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

The Saudi Food and Drug Authority (SFDA) plays a vital role to public health in Saudi Arabia by ensuring the effectiveness and safety of medicines and vaccines. Advances in data science and medical health records systems can significantly contribute to achieving transformation of health care services and regulations in line with the SFDA fourth strategic objectives and ultimately the Saudi Vision 2030. The transformation is likely to impact evidence for medicines applications, in particular the way it is generated, ranked and interpreted. 

As a part of its mission, and to keep current with the transformation in health care, the drug sector at the SFDA is outlining a framework to regulate the use of Real-World Data (RWD) and Real-World Evidence (RWE) in support of medicines marketing authorizations.​​

This framework aims to define RWD and RWE and provide an outlook to demands and issues in this context such as data quality and confounding.  Moreover, it implies SFDA’s views and possible initiatives related to the ecosystem of RWE/RWD to improve its utilization, appropriateness and quality.​​

Real-World Evidence (RWE) is generated from the analysis of fit-for-purpose RWD. The analysis usually aims to answer a clinical question of interest on the effectiveness and safety of a human medicinal product or vaccine in a clinical condition. These questions are typically framed using well-defined parameters “e.g. etimands" that precisely describe the treatment effect to be estimated. ​​

​Real-World Data (RWD) is defined as any data routinely collected from patients concerning their health conditions, the healthcare services they receive, or data that supports the healthcare services provided to them. ​​

Fit-for-purpose RWD refers to a source of RWD that captures accurate longitudinal information, meeting the needed sample size, on consistent time-fixed or time-varying treatments, necessary covariates to reduce confounding, and outcomes of interest considering the clinical question in study. Quality assurance of the data source must be checked, and it should have acceptable generalizability of the study sample to the target population.​​

A clinical trial is any investigation in human subjects intended to discover or verify the clinical, pharmacological and/or other pharmacodynamic effects of an investigational product(s), and/or to identify any adverse reactions to an investigational product(s), and/or to study the pharmacokinetics of an investigational product(s) to ascertain its safety and/or efficacy.

In Randomized Clinical Trials (RCTs), participants are randomly assigned to the different treatment groups ensuring comparable numbers and characteristics at baseline. Randomization, blinding, and the use of control groups are approaches to minimize bias and confounding when conducting a clinical trial.

On the other hand, in non-randomized clinical trials researchers tend to assign participants to treatment groups according to other factors rather than by random. Although this study design might be useful in rare cases where RCTs are unethical or inapplicable, it introduces a higher risk of bias and confounding. ​​

5.1. Examples of Appropriate Data Sources of RWD

Generally, any data source used to generate RWE must be of high quality, reliable, and relevant “fit-for-purpose” to capture the outcome of interest. Some examples of appropriate data sources are Electronic Health Records (EHR), medical claims data, patient/ disease registries, wearable devices and Patient- Reported Outcomes (PRO). 

5.2. Examples of Study Designs for RWE

5.2.1. Observational Studies

Observational studies are non-interventional studies that are generally designed to collect and observe a case or a series of cases and identify the associations among the observed variables. It can be retrospective or prospective in design (including cohort studies, case-control studies, etc.). Observational studies can provide data assessments with high applicability and generalizability to the real-world clinical practice. However, data obtained from this study design is highly susceptible to bias and confounding. 

5.2.2. Target Trial Emulation

One of the approaches that can be used to address causality questions in observational studies is the emulation of a hypothetical clinical trial “target trial” using observational datasets and an appropriate methodology. This type of study design could help with enhancing the validity of causal inferences drawn from RWD and may result in clinically interpretable findings.

5.2.3. Hybrid Trials

Hybrid trials combines the elements of traditional clinical trials and real-world evidence. Initially, the data are collected through case report forms (CRFs), then for the remaining of the study data are extracted from real-world sources such as electronic health records, and medical claims. Hybrid trial designs tend to have stronger external validity and generalizability due to the inclusion of a large sample size and heterogeneous population. However, hybrid trials are prone to risk of lower data quality in comparison to data actively collected in clinical trials.

