Real world evidence (RWE) in medicine is evidence on the use, safety, effectiveness, and cost of health technologies, which is observational data obtained outside the context of randomized controlled trials (RCTs) and generated during routine clinical practice. RWE is generated by analyzing data obtained from patient registries, medical records, or in some cases hybrid trials, pragmatic trials, and late-phase trials [1, 2]. As part of an effort to accelerate medical product development and bring innovations faster and more efficiently to patients who need them, the 21st Century Cures Act was signed into law on December 13, 2016, with the US FDA issuing its “Framework for FDA’s Real-World Evidence Program” in December 2018 [3]. Since then, the FDA has updated the guidance multiple times to assist with using RWE to approve a new indication for a drug [4, 5]. Health Canada followed suit and, on April 16, 2019, published the document “Optimizing the Use of Real-World Evidence to Inform Regulatory Decision-Making,” acknowledging that the use of RWE in regulatory decisions is increasing globally in the assessment of drug safety, efficacy, and effectiveness [6]. However, upon review of regulatory reports supporting approvals, Health Canada’s use of RWE in regulatory decision-making was considerably lower than that of European Medicine (EU) and US FDA [7]. Recently, on April 2023, the Canadian Drug Agency (CDA-AMC) and Health Canada jointly published a submission-ready RWE guidance outlining a step-by-step path to prepare quality documents for submission [8]. The Canadian guidance document, more so than other regulatory guidance, stresses the importance of “Transparency” in every step of the submission preparations. In the United States, the availability of sophisticated databases from electronic health and medical records provides rich platforms for generating RWE based on routine clinical practice outside clinical trials [4, 5]. While the utility of provincial, institutional datasets and registries can provide valuable data sources for RWE, the absence of a unified, pan-Canadian database underscores the importance of PSP capturing nation-wide patient data. Canada and Europe struggle to access large national databases for real-world data to generate RWE. The lack of robust common databases and the strength of patient privacy laws appear to be the most prominent obstacles [9]. In Canada, the arrival of specialty drugs in the past decade has promoted the growth of company-supported, nationwide PSPs. The goal of PSP is to help patients navigate the complex path of the drug reimbursement environment, assist with dosing administrations (including infusions/injections and training), manage side effects, and provide general patient support [10]. Most programs collect patient demographic data, dosing information at initiation and subsequent changes, and critical laboratory data essential to patient management and disease activities. As drug utilization is a key measurement yardstick, most programs collect individual patients’ data from the start to the end of treatment, providing essential data for drug adherence and persistence [11]. Moreover, Patient Support Programs (PSPs) offer valuable insights into the early effectiveness and safety profiles of drugs. These programs often provide critical data points, such as optimal treatment duration and potential signals for early withdrawal. This information not only aids in refining treatment protocols but also supports healthcare providers in making evidence-informed decisions that enhance patient outcomes.

With the lack of national and, in some cases, even provincial databases, will pan-Canadian PSP databases be able to fill the gap for the generation of RWE? Several recent publications advocate using PSP data as a source of RWD and assess whether the RWE generated meets regulatory/reimbursement requirements [10,11]. This publication reviewed peer-reviewed studies from 1 January 2020, to 28 February 2025, based on data from PSPs enrolling patients with chronic diseases. The recently published CDA RWE guidance was used to evaluate the appropriateness of the data source and subsequent analysis for submission purposes.

The group under interventional studies had the highest potential for regulatory or reimbursement submissions as they evaluated interventions that could support claims of efficacy or cost-effectiveness [4]. In addition to assessment using the CDA-AMC checklist, the validity of the studies was further examined under “Common Mistakes in Biostatistics” for cohort studies. The non-interventional studies focused on longitudinal follow-up of patients remaining on the drug or the program, and these patient persistence data are generally not candidates for regulatory submissions.

Table 1 lists the nine studies selected for evaluation. The first four studies were grouped under interventional and the last five as noninterventional.

In the past few decades, the rise of specialty drugs, often requiring complex drug administrations such as infusions/injections or patient-safety monitoring, necessitated specialized patient assistance [10, 23]. In Canada, Patient Support programs (PSP) grew quickly in terms of numbers and level of service offered. In addition to drug administration and safety follow-up, staff from PSPs also help patients access reimbursement, provide education regarding the disease and the drug, and collect patient-level data to facilitate better patient management. An early report suggests that there are over 400 PSPs in Canada, each with the infrastructure to administer specialty drugs, provide patient care, and collect patient-level data [21]. A more in-depth analysis [24] at the drug level indicates that up to Aug 2023, of the 2556 prescription drugs marketed by 89 companies in Canada, 256 (10.0%) had a patient support program. Some drugs had multiple PSPs, and they were mostly managed by outside vendors.

