Statistical evaluation of stability data: criteria for change-over-time and data variability

Area of Science:

• Pharmaceutical Sciences
• Statistical Analysis
• Drug Stability Evaluation

Background:

• ICH Q1E guidance emphasizes statistical evaluation of drug stability data.
• Current guidance lacks specific measures for data change-over-time and variability.
• Objective assessment of stability data is crucial for shelf-life and retest period determination.

Purpose of the Study:

• To demonstrate the application of simple statistical parameters for evaluating drug stability data.
• To propose p-value and Cpk as measures for change-over-time and variability, respectively.
• To facilitate a more unified approach to stability data evaluation.

Main Methods:

• Linear regression analysis of stability data.
• Utilizing the p-value of the slope to assess change-over-time.
• Calculating the minimal process capability index (Cpk) from the regression line's standard deviation to measure variability.

Main Results:

• A p-value threshold of 0.25 was suggested for insignificant change.
• A Cpk threshold of 2.5 was proposed for insignificant variability.
• Demonstrated effectiveness on refrigerated drug product and drug substance data.

Conclusions:

• The proposed statistical parameters (p-value and Cpk) effectively characterize stability data.
• These parameters enable a unified and objective evaluation of stability data according to ICH guidelines.
• Facilitates consistent shelf-life and retest period determination.