Levels of evidence in endoscopic ultrasonography: a systematic review

Abstract

Background and aims: A prolific trend currently designates endoscopic ultrasonography (EUS) literature. We aimed to record all EUS-studies published during the past decade and evaluate them in terms of scientific quality, creating a stratification based on levels of evidence (LE).

Methods: All articles on EUS published between January 2001 and December 2010 were retrieved using a Pubmed and Cochrane Library search. Inclusion criteria were: original research papers (randomized controlled trials-RCTs, prospective and retrospective studies), meta-analyses, reviews and surveys pertinent to gastrointestinal EUS. Levels of evidence (LE) were assessed using the North of England evidence-based guidelines.

Results: Overall, 1,832 eligible articles were reviewed. The majority (46.1%) of reports comprised retrospective descriptive studies (LE III). Expert reviews and committee reports (LE IV) accounted for 28.9%, well-designed quasi-experimental studies (LE IIb) for 20.1%, RCTs (LE Ib) for 2.4%, prospective controlled trials (LE IIa) for 1.4%, and meta-analyses (LE Ia) for 1.1% of the total. High LE (Ia-Ib) were assigned to loco-regional staging of luminal gastrointestinal cancers; mediastinal staging of lung cancer; diagnostic work-up of solid pancreatic tumors, suspected biliary obstruction and choledocholithiasis; celiac plexus neurolysis; and pancreatic pseudocysts drainage. Intermediate to low LE (IIa-IV) were assigned to submucosal tumors, pancreatic cysts, chronic pancreatitis and novel therapeutic applications (pancreato-biliary drainage, tumor ablation).

Conclusions: Diagnostic and staging EUS has matured and has proven its clinical impact on patient management. Therapeutic or interventional EUS is still evolving and more quality research and data are needed to establish EUS as the best next intervention to perform once firm evidence is available.