Real world evidence (RWE) is increasingly accepted by FDA to support regulatory decisions, including new indications, post-market surveillance, and label expansions. But FDA’s acceptance of RWE is conditional on data quality. Your QMS is the system of record that determines whether your real-world data meets FDA’s standards. This article explains what FDA’s guidance requires, where quality systems fall short, and what regulated companies need to do before their RWE becomes a liability during an inspection or submission review.

What is real world evidence and why does it matter?

Real world evidence is clinical evidence derived from the analysis of real-world data (RWD): health information collected outside the controlled setting of a traditional randomized clinical trial. Sources include electronic health records, claims databases, disease registries, post-market surveillance data, and patient-generated data from wearables and apps.

FDA has authority under the 21st Century Cures Act to use RWE to support approval of new indications and to satisfy post-approval study requirements. In practice, the agency has accepted RWE for a growing number of decisions, particularly in oncology, rare disease, and medical device post-market studies. A January 2026 analysis of FDA’s RWE experience noted that both the scope and frequency of RWE-based submissions have grown significantly since 2017 (HRP Consulting Group, 2026).

For quality teams, this matters because RWE does not exist in a vacuum. It is generated, collected, stored, and analyzed through processes and systems that a quality system must govern. An FDA reviewer assessing an RWE submission will ask the same data integrity questions they would ask about any other study data: Is this data attributable? Is it accurate? Is the collection process documented and controlled?

What FDA requires for RWE data quality

FDA’s framework for RWE quality draws on several guidance documents, with the most directly relevant being the agency’s updated RWE guidance finalized in late 2025, which reflects FDA’s evolved position from the 2017 draft (HRP Consulting Group, 2026). The European Medicines Agency has published parallel requirements in its Data Quality Framework for EU Medicines Regulation (EMA, 2023).

Across these documents, FDA and EMA converge on several core requirements for RWE data quality.

Fitness for purpose

RWD is only useful as evidence if it captures the variables relevant to the question being asked. A claims database designed to process insurance payments was not built to capture disease severity scores or adverse event nuances. Before using any RWD source to generate RWE, companies must assess whether the data captures what they need at the resolution they need it. This assessment is a documented quality activity, not a judgment call made in a statistics team meeting.

Data provenance and traceability

FDA expects RWE submissions to document where the data came from, how it was transformed, and what quality controls were applied at each step. This is the data equivalent of an audit trail: a complete, reconstructible record of how raw real-world data became the evidence submitted to FDA. Any break in that chain, such as an undocumented data transformation or a cleaning step with no version control, is a data integrity problem that can invalidate the submission.

Consistency and standardization

RWD comes from sources that use different coding systems, time formats, and data structures. A patient diagnosed in one system as ICD-10 code E11.9 and in another as “Type 2 diabetes, unspecified” represents the same condition, but combining those records requires documented mapping rules. FDA reviewers expect to see the harmonization methodology documented, with rationale for every mapping decision and a record of who approved it.

Missing data handling

Real world data always has gaps. Patients miss appointments, EHR systems fail to capture certain lab results, and claims data misses care delivered outside the covered network. FDA expects sponsors to document how missing data was handled, including the analytical assumptions made and their potential effect on the conclusions. A missing data plan is a quality document, not just a statistical one.

Pre-specification of analyses

One of FDA’s consistent concerns about RWE is the potential for post-hoc analysis: cherry-picking the analysis approach after seeing the data. FDA’s guidance strongly favors pre-specified analysis plans that are locked before data access. This mirrors the protocol discipline of a randomized trial. From a quality standpoint, a pre-specified analysis plan is a controlled document that must go through change control if it changes after data access begins.

Where QMS practices fail to support RWE requirements

Most quality systems were designed for manufacturing and device quality, not for data governance at the scale RWE requires. Several gaps appear consistently when quality teams try to support RWE programs.

No data lifecycle management

Traditional QMS document control manages SOPs and records. RWE requires governing data itself through its full lifecycle: ingestion, cleaning, transformation, analysis, and archival. Companies that lack a data lifecycle management process end up with RWE datasets that cannot be traced back to their sources, cannot demonstrate chain of custody, and cannot satisfy FDA’s provenance requirements.

Vendor oversight gaps

RWE programs typically involve data vendors, contract research organizations, and analytics platforms. Each is a supplier under your supplier quality management program. In practice, many companies treat data vendors as IT vendors with minimal quality oversight, not as suppliers whose outputs affect the integrity of a regulatory submission. FDA expects the same qualification, contracting, and audit oversight for data vendors as for any other critical supplier.

Uncontrolled data transformations

Data cleaning and transformation scripts are not always treated as controlled documents. A Python script that removes duplicate records or recodes a variable is, in the context of RWE, a critical step in the data provenance chain. If that script is not version-controlled, not subject to change control, and not validated for its intended function, the data it produces cannot satisfy FDA’s traceability requirements.

