Real-World Evidence Strategy: Building RWE Capabilities for Regulatory and Commercial Decision-Making

The pharmaceutical industry’s approach to evidence generation is undergoing a fundamental shift. For decades, the randomized controlled trial has served as the gold standard for demonstrating the safety and efficacy of pharmaceutical products, and it remains the cornerstone of regulatory approval processes worldwide. However, the limitations of clinical trials, including their controlled patient populations, limited duration, restricted comparators, and high cost, have created a growing recognition that trial-based evidence alone is insufficient for the full range of decisions that must be made across a product’s lifecycle. Real-world evidence, derived from the analysis of data collected during routine clinical practice and patient experience outside the controlled environment of clinical trials, offers a complementary evidence source that can address questions that trials cannot answer efficiently or ethically, support regulatory decisions beyond initial approval, inform commercial strategy and market access, and provide the continuous evidence generation that modern healthcare systems increasingly demand.

The development of RWE capabilities in pharmaceutical organizations has accelerated dramatically in recent years, driven by the convergence of three forces. First, regulatory agencies including the FDA and EMA have established formal frameworks for the use of real-world evidence in regulatory decision-making, moving beyond aspirational statements to concrete guidance on how RWE can support label expansions, post-market safety monitoring, and in some cases, primary evidence for regulatory approval. Second, the digital transformation of healthcare has created unprecedented access to real-world data through electronic health records, administrative claims databases, patient registries, wearable devices, and genomic databases that capture the complexity of real-world clinical practice at scale. And third, advances in data science, causal inference methods, and computing infrastructure have provided the analytical capabilities needed to extract meaningful evidence from the inherently messy, incomplete, and confounded data that characterizes real-world clinical practice.

This article presents a strategic framework for building RWE capabilities in pharmaceutical organizations, addressing the data sources, regulatory considerations, analytical methods, technology infrastructure, organizational design, and governance practices that determine whether RWE programs generate evidence that is credible, actionable, and valued by regulators, payers, providers, and patients.

The Real-World Evidence Transformation in Pharma

The transformation from RWE as a niche function within outcomes research departments to a strategic capability that informs decisions across the product lifecycle represents one of the most significant organizational changes in pharmaceutical operations.

From Post-Market Surveillance to Lifecycle Evidence

Historically, pharmaceutical companies’ engagement with real-world data was limited primarily to post-market safety surveillance and pharmacovigilance activities required by regulatory authorities. Real-world evidence in this context was a compliance obligation, focused on detecting safety signals in marketed products through spontaneous adverse event reporting and post-authorization safety studies. The modern RWE paradigm expands this scope dramatically to encompass the entire product lifecycle. In early development, real-world data informs disease understanding, patient population characterization, and clinical trial design. During clinical development, RWE supports external control arms, patient recruitment optimization, and site selection. At regulatory submission, RWE provides supplementary evidence for benefit-risk assessment. For market access, RWE demonstrates real-world effectiveness and health economic value to payers and health technology assessment bodies. And throughout the post-market period, RWE supports label expansion, comparative effectiveness research, outcomes-based contracting, and continuous safety monitoring.

The Evidence Credibility Challenge

The central challenge for RWE is establishing credibility sufficient to influence high-stakes decisions about patient treatment, regulatory approval, and healthcare reimbursement. Unlike randomized controlled trials, which control for confounding through randomization, real-world studies must address confounding through study design and analytical methods that can never fully eliminate the possibility of unmeasured confounding or selection bias. This inherent limitation means that the credibility of real-world evidence depends on transparent documentation of study methods, rigorous assessment of data quality and fitness for purpose, appropriate application of causal inference methods, sensitivity analyses that test the robustness of findings to methodological assumptions, and clear communication of the limitations and uncertainties associated with the evidence. Building organizational capabilities that consistently produce credible real-world evidence requires investment in methodological expertise, data quality assessment processes, and governance structures that ensure scientific rigor is maintained across the RWE portfolio.

Real-World Data Sources and Their Characteristics

The landscape of real-world data sources available to pharmaceutical organizations has expanded dramatically, creating both opportunities and challenges for evidence generation.

