Saudi Food and Drug Authority finalizes real-world evidence framework to support regulatory decision-making

Saudi Arabia’s medicines regulator has finalized a framework outlining key concepts, methodological expectations, and regulatory applications for real-world data (RWD) and real-world evidence (RWE) to support medicinal product approvals.

The Saudi Food and Drug Authority (SFDA) has released Version 1.0 of its ‘Framework on the Use of Real-World Data (RWD) and Real-World Evidence (RWE) to Support Effectiveness and Safety for Marketing Authorization of Medicinal Products’, providing structured guidance on how these data sources may be used in regulatory decision-making. The release forms part of broader regulator‘s efforts to modernize evidence generation and evaluation in response to advances in digital health systems and data analytics. The finalized version follows public consultation on the draft guidance released last year.

According to the SFDA, developments in electronic health record (EHR) systems and analytical methods are expected to reshape how clinical evidence is produced and assessed. The authority notes that these advances are “likely to impact evidence for medicine applications, in particular, the way it is generated, ranked and interpreted,” highlighting the growing importance of consistent and transparent approaches to RWE.

A central feature of the framework is the requirement that datasets used to generate evidence must be fit-for-purpose. The document describes such data as capturing “accurate longitudinal information, meeting the needed sample size… necessary covariates to reduce confounding, and outcomes of interest considering the clinical question in the study.” These criteria are intended to ensure that analyses remain aligned with clearly defined research objectives and are capable of supporting reliable regulatory conclusions.

The guidance also underscores the importance of clearly defined analytical parameters, including estimands that specify the treatment effects under evaluation. By structuring analyses around predefined clinical questions, the SFDA aims to improve transparency and promote consistent interpretation of results derived from real-world sources.

The framework identifies several study designs that may contribute to the generation of RWE, including observational studies, pragmatic trials, hybrid trials, registry-based studies, and target trial emulation (TTE). TTE is highlighted as a methodological approach that “could enhance the validity of causal inferences drawn from RWD and yield clinically interpretable findings,” particularly in situations where randomized clinical trials are not feasible.

Data quality and methodological rigor are emphasized throughout the document. The SFDA states that selected datasets should be “fit-for-purpose, relevant and accurately representing and addressing the research question,” while also demonstrating completeness, consistency, and timeliness. Standardization is identified as a key strategy for improving reliability, with common data models cited as one mechanism to support consistent data structure and quality control.

The document also addresses methodological challenges associated with observational research, particularly bias and confounding. Recommended mitigation strategies include stratification, regression modeling, propensity score methods, and patient matching, all of which are intended to strengthen the validity of findings and support confidence in evidence generated outside traditional randomized clinical trials.

Several regulatory use cases for RWE are outlined. These include supporting hypothesis generation during early drug development, informing statistical assumptions, and evaluating outcomes in populations that may be under-represented in randomized studies. The framework also highlights the use of RWD as external comparators in single-arm trials, noting that “historical and external controls may be used in single-arm trials to develop orphan drugs when clinical trials are not feasible or unethical in the proposed settings.”

In addition, the framework highlights the role of local data in interpreting findings from multi-regional clinical trials. By providing insight into population-specific characteristics, local datasets may help translate observed treatment effects to national settings and enable more context-specific regulatory assessments.

The SFDA described the framework as part of an ongoing effort to strengthen the use of emerging data sources in regulatory evaluation. The authority noted that it is being introduced “as part of a continuous effort to guide further establishment of an understanding of utilization, potential uses, and to optimize the evaluation of such data, with a view to informing future regulatory decisions.”