No formal RWD source qualification

Just as a pharmaceutical manufacturer qualifies raw material suppliers before using their inputs in a product, companies using RWD should formally qualify their data sources. Qualification includes assessing completeness, accuracy, consistency, and fitness for the intended analytic purpose. This is an emerging quality practice that most companies have not yet formalized into their QMS.

Building RWE data quality into your QMS

Establish a data governance policy

Define what constitutes a controlled data asset in your organization, who is responsible for its quality, and what quality controls apply at each stage of the data lifecycle. This policy sits alongside your document control and CAPA policies as a foundational quality document.

Apply supplier qualification to data vendors

Classify data vendors, analytics platform providers, and CROs managing RWD as critical suppliers. Require quality agreements that specify data standards, change notification requirements, and audit rights. Conduct periodic audits or assessments of their data quality controls, and document the results in your supplier qualification records.

Version-control all data transformation code

Treat data processing scripts the same as controlled documents. Every version must be archived, every change must go through change control with a documented rationale, and the current version must be unambiguously identified. This does not require a specialized software repository; your existing document control system can manage scripts as controlled files if your change control workflow is applied consistently.

Pre-specify and lock analysis plans

Create an SOP for pre-specification of RWE analysis plans. The SOP should define when an analysis plan must be finalized, what approval is required before data access, and what change control process applies if the plan must be amended after data access begins. Any amendment after data access is a significant quality event that should be documented with full rationale and tracked through your deviation management system.

Document fitness-for-purpose assessments

Before using any new RWD source, complete a formal fitness-for-purpose assessment. Document the data source characteristics, the analytic question it will be used to answer, the assessment of whether the source captures the required variables at the required resolution, and the conclusion with supporting evidence. This document should be maintained as a controlled quality record and referenced in any submission that uses that data source.

The connection between RWE quality and post-market surveillance

For medical device companies, RWE is closely tied to post-market clinical follow-up (PMCF) and post-market surveillance programs required under EU MDR and FDA’s QMSR. The data collected through post-market studies is real-world data by definition. If your post-market surveillance system does not meet the data quality standards FDA and EMA apply to RWE, the data you collect will not support the regulatory decisions those programs are designed to enable.

This is not a theoretical risk. FDA has begun scrutinizing the data quality practices behind post-market submissions with the same rigor it applies to pre-market submissions. Companies that built their post-market programs on informal data collection processes are finding that those processes are inadequate when a label expansion or comparative effectiveness claim requires RWE-grade data quality.

What this means for your eQMS

Supporting RWE data quality requires a QMS platform capable of managing the full scope of quality activities these programs generate: vendor qualification records, data governance policies, version-controlled analytical plans, fitness-for-purpose assessments, change control records for data transformations, and audit trails for every step in the data lifecycle.

Cloudtheapp’s platform gives regulated companies the infrastructure to manage all of these activities within a single, validated system. With 60+ configurable quality applications covering supplier qualification, document control, deviation management, and risk assessment, Cloudtheapp provides the quality management foundation that RWE programs require, without the complexity of building a separate data governance system from scratch. Request a demo to see how it works.

Frequently asked questions

Does my QMS need to change if we’re not currently submitting RWE to FDA?

If your company collects any post-market data, patient-reported outcomes, registry data, or real-world data that might eventually support a regulatory decision, yes. Building RWE-compatible data quality practices into your QMS now is far less expensive than retrofitting them after FDA questions the quality of data in a submission. The qualification, traceability, and governance practices described here are good quality practice regardless of their regulatory purpose.

How does FDA distinguish acceptable RWD sources from unacceptable ones?

FDA assesses fitness for purpose based on whether the data captures the variables, time periods, and population characteristics relevant to the specific regulatory question. There is no list of pre-approved data sources. Each use of a data source must be justified based on its characteristics and the question it is being used to answer. That justification is a quality document that FDA can request during a submission review.

Are electronic health records automatically suitable as RWD for FDA submissions?

No. EHR data quality varies significantly across institutions, and the data was collected for clinical care, not research. Before using EHR data for RWE, sponsors must assess completeness, coding consistency, missing data rates, and how well the EHR captures the specific clinical variables relevant to their question. That assessment is a documented quality activity with a formal conclusion.

Conclusion

Real world evidence is reshaping how FDA evaluates post-market performance and supports regulatory decisions. Quality teams that build RWE-compatible data governance into their QMS now will have a strategic advantage: data that FDA trusts, submissions that withstand scrutiny, and post-market programs that generate genuinely useful evidence. Quality teams that treat RWE as a statistics problem rather than a quality problem will discover the gap when a submission is questioned or a post-market commitment cannot be satisfied with the data they collected.

This post created by and appeared first on Cloudtheapp