Unit quiz: Reading treatment evidence and reasoning about services

Check understanding, separately from reading

5 questions · 80% to pass · no timer. This quiz checks the lessons in this unit. These authored questions assess recognition and application of the taught distinctions, not professional qualification. You may review the lessons and retry. Repeat attempts reuse the question bank; a remembered answer is not proof of transfer to a new situation.

Saving a lesson records reading only. Previous lesson completions have not been converted into passing scores. You can continue reading without a pass; the assessment remains unpassed.

Question 1

A fictional service wants to use a trial's finding that an intervention reduced relapse by 25% to justify targeting it at 'those most likely to benefit'. The trial's only heterogeneity analysis was an exploratory interaction with symptom duration. What should the service do instead, on this level's evidence?Use the trial's risk model, if it has one validated for a comparable population, to disaggregate by predicted baseline risk and compute the absolute benefit at each level; if it has none, treat differential benefit as unestablished and base targeting on baseline risk from local data while stating the assumption.Re-run the interaction analysis with a more lenient threshold for significance, because the exploratory analysis likely missed a real effect through lack of power and a lower threshold will recover it.

Question 2

A fictional journal club discusses a new meta-analysis of nine trials of a treatment for a specific presentation, pooled effect 0.45, and the presenter says the review 'rules out publication bias'. On this level's evidence, what is wrong with that statement, and what would you look for instead?Nothing is wrong, provided the funnel plot was inspected visually, since visual inspection of a funnel plot remains valid at any number of studies even when the formal tests are underpowered.A nine-study meta-analysis cannot support that conclusion: tests for funnel plot asymmetry should not be used with fewer than about ten studies because power is too low, and asymmetry has causes other than publication bias; look instead for whether unpublished or registered trials were sought, whether the estimate is inflated relative to a complete record, and whether a certainty rating accounts for reporting bias.

Question 3

Two fictional trials of the same psychological treatment are reported. Trial P compares it with a waiting list and reports an effect size of 0.8; trial Q compares it with an active control matched for appointment frequency and clinician contact and reports 0.2. A guideline author asks which estimate applies to a service that already offers regular supportive appointments. What is the best answer?Trial Q's estimate, because the service's realistic alternative already contains the non-specific components, attention, structure and contact, so the relevant question is what the specific package adds to those, and the two trials estimate different quantities rather than the same quantity twice.An average of the two estimates, since both are randomised comparisons of the same treatment and averaging them produces a single best estimate of its effect.

Question 4

A fictional commissioning group is told that a new specialist team will treat 500 patients a year in a catchment of 250,000 adults, where the expected prevalence of the target condition is about 4%. The group also hears that the team's evaluation showed a 30% reduction in admissions to the local ward. Which two questions does this level's systems reasoning require before the claim can be accepted?Whether the 30% figure was statistically significant in the evaluation and whether the team's clinicians were trained by the model's developers, since these determine whether the effect is real and transferable.What proportion of the 10,000 expected cases 500 treated patients represents, and where displaced activity would appear, including beds used out of area, emergency department attendances and readmissions to other providers, since a fall in one setting's admissions may be redistribution rather than reduction.

Question 5

A fictional appraisal of a treatment claim ends: 'The evidence base is heterogeneous and limited; on balance the intervention is probably effective and further research would be welcome.' Which single addition would most improve it, in this level's terms?A named unknown phrased so it could be measured, together with a revision condition, for example, the comparator-matched effect in a population like this service's, and the finding that would falsify the conclusion, so that later evidence can move the judgement rather than be absorbed as confirmation.A certainty rating using all four GRADE categories, because a formal grade would supply the precision that the sentence lacks and make the appraisal complete.

Continuing does not mark this assessment passed.

Progress stays in this browser, on this device, not an account or cloud backup. Clearing browser data removes it. Visits never count as study. Previous untimed legacy completions are not invented as study days.