AVMA ABVS-recognised specialty boards Veterinary Research Methodology & Evidence-Based Veterinary Medicine: practice questions

1 · Research design · hard

An observational study reports that cats fed a commercial renal diet live longer than cats on maintenance diets. Investigators adjust for serum creatinine measured after six months on the diet. Why is this adjustment a problem?

1. Creatinine after six months lies on the causal path, so adjusting removes part of the diet effect
2. Creatinine is a confounder that must always be adjusted for, so the adjustment is correct
3. Creatinine is affected by muscle mass, so the adjustment only biases the result towards harm
4. Adjustment for any laboratory variable is invalid in observational studies of nutrition

2 · Research design · medium

A new non-steroidal anti-inflammatory drug is cheaper than the standard one. Investigators want to show it is not worse than the standard by more than a pre-set margin for pain scores. Which trial design is this?

1. A superiority trial
2. An equivalence trial
3. A crossover bioequivalence study
4. A non-inferiority trial

3 · Critical appraisal · medium

Which of the following is a domain of the Cochrane Risk of Bias tool version 2 (RoB 2) for randomised trials?

1. Bias due to the journal's impact factor
2. Bias due to missing outcome data
3. Bias due to the size of the funding grant
4. Bias due to the country of the trial site

4 · Systematic reviews · medium

A meta-analysis of 8 studies reports Cochran's Q = 5.0. What is I-squared?

1. 38%
2. 0%
3. 63%
4. -40%

5 · Systematic reviews · hard

A fixed-effect meta-analysis gives a pooled log odds ratio of 0.48 with a total inverse-variance weight of 31.25. What is the 95% confidence interval for the pooled odds ratio?

1. 1.52 to 1.72
2. 1.14 to 2.29
3. 0.13 to 0.83
4. 1.02 to 2.56

6 · Systematic reviews · medium

How does a scoping review differ from a systematic review of effectiveness?

1. It searches only a single database to save time and reports no strategy
2. It pools effect sizes in a meta-analysis but skips any formal search step
3. It maps the extent and nature of evidence and usually does not appraise or pool
4. It is restricted to randomised trials and requires a GRADE certainty rating

7 · Grant writing · medium

A proposal has three aims; aims 2 and 3 can only start if aim 1 produces a positive result. Why do reviewers often criticise this?

1. Reviewers expect exactly two aims, so a third aim always counts against a bid
2. Sequential aims make the budget impossible to divide into separate cost lines
3. Funders forbid any dependency between aims in their published call guidance
4. If aim 1 fails, the whole project fails, so the aims should be able to stand alone

8 · Grant writing · hard

Which section of a proposal most directly justifies the number of animals to an animal-welfare reviewer and to a scientific panel at the same time?

1. A statement that the same number was used in an earlier published study on the same topic
2. A sample-size calculation based on the primary outcome, expected effect and variability
3. A budget line showing the cost of animals is within the funder's maximum per project
4. A Gantt chart showing that all animals can be housed in the facility at the same time

9 · Research ethics · medium

Under EU Directive 2010/63/EU, what must a project authorisation for animal research include?

1. A harm-benefit analysis and classification of procedures by severity
2. A guarantee that no animal will experience any pain at all in the project
3. Approval from the journal in which the results will later be published
4. Evidence that the same work has already been done in another EU country

10 · Grant writing · medium

Which statement about indirect costs (overheads) in research grants is correct?

1. They are always fixed at exactly 20% of direct costs by international agreement
2. They must be spent on consumables used directly in the animal experiments
3. They are paid only by industry sponsors and never by public research councils
4. Funders calculate them differently, so the call's own rules must be checked