Real-world data (RWD) – e.g. electronic health records, claims, registries, and patient-generated data – are increasingly leveraged to generate real-world evidence (RWE) about medical products and care delivery. RWE (clinical evidence derived from RWD) can accelerate drug development, inform care decisions, and improve patient outcomes beyond traditional trials. Policy and regulatory momentum over 2020–2025 has fostered this shift, and experts project even broader integration by 2030 (Table 1).
Policy and Regulatory Trends (2020–2025, with 2030 Outlook)
Legislative Drivers: The 21st Century Cures Act (2016) mandated FDA to develop an RWE framework. In 2020, ONC issued its Cures Act Final Rule, banning information blocking and requiring standardized APIs (USCDI/FHIR) to boost interoperability. By 2023, ~70% of U.S. hospitals routinely engaged in all four core data-exchange domains. CMS’s 2020 Patient Access Final Rule similarly pushed insurers to share claims/EHR data via APIs.
FDA Initiatives: Since 2021, FDA has released multiple guidance documents on RWD/RWE for drugs and biologics. PDUFA VII (2022) formally committed FDA to an “Advancing RWE Program” (improving quality/acceptability of RWE for label expansions and post-market studies). In 2023, FDA funded research grants (U01s) to improve RWD quality for regulatory use. The FDA’s Digital Health Center of Excellence and recent draft guidances on AI/ML in devices further signal regulatory support for AI-enabled RWD analysis.
CMS Actions: CMS is exploring RWE in coverage decisions. In late 2024 it proposed guidance on RWD-based study protocols for Medicare’s Coverage with Evidence Development (devices under NCDs). The 2022 Inflation Reduction Act requires CMS to use robust evidence (including comparative effectiveness/RWE) in Medicare drug price negotiations. CMS’s Innovation Center also pilots models that rely on real-time data (e.g. remote patient monitoring).
Interoperability Efforts: ONC’s ongoing work – including finalizing the Trusted Exchange Framework and Common Agreement (TEFCA) and updating information-blocking rules – aims to create a nationwide “network of networks” for health data exchange. By 2024, millions of patients routinely access their records via apps, enabling new RWD sources (wearables, patient portals).
Projections to 2030: Experts anticipate learning health systems where every clinical encounter feeds RWD into continuous analytics. By 2030, RWE may be routinely used for initial drug/device approvals (especially for rare diseases) and to drive personalized care. Broad AI regulation frameworks (proposed FDA AI action plans) and global harmonization (ICH’s 2024 RWD reflection paper) are expected to be in place. In sum, future policies will likely emphasize data quality standards, patient privacy, and algorithmic transparency.
Table 1. Key policy/regulatory milestones (2020–2024) and future outlook for RWD/RWE.
Implementation of RWD/RWE in Clinical Settings
Healthcare systems and industry are actively deploying RWD/RWE in practice:
Hospitals and Health Systems: Large networks harness EHR and claims data for quality improvement, clinical research, and performance benchmarking. For example, multicenter hospital collaboratives (PCORnet, H2O) use longitudinal clinical data to study outcomes. Biosensors and wearables (e.g. continuous monitors in ICUs) feed live data into hospital databases. Physicians routinely consult RWE (e.g. registry-based treatment comparisons) to inform care. In one national survey, 65% of U.S. hospitals reported using AI/predictive models integrated with their EHR to guide clinical and administrative decisions. (These models predict patient deterioration, optimize resource use, etc.) Even so, only ~61% of those hospitals locally test model accuracy and 44% test for bias.
Pharmaceutical Industry: Companies are increasingly using RWD in drug development. RWD from prior treatment episodes and registries can supplement clinical trials (e.g. external control arms for oncology), informing benefit–risk profiles. RWE has enabled label expansions, especially in rare/ultra-rare diseases (regulators have accepted registry data in orphan drug approvals). AI-driven analytics on RWD accelerate target discovery and patient cohort identification. RWE also plays a role in value assessments and post-market safety monitoring. Collaborative initiatives like TransCelerate’s Real World Data Initiative and FDA’s RWE grants encourage data-sharing among biopharma. Table 2 compares sector trends.
MedTech and Digital Health: Device manufacturers leverage RWD from implantable devices, wearables, and home monitoring. For instance, cardiac device registries and telemetry data inform long-term device performance and help tailor therapies. FDA’s Digital Health initiative has encouraged using real-world registries in device evaluations. The Medical Device Innovation Consortium (MDIC) reports that embedding RWE throughout product development “offers more reliable evidence than traditional models,” speeding access and improving care. MDIC has published an RWE strategy guide to help firms use RWD end-to-end. Telehealth data (e.g. remote blood pressure logs) and smartphone health apps are emerging RWD sources in this sector.
Figure: Illustration of AI-assisted patient monitoring capturing real-time health data (illustration). AI systems can continuously analyze vital signs to alert clinicians about conditions like sepsis.
Table 2. Comparison of RWD/RWE and AI use across sectors.
AI Integration for Real-Time Decision-Making
Artificial intelligence (AI) is increasingly driving real-time clinical decisions using RWD. AI-powered decision support systems are being embedded in EHRs and medical devices to analyze data on the fly:
Clinical Decision Support: In critical care, AI models can continuously analyze streams of vitals and labs. A prominent example is the Johns Hopkins “Targeted Real-time Early Warning System” (TREWS) for sepsis: it scanned EHR data to identify sepsis hours before standard clinical recognition, reducing patient mortality by ~20%. The system was used by thousands of clinicians at the bedside, with explainable alerts that improved treatment timing. Many health systems now deploy similar predictive alerts (e.g. for acute kidney injury, cardiac arrest), enabled by improved data capture. A 2023 survey found AI models integrated into EHRs are used for both care decisions and administrative tasks by 65% of hospitals.