The large and growing number of PSPs in Canada should serve as a gold mine for the generation of pan-Canadian RWE. Regulatory and HTA agencies consider data generated as part of a PSP to be a low level of evidence [8]. Most of these programs are tied to special drugs or companies, and the lack of public transparency in their operations is of high concern [10, 11]. Data quality remains obscure, including how missing data were handled, data completeness, data governance, and patient privacy. Databases used in PSPs are designed to collect information for patient management and not prospectively address scientific questions. How can that be used retrospectively to generate new effectiveness and safety information? Grundy et al., after analyzing a comprehensive database of PSPs in Canada up to August 2023, saw a strong correlation between drug prices and availability of PSPs, suggesting a commercial motive in the setting up of these PSPs besides patient care [24]. Due to these limitations, most PSP data publications focus on patients’ persistence and adherence to the drug tied to the specific PSP. For RWE from PSP data to be acceptable to regulatory agencies, additional rigor in transparency in design, data validation, and analysis are critical parameters.

With RWE gaining momentum in decision-making globally, manufacturers are interested in capturing and utilizing data from PSPs to gain additional insight into their products post-launch, such as effectiveness studies for reimbursement or supporting new indications. They are also aware that the databases currently structured are inadequate, and the data collection methodology would have to be revamped to meet the quality and transparency requirements matching those of clinical trials.

In collaboration with academic partners, regulatory agencies recently published checklists and templates to guide the generation of fit-for-purpose RWE intended for decision-making. The HARPER template [25], the STROBE checklist [26], and the CDA-AMC RWE checklist [8] were among the most recent ones. The HARPER template provides a set of core recommendations for clear and RWE protocols, whereas the STROBE checklist includes a list of high-level items that should be included in reports of cohort studies. The RWE checklist guide published by CDA-AMC is more prescriptive, specifying in detail the specific information required for each section. The CDA-AMC checklist has 12 sections, each with 2–12 items plus some with subitems. This publication evaluated nine peer-reviewed publications based on Canadian PSP databases using the CDA-AMC RWE checklist of 103 items (see Methods). The nine publications of cohort studies were divided into two groups, interventional and non-interventional, according to FDA RWE guidance [27]. The scores of NR (not reported) and NA (not applicable) were compared across studies. The NA scores serve as an indicator of the type of study design, highlighting that similarly designed studies will exhibit the same NA. On the other hand, NR scores measure the lack of compliance to critical parameters, which are deemed essential for a robust RWE study according to the CDA.

The infliximab study in IBD patients demonstrated the highest compliance with the lowest NR score among the interventional studies. The study was based on a protocol prospectively designed before the analysis (after data collection), with quality data checked and validated by a third party, a formal governance committee, ethics board approval, and justifications for missing data management. The study population and variables were well-defined and controlled, and study windows and gaps were specified. The Cox Proportional Hazards Model with the intervention time as the covariant was used to avoid immortal time bias. To account for measured and unmeasured confounders, quantitative bias analysis was used to evaluate critical endpoints. The study also completed a STROBE checklist and is in good standing. The study might still be inadequate in meeting all Health Canada and CDA-AMC requirements due to the lack of data transparency, as data extraction methods, code, algorithm, and data dictionary were not provided in the publication. The company probably considers the data information proprietary and would not disclose it in a publication, but would submit confidentially to the regulatory agencies. The rest of the interventional studies (vedolizumab in IBD, mepolizumab in asthma, and brodalumab in psoriasis) failed stringent data quality requirements, lacked statistical vigor, did not account for bias or confounders, and scored inadequately against the checklist. As for the non-interventional studies Ustekinumab in IBD, ixekiuzmab in psoriasis, dimethyl fumarate in MS, tofacitinib in RA, and erenumab in migraine, drug persistence was the objective of the studies, and endpoints were descriptive. The lack of control for bias and confounders might have skewed the data to be more favorable for the study drug; there was insufficient information for concluding, and these publications did not score favorably against the CDA-AMC checklist.

Limitations of the study: The most significant limitation was that the evaluations were performed on publications rather than study reports. These publications often omits details deemed confidential to manufacturers, restricting an assessment based on the checklist. The current paper attempts to perform the most precise evaluations based on publicly available information. The author believes that peer-reviewed publications serve as a good proxy for the quality of the study. Despite the need to protect confidential information, data quality validation, considerations of bias and confounders, and measures to control all key variables should be transparent in the publications. Without these details, the information cannot be used for decision-making. It is also acknowledged that most of the publications were written before the issuance of the CDA RWE guidance; however, the guidance was developed to align with established global standards, key principles for generating RWE were available for reference at the time these studies were conducted and should not have major impact on the quality of the studies.

Nine published studies using PSP databases were analyzed using the CDA-AMC checklist. Only one study scored adequately to be a potential submission candidate for decision-making. For successful regulatory considerations, data quality validation, data source transparency, validated methodology to manage study bias, measured or unmeasured confounders, and robust outcome analysis, including sensitivity and quantitative bias analysis are critical factors to consider.