4.13 - How medicines are tested

Withering and foxglove

William Withering's investigation began with a traditional remedy rather than a purified drug.

In his 1785 account, Withering described being asked in 1775 about a family remedy associated with an unnamed woman in Shropshire. It contained twenty or more herbs and was used for dropsy, a historical description of swelling associated with fluid accumulation. Dropsy included conditions with different causes; it was not a precise modern diagnosis of heart failure. Withering identified foxglove, Digitalis purpurea, as the likely active herb.

The specification's phrase “digitalis soup” refers to this historical herbal-remedy context. Distinguish the original mixture from Withering's subsequent investigations of foxglove preparations, including liquids made from the leaves and dried leaf powder. He did not isolate a pure modern medicine or simply establish the effectiveness of every ingredient in the original mixture.

Withering recorded cases, preparations, treatment amounts, responses and failures. Increased urine production could accompany reduced fluid accumulation. He also recorded sickness, vomiting and a dangerously slowed pulse. His observations showed that an apparently beneficial substance could also be poisonous, and that preparation, amount and the timing of repeated administration mattered. Plant material could vary in strength, so equal quantities of crude plant material need not contain equal quantities of active substances. These are historical observations, not preparation or dosing instructions. Withering's original account supplies the historical evidence.

Systematic case records were an advance over an unsupported claim that a remedy worked. Nevertheless, a person improving after treatment does not by itself show that treatment caused the improvement. The illness might fluctuate, another treatment might help, and the patients selected might differ from those not treated. Withering's reported cases were not a modern randomised, blinded comparison with a concurrent control group. They helped identify a promising treatment and important hazards, but could not estimate its benefits and risks with contemporary trial safeguards.

Preclinical research and three clinical phases

Contemporary development investigates both whether a candidate medicine helps and whether its harms are acceptable for its intended use.

Before human trials, preclinical research uses laboratory systems, such as cells or tissues, and appropriate animal studies to investigate biological effects and toxicity. Researchers examine how effects change with exposure and gather evidence relevant to a cautious first human study. Results from another species or isolated cells cannot establish safety or efficacy in people. The tested material must also have a sufficiently controlled identity, composition and quality: an undefined herbal mixture makes it harder to know what exposure produced an effect. FDA preclinical guidance describes this stage.

A clinical trial is an investigation involving people. Its written protocol specifies the question, eligible participants, treatment arrangements, outcome measurements, follow-up and analysis before the results are known. A dose is the amount administered at one time; a treatment regimen also specifies timing and duration. The typical three-phase progression has changing main purposes; safety is monitored throughout.

The phases describe a typical development pathway, not fixed sample sizes or a rule that every study has an identical design. FDA clinical-research guidance supports these distinctions.

Independent ethical and regulatory review considers the scientific design and balance of anticipated benefit, risk and uncertainty. Participants should give informed consent: voluntary agreement after understandable information about the research, its uncertainties, possible harms, alternatives and what participation involves, with opportunity to ask questions. They can withdraw from participation without penalty. Consent does not mean that a trial is risk-free or that researchers may ignore emerging harm. Safety monitoring, reporting and stopping arrangements can require a treatment or trial to be paused. These protections are supported by MHRA trial-approval guidance, MHRA safety guidance and the UK participant-withdrawal provisions.

After the clinical evidence is assessed, a regulator may authorise an intended use. Monitoring continues after authorisation because wider or longer use can reveal additional harms. Completing phase 3 is not a guarantee of authorisation or absolute safety.

Controls and random allocation

A fair comparison needs a suitable control group as well as a group receiving the candidate medicine.

Symptoms can change because of expectations, attention from staff, natural variation or usual care. An improvement in the placebo group is therefore not evidence that the placebo contains the test drug, and not every improvement in that group is caused by expectation alone. Comparing outcomes between groups can estimate the additional effect of the test treatment under the trial conditions.

A placebo is not always an acceptable replacement for effective care. If withholding an established treatment risks serious harm, a trial can compare the candidate with an active control, meaning an established treatment. Alternatively, where appropriate, both groups can continue usual effective care, with the new medicine added in one group and a matching placebo added in the other. The question then concerns the additional benefit of the new medicine, not its effect in the absence of usual care. Control choice must be ethically and scientifically justified. ICH E10 control-group guidance explains these choices.