CPV Trend Limit | Statistical control limit, process control limit, CPV limit | Limit derived statistically, or justified scientifically, for use in process trending. Limit is meant to predict future process performance based on past performance experience and is not necessarily linked to process or patient requirements. In a capable process, CPV trend limits will be tighter than other limits, ranges, or specifications that are required by the molecule's control strategy. |
Initial CPV | Early CPV | Initial implementation of a CPV plan, using available data, which is often limited. Acknowledging the uncertainty that is a result of limited data, accommodations may be made during initial CPV in the method of limit calculation, the types of statistical signals that are evaluated, or the required steps for signal evaluation. |
Late CPV | Long-term CPV | Period marked by the transition to and ongoing execution of a CPV plan, that is based on adequate data to predict long-term performance. |
Long-Term CPV Trend Limit | Permanent limit | Revised limits set after adequate process data have accumulated to predict long-term performance. Long-term limits are not expected to need revision in the absence of process change, but may need to be updated as changes occur. |
Non-conformance | Deviation, investigation | A breach of a CPV trend limit or good practice that results in a formal investigation managed by the quality system. A non-conformance typically must be resolved prior to batch disposition. |
Preliminary CPV Trend Limit | Temporary limit, initial limit | Limits set as part of the initial CPV plan when limited data are available to predict long-term process performance. |
Statistical Signal | Flag, rule violation, statistical event, statistical alert | Indication of unexpected process variation that is triggered by a violation of a statistical rule in a parameter or attribute being tracked in a CPV plan. |