Zero Acceptance Number Sampling Plans and the FDA

Pharmaceutical sampling


There has been some debate over using the MIL-STD-1916 acceptance sampling plan over the ANSI/ASQ Z1.4-2003 (R2013) sampling plans.  The opinion is that the ANSI/ASQ Z1.4-2003 (R2013) is outdated and no longer an acceptable method of determining a qualification sample plan and the MIL-STD-1916 should be used in place of ANSI/ASQ Z1.4-2003 (R2013). Do you have information around this debate over which sampling plans are acceptable by the FDA?


FDA does not (and can not) tell you what sampling plan is to be used.  The FDA requirement is that the plan be statistically valid.  As long as you follow the regulation, you are meeting FDA requirements.

In medical device manufacturing the key point is to have the plan accept on zero defectives.  This point is not FDA but legalese.  It is based on past lawsuits.  The plan “Zero Acceptance Number Sampling Plans” by Nicholas L. Squeglia (available from ASQ) has been widely adopted for this reason.

ANSI/ASQ Z1.4 in not outdated and continues to be widely used.  It is the American National Standard Institute (ANSI) version of MIL-STD-105 which the government discontinued maintaining, allowing ANSI to maintain it along with many, many other MIL-STD’s as a government cost reduction.

MIL-STD-1916 can be used but it is not widely used because of its difficulty and practical use.

James Werner

AQL Clarifications

Automotive inspection, TS 16949, IATF 16949


I am confused about the values used for AQLs. For example in Table II-A the AQL values range from 0.010 to 1000. Where do these values come from and what do they mean?

The table states, “AQLs, in Percent Nonconforming Items and Nonconformities per 100 Items .” At first I thought the values were percentages, but how can you have more than 100, as in 100%, as the values go up to 1000? Also how can there be more than 100 nonconformities per 100 items, unless one part can have multiple nonconformities?

Just looking for clarification on the AQL numbers, what they mean, and how to interpret them.


Let’s start with the definition of Acceptable Quality Level (AQL).  From Z1.4, the AQL is the quality level that is the worst tolerable process average when a continuing series of lots is submitted for acceptance sampling.  Although individual lots with quality as bad as the AQL can be accepted with fairly high probability, the designation of an AQL does not suggest that this is necessarily a desirable quality level. The AQL is a parameter of the sampling scheme and should not be confused with a process average which describes the operating level of a manufacturing process. It is expected that the product quality level will be less than the AQL to avoid excessive non-accepted lots.

The columns with percentages greater than 100% should not be included in the standard, but remain as indication of how to interpret lots where the entire sample is defective.  It has some statistical relevance with use of the switching rules, but for the general practitioner, it should be ignored.

Hope this helps.

Steven Walfish



My question concerns the process performance metric DPMO (defects per million opportunities). I want to use this to quantify a particular supplier’s performance. My question is, is the number of defects referred to in the calculation the number of defects produced by the supplier (in which case it would involve data I don’t have access to), or is it the number of defects experienced by the customer (which is us)? I of course can count the number of defects we receive from the supplier, but if this metric is supposed to be based on the number of defects produced by an organization, I would have no way of knowing how many defects are produced by the supplier’s process, but contained within the supplier’s facility. My hope is to be able to characterize the supplier’s process performance in terms of sigma level.

The DPMO metric is not usually considered a point estimate of the true percent defective in the lot (either at the supplier or customer site).  It is a relative performance metric used to equate the observed percent defective from a sample to defective units per million opportunities.  If a supplier culls out all the defective units before shipping to you (i.e. perfect inspection system), your internal DPMO would be 0, even if the supplier DPMO is high. If your goal is to characterize the supplier’s process performance in terms of sigma level, you would need their data, as the data you collect internally is just an estimate for the average outgoing quality from the supplier and not their process performance.

Steven Walfish

Document Revision Criteria

ISO documentation practices, requirements


Is there any criteria available for the frequency of document revision in ISO 9001 or ISO 13485?  Some organization don’t revise the documents for a period of more than 2-3 years.  The reason provided by the organization is that there were no changes during this period. Do ISO standards mandate the revision of documents within a certain time frame? Can we treat this as non-compliance, if the documents are not revised over a period of 2-3 years ?

There are no criteria nor a requirement for document revision in ISO 9001:2015, 7.5.

ISO 13485:2016, 4.2.4, states, “review, update as necessary and re-approve documents.” This leave the review to the discretion of the organization.

Thus, there is no mandatory review frequency and no non-conformance if documents are not revised within a determined time frame.  ISO 13485 does require a review, however. But, the frequency of the review is not mandated.

George Hummel