Complete Case or Full Dataset? Choosing the Right Analysis Approach for Missing Data

Handling missing data is a critical decision in clinical trial analysis. Two commonly considered approaches are Complete Case Analysis (CCA) and Full Dataset Modeling (e.g., MMRM or Multiple Imputation). Choosing between them requires understanding the underlying assumptions, data structure, regulatory expectations, and impact on validity.

This guide explores when it is appropriate to use complete case analysis versus full dataset methods in biostatistical evaluations. We’ll also discuss the regulatory context from agencies like the USFDA and EMA, and offer practical recommendations to guide your decision-making process.

Understanding Complete Case Analysis (CCA)

Complete Case Analysis involves analyzing only those subjects for whom all relevant data are available. Any patient with missing data on the outcome or a key covariate is excluded from the analysis.

Advantages of CCA:

• Simple to implement and interpret
• Works with standard statistical tools
• No modeling assumptions about the missing data

Limitations of CCA:

• Leads to loss of sample size and statistical power
• Results may be biased if data are not Missing Completely at Random (MCAR)
• Cannot be used when missingness is high or systematic

When to Use CCA:

• When the proportion of missing data is low (<5%)
• When data are MCAR (i.e., probability of missingness is unrelated to both observed and unobserved data)
• When conducting exploratory or supportive analyses

CCA may be acceptable under specific circumstances, but its limitations must be clearly stated in the trial documentation.

Understanding Full Dataset Analysis

Full Dataset Analysis refers to techniques that incorporate all available data, including cases with partial information. Examples include:

• MMRM (Mixed Models for Repeated Measures): Accommodates MAR (Missing at Random) data
• Multiple Imputation: Uses observed data to predict and fill in missing values
• Maximum Likelihood Estimation: Accounts for partial data without explicit imputation

Advantages of Full Dataset Methods:

• Preserves statistical power by using all available information
• Yields unbiased estimates under MAR assumptions
• Widely accepted by regulatory agencies

Limitations:

• Requires correct specification of the model
• May be computationally intensive
• Assumptions (like MAR) must be justified

These methods are favored in regulatory reviews, especially for primary endpoints. Their inclusion in the Statistical Analysis Plan reflects best practice in handling missing data.

Regulatory Guidance: CCA vs Full Dataset

Regulators discourage CCA as a primary analysis method unless MCAR can be assumed and justified. For pivotal trials, agencies like the FDA and EMA recommend full dataset approaches with appropriate sensitivity analyses.

Key Guidelines:

• FDA Guidance on Missing Data (2010): Emphasizes pre-specification and avoidance of CCA
• ICH E9(R1): Introduces estimands that define the role of intercurrent events like dropout
• EMA Guideline on Missing Data: Encourages model-based analyses with sensitivity checks

Documentation of methods and justification of assumptions is critical for regulatory compliance.

Practical Comparison: When to Choose What

Best Practices for Implementation

• Include both CCA and full dataset methods in SAP as primary and supportive analyses
• Clearly define assumptions about missing data mechanisms
• Perform and report sensitivity analyses (e.g., tipping point, delta adjustment)
• Use statistical software with validated imputation modules
• Document rationale and results per SOPs and in the CSR

Conclusion

The decision to use complete case analysis or full dataset modeling should be driven by data characteristics, missingness mechanisms, and regulatory requirements. While CCA is easy to apply, it is limited to rare MCAR situations and should only be used as supportive analysis. Full dataset approaches like MMRM and multiple imputation offer robust solutions under MAR and are preferred in regulatory submissions. Incorporating both strategies—alongside transparent assumptions and sensitivity analyses—ensures your trial results remain valid and defensible.

How to Meet Regulatory Expectations for Missing Data in Clinical Trials

Missing data in clinical trials can threaten both the credibility and regulatory acceptability of your study results. Regulatory authorities such as the USFDA, EMA, and CDSCO expect sponsors to proactively plan for, minimize, and transparently report all aspects of missing data. Failure to do so can lead to delayed approvals, requests for additional trials, or outright rejection.

This tutorial provides a comprehensive overview of regulatory expectations regarding missing data—covering how to document, analyze, and justify your approach. It also discusses strategies to align with key guidelines such as ICH E9(R1) and the FDA’s “Guidance for Industry on Missing Data in Clinical Trials.”

