Answers are grounded in this paper and cite their sources.
Drug discovery and development require years of exhaustive studies to adequately test the candidate substances for efficacy and safety before they reach the market.1 These tests include laboratory work and clinical evaluation under specific conditions pre-established by the researchers. However, drugs are mainly used as therapeutic or diagnostic agents in real-life conditions that present important challenges. Randomized clinical trials may not provide all the answers about the therapeutic value of new drugs, largely due to feasibility,2 especially in an ageing population with multiple chronic conditions.3 This situation, referred to as multimorbidity, increases the risk of complex interactions between diseases, inappropriate polypharmacy, drug-related side effects and adverse reactions and drug–drug and drug-disease interactions, which can considerably affect the clinical outcomes and the overall performance of drugs. To properly assess the real-world performance of new and old drugs, it is essential to conduct large-scale pharmacoepidemiologic studies using high-quality, RWD. Clinical trials are usually performed in relatively homogenous cohorts. The high socio-demographic and clinical heterogeneity between individuals in real-life conditions should be considered when evaluating drug safety and effectiveness, especially when a medicinal preparation is tested for use in new indications for different population groups.4 Studies using RWD can be an appropriate alternative in situations where the realization of clinical trials is not possible, such as for ethical reasons, low prevalence of the disease or commercial lack of interest.5 Personalized precision medicine adds complexity to the design of “classic” randomized clinical trials, and adaptive umbrella designs can be extremely complex in practice. Real-world evidence (RWE) is fundamental in pharmacovigilance. In postmarketing surveillance, RWE on the effectiveness and safety profile supplements the existing evidence from preclinical studies and clinical trials and can provide post-approval guidance.6 Once a drug has reached the market, vigilant monitoring is vital to detect any possible adverse reactions timely. It is imperative to study how drugs perform in real life, especially in groups of patients that are under-represented in clinical trials, such as older adults, paediatric populations, pregnant or breastfeeding individuals, and people with multiple chronic and complex conditions. RWD is a powerful tool both in post-marketing surveillance and in generating a rapid response to safety alerts that require identifying a specific group of patients. Furthermore, RWD can be useful in identifying unintended impacts of pharmacovigilance regulatory interventions, such as undesirable use of drug substitutes or the decrease in the use of drugs from the same therapeutic subgroup, among others.7 RWD allows us to study both appropriate and inappropriate use of drugs and healthcare services, as well as the impact and appropriate implementation of regulations and special advisories issued by competent authorities. When appropriately used, RWD may have great potential and implications in medical research and clinical practice. They are a tool to identify signals of possible adverse reactions but also signals of potentially unexpected clinical advantages; a fascinating example is drug repurposing. Data from electronic health records have generated hypotheses for new uses outside the scope of the original indication for several marketed drugs.1 It is necessary to continuously conduct studies using RWD, considering substantial changes in the population's demographic, clinical and socio-behavioural characteristics over time. Today's population age structure, most common and lethal diseases, commonly prescribed drugs, life expectancy, lifestyle and socio-economic characteristics are different from those a few decades ago, and they will be different from those a few decades in the future. Climate change will impact people's health differently depending on their place of residence, due to multiple variables (environmental temperature, air, water and food quality, geographic distribution of vectors, etc.).8 Therefore, the population's health profiles are not static but dynamic, and it is vital to study their evolution over time. Pharmacological approaches aimed to offer appropriate clinical management should follow the dynamic evolution of the clinical profiles, and public health systems should be prepared to give an agile and appropriate response to changing health profiles. RWD can be used to estimate the safety and cost-effectiveness of drugs in different clinical and geographic contexts. This can lead to better recommendations for patients in their respective communities. RWD also presents exciting possibilities for analysing different patient groups, identifying therapeutic patterns, detecting risks and generating new research hypotheses. Furthermore, RWD can be used to evaluate health policies and identify risks related to drug shortages, preventing negative consequences for healthcare systems and poor patient outcomes.9 Despite these benefits, studies using RWD face many challenges. They