Process Capability and Statistical Process Control

Abstract

Pharmaceutical drug substances and drug products are manufactured following a well-defined process in units of batches. Drug product batches consist of a large number of dosage units (tablets, capsules, patches, syringes, vials, etc.) Such processes have inherent variability in critical quality attributes (CQA) as a consequence of incipient differences in starting materials, environmental conditions, equipment control setting and other uncontrolled sources of variability at both the batch and dosage unit levels. Regulatory and business considerations intended to safeguard patient health, require careful statistical monitoring of the batch’s CQAs. Adherence to specifications of every batch of drug product is a regulatory requirement for marketing. This chapter summarizes the statistical methods used to carry out the statistical characterization of the manufacturing process’s ability to consistently produce acceptable product (process capability) and associated monitoring tools (statistical process control). For reporting purposes, a process capability index (e.g. C_pk) or process performance index (e.g. P_pk) is calculated based on two quantities: (1) the variability of the CQA (standard deviation, σ), and (2) the comparison of that variability with the specification, typically as the quotient of the specification width to 6σ. Statistical process control refers to graphical methods that permit detection of departures from target over time of manufacture by revealing process shifts greater than capability or inherent variability alone would allow. Details of these charts are described along with additional rules for increased vigilance in detecting assignable causes (beyond capability) of variation. Alternative charts for serially correlated data are also shown, and recommendations for the use of these tools in practical applications and for reporting quality indices are provided.