Regulation of active ingredient content in tablets

We began to use individual tablets of over-the-counter (OTC) consumer products as samples for some introductory analytical chemistry experiments in 2006.¹ The four sample types used were soluble calcium citrate (gravimetrically as calcium oxalate monohydrate), calcium carbonate (by EDTA titration), zinc gluconate (by EDTA titration) and Aspirin®² (by visible wavelength spectrometry). The intent was to cut costs and do away with accuracy as a grading criterion for some experiments in the introductory lab course. We found that using OTC tablets had these advantages:

• Introduces the students to quality/process control analysis
• Removes analysis-accuracy-anxiety for introductory course students and instead evaluates them on obtaining and recording data, data precision, and pre- and post-laboratory calculations.
• Cuts costs. It is very expensive to purchase³ or prepare in-house reference standard mixtures, and to dispense them.

Still, using pharmaceutical tablets² raised the question: What is the allowed variability of the active ingredient in such tablets?

To add to the curiosity, an article appearing in the Toronto Globe and Mail on February 13, 2013, reported a medical study finding that a selection of commercial vitamin D tablets from numerous manufacturers contained as little as 9% and as much as 140% of the label content.⁴

None of the manufacturers contacted answered this allowed variability question. Nor was the response from Health Canada useful. After years of sporadic searching, useful information was eventually located online by using the statistical term ‘standard deviation’ in the search string. A more detailed discussion follows below, but in summary:

• It is a significant technical challenge to produce and dispense uniform solid-state mixtures.
• The analysis of multiple pharmaceutical tablets of the same batch for their active ingredient content will produce results that form a normal distribution.² (Note 1)
• Under US regulation, a maximum percent relative standard deviation (%RSD) of 7.5 is allowed for the results in a batch.
• Canada has no such regulation, so manufacturers here use the US regulation as a guideline.

Take two specific pharmaceuticals as examples. Suppose that 100 μg thyroid hormone tablets and 250 mg vitamin C tablets are both produced in tablet batches with a mean content of 100% of the label value and a %RSD of 5.0 for the scatter of the analysis values of the two pharmaceuticals (Note 1). Table 1 summarizes what this means for the two batches of tablets. The three content ranges in Table 1 are respectively one, two and three standard deviations broad. Someone taking these tablets could expect that about 95% of their tablets would contain within ± 10% of the label content value. However, about 3 in 1000 (0.3%) of the tablets would contain an amount more than 15% above or below the label value.

Canadian regulations

What is the Canadian regulation on pharmaceutical content? An official of Health Canada⁵ pointed out the appropriate section of Canada’s Food and Drug Regulations,⁶ which states that with certain exceptions, a pharmaceutical manufacturer must not sell a tablet whose active content is more than ± 10% from the label value. But in the real world no matter how accurate and precise the manufacturing process, there is always a statistical probability that this will occur. The US regulation recognizes this reality.

FDA 2004 pharmaceutical content proposal

The most useful and authoritative source found on this topic is a 51-page draft proposal document⁷ issued on January 21, 2004 by the Food and Drug Administration (FDA)² of the US government. To summarize, the analysis results for the active component of tablets in a batch fall on a normal distribution. The variability of the active ingredient derives from variability in: the uniformity of the powder blend; the tablet manufacturing process; the tablet masses; and analysis method.

This document indicated that manufacturers should aim for a mean active ingredient content of 100.5% of the label value and a maximum %RSD of about 2.4. This value was evidently not adopted. A 2008 Wiley publication for US manufacturers⁸ states that the requirement in force at that time was a %RSD maximum of 5. In 2011 the US requirement standard was further lowered.

