The Importance of Data Analysis: The Foundation of Quality Assurance in Medical Device and Pharmaceutical Companies

The Importance of Data Analysis: The Foundation of Quality Assurance in Medical Device and Pharmaceutical Companies

Introduction

Regulatory authorities worldwide, including the FDA, require pharmaceutical and medical device companies to apply statistical methods (data analysis). This article examines in detail why data analysis is essential for quality assurance activities from a regulatory requirements perspective.

Why Collect Data?

The fundamental purpose of data collection is to enable quality assurance activities based on “objective evidence” rather than “intuition” or “rules of thumb.” Let us examine this difference through concrete examples.

Example 1: Evidence-Based Approach (Recommended) “Data analysis has confirmed that the top three causes of complaints for this model are A, B, and C. Therefore, resources should be allocated prioritized to A, B, and C, with focused countermeasures implemented.”

Example 2: Intuition-Based Approach (Not Recommended) “Lately, it feels like there have been many complaints caused by D. I’ve also been noticing complaints related to E and F recently. So, we should focus our countermeasures on D, E, and F.”

Example 1 is based on objective data analysis, ensuring the validity and effectiveness of countermeasures. In contrast, Example 2 is based on subjective impressions and may diverge from the actual priority of problems, potentially leading to inefficient resource allocation. Such unfounded countermeasures are unlikely to yield effective results.

What is Data?

Data refers to “unorganized information,” or “raw data.” Data itself is merely a collection of numbers or observations and is difficult for humans to directly understand with their cognitive abilities.

Therefore, it is necessary to transform Data into Information. Statistical methods play a crucial role in this transformation process.

The DIKW Model

Generally, data creates value through the following hierarchy, known as the DIKW Model:

LevelContentCharacteristics
DataUnorganized raw facts or observationsHas no meaning in isolation
InformationProcessed and organized dataHas meaning within context
KnowledgeIntegrated and systematized informationUnderstanding of patterns and relationships
WisdomAbility to apply knowledge to practical judgmentLeads to decision-making and action

Through this process, mere collections of numbers are transformed into practical wisdom.

Why Statistics Are Necessary

The most important function of statistical methods is to visualize data that cannot be understood by the human mind and make it interpretable. By statistically processing numbers and visualizing them as graphs and charts, the trends and patterns indicated by the data become clear.

Two Major Categories of Statistical Methods

1. Descriptive Statistics

This method involves collecting a data set, creating tables and graphs, and calculating statistical measures such as mean, median, and variance to understand the characteristics of the data.

Main Applications:

  • Monitoring and tracking manufacturing process status
  • Tracking temporal trends in complaint rates and non-conformance rates
  • Classification and frequency analysis of complaint causes
  • Quantitative evaluation of process performance

2. Inferential Statistics

This method involves extracting a sample from a population, estimating the characteristics of the entire population from the sample characteristics, and statistically testing the validity of that estimation.

Main Applications:

  • Design and implementation of sampling inspection in manufacturing processes
  • Design Verification involving sampling
  • Design Validation
  • Process Validation

The Need for “Appropriate” Statistical Methods in Regulatory Requirements

Historical Background: Establishment of 21 CFR Part 820

In the Federal Register dated October 7, 1996 (Volume 61, Number 195, Pages 52602-52662), the FDA promulgated the Quality System Regulation for medical devices (21 CFR Part 820). This regulation establishes comprehensive quality control requirements for the design, manufacture, packaging, labeling, storage, installation, and servicing of medical devices.

Statistical Method Requirements in 820.100

Section 21 CFR 820.100, “Corrective and Preventive Action,” subsection (a)(1), stipulates:

“Each manufacturer shall analyze processes, work operations, concessions, quality audit reports, quality records, service records, complaints, returned product, and other sources of quality data to identify existing and potential causes of nonconforming product, or other quality problems. Appropriate statistical methodology shall be employed where necessary to detect recurring quality problems.

Important Points in the FDA Preamble

In the preamble (commentary section) of the Federal Register (61 FR 52654, October 7, 1996), the FDA emphasized the following points:

  1. Use of Appropriate Tools: The FDA strongly emphasizes that when statistical methods need to be used, appropriate statistical tools should be employed.
  2. Misuse of Statistics: The FDA has repeatedly observed the misuse of statistics by manufacturers. Some manufacturers have used statistical methods not to genuinely address problems but to minimize regulatory compliance efforts. The FDA clearly states that such misuse of statistics constitutes a violation of Section 820.100.

This guidance indicates that statistical methods are not merely formal requirements but must genuinely contribute to the detection and prevention of quality problems.

