|Level of evidence and type of study|
|Therapy /Prevention, Aetiology/Harm||Prognosis||Diagnosis||Differential diagnosis/ symptom prevalence||Economic and decision analysis|
|Investigating the effect of patient characteristic on the outcome of disease||Investigating a diagnostic test. Is this diagnostic test accurate?|
|Level 1||Systematic Review of randomized trials(RT)||Systematic Review of inception cohort studies||Systematic Review of level 1 diagnostic studies||Systematic Review of prospective or classic cohort||Systematic Review of level 1 economic studies|
|High quality RT(e.g.:> 80% follow up, narrow confident interval)||Individual cohort study with > 80% follow up, all patient enrolled at the same time||Level 1 diagnostic studies or Validating studies which test the quality of a specific diagnostic test, previously developed, in series of consecutive patients with reference “gold” standard||Prospective or classic cohort studies with good follow up (>80%)||Level 1 studies (analysis based on clinically sensible costs or alternative, values obtained from many studies, and including multiway sensitive analysis|
|Level 2||Systematic Review of cohort studies||Systematic Review of either historical cohort study or untreated control groups (control arm) in RCTs||Systematic Review of level 2 diagnostic studies||Systematic Review of level 2 studies||Systematic Review of level 2 studies|
|Lesser quality RT (e.g.: <80% follow up, wide confident interval, no clear randomization, problems with blinding, etc.)||Historical (retrospective) cohort study or control arm from a RCT||Level 2 diagnostic studies or Exploratory studies which collect information, trawl data to find which factor are significant (e.g.: using regression analysis)||Level 2 studies (retrospective or historical cohort study or with follow up <80%)||Level 2 studies (analysis based on clinically sensible cost or alternative from limited studies, and including multiway sensitivity analysis.|
|Individual Cohort study, including matched cohort studies (prospective comparative studies)||Ecological Studies|
|Level 3||Systematic Review of case–control studies||Systematic Review of level 3 studies||Systematic Review of level 3 studies||Systematic Review of level 3 studies|
|Individual case–control study||Level 3 diagnostic studies or studies in non-consecutive patients and without consistently reference “gold” standards||Level 3 studies (non-consecutive cohort or very limited population)||Level 3 studies (analysis based on poor alternative or costs, poor quality estimates of data, but including sensitivity analysis|
|Level 4||Case-series||Case-series||Case–control study||Case-series||No sensitivity analysis|
|Poor quality cohort and case–control studies*||Poor quality cohort and case–control studies*||Poor or non independent reference standard|
|Level 5||Expert opinion||Expert opinion||Expert opinion||Expert opinion||Expert opinion|
A systematic review (SR) is generally better than an individual study. Experimental study (e.g.: good quality RCT) is generally better than any observational study. For observational studies : cohort study is generally better than any case–control study . A case- control study is generally better than any case- series study. * By poor quality cohort study we mean a cohort study that failed to clearly define comparison groups and/or failed to measure exposures and outcomes (preferable blinding) in the same objective way in both expose and non-exposed individuals and/or failed to identify control known confounders and/ or poor follow up. The same for poor quality case–control study except that the patients are identified based on the outcomes in this design ( e.g.: failed replant) called “cases” are compared with those who did not have the outcome (e.g.: had a successful replant) called “controls” and consequently we do not have “exposed and non-exposed” and “longitudinal follow up”. Ecological studies and Economic/decision analysis studies are very uncommon in hand surgery. This chart was adapted from material published by the Centre for Evidence-Based medicine, Oxford, Uk. March 2009.
Rosales et al.
Rosales et al. BMC Research Notes 2012 5:665 doi:10.1186/1756-0500-5-665