ECA Survey Results II: How is Statistical Process Control used in the Pharmaceutical Industry?

Phase 3 (ongoing/continued process verification) is the longest phase in the validation life cycle. This phase is intended to demonstrate that the process continues to run under validated conditions. The aim is to identify “drifts” and “shifts.” Many companies use statistical process control (SPC) for this purpose. Concept Heidelberg conducted a survey for the ECA specifically on this topic. Among other things, the ECA wanted to know whether SPC is also used in other areas, which control charts are used, what limit values exist for process capability indices, etc.
 
34 participants took part in the survey. Not every participant answered every single question: sometimes several answers were given by the same person, and rounding also contributes to the fact that the total sum of the answers does not always add up to 100%. Below you  will find part II of the survey results.

Biographical Data

Most participants came from companies with more than 500 employees (44%), 38% of participants work in companies with 101-500 employees, 12% come from companies with 51-100 employees, and 14% from companies with 1-50 employees.

The majority of participants (74%) manufacture pharmaceuticals. Chemical active ingredient manufacturers accounted for 32% and manufacturers of biological pharmaceuticals for 21%. Manufacturers of medical devices also accounted for 18%, and 3% came from excipient manufacturers.

Survey results Part II

Which control charts are used? Here, too, multiple answers were possible.
The use of X/S charts (50%) and I/MR charts (49%) were mentioned almost equally. Control charts for attributive characteristics are also used at a comparably high rate (47%). Levey-Jennings charts (13%) and Cusum charts (9%) are also used.

Surprisingly, many participants have no problems with SPC (90%). Consequently, only 10% have problems with SPC. Biological products and attributive data were mentioned as problem areas.
 
The answers to the question of which statistical methods are used were also interesting. With 55%, the Nelson rules were in first place, followed by the Western Electric rules (24%). None of the participants use the Westgaard rules. At 21%, the “Other” category is also relatively well represented. The following were mentioned here: control limits vs. specification limits, use of default rules in the Minitab statistics tool, a combination of different rules, Deming rules.

And what about the use of process capability indices?

85% of participants use them. Most of them (41%) have a limit between 1.0 and 1.33, while 28% go even further and have chosen a limit of 1.33–1.66. On the other hand, 9% also accept values of 1.0.Comments on “Other” (22%) were: >1.33, depends on the amount of data, not used in PQR or production, >= 1.0

 
And what software is used for SPC?

Minitab is clearly in first place with 44%. Statgraphics and the internal LIMS system are also mentioned, each with 6%. The “Other” response is surprisingly high (41%). Excel is mentioned as a tool seven times, which is strikingly frequent.
 
Conclusion: SPC is widely used in the pharmaceutical industry without any major problems. Different control charts are used, and the limits for process capability indices are also chosen differently.