Critical appraisal

• Systematically examining research evidence to determine trustworthiness, value and relevance
• Necessary for evidence-based medicine
• Apply suitable list of questions related to quality
• Formulate an overall opinion of quality and relevance of given research

Validity

• Internal: strength of evidence for causality, degree to which results attributable to intervention rather than other factors
• External: generalisability, degree to which study results apply to wider populations/settings

ROSSINI trial

• Research question

• Appropriate design

• Randomisation

• Allocation concealment

• Baseline balance

• Blinding

• Follow-up

• Outcome reporting

• Applicability

• Clinical significance

• Good internal validity

• Broadly generalisable

• Low risk of major bias

• No meaningful evidence of benefit

Bias

• Selection: differences between groups from when participants enter trial
• Performance: difference in care or behaviour of participants between groups
• Detection: differences in how outcomes are assessed
• Attrition: differences caused by withdrawals

Tools

• Formalising critical appraisal
• Critical Appraisal Skills Programme (CASP) checklists
• Cochrane risk of bias 3 (RoB 2)
  • Answer series of questions
  • Apply algorithms to judge risk of bias overall on specific domains
• SIGN and NICE checklists