Skip to main content

Stony Brook University

Evidence-Based Medicine: Levels of Evidence

Using the best current evidence for patient decision making.

Grades of Recommendation

Grade of
Level of
Type of Study
A 1a    Systematic review of (homogeneous) randomized
controlled trials
A 1b Individual randomized controlled trials (with narrow
confidence intervals)
B 2a Systematic review of (homogeneous) cohort studies
of "exposed" and "unexposed" subjects
B 2b Individual cohort study / low-quality randomized
control studies
B 3a Systematic review of (homogeneous) case-control studies
B 3b Individual case-control studies
C 4 Case series, low-quality cohort or case-control studies
D    5 Expert opinions based on non-systematic reviews of
results or mechanistic studies

Levels of Evidence

Critically-appraised individual articles and synopses include:

Filtered evidence:

  • Level I: Evidence from a systematic review of all relevant randomized controlled trials.
  • Level II: Evidence from a meta-analysis of all relevant randomized controlled trials.
  • Level III: Evidence from evidence summaries developed from systematic reviews
  • Level IV: Evidence from guidelines developed from systematic reviews
  • Level V: Evidence from meta-syntheses of a group of descriptive or qualitative studies
  • Level VI: Evidence from evidence summaries of individual studies
  • Level VII: Evidence from one properly designed randomized controlled trial

Unfiltered evidence:

  • Level VIII: Evidence from nonrandomized controlled clinical trials, nonrandomized clinical trials, cohort studies, case series, case reports, and individual qualitative studies.
  • Level IX: Evidence from opinion of authorities and/or reports of expert committee

Two things to remember:

1. Studies in which randomization occurs represent a higher level of evidence than those in which subject selection is not random.

2. Controlled studies carry a higher level of evidence than those in which control groups are not used.

Strength of Recommendation Taxonomy (SORT)