Evidence / Stats

Question 1

STRENGTH OF EVIDENCE

Answer 1

I - High quality diagnostic studies, prospective or RCT

II - lesser-quality diagnostic studies, prospective or RCT (weaker diagnostic criteria & reference standards, improper randomization, no blinding, etc)

III - Case controlled or retrospective studies

IV - Case series

V - Expert opinion (consensus)

Question 2

EFFECT SIZE

Answer 2

Small 0.2 (58%)

Moderate 0.5 (69%)

Large 0.8 (79%)

Question 3

LIKELIHOOD RATIOS

Answer 3

+LR>10 -LR <.1 large & often conclusive

+LR 5-10 - LR <.1-.2 moderate shifts

+ LR 2-5 -LR 0.2-0.5 small, sometimes important shifts in probability

+LR 1-2 -LR 0.5-1 small & rarely important degree

Question 4

STRENGTH OF RECOMMENDATIONS

Answer 4

A - strong (preponderance of level I / level II support - MUST include at least one level 1)

B- moderate (single high-quality RCT or preponderance of level II)

C - weak (single level II or preponderance Level III,IV)

D - conflicting (high quality studies diagree)

E - theoretical or foundation (animal/cadaver/basic science)

F - expert opinion (best practice based on guideline development team)