The United States healthcare system is experiencing a profound shift towards delivering care in the home, a trend accelerated by an aging population, patient preference, and policy incentives aimed at reducing costs and improving outcomes. Within this context, home health providers are at a critical juncture, compelled to innovate beyond traditional service delivery models. The integration of advanced technologies, particularly artificial intelligence (AI) combined with remote patient monitoring (RPM), presents a pivotal opportunity to redefine care delivery. AI-enabled RPM systems can continuously collect and analyze patient data, offering the potential to prevent adverse events, personalize treatment, and enhance operational efficiency.
However, the mere adoption of technology is insufficient for sustainable success. Providers must fundamentally innovate their business models to translate technological capabilities into clinical and financial value. This paper addresses a critical gap in the literature by conducting a comparative analysis of how U.S. home health providers are innovating their business models through AI-enabled RPM. The study focuses specifically on the evolution of the value proposition and the corresponding adaptation of revenue streams, which are foundational to any viable business model. The objectives are to identify the primary AI applications driving this change, analyze how new value is created and captured financially, and discuss the significant challenges that providers face in this transition.
The success of technology-driven healthcare services hinges on well-defined business models. A systematic review of telehealth business models underscores that the value proposition and financial components, such as revenue streams, are the most critical factors for viability. Many telehealth initiatives fail not because of technological shortcomings but due to an inability to create and demonstrate tangible value within specific social and economic contexts.1 AI introduces a powerful mechanism for value creation in healthcare by generating new medical insights from data and by automating key activities across the value chain. This enables business model innovation, facilitating new revenue opportunities like Software as a Service (SaaS) or data licensing, though it also introduces new cost structures related to specialized personnel and infrastructure.2
AI-driven RPM has demonstrated a significant economic impact. Studies show its implementation can reduce hospital stay expenses by at least 30% and emergency room usage by 25%, with readmission rates for chronic disease patients decreasing by up to 30% annually.3 The business case is further supported by direct revenue opportunities; providers can bill over $1,300 per patient per year using established RPM CPT codes, while simultaneously improving clinical outcomes. For instance, one cardiology practice reduced 30-day heart failure readmissions from 15% to 7% after implementing an RPM program.4 Predictive AI platforms have been shown to reduce hospitalizations by 40% and falls by 69% in senior living communities.5 Similarly, a nurse-led RPM program at Mayo Clinic for nearly 22,000 patients demonstrated that engaged participants had significantly lower rates of 30-day hospitalization, ICU admission, and mortality.6
This technological shift aligns with major policy changes, most notably the nationwide expansion of the Home Health Value-Based Purchasing (HHVBP) Model. This model, which generated an average of $141 million in annual Medicare savings, directly rewards agencies for quality improvements and cost reductions, particularly in preventing unplanned hospitalizations.7 Consequently, technologies like RPM are considered essential tools for providers to succeed under HHVBP.8 Research indicates that for-profit and non-profit agencies respond differently to these incentives, with for-profits showing greater increases in care quality metrics and non-profits showing greater gains in patient experience scores.9
Despite the potential, the implementation of AI-enabled RPM is fraught with challenges. A systematic review identified major barriers perceived by practitioners, including increased workload from large data volumes, data inaccuracy, poor integration with electronic medical records (EMRs), high costs, and patient privacy concerns.10 Other significant obstacles include ensuring the safety and accuracy of algorithms, navigating strict regulatory compliance, and addressing ethical issues like algorithmic bias.11 From an operational perspective, providers struggle with patient adherence to device protocols and the complexities of reimbursement requirements.12 The successful integration of AI into RPM systems requires overcoming these multifaceted challenges related to technology, workflow, security, and reimbursement models.13
This study employs a multi-case comparative analysis framework to investigate business model innovation among U.S. home health providers adopting AI-enabled RPM. This approach allows for a nuanced examination of how different organizational and market contexts influence strategic choices and outcomes. The analysis is guided by a multi-faceted set of selection criteria designed to capture the heterogeneity of the U.S. home health market.
The core of the methodology involves comparing providers based on the following key criteria:
The analytical lens focuses primarily on two interconnected components of the business model: the evolution of the value proposition and the adaptation of revenue streams. The investigation centers on three specific AI applications within RPM: predictive analytics for patient deterioration, AI-driven personalization of care plans, and Natural Language Processing (NLP) for patient communication and data intake.
The analysis of U.S. home health providers reveals that AI-enabled RPM is a catalyst for significant business model innovation, primarily by transforming the core value proposition, creating new revenue streams, and optimizing cost structures.
The most fundamental innovation is the shift from a reactive, episodic model of care to a proactive, continuous, and personalized one. This enhanced value proposition is delivered through specific AI applications.
Providers are using AI to analyze continuous data streams from RPM devices to forecast patient deterioration and enable preemptive interventions. For example, platforms like CarePredict have demonstrated a 40% reduction in hospitalizations and a 69% decrease in fall rates by tracking daily activity patterns.5 Other platforms, such as Revella Health’s ARIA, use AI to analyze clinical and social determinants of health data to reduce avoidable emergency room visits and hospitalizations, directly improving performance on quality metrics critical for value-based care contracts.16 This capability is crucial for success in the HHVBP model, where reducing acute care hospitalizations is a key performance indicator.8
AI enables the move beyond static, one-size-fits-all care plans. Machine learning algorithms analyze a patient’s unique data—including vitals, lifestyle, and treatment responses—to dynamically tailor recommendations. For instance, value-based cardiology providers like Ventricle Health utilize RPM alongside health coaching and digital touchpoints to manage heart failure patients with personalized, guideline-directed medical therapies.16
Generative AI and NLP are enhancing both clinical workflows and patient communication. Ambient clinical intelligence tools are automating documentation, with solutions from Abridge reportedly reducing physician charting time by 74%.17 This automation of administrative tasks frees up clinical time for direct patient care.13 Furthermore, AI-powered chatbots can provide 24/7 patient support, answer questions, and collect symptom data, creating a more complete and usable health profile.