GenAI and Workflow: Generative AI models (large language models) are being tested for summarizing patient charts, drafting notes, and retrieving medical knowledge. Recent surveys show 68% of U.S. physicians believe GenAI can save time by rapidly searching literature, and 59% say it could summarize EHR data for clinical insights. About 54% expect a ≥20% time savings on care processes, and 81% think GenAI will improve team–patient communication by providing timely information. These tools are in pilots, but clinicians emphasize the need for transparency – e.g. knowing the source of AI-generated content.
MedTech and Imaging: AI algorithms are embedded in medical devices and equipment. For example, AI-enhanced imaging (MRI, CT, X-ray) can provide real-time interpretation to radiologists. Smart infusion pumps and ventilators adjust settings based on continuous analytics. Robotics (surgical robots) increasingly incorporate AI for real-time guidance. In all cases, the underlying RWD – sensor outputs, telemetry – feed the AI in closed-loop decision systems.
Challenges: Broad AI adoption brings challenges of data quality, interoperability, and bias. Many hospitals lack mature processes for validating models on local data. Regulators (FDA) are drafting guidance to ensure transparency and safety of AI/ML-based medical software. Nonetheless, the trend is clear: AI is moving from pilot to practice, enabling decisions (e.g. sepsis alerts, image diagnosis) in real time.
Impact on Patient Outcomes
Evidence to date suggests RWD/RWE and AI can improve patient outcomes:
Improved Survival and Safety: The sepsis AI example demonstrates how RWD-driven AI can save lives (20% mortality reduction). In chronic disease management, remote monitoring algorithms detect exacerbations early (e.g. predicting heart failure decompensation), reducing hospitalizations. RWE has supported expanding indications of effective therapies (e.g. oncology drugs) to broader populations by confirming benefits in routine care settings. Overall, experts argue that integrating RWE yields “faster delivery” of innovations and “improved patient outcomes”.
Faster Access to Therapies: Using RWE can shorten development times. For instance, external control arms (using real patients’ past data) are reducing the need for large placebo groups in rare disease trials. This accelerates market access for new treatments. Similarly, RWE from registries helps identify subgroups that benefit from a drug, leading to personalized care.
Value and Costs: By informing care decisions, AI and RWE can improve quality and efficiency. Physicians expect GenAI to save time on routine tasks and hospitals report more efficient resource use. Payers and health systems anticipate that predictive analytics will optimize interventions (e.g. targeting high-risk patients for preventive care), ultimately lowering costs. While long-term studies are ongoing, early signals (like reduced lengths of stay or readmissions through predictive alerts) suggest a positive impact on the care “quadruple aim.”
Regulatory Developments (FDA, CMS, ONC)
Federal agencies continue to codify the shift toward RWD and real-time data:
FDA: In addition to RWE guidances, FDA has issued multiple frameworks around AI. In 2022, FDA released an action plan for AI/ML-based medical software (including periodic reviews of adaptive algorithms). The agency’s “Advancing RWE Program” (PDUFA VII) explicitly aims to integrate RWE into labeling and post-market studies. FDA’s oncology center (OCE) and others regularly accept registry data and modeling in new drug applications. Guidance is also evolving for electronic data (e.g. EHR in clinical investigations).
CMS: CMS has formally embraced evidence development using RWD. It recently proposed clear protocols for Medicare coverage decisions that rely on RWD studies. CMS is also incorporating RWE into alternative payment models and quality programs. With IRA drug negotiations commencing, CMS will require manufacturers to present evidence (potentially RWE) on comparative effectiveness. Separately, CMS’s Interoperability Rule enables patients to receive (and share) their data, indirectly supporting RWD use.
ONC and HHS: The Office of the National Coordinator enforces interoperability rules (e.g. 21st Cures Act regs). Its annual reports show steep increases in data exchange: by 2023, 70% of hospitals routinely exchanged clinical data with external providers. ONC also published an interoperability progress report (2023) tracking growing API use. Upcoming ONC priorities (TEFCA implementation, HITAC 2023 recommendations) further embed RWD flow across healthcare. HHS’s Office for the Advancement of Telehealth and ONC’s recent roadmaps emphasize using RWD for pandemic/epidemic preparedness as well.
Overall, regulators have moved from exploratory (pre-2020) to operationalizing RWE. The ongoing issuance of guidances and rules by FDA, CMS, and ONC reflects a trajectory toward data-driven oversight. Healthcare leaders should watch for finalized FDA AI/ML guidances and CMS RWE coverage rules in the coming years.
In conclusion, the past five years have seen a substantial evolution in RWD/RWE policies and technologies in U.S. healthcare. Hospitals, life-science companies, and device manufacturers are increasingly integrating clinical data and AI into their workflows, as outlined in Table 2. These trends are already improving care processes and outcomes. Looking ahead to 2030, one can expect even more pervasive use of real-time data: interoperable data networks and robust AI decision-support will transform hospitals, expand precision therapeutics in pharma, and enable smarter medtech. Stakeholders must prepare by investing in data infrastructure, workforce training, and cross-sector collaborations to fully realize the promise of a data-driven, patient-centered healthcare system.
Sources: Authoritative industry and academic reports, federal agency publications, and peer-reviewed studies (cited above). Tables summarize key initiatives and trends identified in these sources.