Aili Cheng, PhD, is director of Pfizer’s department of pharmaceutical science and manufacturing statistics.

Aili Cheng

	Determining critical process parameters (CPPs) is vital to defining the control strategy for drug substance and drug product manufacturing processes (1). Deciding the criticality of a process parameter, however, can often become a subjective exercise, resulting in long discussions within product development teams as well as back-and-forth communication with government regulators during new drug application review. 
	
	A data-driven statistical approach was developed (2) to help reduce subjectivity and debate in determining the criticality of process parameters in the manufacturing of drug substance and drug product. The approach uses the distance of a critical quality attribute (CQA) from its quality limit, together with the estimated parameter effect size, to designate the criticality of a process parameter. 

	The method relies on straightforward calculations. However, performing these calculations manually can be time-consuming and error-prone when the statistical model involves multi-factor interactions and higher order parameter effects. To incorporate this systematic approach into daily practice, statisticians have developed a web-based computational tool that uses this method.

	This article examines how the algorithm was developed, focusing on the calculation of process parameter effect size on a CQA. It also discusses how analytical and numerical solutions of parameter effect size resulted from conventional linear models. 

	Submitted: July 31, 2017.
	Accepted: August 11, 2017

 (ke.wang2@pfizer.com), is director, and 

, is director, all within Pfizer’s department of pharmaceutical science and manufacturing statistics. 

To whom correspondence should be addressed


	Vol. 42, No. 1
	January 2018
	Page: 46-54

	When referring to this article, please cite it as F. Li 

., "Removing Subjectivity from the Assessment of Critical Process Parameters and Their Impact," 

Articles

A new algorithm uses a statistical approach to critical process parameter assessment, allowing for faster, more consistent, and less subjective critical process parameter quantification, visualization, and documentation. 