The enhanced value proposition creates opportunities for new and diversified revenue streams, moving providers beyond traditional per-visit reimbursement.
The Centers for Medicare & Medicaid Services (CMS) has established specific CPT codes for RPM, allowing providers to bill for services like initial device setup (CPT 99453), monthly data monitoring (CPT 99454), and clinical staff time (CPT 99457 and 99458).18 This creates a direct, recurring revenue stream that can exceed $1,300 per patient annually, providing a clear business case for RPM adoption.4
AI-enabled RPM is a critical tool for succeeding in VBC arrangements. By demonstrably reducing hospital readmissions and improving quality metrics, providers can earn bonuses and share in cost savings. Companies like DispatchHealth partner directly with risk-bearing entities to provide in-home acute care, leveraging their model to reduce costly hospitalizations.16 The HHVBP model itself, which led to $141 million in annual Medicare savings, institutionalizes this link between quality outcomes and financial reimbursement.7
Policy innovations are creating entirely new business models. The CMS ‘Acute Hospital-At-Home’ program, for example, allows hospitals to treat acute conditions at home while receiving payment equivalent to traditional inpatient care, establishing a new financial framework for advanced home-based services.19
Beyond revenue generation, AI-RPM drives financial viability by significantly reducing costs. Studies indicate that home health AI can lower a practice’s operational costs by nearly 40%20 while cutting hospital stay expenses by at least 30% and ER use by 25%.3 Economic evaluations have quantified a positive return on investment (ROI). A study of an RPM program for hypertension found a 22.2% ROI, even after accounting for significant costs related to data review by nurse practitioners.21 In another case, a remote care program for congestive heart failure patients reduced 30-day readmissions to zero for the enrolled group, resulting in a cost avoidance of $216,000.22
The findings illustrate that AI-enabled RPM is not merely an incremental technological upgrade but a disruptive force reshaping the fundamental business models of home health providers. The innovation pivots on a re-envisioned value proposition: care that is continuous, predictive, and personalized. This shift directly addresses the core tenets of value-based care, enabling providers to move away from a model compensated by volume of visits to one rewarded for quality of outcomes. The analysis confirms that the value proposition and revenue streams are intrinsically linked; the ability to prevent hospitalizations and manage chronic conditions more effectively (value creation) unlocks new fee-for-service billing opportunities and facilitates success in risk-bearing contracts (value capture).
The differentiation in how for-profit and non-profit agencies respond to VBC incentives suggests that ownership models may influence the pace and focus of AI-RPM adoption.9 For-profit agencies, which constitute a growing share of the market,15 may be quicker to adopt technologies that demonstrate a clear ROI and improve performance on financially weighted quality metrics. Non-profits may prioritize investments that enhance patient experience and community health outcomes. This divergence warrants further investigation into how strategic goals shaped by ownership status affect technology implementation and business model design.
While the potential is clear, providers must navigate significant implementation barriers to realize the benefits of AI-RPM. The research points to several key strategies for mitigating these challenges.
A primary complaint from practitioners is being overwhelmed by the sheer volume of data generated by RPM devices.10 A recommended strategy is to implement a tiered or “drill-down” data presentation framework. This approach provides clinicians with a high-level summary and critical alerts first, allowing them to access more granular data only when necessary, thereby prioritizing actionable insights and preventing cognitive overload.23
The lack of seamless integration with existing EMR systems is a major operational hurdle.10 The most effective solution is to embed RPM data directly into the EMR workflow. This streamlines processes, reduces the need for separate logins, and enables the creation of powerful clinical decision support tools that leverage a unified patient dataset.23
Programs are often hindered by patients who forget to use or charge devices.12 To improve adherence, providers should involve patients in the co-design of their monitoring plans and prioritize passive data collection methods (e.g., wearable sensors) over active ones that require frequent patient input, as passive methods are less likely to cause patient burnout.23
The high upfront cost and complex reimbursement landscape remain significant barriers.12 However, policy is evolving to address this. For example, CMS has expanded reimbursement pathways for safety-net providers like Federally Qualified Health Centers (FQHCs) and Rural Health Clinics (RHCs), allowing them to bill for RPM services.24 Furthermore, providers can bill for RPM concurrently with other care management programs, such as Chronic Care Management (CCM), creating bundled revenue opportunities, provided that time is not double-counted.25
The integration of artificial intelligence into remote patient monitoring is fundamentally reshaping the business models of U.S. home health providers. By enabling a transition to proactive, personalized, and continuous care, AI-RPM enhances the core value proposition offered to patients and payers. This innovation unlocks new financial viability through direct fee-for-service revenue streams and, more critically, provides the essential toolkit for succeeding in a healthcare ecosystem increasingly dominated by value-based payment models. While formidable challenges related to workflow integration, data management, patient adherence, and cost remain, strategic implementation can mitigate these obstacles.
For home health agencies to thrive, they must move beyond viewing AI-RPM as a standalone technology and instead embrace it as the cornerstone of a new, data-driven business model centered on delivering and demonstrating value. Future research should focus on longitudinal studies tracking the long-term ROI of these models, conducting more granular comparisons of adoption patterns and outcomes between for-profit and non-profit providers, and exploring the emerging impact of generative AI on both administrative efficiency and the quality of patient-provider interactions.
