Real-World Evidence in Health Canada Drug Submissions: What Canadian Pharmaceutical Sponsors Actually Need to Know

Health Canada’s 2019 Real-World Evidence Framework landed without much fanfare — a 30-page guidance document published as regulators worldwide were grappling with how to incorporate non-traditional data into drug review. By 2023, that framework was being cited in pre-submission meeting transcripts, product monograph discussions, and information requests across multiple therapeutic areas. By 2026, real-world evidence has become a routine expectation in post-market commitments and a legitimate pathway for certain label expansions and rare disease submissions.

Most Canadian pharmaceutical sponsors haven’t caught up. They either treat RWE as a post-market formality — something to satisfy a term and condition — or they import methodologies from FDA submissions without checking whether Health Canada’s specific documentation and pre-specification requirements actually align. Both approaches create friction at the review stage, and that friction translates directly into delayed approvals.

What Health Canada Actually Means by “Real-World Evidence”

The terminology matters here. Real-world data refers to the source: health information collected outside a controlled research setting. Electronic health records, provincial drug claims databases, patient registries, mobile health devices, and spontaneous adverse drug reaction reports are all examples. Real-world evidence, by contrast, is what you derive from that data through rigorous analysis — the clinical insights extracted to answer a specific regulatory question.

Health Canada distinguishes three broad categories of real-world data acceptable for regulatory purposes: administrative and claims data (including provincial drug plan databases), clinical data from EHR systems, and patient-reported outcomes collected through registries or prospective observational studies. Each carries different strengths and limitations depending on the regulatory question you’re trying to answer.

The Food and Drug Regulations — specifically sections C.08.002 and C.08.004 — outline the evidence requirements for New Drug Submissions and Supplemental New Drug Submissions respectively. RWE doesn’t replace these requirements. It supplements or, in some circumstances, satisfies portions of them. The agency’s position, articulated through its 2019 framework and refined in a 2023 discussion paper on pragmatic clinical trials, is that RWE acceptability depends almost entirely on the regulatory question being asked and the quality of the underlying data.

One number worth anchoring to: Health Canada’s standard review timeline for a New Drug Submission runs 300 days for standard submissions and 180 days under Priority Review designation. RWE packages that aren’t methodologically airtight don’t just get rejected — they generate lengthy information requests that can consume weeks of that clock on both sides of the file.

Where RWE Legitimately Fits in Canadian Drug Submissions

The most common application is post-market. Many Canadian market authorizations include terms and conditions requiring sponsors to conduct post-authorization safety studies (PASS) or post-authorization efficacy studies (PAES). Real-world data, particularly from provincial drug plan linkages, is often the most practical way to fulfill these obligations without running a new randomized trial. Provincial databases offer longitudinal, population-level follow-up at a scale that dedicated clinical studies rarely match.

Label expansions are the second major use case. If a sponsor wants to add a pediatric indication, extend an approved age range, or modify a dosing schedule, observational data from real-world populations can sometimes be sufficient — particularly when the underlying mechanism is well understood and the safety profile is established. Health Canada reviewed approximately 150 Supplemental New Drug Submissions in a recent fiscal year, and a meaningful proportion of those now include at least some real-world data component.

Project Orbis deserves specific mention for oncology sponsors. Canada joined this international collaborative review program — which includes FDA, EMA, TGA, Swissmedic, and Singapore HSA — and submissions under this framework increasingly feature RWE, particularly for rare cancer subtypes where RCT enrollment is constrained by small patient populations. Health Canada’s participation in Project Orbis has accelerated its institutional familiarity with external RWE packages previously validated by partner agencies.

Rare disease applications represent the most consequential frontier. For conditions with fewer than 1 in 10,000 prevalence in the Canadian population, recruiting adequate RCT populations is often structurally impossible. Health Canada’s Special Access Program and Priority Review designation both acknowledge this reality. Sponsors pursuing these pathways should engage the Therapeutic Products Directorate early — a pre-submission meeting at the 60-day protocol stage can save months of back-and-forth before the formal submission clock starts.

Data Quality Standards Health Canada Applies to RWE

This is where many sponsors underestimate the regulatory bar. Health Canada has adopted FAIR data principles — Findability, Accessibility, Interoperability, and Reusability — as its baseline for assessing RWD quality. These aren’t just boxes to check; they inform how reviewers evaluate every step in the data lifecycle from collection through analysis.

More specifically, Health Canada’s RWE guidance aligns with the ICH E9(R1) addendum on estimands, formally adopted by Canada in 2021. The estimand framework requires sponsors to pre-specify precisely what treatment effect they’re trying to measure, how they’ll handle patients who discontinue, and how intercurrent events will be managed. Getting the estimand wrong means your RWE analysis is technically answering a different question than your submission claims. Reviewers notice this, and information requests citing estimand misalignment are among the more time-consuming to resolve.

