Drug development and clinical trials – preclinical tests, phases, placebo and double-blind trials

BiologyHealth & DiseaseAges 16–17Virtual lab

Use with my class

Take six made-up candidate compounds through tests on cells and animals and then phase 1, 2 and 3 clinical trials to a licence, and see where and why each one is stopped. Design your own trials of three made-up drugs: choose the sample size, the control group and the blinding, then compare the measured effect and its 95% confidence interval to see how bias and sample size change the result. A card sort separates painkillers from antibiotics.

Lesson: Discovery and development of drugs: clinical trials

What it shows

A new medicine is tested first in the laboratory, on cells, tissues and live animals, for toxicity, efficacy and dose. Clinical trials follow: a few healthy volunteers in phase 1 for safety, then patients in phases 2 and 3 to see whether it works and which dose is best. A fair trial compares the drug with a placebo and is double-blind, so expectations cannot bias the result, and a larger trial gives a narrower confidence interval. The candidates, drugs and effect sizes here are made up for learning.

How to use

On From lab to licence, pick a Candidate and press Next stage until it is stopped or licensed. On Design a trial, choose the Drug, Patients in each group, Control group and Blinding, then press Run the trial or +20 trials and read the measured effect and its interval. On Painkillers or antibiotics?, put each card in a box and press Check.

Parameters you can change

  • Screen From lab to licence, Design a trial, Painkillers or antibiotics?
  • Candidate compound Candidate 1, Candidate 2, Candidate 3, Candidate 4, Candidate 5, Candidate 6
  • Drug Drug A, Drug B, Drug C
  • Patients in each group 10–2000
  • Control group None, No treatment, Placebo
  • Blinding Open (nobody blinded), Single-blind, Double-blind

Questions to explore

  1. Why does a fair trial give the control group a placebo instead of no treatment at all?
  2. How does the width of the 95% confidence interval change when you use four times as many patients?
  3. Why can a drug with no real effect sometimes look effective in a well-designed trial?