or participate more fully in employment and community life.
These outcomes can be central to the purpose of HCBS and disability services, even where they do not produce immediate reductions in hospital or emergency department spending.
Purchasers therefore need a broader understanding of value. This may include participant-defined goals, caregiver impact, continuity, community participation, housing stability, workforce sustainability and equitable access.
The distinction is particularly important when evaluating community impact and wider system value. An intervention that supports independence may be worthwhile even where its direct financial return is uncertain.
Organizations can use the Community Impact Report Builder to organize evidence about access, outcomes, partnerships and community benefit. Such reporting can strengthen grant applications, payer discussions and strategic investment cases, but it should distinguish verified results from projections.
Caregiver outcomes also require careful treatment. A digital monitoring system may provide reassurance and reduce some coordination tasks. It should not be used to justify transferring essential responsibilities to unpaid relatives without agreement, appropriate support or consideration of caregiver burden.
Meaningful evaluation asks whether people experience greater control over their lives, not simply whether services become less expensive to deliver.
14. Scenario: A Rural Aging Services Network Builds a Sustainable Funding Case
Consider an illustrative rural aging services network working with home care agencies, a community health organization and a state-funded aging program. Travel distances make routine in-person monitoring expensive, and some older adults experience delays when support needs change.
The network pilots a combination of accessible communication devices, scheduled remote check-ins and improved coordination between participating organizations. Participants can choose whether to use the technology, and in-person visits remain available where needed.
After nine months, the evaluation identifies faster responses to nonemergency concerns and improved communication with caregivers. It also reveals substantial differences between communities with reliable broadband and those with poor connectivity.
The network initially considers expanding the model through a single technology contract. Further analysis shows that devices alone are insufficient. The intervention depends on local staff who can respond, maintain equipment and arrange in-person assistance when remote contact is unsuitable.
Its funding proposal therefore distinguishes technology costs, workforce costs and potential savings from more efficient coordination. It identifies which activities might be supported through existing authorized services and which would require additional state, grant or other funding.
The network also establishes a common evaluation framework so participating agencies can compare results without ignoring differences in geography and participant need.
The example demonstrates why rural access and service equity should influence payment design. A model that performs well in urban settings may require different rates, implementation support or service arrangements to remain viable in rural communities.
15. Turning Evidence Into a Purchasing and Implementation Decision
Evidence does not create a payment pathway by itself. It informs a series of decisions about authority, affordability, service design, contractual responsibility and acceptable risk.
Before committing to ongoing funding, a state agency, MCO or other purchaser should be able to explain why the intervention is needed, what alternatives were considered and how the proposed arrangement fits the relevant program.
Providers should be able to demonstrate that the intervention can operate beyond the original pilot team. That includes supervision, training, technical support, incident management, participant communication, data governance and service continuity when equipment or connectivity fails.
A credible purchasing decision normally requires four connected forms of assurance:
- Evidence assurance: the claimed outcomes are supported by appropriate and sufficiently reliable data.
- Financial assurance: the full cost, affordability and payment assumptions are transparent.
- Operational assurance: providers can deliver the model safely and consistently.
- Rights and equity assurance: participants retain meaningful choice, accessibility and appropriate protections.
The Governance Maturity Assessment can help provider leaders examine whether accountability, escalation and assurance arrangements are sufficiently developed to support expansion.
Provider boards should receive more than adoption statistics and favorable testimonials. They need information about actual outcomes, cost variation, unresolved incidents, participant concerns, supplier dependence and the conditions under which the organization would pause or withdraw the intervention.
For purchasers, the same discipline should shape contract monitoring. Performance expectations should be achievable, proportionate and linked to outcomes that the contracted organizations can reasonably influence.
16. The Next Stage: Evidence-Linked Funding Rather Than Permanent Pilot Cycles
The future opportunity is not simply to create more care technology pilots. It is to develop clearer routes through which useful innovations can be evaluated, improved and considered for sustainable payment.
Some states and purchasers may increasingly use staged investment models: initial feasibility funding, a defined evaluation period, conditional expansion and longer-term purchasing where evidence and authority support it. This is a plausible direction for system development, not a uniform national policy.
More sophisticated approaches could connect integrated funding pilots with outcome measurement, implementation cost analysis and coordinated decisions across health and human services. However, separate funding authorities, federal requirements and organizational responsibilities will continue to constrain what can be combined.
Interoperability may also improve the evidence available to purchasers. Better connections between service records, claims, encounter data and participant-reported outcomes could help identify where technology adds value. Data-sharing arrangements must nevertheless respect privacy, consent, data quality and the limits of attribution.
Future payment models may place greater emphasis on maintaining outcomes over time rather than rewarding short-term improvements. This would encourage purchasers to examine whether benefits persist after supplier support is reduced, staff change and participants' needs evolve.
AI and predictive analytics may support earlier identification of operational risks, but their use in payment decisions will require careful validation, transparency and oversight. Automated analysis should not become an opaque mechanism for restricting access or shifting financial risk onto people with greater support needs.
One of the strongest opportunities lies in designing evidence requirements before a pilot begins. When evaluators, providers, purchasers and people receiving services agree what success means, the eventual funding decision becomes more informed and less dependent on persuasive but incomplete claims.
Conclusion
The movement from care technology pilots to sustainable payment is fundamentally a question of evidence, authority and accountability. A promising intervention may improve everyday support without fitting an existing reimbursement structure. Equally, a technically successful pilot may fail to demonstrate sufficient value once its full costs, implementation demands and effects on different populations are examined.
Across the United States, federal Medicaid frameworks establish important boundaries, while state program design, approved authorities, managed care contracts and provider arrangements determine how technology-enabled services can be funded in practice. Medicare and other funding sources create additional possibilities, but none provides a universal solution.
The strongest approach connects participant-defined outcomes with credible evaluation, realistic cost modeling and operational assurance. It recognizes that technology does not operate independently of workers, supervisors, care coordinators, clinicians or the people whose lives it is intended to improve. Payment arrangements should reward demonstrable benefit without encouraging inappropriate service reduction, exclusion or unsupported claims of savings.
For providers, this means treating evaluation as part of service design rather than an exercise completed at the end of a grant. For states and health plans, it means developing purchasing decisions that recognize both the value and the limitations of emerging models. For boards and executives, it means challenging evidence, understanding financial exposure and retaining accountability as innovations scale.
The most significant change will occur when evidence no longer serves merely to celebrate successful pilots, but helps determine which interventions deserve lasting investment, under what conditions and for whose benefit. That is how care technology can move from temporary experimentation toward sustainable, person-centered community support.