Why Regulatory Authorities Prioritize Missing Data

Regulators require clarity on how missing data may have influenced study conclusions. They expect the sponsor to:

• Plan for missing data prevention and mitigation in the protocol
• Analyze the potential impact of data loss on trial outcomes
• Conduct appropriate sensitivity analyses
• Document everything in the SAP and Clinical Study Report (CSR)

In short, missing data isn’t just a statistical issue—it’s a matter of trial integrity, reliability, and ethical responsibility.

1. Documenting Missing Data in Protocol and SAP

Both the clinical protocol and the Statistical Analysis Plan (SAP) should address missing data explicitly. According to ICH E9(R1), this includes:

• Identifying the estimand and how intercurrent events like dropout affect it
• Describing strategies for preventing missing data (e.g., flexible visit windows, retention efforts)
• Pre-specifying statistical handling approaches (e.g., MMRM, Multiple Imputation, LOCF)
• Defining sensitivity analysis plans to assess robustness under MNAR assumptions

Failure to specify these elements may raise red flags during regulatory review and compromise GMP compliance.

2. Analysis Requirements in the CSR

Clinical Study Reports (CSRs) submitted to regulators must clearly report:

• Extent and reasons for missing data
• Number of missing observations by treatment arm and timepoint
• Statistical models used for handling missingness
• Sensitivity analysis results and interpretation

Transparency is critical. Sponsors should avoid selective reporting or retrospective justifications for missing data handling.

3. Regulatory Preference for Certain Statistical Methods

Acceptable Approaches:

• MMRM (Mixed Models for Repeated Measures): Appropriate under MAR assumptions
• Multiple Imputation (MI): Widely supported if implemented correctly
• Pattern-Mixture Models: Useful for MNAR sensitivity analysis

Discouraged Methods:

• LOCF (Last Observation Carried Forward): Discouraged as a primary method due to unrealistic assumptions
• Complete Case Analysis: Acceptable only under MCAR, which is rare

To demonstrate compliance with regulatory standards, sponsors should include sensitivity analysis methods aligned with ICH stability principles and current statistical practices.

4. Reporting Missing Data by Reason and Mechanism

Regulators expect missing data to be classified by reason (e.g., AE, withdrawal of consent, lost to follow-up) and potentially by missingness mechanism:

• MCAR: Missing Completely at Random
• MAR: Missing at Random (most common)
• MNAR: Missing Not at Random (most difficult to handle)

Although the missing data mechanism is untestable, the classification provides a framework for sensitivity analysis and modeling choices.

5. Regulatory Guidelines on Missing Data

Key Guidance Documents:

• FDA Guidance on Missing Data (2010)
• EMA Draft Guideline on Handling Missing Data
• ICH E9(R1) Addendum on Estimands

These guidelines stress the importance of planning, pre-specification, and transparency in handling missing data. Non-compliance may lead to major findings during regulatory audits.

6. Sensitivity Analysis Expectations

Sponsors must demonstrate that their results are robust under alternative missing data assumptions. Typical methods include:

• Delta-adjusted multiple imputation
• Tipping point analysis
• Pattern mixture models

These analyses help reviewers assess whether conclusions hold if missing data mechanisms differ from assumptions used in primary analysis.

7. Real-World Example: EMA Rejection Due to Missing Data

In a 2019 case, EMA declined approval of a CNS drug because the trial failed to appropriately handle high dropout rates. The sponsor used LOCF as the primary imputation strategy without sensitivity analyses, leading to doubts about the treatment’s efficacy. This underscores the need for regulatory-aligned strategies.

8. Internal SOPs and Training

To ensure compliance, sponsors should develop internal SOPs that mandate:

• Inclusion of missing data strategies in protocol/SAP
• Documentation of all imputation methods
• Clear communication with CROs and vendors
• Regular training on evolving regulatory guidance

Integrating these steps into validation protocols also ensures inspection readiness and internal consistency.

Conclusion

Regulatory expectations for missing data are stringent and evolving. Sponsors must anticipate and prevent data loss wherever possible, document their assumptions, and transparently analyze and report missing data in compliance with global standards. By adhering to ICH, FDA, EMA, and CDSCO guidance, and by embedding these practices into trial design and reporting systems, sponsors can significantly improve their chances of regulatory success.