Electronic Health Records

Electronic health records capture the clinical documentation generated during routine patient care, including diagnoses, procedures, medications, laboratory results, vital signs, clinical notes, and imaging reports. EHR data provides a rich, longitudinal view of patient health and treatment patterns that is particularly valuable for understanding disease progression, treatment sequencing, and clinical outcomes in real-world practice. The challenges of EHR data for evidence generation include the variability of clinical documentation practices across providers and health systems, the prevalence of unstructured free-text data that requires natural language processing to extract, the incompleteness of records when patients receive care across multiple health systems, and the difficulty of identifying clinical endpoints that are not routinely captured in structured EHR fields.

Administrative Claims Data

Administrative claims databases capture the billing records submitted by healthcare providers to insurance companies and government payers for reimbursement of healthcare services. Claims data provides comprehensive capture of healthcare utilization, including diagnoses, procedures, medication dispensing, hospitalizations, and emergency department visits, with the advantage of capturing services across all providers within the payer’s network. The limitations of claims data include the absence of clinical detail beyond diagnosis and procedure codes, the potential for coding inaccuracies driven by billing incentives rather than clinical precision, the limitation to insured populations that may not represent the full patient population, and the lag between care delivery and claims availability that can delay real-time evidence generation.

Patient Registries

Disease registries and treatment registries collect standardized clinical data on defined patient populations, often with greater clinical depth and data quality than EHR or claims data. Registries are particularly valuable for rare diseases where other data sources may have insufficient patient numbers, for chronic conditions where long-term follow-up is essential, and for conditions where clinical outcomes require specialized assessment. The Cystic Fibrosis Foundation Patient Registry, the National Cancer Database, and therapeutic area-specific registries maintained by professional societies provide examples of registry data sources that support high-quality real-world evidence generation. The limitations of registry data include potential selection bias if participation is voluntary, the cost and effort of maintaining registry operations, and the limited scope of data collection compared to the comprehensive capture of EHR systems.

Emerging Data Sources

Several emerging data sources are expanding the scope and resolution of real-world data available for evidence generation. Digital health technologies including wearable devices, connected medical devices, and smartphone health applications generate continuous, objective measurements of physical activity, physiological parameters, medication adherence, and patient-reported outcomes that complement traditional clinical data sources. Genomic and molecular databases link genetic variation to disease outcomes and treatment response, enabling pharmacogenomic analyses that inform precision medicine strategies. And social determinants of health data from sources including census data, environmental databases, and socioeconomic indicators provide contextual information that enriches clinical and claims data with the social and environmental factors that influence health outcomes.

The Evolving Regulatory Framework for RWE

Regulatory agencies have moved from cautious exploration of real-world evidence to the development of formal frameworks that define when and how RWE can be used to support regulatory decisions.

FDA Real-World Evidence Framework

The FDA’s Real-World Evidence Program, established under the 21st Century Cures Act, has evolved from a conceptual framework into a progressively more concrete set of guidance documents, pilot programs, and regulatory precedents. The FDA’s framework distinguishes between real-world data, which is the data collected from sources outside of traditional clinical trials, and real-world evidence, which is the clinical evidence derived from the analysis of real-world data. This distinction is important because it establishes that the mere availability of real-world data does not constitute evidence; data must be analyzed using rigorous methods appropriate to the research question to generate evidence that meets regulatory standards. The FDA has issued guidance on the use of real-world evidence to support regulatory decisions for drugs and biologics, addressing considerations for study design, data quality, and analytical methods. The agency has also established the RWE Program’s demonstration projects that explore specific use cases for RWE in regulatory decision-making, generating practical experience that informs future guidance development.

EMA and European Initiatives

The European Medicines Agency has pursued a complementary approach to RWE through its Regulatory Science Strategy, its network strategy for data analytics, and its leadership in the European Health Data Space initiative. The EMA’s work on the Data Analytics and Real-World Interrogation Network aims to create a sustainable framework for accessing and analyzing real-world healthcare data across European member states to support regulatory decision-making. The European Health Data Space regulation, which establishes a framework for the secondary use of health data across Europe, is expected to significantly improve the availability and accessibility of real-world data for pharmaceutical evidence generation when fully implemented. The EMA has also been active in developing methodology guidance for real-world studies, including guidance on registry-based studies and observational study design that helps pharmaceutical companies understand European regulatory expectations for RWE quality and rigor.