Additional Requirements in 820.250 for Statistical Techniques

Furthermore, 21 CFR 820.250, “Statistical Techniques,” states:

(a) Where appropriate, each manufacturer shall establish and maintain procedures for identifying valid statistical techniques required for establishing, controlling, and verifying the acceptability of process capability and product characteristics.

(b) Sampling plans, when used, shall be written and based on a valid statistical rationale. Each manufacturer shall establish and maintain procedures to ensure that sampling methods are adequate for their intended use and to ensure that when changes occur the sampling plans are reviewed.

Latest Regulatory Developments: Transition to QMSR

Quality Management System Regulation (QMSR)

On February 2, 2024, the FDA promulgated a historic regulatory amendment. The new regulation, called the Quality Management System Regulation (QMSR), will become effective on February 2, 2026.

Key Features of QMSR

  1. Incorporation of ISO 13485:2016 by Reference The QMSR incorporates by reference ISO 13485:2016, the internationally recognized standard for medical device quality management systems.
  2. Promotion of Global Harmonization This amendment harmonizes U.S. regulatory requirements with international regulations such as those of the European Union, Canada, and Japan.
  3. Simplification of 21 CFR Part 820 The regulation is reduced from 15 subparts (A-O) to just 2 subparts (A-B), with most requirements replaced by references to ISO 13485:2016.
  4. Continuing Importance of Statistical Methods In the QMSR, the application of statistical methods continues to be maintained as an important requirement. ISO 13485:2016 Clause 8.2.5 “Monitoring and measurement of product” and Clause 8.4 “Analysis of data” also require the collection and analysis of appropriate data to demonstrate the effectiveness of the quality management system.

Preparation for QMSR Transition

Medical device manufacturers need to prepare for the effective date of February 2, 2026, by:

  • Conducting gap analysis between ISO 13485:2016 requirements and current QMS
  • Strengthening risk management methods (considering ISO 14971)
  • Reviewing documentation and traceability requirements
  • Ensuring appropriate application and documentation of statistical methods

Alignment with International Standards

Statistical Requirements in ISO 13485:2016

ISO 13485:2016 specifies requirements related to statistical methods in the following clauses:

Clause 8.2.5 Monitoring and measurement of product: The characteristics of the product must be monitored and measured to verify that product requirements have been met.

Clause 8.4 Analysis of data: Appropriate data shall be determined, collected, and analyzed to demonstrate the suitability and effectiveness of the quality management system and to evaluate where continual improvement of the effectiveness of the quality management system can be made.

Medical Device Single Audit Program (MDSAP)

MDSAP is a single audit program participated in by multiple regulatory authorities (FDA, Health Canada, Brazil ANVISA, Japan MHLW/PMDA, Australia TGA). MDSAP is based on ISO 13485, and appropriate use of statistical methods is an important evaluation item in audits.

Application to Practice

Selection of Effective Statistical Methods

The selection of appropriate statistical methods should be made considering the following factors:

ConsiderationContent
Type of DataContinuous data, discrete data, categorical data
Sample SizeLarge sample or small sample
Data DistributionIs the normal distribution assumption valid
Purpose of AnalysisDescription, comparison, prediction, elucidation of relationships
Risk LevelProduct risk class and required reliability

Common Statistical Tools

ToolPurposeApplication Examples
Control ChartsProcess stability monitoringStatistical process control in manufacturing
Pareto ChartProblem prioritizationIdentification of non-conformance causes
HistogramData distribution visualizationProcess capability assessment
Scatter PlotRelationship analysis between variablesCorrelation detection
Hypothesis TestingVerification of statistical significanceConfirmation of process improvement effects
Design of Experiments (DOE)Search for optimal conditionsProcess optimization

Conclusion

The application of data analysis and statistical methods is the foundation of quality assurance in modern medical device and pharmaceutical companies. Regulatory authorities require the application of statistical methods that genuinely contribute to the detection and prevention of quality problems, not merely as formal requirements.

In preparation for the effective date of QMSR in February 2026, companies need to build more advanced and effective quality management systems while ensuring alignment with international standards. By appropriately utilizing statistical methods, decision-making based on objective evidence becomes possible, enabling continuous improvement in product safety and effectiveness.

References

  • FDA 21 CFR Part 820 Quality System Regulation (1996)
  • Federal Register, Vol. 61, No. 195 (October 7, 1996)
  • FDA Quality Management System Regulation (QMSR), Final Rule (February 2, 2024)
  • ISO 13485:2016 Medical devices — Quality management systems — Requirements for regulatory purposes
  • ISO 14971 Medical devices — Application of risk management to medical devices
  • FDA Guidance Documents on Statistical Quality Control

Related post

Comment

There are no comment yet.