USP test of pharmaceutical content (2011)

In 2011 the United States Pharmacopeia (UPS) (Note 2) established the current test for the pharmaceutical content of a tablet, capsule or other unit dose pharmaceutical.⁹ The pharmaceutical manufacturer is responsible for applying this quality control test. The test defines a score that combines both the accuracy and the precision² of the manufacturing process. In simplified terms, the test is:

• Randomly select 30 tablets; randomly analyze 10 tablets.
• Determine the mean result as a percentage of the label value, and the %RSD of the set of 10 values.
• Calculate the acceptance value AV₁₀ for the 10 tablets:

AV₁₀  =  |100 - Mean| + 2.4*%RSD

• The allowed maximum of AV₁₀ is 15, which corresponds to a maximum %RSD = 6.25 if the mean content of the tablets is exactly 100 % of the label value.
• If the AV₁₀ exceeds 15, analyze the remaining 20 tablets and recalculate the AV₃₀ for all 30 tablets with a lower multiplier:

AV₃₀  =  |100 - Mean| + 2.0*%RSD

• The allowed maximum of AV₃₀ is also 15, which corresponds now to a maximum %RSD of 7.5.
• Any batch is to be rejected if any one of the tablets analyzed is more than 25% above or below the batch mean content. No rationale is given for this or any other criterion.
• The text of the test implies that the mean content is normally to be between 98.5% and 101.5% of the label value, but there is no specific penalty for being outside this range.

Pharmaceutical — tablet making

Pharmaceutical tablet manufacturing is a major industry. The manufacture, analysis and regulation of all pharmaceuticals is based on what is termed Current Good Manufacturing Practice or CGMP^2,10, which is based on record keeping, ingredient specifications, statistics and statistical process control² and the related six-sigma method.²

Of the various unit operations in tablet manufacturing, mixing,² granulation² and tableting² are the most relevant unit operations for our discussion because they affect the uniformity and amount of the delivered powder blend, and therefore the variability of the tablet content.

Mixing or blending of a mixture of pharmaceutical solids to uniformity is a vast topic with elements in common to food technology, and also to the production of reference standard mixtures for industry, government and academia.³An interesting phenomenon that greatly affects pharmaceutical production is granular convection² or segregation,¹¹ more commonly known as the ‘Brazil Nut Effect’. Segregation is the tendency of a heterogeneous mixture of solids, however uniformly blended, to un-mix when jostled, due to gravity.

One component of mixing of interest is the use of binders, fillers, glidants², lubricants and disintegrants in formulations. These substances are all excipients² in pharmaceuticals. Fillers, glidants and lubricants are also ingredients in formulations of polymer resins for injection molding, a process with certain similarities to making pharmaceutical tablets. Table 2 has data for a number of pharmaceutical tablets. The percent by mass values in the table are not fully accurate for the powder blends, since all of the tablets are coated in some manner. Distributing the active ingredient uniformly in a powder blend when it is present in very small proportion, as in the cases of the thyroid hormone and the vitamin D in the table is a severe technical problem and adds cost.

Question for students

1.   According to Wikipedia, the tablet is just one of the many forms that an oral drug can take. Other forms are syrups, elixirs, suspensions and emulsions. Define ‘tablet’ and each of these alternatives. Describe what could happen to dose control and drug safety in homes, nursing homes or hospitals if all drugs were dispensed in any of those alternative forms instead of in tablet form.

2.   Why are tablets made in a wide range of colours and shapes? Can a tablet be too large or too small? Why are tablets coated? Who cannot swallow a tablet, and what do they take instead?

3.   Verify the mass percent proportion calculation for the calcium carbonate and the zinc gluconate tablets in Table 2. Gluconate is the anion of gluconic acid.²

4.   What is the function of each of these excipients: binder; filler; glidant; lubricant; and disintegrant?

5.   List and describe three examples where you observe granular convection (‘Brazil Nut Effect’) in everyday life.

6.   Suppose any single thyroid hormone tablet has a 3 in 1000 (0.3%) chance to be more than ± 15% from the label amount. What is the chance of taking two successive tablets that are more than 15% below the label value?  (Answer: a chance of 9 in 4 × 10⁶)

7.   Select any 30 tea bags from a box of tea bags. Open 10 bags and measure the mass of tea inside each. Using the statistical functions of Excel or other software, apply the test for AV₁₀. Does the batch of tea bags pass the test? If not, continue and apply the test for AV₃₀.

Note 1. Statistical functions can be applied to any normal distribution data set to determine the mean and the percent relative standard deviation (%RSD) of the set. For all normal distributions the proportions of the population falling within one standard deviation (± σ), within two standard deviations (± 2σ), and within three standard deviations (± 3σ) of the mean value are as given in Table 3.