Health Technology Assessment Integration

Health technology assessment bodies, which evaluate the clinical and economic value of new therapies to inform reimbursement decisions, are among the most important consumers of real-world evidence. HTA agencies including NICE in the United Kingdom, IQWIG in Germany, HAS in France, and CADTH in Canada increasingly request or require real-world evidence as part of their assessment processes, particularly for technologies that received conditional regulatory approval, for treatments in therapeutic areas where long-term outcomes data is essential, and for assessments of comparative effectiveness against therapies that were not directly compared in clinical trials. The growing integration of RWE into HTA processes creates additional incentive for pharmaceutical organizations to invest in RWE capabilities, because the quality and availability of real-world evidence increasingly determines the speed and terms of market access for new therapies.

RWE Study Design and Methodological Rigor

The credibility of real-world evidence depends fundamentally on the rigor of the study designs and analytical methods used to generate it.

Target Trial Emulation

The target trial emulation framework has emerged as a leading methodological approach for designing rigorous observational studies using real-world data. The approach begins by specifying the hypothetical randomized controlled trial that would ideally answer the research question, including its eligibility criteria, treatment strategies, outcomes, and follow-up period. The observational study is then designed to emulate this target trial as closely as possible using available real-world data, with explicit identification and documentation of the ways in which the observational study deviates from the target trial and the potential biases these deviations introduce. Target trial emulation provides a disciplined framework for study design that reduces the risk of common methodological errors in observational research, including immortal time bias, prevalent user bias, and inappropriate analysis of time-varying treatments.

Causal Inference Methods

Extracting causal conclusions from observational data requires analytical methods that address the confounding inherent in non-randomized treatment comparisons. Propensity score methods, including matching, stratification, inverse probability weighting, and doubly robust estimation, are widely used to reduce confounding by balancing measured patient characteristics between treatment groups. Instrumental variable methods exploit natural variations in treatment assignment that are independent of patient characteristics to estimate causal treatment effects. Regression discontinuity designs leverage threshold-based treatment assignment rules to create quasi-experimental comparisons. And difference-in-differences methods compare changes in outcomes before and after treatment introduction across treatment and comparison groups. Each method has specific assumptions, strengths, and limitations that must be evaluated for each research question, and the choice of method should be determined by the research question, the available data, and the expected sources of confounding rather than by analytical convenience.

Sensitivity Analysis and Transparency

Because no analytical method can completely eliminate the possibility of unmeasured confounding in observational studies, sensitivity analyses that assess the robustness of study findings to potential unmeasured confounding are essential for establishing evidence credibility. E-value analysis quantifies the minimum strength of association that an unmeasured confounder would need to have with both the treatment and the outcome to explain away an observed treatment effect. Negative control analyses test whether the analytical method produces expected null results for exposure-outcome pairs where no causal effect is expected, providing a diagnostic check on the method’s ability to control confounding. And quantitative bias analyses model the potential impact of specific unmeasured confounders on study results, providing explicit estimates of how findings might change under different confounding scenarios.

Data Infrastructure for RWE Generation

Generating real-world evidence at scale requires a data infrastructure that can ingest, process, standardize, and analyze diverse real-world data sources efficiently and reproducibly.

Data Acquisition and Partnerships

Pharmaceutical organizations access real-world data through a combination of direct data acquisitions, licensing agreements with data vendors, research collaborations with health systems and academic institutions, and participation in data networks and consortia. Major commercial RWD vendors including IQVIA, Optum, Flatiron Health, and Veradigm provide access to large-scale claims, EHR, and specialty datasets through licensing agreements. Academic medical centers and integrated delivery networks provide access to detailed clinical data through research collaborations. And multi-stakeholder data networks such as the FDA Sentinel System, PCORnet, and the Observational Health Data Sciences and Informatics consortium provide access to distributed data networks that enable large-scale observational research without centralizing patient-level data. The data acquisition strategy should be driven by the organization’s therapeutic area focus, research priorities, and regulatory strategy, and should balance the breadth of population coverage needed for certain research questions against the clinical depth needed for others.

Data Processing and Curation

Raw real-world data requires extensive processing and curation before it can support rigorous evidence generation. Data processing workflows include data profiling that characterizes the content, quality, and completeness of incoming data, data cleaning that addresses format inconsistencies, coding errors, and data quality issues, data standardization that transforms diverse data sources into common data formats and coding systems, data linkage that connects records across multiple data sources to create longitudinal patient profiles, and data enrichment that adds derived variables, phenotype definitions, and clinical context to the processed data. These processing workflows should be automated, versioned, and documented to ensure that evidence generation is reproducible and that the data processing steps applied to any given analysis can be fully described in study documentation.

