Registered, approved, and still failing.

Powder that won't flow gives uneven doses.

A tablet press fills each die with a fixed volume of powder, not a fixed weight. If the powder flows smoothly, every die receives about the same amount, so every tablet weighs about the same and contains about the same dose.

Fine, light powders tend to clump and flow unevenly, so some dies get more powder than others. Granulation fixes this by binding fine particles into larger, denser granules that pour freely, like sugar compared with flour.

How a tablet is pressed.

A tablet is formed in a die (a hole in a steel plate) between an upper punch and a lower punch. On a rotary press, dozens of punch sets travel round a turntable, and each set repeats the same four steps:

1. 1Fill: powder flows into the die above the lower punch.
2. 2Pre-compress: a light press pushes out trapped air.
3. 3Compress: full force bonds the particles into a tablet.
4. 4Eject: the lower punch rises and pushes the tablet out.

The amount of powder in step 1 sets the tablet's weight, and the force in step 3 sets its hardness. Both are checked during production.

Hard capsules: fill the body, close the cap.

The empty two-piece shells are bought ready-made. A filling machine then:

Shells are usually made of gelatin. Plant-based HPMC (hydroxypropyl methylcellulose) shells were introduced in 1989; gelatin remains more common because it costs less. Because capsule filling avoids compression, it can suit drugs that are hard to make into tablets. (Capsule (pharmacy))

Bigger number, smaller capsule.

Hard capsule sizes run backwards: size 000 is the largest, holding about 1.36 mL, and size 5 is the smallest, about 0.13 mL.

The formulator picks the smallest size that holds the full dose plus excipients, because smaller capsules are easier to swallow. A very large dose may need a big capsule or two capsules per dose, which is one reason some products are made as tablets instead.

Capsule (pharmacy)

Softgels are formed, filled and sealed in one step.

Two continuous ribbons of warm gelatin, softened with a plasticiser such as glycerin, are fed between two rotating die rolls. Each roll has pockets cut into its surface.

As matching pockets meet, a pump injects a measured dose of fill between the ribbons, and the rolls press the ribbons together, sealing and cutting out each softgel. The softgels are then dried to firm up the shell.

Softgels suit drugs that are dissolved or suspended in oil, including drugs that dissolve poorly in water.

Weigh 20 tablets. How far can they stray?

Take 20 units at random, weigh each one, and calculate the average (mean) mass. Then compare every unit with that average. (Ph. Eur. 2.9.5)

The chart shows tablets averaging about 250 mg, where the limit is 5%. Green band: within 5%. Dashed red lines: twice the limit (10%). Press the button to simulate new batches; the data are made up.

Same weight does not mean same dose.

Weighing only works as a check on dose when the drug makes up a large part of each unit. If the drug is a tiny fraction of the tablet, it may be spread unevenly through the blend, so two tablets can weigh the same yet contain different amounts of drug. Then the drug in each unit must be measured directly.

The 25/25 rule applies to hard capsules and uncoated or film-coated tablets; other dosage forms need content uniformity. (USP <905>)

Tumble, weigh, and measure what is lost.

Friability tests whether tablets chip or wear away when knocked about, as they are during packing and transport. (USP <1216>)

1. 1Take whole tablets totalling about 6.5 g (if each weighs 650 mg or less), or 10 tablets (if heavier). Remove loose dust and weigh.
2. 2Rotate them in the drum 100 times at 25 rpm.
3. 3Remove dust, weigh again and calculate the percentage weight lost.

How long before it breaks up?

Six units are placed one per tube in a basket with a mesh floor. The basket moves steadily up and down in a liquid medium (usually water) kept at body temperature, 37 °C.

A unit counts as disintegrated when no firm core is left on the mesh; only soft remains or bits of shell or coating may stay behind. The test passes when all six units have disintegrated within the time limit for that dosage form.

Passing shows the unit breaks up in time. It does not prove the drug dissolves; that needs the dissolution test.

Different forms, different deadlines.

The time limit reflects how each form is designed to behave. Dispersible and effervescent tablets are meant to break up in a glass of water before they are taken, so they must work within minutes.

Coatings and capsule shells must dissolve before the contents can break up, so film-coated tablets and capsules are allowed 30 minutes, compared with 15 minutes for a plain, uncoated tablet.

BP/Ph. Eur., Tablets; BP/Ph. Eur., Capsules

Enteric tablets must pass two opposite tests.

An enteric (gastro-resistant) coat is designed to survive stomach acid and dissolve only in the less acidic intestine. So the test checks both halves of that promise:

1. 10.1 M hydrochloric acid for 2 hours (mimicking the stomach): no tablet may disintegrate or crack.
2. 2pH 6.8 phosphate buffer (mimicking the intestine): the tablets must disintegrate within 60 minutes.

Failing stage 1 means the drug could be released, or damaged, in the stomach. Failing stage 2 means it might never be released at all.

BP/Ph. Eur., Tablets

Breaking up is not enough. The drug must dissolve.

A tablet can pass disintegration while its drug still dissolves too slowly. The dissolution test measures how much of the drug has dissolved after a set time. (USP <711>)

Each unit sits in its own vessel of medium at 37 °C, stirred by a rotating basket (Apparatus 1) or paddle (Apparatus 2). Samples of the medium are taken at set times and analysed for drug.

The product's monograph sets Q: the percentage of the labelled dose that must be dissolved by the test time (see a real example in the USP monograph). In the chart, batch A clears the S1 line at 30 minutes; batch B dissolves too slowly.

The right drug, the right amount, and little else.

Most chemical tests use high-performance liquid chromatography (HPLC), which separates the substances in a sample so each appears as its own peak.

1. 1Identity: the drug peak must appear at the same time as a reference standard of the drug. Infrared spectroscopy is another common identity test.
2. 2Assay: the size of the drug peak shows how much drug is present, compared with the amount stated on the label.
3. 3Related substances: small extra peaks are impurities, such as leftovers from making the drug or products of it breaking down. Each must stay below its limit.

Solvents left behind.

Organic solvents are used to make drugs and coatings, and traces can remain in the final product. The international guideline ICH Q3C sorts them into three classes by how harmful they are: (ICH Q3C(R5))

• 1Class 1, avoid: known or suspected carcinogens or environmental hazards, e.g. benzene (2 ppm).
• 2Class 2, limit: significant toxicity, e.g. methanol (3000 ppm).
• 3Class 3, low toxic potential: e.g. ethanol; up to 5000 ppm (0.5%) is accepted without justification.

Residual solvents are commonly measured by headspace gas chromatography (GC): the sample is warmed in a sealed vial, and the vapour above it is injected into the GC.

Quality does not stop at the factory gate.

After release, NPRA's Surveillance and Complaints Section keeps watching registered products on the market. In 2025 it received 888 complaints, most about product quality and packaging, and 30 registered products were recalled, mainly for quality defects. (NPRA)

Tablets that are printed.

In 2015, Spritam (levetiracetam, for epilepsy) became the first 3D-printed drug product approved by the FDA. It is built up in thin layers into a porous tablet that breaks up rapidly with a sip of liquid, even at doses up to 1000 mg. (PharmTech, 2015)

Technologies change, but the quality questions stay the same: is every unit alike, does it contain the right drug in the right amount, and does it release that drug as intended?