FAQs: <905> Uniformity of Dosage Units

1. We are testing a multiple component tablet for Uniformity of Dosage Units. One of the components meets the requirement for testing by the Weight variation procedure. If we test this one component and it passes the criteria in the chapter can we assume that the other components will pass too?

<905> applies to each drug substance in the product. <905> procedures can be applied to each substance individually so if one or more meet the conditions summarized in Table 1, it can be tested using the Weight variation procedure.

2. The Acceptance Value calculated for the Content uniformity procedure with 10 units tested is 16.4% (average is 105.9, SD is 5.0). This result is more than the limit L1% (15.0%) and we tested 20 more units. The Acceptance Value is still 16.4%. Does this pass the test because the Acceptance Value is less than L2% (25.0%) and no result is less than (1 – (0.01)(L2))M, (76.125%), and no result is greater than (1+(0.01)(L2))M, (126.875%)?

The limit for the Acceptance Value is 15.0% at both levels of the test. L2 is only used to determine the acceptable range of individual values observed when 30 units have been analyzed. Where the average result is more than 101.5% then M has the value of 101.5% and the range is calculated as in the question.

3. Can the results obtained from <905> be used to calculate the Assay results?

Section 5.70 of the General Notices says that when the same analytical methodology is used for the Assay and the Content uniformity with appropriate allowances made for differences in sample preparation, the average of all the individual-unit content uniformity determinations may be used as the Assay value.

4. The calculations for the Uniformity of Dosage test by the Weight variation procedure use the average of the individual weights in the formula for calculating the estimated contents of the units tested. Is this average from the units tested for weight variation or from the average weight of the units used in the assay that determines the value, A, used in the calculation?

The average weight of the units being tested by the Weight variation procedure is used in determining their estimated content.

5. Following the procedure in <905> we get 10 or 30 individual results. Must all these results be within the Assay limits?

The Assay is a separate test from <905>. It uses the analysis of a composite from multiple units. <905> evaluates uniformity of dose with the limit on the Acceptance Value and an acceptable range of individual values if testing proceeds to 30 units. In neither of these tests is there a requirement that individual results conform to the Assay limits, typically 90% to 110% of label claim.

6. We have an uncoated tablet with more than 25 mg of drug and the content is more than 25% of the tablet weight. We performed the Content uniformity procedure. Must we now perform the Weight variation procedure?

The Weight variation procedure may be used to estimate the contents of the individual units under a few defined conditions.  The results of a single procedure, Content uniformity or Weight variation, is sufficient to satisfy the test requirement in the chapter.

7. Before 2006, <905> had separate procedures and criteria for transdermal systems, suppositories and inhalations in pre-metered dosage units. These dosage forms have disappeared from the chapter. How are they tested?

As a result of harmonization, the separate procedures and criteria for transdermal systems, suppositories and inhalations in pre-metered dosage units were removed from the chapter. These dosage forms are now classified as Others and require the use of the Content uniformity procedure. This is the same for any dosage form not meeting the conditions for the application of the Weight variation procedure.

8. How can syrup in a multiple dose container be tested for uniformity of dosage units?

<905> is not applied to liquids in multiple dose containers.

9. We have a target value, T, which is 107.5% of the label claim. How is this used in the calculation of the AV?

The value of T is used to modify the upper value of M. T represents a manufacturing target that is more than 100% (see General Notices, Section 4.10.20 Acceptance Criteria). Where the average value of the results is more than T, the Acceptance Value is calculated based on the absolute value of the difference (plus ks). Examples where T has the value of 107.5%, 10 units have been tested and SD is 3 are given in the following table.

Mean T Is mean within 98.5% and T Reference Value Abs Value of the Difference (mean and reference value) ks AV
105.0 107.5 yes 105.0 0 7.2 7.2
109.5 107.5 no 107.5 2.0 7.2 9.2
10. How is the tablet weight used in the Content uniformity procedure?

The values observed from individual units tested by the Content uniformity procedure are in terms of the label claim.  Outside of being part of the analysis of the drug substance content, the mass of the units are not considered.

11. We produce a sterile solid packaged in single unit containers that is prepared from a true solution and freeze-dried in the final container. We test this product by <905> using the Content uniformity procedure, but in <905>, Table 1 shows that the Weight variation procedure can also be used. Can we test this product using the Weight Variation procedure in place of the Content uniformity procedure?

The uniformity of dosage units can be demonstrated by either of the two methods. The Content uniformity procedure may be used in all cases. The Weight variation procedure can be used where permitted as detailed in Table 1. The Weight variation procedure may be used for products for which it is permitted.

12. What is a simple way to express the acceptance criteria outlined in GC 905 in a product specification document?

Assuming L1 = 15.0 and L2 = 25.0, an example is as following:

Test Method Acceptance Criteria
Content Uniformity USP <905>

Level 1: AV (10 units) NMT 15.0%

Level 2: AV (30 units) NMT 15.0%
No Individual unit less than (0.75*M)% or greater than (1.25*M)%


13. When doing the L2 evaluation of large deviations for individual results, why is the criterion set relative to the reference value rather than either the mean value or the target value?

Prior to harmonization, Chapter <905> contained criteria on the number of individual tablets and capsules falling outside 85.0% to 115.0% of label claim, and the zero-tolerance criterion and that no result falls outside 75% to 125% of label claim, and a criterion for the %RSD.  The harmonized <905> test does not include any criteria on individual results falling outside 85.0% to 115.0% label claim but as in the previous test, does include a similar requirement that all individual results are between 75.0%–125.0% of a value, M. M is 98.5% of label claim (LC) if the sample mean is <98.5% LC, 101.5% LC if the sample mean is >101.5% LC, and the sample mean otherwise. The 98.5%-101.5% interval is termed the indifference zone.  In addition, the harmonized test has a criterion on a value called Acceptance Value (AV), which must be no more than 15.0. The AV is a function of deviation of the mean from label claim by more than 1.5% (the indifference zone mentioned above) and the standard deviation of the results.  The harmonized test was shown to be as, at least, conservative than the prior chapter <905>.

14. There are criteria labeled as L1 and L2. When you test 30 dosage units (which is sometimes referred to as Level 2 testing), should the AV be

No, the AV criterion is always AV ≤ L1 which is 15.0% unless otherwise specified.  The L2 criterion is only used to evaluate large deviations of individual results.

15. How do I determine the correct value for T? Our formulation includes a 2% overage of drug, because we have processing losses and this will, on average, give us potency of about 100%. Does the appropriate value of T depend on the actual potency we get in each lot?

In this case the correct value of T will be 100%.  The Target value is the intended manufacturing target for the process not the amount of drug added to the process relative to the label claim.  This value does not change from lot to lot based on the actual results and is not calculated based on the apparent mean.  So, you should be using the reference value from Case 1 in this situation for all lots.

16. What are the units for Acceptance Value (AV)? Should I write it as 15.0 (unitless) or 15.0%?

The units for all the calculations are % of label claim (LC) and it is best to report the AV as either a % or as % of LC.

17. Can you provide some example data and results calculated on that data set so that we can verify we are performing our calculations correctly?

Here are 3 examples [Note: The units for all values in the table is % of LC]:

  Data Set #1 Data Set #2 Data Set #3
Level 1 Results 93.23
Mean 100.40 99.10 98.98
StDev 5.82 6.60 6.50
Reference Value (M) 100.40 99.10 98.98
Acceptance Value (AV) 13.97 15.84 15.60
Pass/Fail Pass Fail Fail
Level 2 Results   96.95
Mean   98.46 98.31
StDev   5.35 7.38
Reference Value (M)   98.50 98.50
Acceptance Value (AV)   10.73 14.94
Large Deviation Results (Sn)   0 1
Pass/Fail   Pass Fail