Note 2.  The annual “United States Pharmacopeia” (USP) is the official pharmacopeia of the United States, published jointly by the “US Pharmacopeial Convention” — also known as USP — with the National Formulary as the USP-NF. Here the standards are set for prescription, over-the-counter medicines, other health care products, food ingredients and dietary supplements sold in the United States.²

Acknowledgement

The author thanks Andy Tallevi, Formulation and Process Consultant, Keller Consulting, Mississauga, Ontario. Andy was a senior pharmaceutics scientist for GlaxoSmithKline in Canada for 18 years, and lectures part-time at the Academy of Applied Pharmaceutical Sciences, a private college in Toronto. He read the article in draft form and contributed some valuable suggestions. In particular, two statements in his letter to me stand out:

“...... I can tell you though that I am familiar with the USP <905> Uniformity of Dosage Units test which you correctly reference and quote and that when we manufacture tablets that is the standard we are aiming to achieve both in a development and commercial production setting. ….... Tableting is a very complicated operation. I have spent a significant amount of time studying this, as have many others, and have tried to understand how to control it to produce good tablets. In particular the die fill volume phase is absolutely critical in determining the final tablet active content. The die fill (with blend or granulation) prior to compression determines the final tablet weight, which determines the final tablet active content. Under-filled die cavities, or worse still, inconsistently filled die cavities can cause havoc on uniformity of tablet content. A major factor here is powder flow. It's curious to note that in the end the amount of drug a patient ingests is determined volumetrically, not gravimetrically.”

References (websites accessed November 2016)

1. Experiments for term 2 of the 2-year programs for health science, environmental, and biotechnology technicians. Experiments EDTA 1, EDTA 2 and Spectroscopy 2:www.uclmail.net/users/dn.cash/experiments.html.
2. www.wikipedia.org for: Aspirin; tablet (pharmacy); normal distribution; Food and Drug Agency; United States Pharmacopeia; accuracy and precision; CGMP; statistical process control; six-sigma; tableting; mixing (process engineering); granulation (process); glidant; excipient; granular convection; gluconic acid.
3. Standard Materials for Analysis: www.thornsmithlabs.com/about-us#why;www.thornsmithlabs.com/images/stories/catalog/thorn.pdf;www-s.nist.gov/srmors/pricerpt.cfm
4. Paul Taylor, Toronto Globe and Mail, Feb 14, 2013: www.theglobeandmail.com/life/health-and-fitness/health-navigator/those-supplements-may-not-be-delivering-the-vitamins-they-claim/article8672497/.   
5. Private communication from Bobbie Chiu, Senior Compliance Officer, Drug Compliance Verification & Investigation, Ontario Operational Centre, Health Products and Food Branch Inspectorate, Health Canada, July, 2006.
6. Canada Food and Drug Regulations, Section C.01.062 (1), page 694 (page 709 of 1019 of pdf file):http://laws-lois.justice.gc.ca/PDF/C.R.C.,_c._870.pdf
7. FDA Document. Guidance For Industry; Powder Blends And Dosage Units — In-Process Blend And Dosage Unit Inspection (Sampling and Evaluation) For Content Uniformity: w
   /dockets/dohrmsww.fda.gov/ailys/04/jan04/013004/03D-0493_emc-000003-01.pdf
8. Pharmaceutical Manufacturing Handbook:Regulations and Quality, Editor Gad, Wiley, 2008, pages 32-35, especially section 12.1.1:
   http://books.google.ca/books?id=l15Y-MvGbxQC&q=32#v=snippet&q=32&f=false.
9. USP Document <905> Uniformity of Dosage Units: www.usp.org/sites/default/files/usp_pdf/EN/USPNF/2011-02-25905UNIFORMITYOFDOSAGEUNITS.pdf.
10. Health Canada, Good Manufacturing Practices:www.hc-sc.gc.ca/dhp-mps/alt_formats/pdf/compli-conform/gmp-bpf/docs/gui-0001-eng.pdf.
11. Holdich, Fundamentals of Particle Technology, especially chapters 12 and 16:5Ftecwww.particles.org.uk/particle%hnology%5Fbook/particle_book.htm.