Table 1.
Strength of Evidence Hierarchy*
| Study design |
| Level 1: Evidence obtained from at least 1 properly designed randomized, controlled trial† |
| Level 2-1: Evidence obtained from well-designed controlled trials without randomization |
| Level 2-2: Evidence obtained from well-designed cohort or case–control analytic studies, preferably from more than 1 center or research group |
| Level 2-3: Evidence obtained from multiple time series with or without the intervention |
| Level 3: Opinions of respected authorities, based on clinical experience, descriptive studies and case reports, or reports of expert committees |
| Internal validity |
| Good: Meets all criteria for study design |
| Fair: Does not meet all criteria for study design but is judged to have no fatal flaw that invalidates its results |
| Poor: Study contains a fatal flaw |
| Criteria |
| Systematic reviews |
| Comprehensiveness of sources/search strategy used |
| Standard appraisal of included studies |
| Validity of conclusions |
| Recency and relevance |
| Case–control studies |
| Accurate assessment of cases |
| Nonbiased selection of cases/controls with exclusion criteria applied equally to both |
| Response rate |
| Diagnostic testing procedures applied equally to each group |
| Appropriate attention to potential confounding variables |
| Randomized, controlled trials |
| Initial assembly of comparable groups (concealment and distribution of potential confounders) |
| Maintenance of comparable groups (includes attrition, crossovers, adherence, contamination) |
| Important differential loss to follow-up or overall high loss to follow-up |
| Measurements: equal, reliable, and valid (includes masking of outcome assessment) |
| Clear definition of interventions |
| All important outcomes considered |
| Intention-to-treat analysis |
| Sample size/power |
| Cohort studies |
| Consideration of potential confounders; consideration of inception cohorts |
| Adjustment for potential confounders |
| Important differential loss to follow-up or overall high loss to follow-up |
| Sample size/power |
| External validity |
| High: Meets criteria |
| Low: Does not meet criteria |
| Criteria |
| Biological plausibility |
| Similarities of the study sample to the target population (risk factor profile, demographics, ethnicity, sex, clinical presentation, and similar factors) |
| Similarities of the test or intervention studied to those that would be routinely available or feasible in clinical practice |
| Clinical or social environmental circumstances in the studies that could modify the results from those expected in a primary care setting |
*
Modified from the U.S. Preventive Services Task Force (4).
†
Observational studies qualify as level 1 if they are used to estimate a parameter that cannot be determined experimentally (for example, mortality rate due to age-, sex-, and race-related causes).