Beyond Data Cleaning: A Strategic Shift for Clinical Data Management under ICH E6 (R3)

Authored by

The completion of ICH E6 (R3) represents a significant development in clinical research, fundamentally altering the approach to defining, implementing, and maintaining quality standards. Clinical Data Management (CDM) is undergoing a major change, moving away from everyday tasks and placing greater emphasis on delivering strategic support.

CDM has traditionally relied on extensive oversight, using multiple edit checks, generating a large volume of queries, and conducting wide-ranging Source Data Verification (SDV). While these practices aimed to ensure accuracy, they often emphasized quantity over meaningful outcomes. Adding more checks didn’t always lead to better data quality; it just increased the workload.

E6 (R3) changes that narrative by introducing proportionate, risk-based oversight. The focus is no longer on reviewing everything, but on prioritizing what truly matters. This begins with identifying Critical-to-Quality (CtQ) factors early in study design and aligning all downstream activities accordingly. This shift forces more deliberate decision-making:

• Which data points directly impact primary endpoints?
• Which processes carry the highest risk to data integrity?
• Where can controls be reduced without compromising reliability?

From Error-Free to Fit-for-Purpose: Redefining Data Quality Goals

A key enabler of this transformation is Quality by Design (QbD). Rather than inspecting quality at the end, CDM must now help build it into the study from the start. This means earlier involvement in protocol development, proactive identification of CtQ variables, and thoughtful design of data flows across systems.

When CDM contributes to the design stage, downstream inefficiencies are significantly reduced. Targeted edit checks replace blanket validations. Reconciliation challenges are addressed in advance. Thanks to steady quality management, database locking is efficient and doesn’t rely on urgent corrections.

The operational impact is immediate. Edit check strategies become more focused, prioritizing critical variables over exhaustive coverage. Query management shifts from volume-driven to signal-driven, emphasizing meaningful discrepancies over minor inconsistencies.

Traditional reliance on 100% SDV is declining, making way for risk-based and centralized review approaches. Static listings are no longer sufficient to detect issues in complex datasets. Instead, CDM teams increasingly rely on data visualization, trend analysis, and real-time monitoring to identify anomalies early.

Perhaps one of the most significant changes is the redefinition of database lock. It is no longer a high-pressure milestone at the end of a study. Instead, it becomes a natural outcome of continuous data quality and ongoing oversight.

At the same time, the data landscape itself has evolved. Clinical trials now operate across multiple systems, EDC, ePRO, laboratory platforms, wearables, and electronic health records. E6 (R3) acknowledges this complexity and raises expectations for data governance across the entire ecosystem. CDM is now responsible for ensuring:

• Systems are validated and fit for purpose
• Data lineage is clearly defined and traceable
• Audit trails meet ALCOA+ principles
• Source data is consistently identified across platforms
• Data is controlled from collection through archival

CDM: From Data Handler to Ecosystem Coordinator

E6 (R3) introduces defensibility as an essential consideration, so meeting compliance standards now involves adhering to procedures and also being able to explain and support decisions made. Teams must be able to clearly explain:

• Why certain controls were applied or removed
• Why specific data points were deemed critical
• How risk assessments influenced strategy

In this new regulatory landscape, effort alone is not visible; only rationale is. Every decision must be documented, risk-based, and inspection-ready. Adding to this transformation are timelines that have become more condensed than ever before. Regulatory requirements, such as results reporting within strict post-completion windows, demand faster database lock and submission readiness. This eliminates the feasibility of traditional, end-stage data cleaning models.

Rather, CDM should facilitate ongoing data monitoring and timely data cleaning. Data quality must be maintained consistently throughout the duration of the study, rather than being addressed solely at its conclusion.

Platforms like Datacise® are designed to support this evolution of CDM from operational execution to strategic oversight. By enabling centralized data review, real-time visualization, and CtQ-focused monitoring across multiple data sources, Datacise® helps teams move beyond volume-driven checks toward proportionate, risk-based decision-making. Instead of chasing late-stage issues, CDM teams can continuously assess data trends, document rationale, and demonstrate control throughout the study lifecycle, aligning data quality activities with the expectations of ICH E6 (R3) while supporting faster, more defensible database lock.

Ultimately, E6 (R3) redefines the role of CDM. It is no longer centered on data cleaning and query resolution, but on data strategy, system design, and lifecycle governance. High-performing CDM teams will be those that can:

• Operationalize CtQ-driven thinking
• Design intelligent, risk-based data collection and validation strategies
• Manage and integrate multi-source data ecosystems
• Ensure full traceability and audit readiness
• Defend every decision with clear, evidence-based rationale

The philosophical shift is clear. CDM is no longer measured by how much is checked, but by how effectively systems are designed to produce reliable data from the outset.

The question has changed. It is no longer ‘did we review everything?’, it is now ‘did we focus on what truly matters and can we prove it?’ That is the standard E6 (R3) sets; meeting it requires CDM to think differently, act strategically, and lead from the front. E6(R3) promotes proportionate, risk-based approaches that enable more efficient use of resources, with the potential to reduce unnecessary operational costs.