Provincial drug claims data is the most commonly used RWD source in Canada, and for good reason. Databases from Ontario (ICES), British Columbia (Population Data BC), and Quebec (RAMQ) offer linked, longitudinal records covering millions of patients at relatively low cost. But they have real limitations. Drug claims data captures dispensing, not adherence. It frequently lacks the clinical granularity — disease severity, comorbidity detail, reason for discontinuation — that regulators need to interpret treatment effects in context. Sponsors who treat claims data as a complete clinical picture routinely receive information requests challenging their exposure and outcome definitions.

Confounding is the central methodological concern. Because patients in real-world settings aren’t randomized, systematic differences between treated and untreated groups can easily masquerade as treatment effects. Health Canada expects sponsors to address this through pre-specified propensity score methods, instrumental variable analyses where feasible, or — for simpler questions — standardization approaches. External control arms, where historical data from a comparable population serves as a comparator, are gaining acceptance but require careful justification of the underlying exchangeability assumption.

A practical benchmark: study protocols for RWE submissions should be registered in ClinicalTrials.gov or the ISRCTN registry before data analysis begins. Health Canada reviewers check this. Protocols registered after analysis is complete raise immediate concerns about selective outcome reporting, and no amount of methodological sophistication in the analysis report will fully resolve that flag.

What Consistently Derails RWE Submissions at Health Canada

Sponsors who haven’t navigated Health Canada’s RWE framework before tend to make the same set of mistakes. Here’s what we see repeatedly in the files we support.

Insufficient data governance documentation. Health Canada expects a complete data lineage — how data was collected, transformed, linked, and analyzed. Gaps are treated as potential data integrity issues. If you’re using linked administrative databases, every linkage step needs a documented error rate and methodological justification.

Mismatch between regulatory question and study design. A sponsor might conduct a high-quality observational study using provincial claims data, only for Health Canada to note that the study population doesn’t generalize to the patient population described in the proposed Canadian label. Real-world data is only as useful as its representativeness of your intended indication.

Treating CADTH and Health Canada as interchangeable. The Canadian Drug and Health Technology Agency conducts health technology assessments for reimbursement recommendations, not market authorization. Sponsors sometimes present CADTH-commissioned RWE studies as evidence of regulatory-grade quality. The standards differ meaningfully. CADTH’s comparative clinical effectiveness framework does not map onto Health Canada’s safety and efficacy evidentiary standard under the Food and Drugs Act.

Late statistical analysis plans. An SAP submitted alongside the study report — rather than before data lock — is a flag reviewers raise almost immediately. Health Canada expects pre-specification of primary endpoints, sensitivity analyses, and subgroup assessments, with a clear justification of the chosen analytical approach relative to alternatives. Submitting a SAP post-hoc is effectively conceding that the analysis could have been shaped by the results.

Before Your Next RWE Package Goes to Health Canada

The single most valuable step is requesting a pre-submission meeting with the Therapeutic Products Directorate or the Biologics and Genetic Therapies Directorate (depending on product type) before committing to a specific RWE design. These meetings are underused across the industry. Health Canada will tell you, in writing, whether your proposed approach is likely to be acceptable for the specific regulatory question you’re trying to answer. That early alignment can eliminate months of post-submission iteration.

If you’re working with provincial data, start the access process at least 18 months before your target submission date. ICES and Population Data BC both run data application processes that take 6 to 12 months before access is granted — and that’s before any analysis has begun. Sponsors who don’t build this lead time into their project plans miss submission windows consistently.

And one point worth making explicit: an RWE package designed for an FDA submission is not automatically fit for Health Canada. The agency’s documentation expectations — particularly around data governance, pre-specification timelines, and estimand framing — have evolved independently. Before repurposing an FDA-destined RWE package for a Canadian filing, plan for a formal gap analysis against Health Canada’s framework. The differences are specific enough that a general-purpose quality review won’t catch them.

The Canadian regulatory environment for real-world evidence is maturing faster than most sponsors realize. The organizations that engage Health Canada early, pre-specify rigorously, and document their data governance completely are consistently the ones that move from submission to approval without getting stuck in extended review cycles.

---

Written by Nour Abochama, Quality & Regulatory Advisor, Androxa. Learn more about our team

Talk to our team about Health Canada compliance. Contact us

Related from our network

• ISO 17025-accredited pharmaceutical testing for North American submissions — How accredited laboratory data packages are structured to support drug submissions with Health Canada and the FDA
• European regulatory strategy for multi-market pharmaceutical submissions — Understanding EMA frameworks for sponsors planning simultaneous EU and Canadian regulatory filings