5.2.4. Pragmatic Trials

Pragmatic trials are conducted to assess the correlation between different interventions in the routine clinical sitting. This design usually explores the outcome and applicability of any intervention in everyday real-world settings. It typically include diverse populations by setting broader inclusion/exclusion criteria making the results more generalizable. 

5.2.5. Registry Based Trials

Registry based trials utilize patient/ disease registries for data collection to investigate a specific research question. Some of the advantages of using registries are their ability in permitting the long-term collection of data and providing prospective data on disease progression. However, one limitation when utilizing data from registries is the lack of harmonization in data format and used terminologies across registries. Moreover, maintaining the quality of the registry is challenging yet crucial in order to increase confidence in the validity and reliability of the extracted data.

​

Data quality is a key element for optimizing the utilization of RWE in regulatory decision-making. The chosen data and its sources must be relevant and accurately representing and addressing the research question, extensive and complete with no missing information, coherent with no differences across the data sets, and timeless providing up-to-date data.

Standardization of RWD help improves the data quality by ensuring the accuracy, consistency and reliability of the extracted data and its source. Many frameworks can be applied to standardize the data, for example, using Common Data Models (CDMs) enable better quality control and offers uniform and consistent data across the different datasets.​

Uncontrolled confounding is a challenge that may affect the validity and limit the interpretation of the results from observational studies. Such confounders arising from the use of large databases (measured and unmeasured confounders) can be handled through the choice of appropriate study design and analysis method of the data (e.g. stratification, standardization, regression models, propensity scores and combining with machine learning tools with high-dimensional propensity score (hdPS) method).

Additionally, biases can possibly arise from the inappropriate handling of data used to generate RWE such as selection bias and information bias. These biases need to be addressed in order to have valid and reliable results from RWE (e.g. matching patients, target trial emulation). ​​

8.1. Hypothesis Generation

RWD and RWE can be helpful in advancing drug discovery. First, repurposing an approved drug to be used in another disease. Second, comparing two strategies before going into randomized clinical trial (RCT) to inform the decision. Third, utilizing genetic data to identify drug leads. This can include data from biobanks for drug discovery.

8.2. Augmenting Single Arm Studies

RWD can augment single-arm studies by serving as an external comparator to the intervention arm. This is usually applied in early phase trials or rare diseases or for testing certain medicinal products. Historical controls and external controls in single-arm trials may be utilized in the development of orphan drugs when clinical trials are not feasible or unethical in the proposed settings. 

8.3. Transportability and Local Data

Considering clinical trials are currently conducted as multi-regional where different regions might have different distribution of effect modifiers, local RWD can be used to transport the causal effect observed in such trial to the target population by providing local data on important covariates.

8.4. Inform Prior Assumptions

In times of growing use of Bayesian statistics, RWD can be valuable in informing which prior assumptions to use to optimize the sample size, treatment dosage, or patient selection criteria. 

8.5. Investigating Under-Represented Populations in RCTs

RWE may be helpful in investigating under-represented populations by including a diverse and broader demographic representation for testing a research question.​

SFDA is initiating this framework as part of a continuous effort to further guide in establishing an understanding of the utilization, potential uses, and optimize the evaluation of such data with the consideration of making future regulatory decisions. ​​

- Disease Registries: a centralized system that collects information about specific diseases and usually focuses on the disease progression and outcomes. 

- Electronic Health Records: an electronic system used to collect, store, and manage patient’s health information such as diagnosis, treatments, laboratory tests, allergies and immunization.  

- Estimand: a precise description of the treatment effect reflecting the clinical question posed by the trial objective. It summarizes at a population-level what the outcomes would be in the same patients under different treatment conditions.

- Medical Claims Data: an administrative record that collect information that healthcare providers submit to insurance payers for reimbursement of medical services.   

- Patient Registry: a centralized system that gathers data on patient’s diseases, exposure, and treatments. 

- Patient Reported Outcomes (PRO): a measurement reported directly from the patients about their health status, symptoms, and quality of life without interpretation by a clinician. 