Common Data Models and Standardization

Common data models provide standardized structures for organizing real-world data that enable consistent analytical approaches across different data sources and facilitate multi-site research collaborations.

OMOP Common Data Model

The Observational Medical Outcomes Partnership Common Data Model, maintained by the Observational Health Data Sciences and Informatics collaborative, has emerged as the most widely adopted common data model for real-world evidence in the pharmaceutical industry. OMOP CDM provides a standardized relational data model that represents patient demographics, conditions, drugs, procedures, measurements, observations, and other clinical events using consistent structures and vocabularies across data sources. The OMOP CDM’s comprehensive vocabulary mapping infrastructure, which maps source data codes to standardized OMOP concepts, provides the semantic interoperability needed to execute consistent analyses across data sources that use different coding systems. The OHDSI community has also developed a comprehensive suite of open-source analytical tools that operate on OMOP CDM-formatted data, including tools for cohort definition, characterization, causal inference, and prediction that enable standardized, reproducible analyses across the global OHDSI network.

Sentinel Common Data Model

The FDA Sentinel System’s Common Data Model is designed specifically for the regulatory use case of active post-market safety surveillance. The Sentinel CDM is structured to support the distributed analytical approach that the Sentinel System uses, where analyses are executed locally at participating data partners and only aggregate results are shared centrally. The Sentinel CDM is more focused in scope than OMOP, targeting the specific data elements needed for safety surveillance, but it has been validated across a large network of data partners covering hundreds of millions of patient lives and has a proven track record of supporting FDA regulatory decision-making.

Regulatory Applications of RWE

The use of real-world evidence in regulatory decision-making is expanding from established post-market applications to increasingly ambitious pre-market and peri-approval use cases.

Post-Market Safety and Effectiveness

Post-market safety monitoring remains the most established regulatory application of RWE, with formal frameworks including the FDA Sentinel System and the EMA’s EudraVigilance providing infrastructure for continuous safety surveillance. Post-authorization safety studies and post-authorization effectiveness studies required as conditions of regulatory approval increasingly rely on real-world data from registries, EHR systems, and claims databases rather than dedicated clinical trials, reflecting both the efficiency advantages of real-world data and the growing regulatory confidence in appropriately designed observational studies. The FDA’s Sentinel System has demonstrated the utility of large-scale, distributed RWD analysis for evaluating post-market safety questions, with multiple Sentinel analyses informing FDA regulatory actions including label changes and safety communications.

Label Expansion and New Indications

The use of RWE to support new indications or label expansions represents a growing frontier in regulatory application. Several precedents have demonstrated that real-world evidence can contribute to label expansion decisions, particularly in situations where randomized trials would be impractical, unethical, or excessively time-consuming. Rare diseases, pediatric populations, and conditions where existing treatments make placebo-controlled trials untenable are areas where RWE has the strongest case for contributing to regulatory evidence packages for new indications. The FDA’s guidance on real-world evidence for drugs and biologics provides a framework for evaluating whether real-world data and the evidence derived from it are sufficient to support regulatory decision-making, considering factors including the quality and reliability of the data, the appropriateness of the study design, and the concordance between real-world and clinical trial results.

External Control Arms

External control arms constructed from real-world data offer a potentially transformative application of RWE by providing comparator data for single-arm clinical trials, reducing the need for concurrent control arms in situations where randomization to control is ethically problematic or operationally impractical. External control arms have been used successfully in regulatory submissions for rare diseases and oncology indications where historical clinical trial data and real-world data provide a credible comparator for single-arm trial results. The methodological challenges of external control arms, including selection bias, temporal trends, and unmeasured confounding, require rigorous study design and transparent documentation to produce evidence that regulators will accept.

Commercial and Market Access Applications

Real-world evidence plays an increasingly central role in commercial strategy and market access, informing pricing and reimbursement negotiations, supporting formulary placement, and demonstrating real-world product value to payers and providers.