are often affected by various types of biases (e.g., selection, information, follow-up, publication bias) and potential confounders.10 The significant heterogeneity among databases also makes meaningful comparisons between studies difficult. To ensure the validity and utility of RWE studies, a pharmacoepidemiological database must meet several requirements, including good quality of records, standardized coding, a common data model (to guarantee interoperability), standardized analysis methodologies (to allow reproducibility), data security and protection, transparency and a robust governance system. There is currently no universally accepted definition for RWD, and international standards for conducting RWD studies are lacking.10 Regulators are at varying stages in developing guidance and frameworks to enhance the quality and transparency of RWD and generate high-quality RWE to support regulatory decisions. This includes the European Medicines Agency (EMA), the US Food and Drug Administration (FDA), Health Canada, the UK's National Institute for Health and Care Excellence (NICE) and the Therapeutic Goods Administration (TGA) in Australia. These are valuable sources for planning, collecting and reporting on RWD. Table 1 summarizes key considerations in selecting and auditing RWD sources6, 7, 11-15; it presents the ‘LPRQ model’, a comprehensive practical approach to essential aspects for optimizing RWE studies, including legal assessment (L), data provenance (P), relevance (R) and quality (Q). These four elements provide a solid framework for researchers and healthcare professionals, ensuring the validity and utility of RWE studies in the context of regulatory decision-making and pharmacovigilance. Nonetheless, users should remain open to supplementing this framework with additional considerations tailored to their specific study contexts and regulatory landscapes. Assess the validity of the data, including whether it conforms to expected formats, such as standardized language or coding for medical diagnoses, medicines and laboratory test results. Determine if the data are unstructured, e.g., captured through free text entry, which may make the data incomplete or limit its usability. Consider if there are any limitations in the classification systems used to standardize the data, such as those used to classify and code a diagnosis. Examine the consistency of the data. Key considerations include whether the data is captured and reported in a uniform way across the study sites and whether values within a data element vary beyond expectations, such as from differences in data coding practices. Researchers, clinicians and policymakers are encouraged to learn from databases that have already standardized and purified their records, such as the UK's Clinical Practice Research Datalink (CPRD; https://www.cprd.com/) and the Pharmacoepidemiological Research Database for Public Health Systems (BIFAP; http://bifap.aemps.es/?lang=en) of the Spanish Agency of Medicines and Medical Devices (AEMPS). Working in interdisciplinary teams with various healthcare professionals endorsed by clinicians, epidemiologists, data scientists, statisticians and software engineers is essential, especially when applying emerging technologies like artificial intelligence and machine learning.16 Fostering and empowering communities to collaborate on generating RWD is crucial to improving the availability of high-quality RWE. Partnerships are most effective when they involve multiple stakeholders from various disciplines to achieve consensus on appropriate governance for ensuring data safety. Standardization of data is essential to promote interoperability, combine information from different sources, identify knowledge gaps and research questions and address the needs of end-users.5 Implementing flexible and scalable infrastructure is crucial to accommodate data collection. Also, it is important to raise awareness about available RWD sources and provide education to optimize its use.5 The Data Analysis and Real-World Interrogation Network (DARWIN EU®; https://www.darwin-eu.org/), coordinated by the European Medicines Agency, is expanding the availability of RWE through established approved healthcare databases with distributed data access across its network. Through the Sentinel Initiative (https://www.sentinelinitiative.org/), the FDA aims to develop new methods for assessing the safety of approved medical products. Through its central coordinating centre at Columbia University, the Observational Health Data Sciences and Informatics (OHDSI; https://www.ohdsi.org/) programme provides open-source solutions to support a collaborative to perform large-scale health data analytics. If methodologies based on randomized clinical trials and RWD learn to collaborate and to be complementary, the future is promising. KA is the secretary of the Drug Utilization Subcommittee of the Australian Pharmaceutical Benefits Advisory Committee (PBAC). The views presented are those of the author and do not necessarily represent those of the Australian Government Department of Health and Aged Care or the PBAC.
No takes yet. Share an insight, caveat, or question.
Ioakeim‐Skoufa et al. (2025) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: