What can the variability of strength be between different brands of supposedly identical medications?

Question

I remember reading that medications are within allowable specs if they contain +/- 30% of the labelled strength. I can't find the source of that information in order to check its validity.

If that is the case, a 100mg pill would be considered to be within allowable tolerances if it has 70-130mg of the active ingredient.

This could mean a 60% change in active ingredient if a patient switches from one brand to another, or if a manufacturer changes its manufacturing process.

What is the allowed variability for prescription medications?

I imagine the law/rule could be different for every country, so please specify for which country your answer applies.

Answer 1

The range where the content of the active substance varies by 60% is way to wide in every country.

In most of the world's pharmacopoeias, both national and international, the allowed variability in the amount of the active substance can be tested either as uniformity of mass or, if the amount of labeled substance is lower than a certain predefined value, as uniformity of content. For certain dosage forms, testing the uniformity of content is a must, no matter the labeled mass of the active substance (e.g. soft capsules filled with emulsions, suspensions or gels).

Pharmacopeias prescribe the testing procedure and the acceptance criteria. Appropriate regulatory authority requests those pharamcopoeial criteria to be met.

As for the uniformity of mass, in the European Pharmacopoeia, Ph. Eur. 7.0. (the newest that can be accessed via browser search) the test is given in the monograph 2.9.5. Uniformity of mass of single-dose preparations.

Weigh individually 20 units taken at random or, for single-dose preparations presented in individual containers, the contents of 20 units, and determine the average mass. Not more than 2 of the individual masses deviate from the average mass by more than the percentage deviation shown in Table 2.9.5.-1 and none deviates by more than twice that percentage. For capsules and powders for parenteral administration, proceed as described below. [emph. mine]

Content mass of capsules and powders for parenteral administration is weighed by difference (the whole preparation is weighed, than the empty shell/primary package and the difference is the content mass).

The table mentioned states that for tablets, e.g:

• average mass is 80 mg or less (allowed) percent deviation is 10
• average mass is more than 80 mg and less than 250 mg percent deviation is 7.5
• average mass is 250 mg or more, percent deviation is 5

As for the uniformity of content, the same source (Ph. Eur. 7.0.) gives the test in the monograph 2.9.6.Uniformity of content of single-dose preparations where, again, the test and acceptance criteria vary by dosage form. To take tablets as an example, we can look at "test A" (for tablets, powders for parenteral administration, ophthalmic inserts, suspensions for injection):

The preparation complies with the test if each individual content is between 85 per cent and 115 per cent of the average content. The preparation fails to comply with the test if more than one individual content is outside these limits or if one individual content is outside the limits of 75 per cent to 125 per cent of the average content. If one individual content is outside the limits of 85 per cent to 115 per cent but within the limits of 75 per cent to 125 per cent, determine the individual contents of another 20 dosage units taken at random.The preparation complies with the test if not more than one of the individual contents of the 30 units is outside 85 per cent to 115 per cent of the average content and none is outside the limits of 75 per cent to 125 per cent of the average content. [emph. mine]

---

As for the other countries, the European pharmacopoeia also gives the test for Uniformity of dosage units in monograph 2.9.40. This test (or most of it) was harmonized between the three pharmacopoeias Ph. Eur. (European), USP (for the US) and JP (Japanese). This test is slightly more complicated than the previous, because it calculates the acceptance value based on the mean of the measured samples, standard deviation of these measurements and a coefficient based on the number of measurements. There are 6 different cases of this formula, but the simplest one is where the acceptance value is the standard deviation multiplied by the given coefficient. This value should not be greater than 15% of the prescribed content.

Regulatory authorities in these countries acknowledge this harmonisation, see the documents issued by EMeA and FDA.

---

So, finally, which of these is to be used in Europe? EMeA states:

[...] the decision on what approach to take is left to the applicant.

---

In the U.S. most of the tests are harmonised, and all are given in the USP, monograph 905 Uniformity of dosage units.

Answer 2

Country - United States

While theoretically possible, a consistent variance of 30% is highly unlikely and it is probable that the actual variance will vary from brand to brand, as well as from production facility to production facility, as well as from batch to batch. The FDA (Food and Drug Administration) mandates a high standard of quality for inspection and adjusts its actual allowed variance based on multiple criteria, including the severity of the effects of a discrepancy in dosage.

Variability is viewed through two perspectives. The rate of absorption and the extent of absorption. These two rates are compared to associated brands of similar medications and within that a difference of 20% is considered significant by the FDA and so a range of approximately 80% to 125% is used for the acceptable window. While this does suggest that a medication can vary by 45% from another brand, the FDA requires that the ratios of the rate and extent both fall within the window as well as a 90% confidence interval to be considered equivalent to another brand. Although theoretically any can vary by a large percentage, if these requirements are met then the practical effect by variance will remain low. Typically, the actual percentage of difference is closer to 10%.

Some further reading:

• An interview on bioequivalence with Dr. Robert Howland, MD
• FDA - A pamphlet on the effects of generic drugs versus brand names
• FDA - A docket explaining terms and procedures for comparison of similar medications.
• FDA - A rather terrific powerpoint on variance regulations in the U.S. as well as the criteria used by other countries. It also has a few graphs demonstrating the allowed window of 80%-125%

Answer 3

Generic drugs do not need to conduct the same phase-3 clinical trials as the initial registration studies, instead they have to meet a few requirements (same active ingredient, same drug concentration etc.) and then show bioequivalence.

Lucky has already provided a good overview regarding the accuracy of the concentration of the active ingredient. Rather more important, and where the +/-30% misconception comes from, is the bioequivalence; meaning how much of the active ingredient ends up in the system of the individual.

The criterion for approval of a generic drug by the FDA, the EMA, the TGA and WHO is that the 90% confidence interval of all key pharmacological parameters (maximum blood concentration, total blood concentration) must lie within 80% to 125% of the original drug.

This is where the +/- 30% misconception comes from. However, as this is a confidence interval, it does not mean that the average of the data can be +/-30% of the original drug - as there is variation in human data even the identical drug will have a larger confidence interval than its average.

In fact, the average difference of key pharmacological parameters between approved branded and generic drugs has been assessed retrospectively, and is around 4%, which is similar to batch differences within the same drug.

Only 3% of drugs had an average difference larger than 10%, and those are drugs where a large intra- and interindividual variety is known (ibid.)