- Wearable Devices: an electronic device that can be worn on the body such as smartwatches and fitness trackers. These devices can collect and monitor various physiological parameters like the heart rate, blood pressure, sleep patterns and physical activity. ​

- Curran, G. M., Bauer, M., Mittman, B., Pyne, J. M., & Stetler, C. (2012). Effectiveness-implementation Hybrid Designs. Medical Care, 50(3), 217–226. https://doi.org/10.1097/MLR.0b013e3182408812

- Li, G., Sajobi, T. T., Menon, B. K., Korngut, L., Lowerison, M., James, M., Wilton, S. B., Williamson, T., Gill, S., Drogos, L. L., Smith, E. E., Vohra, S., Hill, M. D., & Thabane, L. (2016). Registry-based randomized controlled trials- what are the advantages, challenges, and areas for future research? Journal of Clinical Epidemiology, 80, 16–24. https://doi.org/10.1016/j.jclinepi.2016.08.003

- Matthews, A. A., Danaei, G., Islam, N., & Kurth, T. (2022). Target trial emulation: Applying principles of randomised trials to observational studies. BMJ, e071108. https://doi.org/10.1136/bmj-2022-071108

- Matthews, A. A., Young, J. C., & Kurth, T. (2023). The target trial framework in clinical epidemiology: Principles and applications. Journal of Clinical Epidemiology, 164, 112–115. https://doi.org/10.1016/j.jclinepi.2023.10.008

- Olmo, C. A., McGettigan, P., & Kurz, X. (2019). Barriers and Opportunities for Use of Patient Registries in Medicines Regulation. Clinical Pharmacology and Therapeutics, 106(1), 39–42. https://doi.org/10.1002/cpt.1414

- Toews, I., Anglemyer, A., Nyirenda, J. L., Alsaid, D., Balduzzi, S., Grummich, K., Schwingshackl, L., & Bero, L. (2024). Healthcare outcomes assessed with observational study designs compared with those assessed in randomized trials: A meta‐epidemiological study. Cochrane Database of Systematic Reviews, 1. https://doi.org/10.1002/14651858.MR000034.pub3

​- Varga, A. N., Guevara Morel, A. E., Lokkerbol, J., van Dongen, J. M., van Tulder, M. W., & Bosmans, J. E. (2023). Dealing with confounding in observational studies: A scoping review of methods evaluated in simulation studies with single-point exposure. Statistics in Medicine, 42(4), 487–516. https://doi.org/10.1002/sim.9628

- Wang, S. V., & Schneeweiss, S. (2022). Assessing and Interpreting Real-World Evidence Studies: Introductory Points for New Reviewers. Clinical Pharmacology and Therapeutics, 111(1), 145–149. https://doi.org/10.1002/cpt.2398

- Zhu, M., Sridhar, S., Hollingsworth, R., Chit, A., Kimball, T., Murmello, K., Greenberg, M., Gurunathan, S., & Chen, J. (2020). Hybrid clinical trials to generate real-world evidence: Design considerations from a sponsor’s perspective. Contemporary Clinical Trials, 94, 105856. https://doi.org/10.1016/j.cct.2019.105856

- Ford, I., & Norrie, J. (2016). Pragmatic Trials. New England Journal of Medicine, 375(5), 454–463. https://doi.org/10.1056/NEJMra1510059

- Anderson, M. L., Griffin, J., Goldkind, S. F., Zeitler, E. P., Wing, L., Al-Khatib, S. M., & Sherman, R. E. (2015). The Food and Drug Administration and pragmatic clinical trials of marketed medical products. Clinical Trials (London, England), 12(5), 511–519. https://doi.org/10.1177/1740774515597700

- Ambinder, E. P. (2005). Electronic Health Records. Journal of Oncology Practice, 1(2), 57–63.

(Ich-E9-R1-Addendum-Estimands-and-Sensitivity-Analysis-Clinical-Trials-Guideline-Statistical-Principles-Clinical-Trials-Step-5_en.Pdf, n.d.)