Health Technology Assessment Evidence

HTA submissions increasingly incorporate real-world evidence to supplement clinical trial data, particularly for demonstrating long-term outcomes, comparative effectiveness, health economic impact, and patient-relevant endpoints that may not have been captured in pivotal trials. The timing of RWE generation relative to HTA submission timelines is a critical strategic consideration, because the most valuable real-world evidence often requires months or years of post-launch data accumulation, while HTA submissions must be prepared quickly after regulatory approval to secure timely market access. Pre-launch preparation of RWE infrastructure, including the establishment of data partnerships, the development of study protocols, and the creation of analytical capabilities, enables pharmaceutical organizations to generate RWE more rapidly after launch and to incorporate it into HTA submissions within compressed timelines.

Outcomes-Based Contracts

Outcomes-based or value-based contracts, in which reimbursement levels are tied to the real-world performance of pharmaceutical products, depend on reliable RWE capabilities for both contract design and performance measurement. These contracts require the ability to define and measure clinically meaningful outcomes in real-world data, to establish baseline outcome rates against which product performance is measured, to track outcomes for patients receiving the contracted therapy over defined time periods, and to resolve disputes about outcome measurement and attribution that may arise during contract execution. The technology infrastructure for outcomes-based contracts must support patient-level outcome tracking, data sharing between pharmaceutical companies and payer organizations, and transparent analytical methods that both parties can trust.

Data Quality and Fitness-for-Purpose Assessment

The quality of real-world evidence is fundamentally constrained by the quality of the underlying real-world data, making systematic data quality assessment a critical component of any RWE program.

Fitness-for-Purpose Framework

Data quality in RWE contexts is best evaluated through a fitness-for-purpose framework that assesses whether a specific data source is adequate for a specific research question, rather than through absolute quality standards that apply regardless of the intended use. A dataset that is fit for purpose for estimating disease prevalence may not be fit for purpose for evaluating treatment effectiveness if it lacks the clinical detail needed to identify and control for confounding. The fitness-for-purpose assessment should evaluate the relevance of the data source to the patient population and clinical context of interest, the completeness of capture for the exposures, outcomes, and confounders needed for the analysis, the accuracy of the coding and documentation practices that generated the data, the temporal coverage and follow-up duration available in the data, and the potential for systematic biases including selection bias, information bias, and confounding that may affect the analysis.

Data Quality Metrics and Monitoring

Systematic data quality assessment requires defined metrics that can be measured consistently across data sources and over time. Key data quality metrics for RWE include completeness metrics that measure the proportion of expected data elements that are actually captured, conformance metrics that measure adherence to expected data formats, value ranges, and coding standards, plausibility metrics that assess whether data values and patterns are clinically plausible, and temporal metrics that evaluate whether data capture patterns are consistent over time or show artifacts that could bias longitudinal analyses. These metrics should be computed and reviewed as a standard step in every RWE study, with documentation of data quality findings included in study reports to enable readers and reviewers to assess the potential impact of data quality limitations on study conclusions.

AI and Advanced Analytics in RWE

Artificial intelligence and machine learning are transforming RWE capabilities by enabling more efficient data processing, more sophisticated phenotype identification, and more comprehensive evidence synthesis.

Natural Language Processing for Clinical Data

A significant proportion of clinically relevant information in EHR systems is captured in unstructured clinical notes rather than structured data fields. Natural language processing enables the extraction of clinical concepts, including diagnoses, symptoms, medications, procedures, and outcomes, from free-text clinical documentation, substantially expanding the information available for real-world evidence generation. Advanced NLP models can identify nuanced clinical concepts that are not captured in billing codes, including disease severity indicators, functional status assessments, physician decision rationales, and patient preferences that add clinical depth to real-world analyses.

Machine Learning for Phenotyping

Identifying patient cohorts with specific diseases, treatments, or outcomes in real-world data, a process known as phenotyping, is a fundamental step in RWE study design. Machine learning approaches to phenotyping can improve the accuracy and efficiency of cohort identification by learning from expert-validated examples to classify patients based on patterns across multiple data elements, rather than relying on simple rule-based algorithms that may have limited sensitivity or specificity. Transfer learning approaches that adapt phenotyping models trained on one data source to new data sources can reduce the effort required to validate phenotypes across the multiple data sources that many RWE studies employ.

