Home > Testing Treatments > Test your knowledge – the answers

Test your knowledge – the answers

A quackThis section of the site will contain interactive quizzes so that you can test your knowledge of fair tests of treatments and direct you towards resources to explain the ones you got wrong!

For now, we’ve drafted the following questions.  Click the icon to view the answers.

Watch out for interactive quizzes being added throughout 2014.

Questions about fair tests of treatments

1. In a study to assess the effects of a new antiviral agent for flu-like illness, 100 people were given the new drug. One week later the investigators conducting the study observed that 90 of the 100 subjects were better. They concluded that the antiviral drug was highly effective in curing children’s colds.

  • Do you agree with the investigators’ conclusion? NO

2. In an experiment to compare two treatments, patients are allocated at random so that: (mark each True or False)

  • The sample of patients can be referred to a known population FALSE
  • Every patient will get the treatment known to be best suited to him or her FALSE
  • The two groups will be as similar as possible, apart from treatments being compared TRUE
  • Treatments may be allocated according to the characteristics of each patient FALSE

3. Placebos are useful in clinical trials (mark each True or False)

  • To guarantee comparability in non-randomized trials FALSE
  • Because the fact of receiving any treatment may itself have an effect TRUE
  • Because they may help to conceal the patient’s treatment from assessors TRUE
  • When an active treatment is to be compared to no treatment TRUE

Advanced question

4. When evaluating whether you can trust the results of a Randomised Clinical Trial the three most important things you would look for are: (number the three most important as 1, 2, 3)

  • Were all patients who entered the trial accounted for at its conclusion?
  • Were the inclusion criteria clearly stated?
  • Were the patients a random sample from the relevant population?
  • Were the patients and clinicians blind to which treatment was given?
  • Was the randomisation list concealed?
  • Were only patients who fully complied included in the final analysis?
  • Were the outcome measures clearly defined?

Read more about these questions in “Fair tests of treatments“.

Questions about the play of chance in tests of treatments

1. In a small randomized double-blind trial of a new treatment for heart attacks, the number of patients receiving the new treatment who died was half the number who died in the comparison group, but the difference was not statistically significant.  We can conclude that:

  • The treatment is useless FALSE
  • There is no point in continuing to develop the treatment FALSE
  • The reduction in mortality is so great that we should introduce the treatment immediately FALSE
  • We should assess the new treatment in larger clinical trials TRUE

2. Study results are commonly given with a “confidence interval”, for example a survey may say 40% ± 5% of people like apples. Which of the following statements about such confidence intervals are correct?

  • the intervals are larger with smaller sample sizes TRUE
  • the confidence interval depends on the proportion of all the people eligible for inclusion in the study who took part in it FALSE
  • the intervals give a range of values within which the true value is likely to lie. TRUE

Read more in Taking account of the play of chance

Questions about assessing all the relevant, reliable evidence

1. A “systematic review” of the reports of clinical trials is a synthesis of research addressing a specific question. Such systematic reviews:

  • Are based on the consensus of experts FALSE
  • Must obtain informed patient consent before any recruitment of patients FALSE
  • May combine the results of similar studies to produce an overall estimate of effect TRUE
  • Are limited to an extensive search of the largest database FALSE

2. When you are checking whether a review of the research on a specific treatment is trustworthy, which questions would you ask?

  • Did the authors do a thorough search for all relevant research? YES
  • Did the authors work closely with one another? NO
  • Did the authors assess the quality of the research included? YES
  • Did the authors include a balance of men and women? NO

Read more in Assessing all the relevant, reliable evidence.

Questions about how to improve healthcare through fair tests of treatments

1. In a placebo-controlled trial of new drug, Kuritol, 8% of those allocated to Kuritol died compared to 12% allocated to the placebo.

  • What percent of people given placebo survive? 88%
  • What percent of people given Kuritol survive? 92%

2. There are different ways to compare the survival in the two groups. Can you work out:

  • the difference in the likelihood of death for patients given Kuritol?  4%
  • the relative reduction in the likelihood of death with Kuritol? 33%
  • how many similar patients need to be treated with Kuritol to prevent one of them dying? 25

Read more in So what makes for better healthcare?