Causal Machine Learning

The emerging field of causal machine learning combines the predictive power of machine learning with the principles of causal inference to enable more flexible and data-adaptive approaches to causal effect estimation from observational data. Techniques including targeted learning, double machine learning, and causal forests provide methods for estimating treatment effects that are more robust to model misspecification than traditional parametric approaches. These methods are particularly valuable in pharmaceutical RWE because they can accommodate the high-dimensional, complex relationships between patient characteristics, treatments, and outcomes that characterize real-world clinical data.

Building Organizational RWE Capabilities

Generating high-quality real-world evidence consistently requires dedicated organizational capabilities that span data management, scientific methodology, regulatory strategy, and technology infrastructure.

The RWE Center of Excellence

Many pharmaceutical organizations have established RWE centers of excellence that provide centralized expertise and infrastructure for real-world evidence generation across the organization. The center of excellence model concentrates specialized capabilities including epidemiological and biostatistical expertise, data science and engineering skills, regulatory and methodological knowledge, and technology platform management within a dedicated team that serves as an internal resource for RWE projects across therapeutic areas and business functions. The center of excellence should also be responsible for establishing and maintaining organizational standards for RWE quality, including study design templates, analytical protocols, data quality assessment procedures, and reporting standards that ensure consistency and rigor across the RWE portfolio.

Cross-Functional Integration

Effective RWE strategy requires close collaboration between functions that traditionally operate independently. Regulatory affairs must be involved in RWE study design to ensure that evidence generation is aligned with regulatory expectations and submission strategies. Medical affairs must contribute clinical expertise and ensure that RWE findings are communicated appropriately to healthcare professionals. Commercial strategy must articulate the evidence needs that drive market access and competitive positioning. And clinical development must integrate RWE into the broader evidence generation strategy that spans randomized trials and observational studies. This cross-functional integration is best achieved through an integrated evidence planning process that identifies evidence needs across regulatory, commercial, and medical functions early in the product lifecycle and designs an evidence generation strategy that addresses these needs efficiently through a combination of clinical trials and real-world studies.

The Future of Real-World Evidence in Pharma

The trajectory of real-world evidence in pharmaceutical decision-making points toward an increasingly integrated, technology-enabled, and regulatory-embedded future.

Continuous Evidence Generation

The future of pharmaceutical evidence generation is shifting from episodic studies that generate point-in-time evidence to continuous evidence generation platforms that monitor product performance, detect emerging signals, and update evidence synthesis in near-real-time. This shift is enabled by advances in data infrastructure that support streaming data ingestion and analysis, by statistical methods for sequential analysis that can update conclusions as new data accumulates, and by regulatory frameworks that are evolving to accommodate living evidence that is continuously updated rather than generated through discrete studies.

Regulatory Integration

Regulatory agencies are progressing toward deeper integration of real-world evidence into the regulatory framework, with expanding use cases, more specific methodological guidance, and increasingly formal processes for evaluating and incorporating RWE in regulatory decisions. The development of regulatory-grade real-world data networks, the establishment of pre-certification programs for RWD sources and analytical methods, and the growing body of regulatory precedents for RWE use are all accelerating this integration. For pharmaceutical organizations, this regulatory evolution means that RWE capabilities are becoming not merely a competitive advantage but an essential operational capability that is as fundamental to regulatory strategy as clinical trial capability.

Patient-Centricity and Digital Endpoints

The growing availability of patient-generated health data from wearable devices, connected medical devices, and digital health applications is creating opportunities for real-world evidence that captures outcomes that matter most to patients, including functional status, symptom burden, quality of life, and treatment experience. Digital endpoints derived from continuous sensor data offer the potential for more sensitive, objective, and patient-relevant outcome measurement than traditional clinical endpoints, enabling evidence generation that better reflects the real-world impact of pharmaceutical products on patients’ lives.

Real-world evidence has evolved from a marginal complement to clinical trial data into a strategic imperative that shapes regulatory decisions, market access outcomes, and competitive positioning across the pharmaceutical industry. The organizations that invest in building comprehensive RWE capabilities, that establish the data infrastructure and analytical expertise needed to generate credible evidence, and that integrate RWE into their overall evidence strategy will be best positioned to navigate the evolving regulatory and commercial landscape. Those that treat RWE as an afterthought or a purely defensive capability will find themselves at an increasing disadvantage as regulators, payers, providers, and patients all demand the real-world evidence that demonstrates pharmaceutical products deliver value beyond the controlled environment of clinical trials.