Background: Dacryocystorhinostomy (DCR) procedures can be performed using external or endonasal approaches. The comparative success rates of these procedures are unknown.
Objectives: To compare the success rates of external and endonasal approaches to DCR.
Search strategy: We searched CENTRAL (which contains the Cochrane Eyes and Vision Group Trials Register) (The Cochrane Library 2010, Issue 11), MEDLINE (January 1950 to December 2010), EMBASE (January 1980 to December 2010), Latin American and Caribbean Literature on Health Sciences (LILACS) (January 1982 to December 2010), the metaRegister of Controlled Trials (mRCT) (www.controlled-trials.com) (December 2010), ClinicalTrials.gov (www.clinicaltrials.gov) (December 2010) and Web of Science Conference Proceedings Citation Index- Science (CPCI-S) (January 1990 to December 2010). There were no language or date restrictions in the search for trials. The electronic databases were last searched on 7 December 2010. We requested or examined relevant conference proceedings for appropriate trials.
Selection criteria: We included all randomised controlled trials (RCTs) comparing external and endonasal dacryocystorhinostomies.
Data collection and analysis: Two review authors independently performed data extraction and assessment of quality with a predefined form. We contacted investigators to clarify the methodological quality of the studies.
Main results: We identified one trial that fulfilled the inclusion criteria. This trial compared 64 DCR procedures (32 external and 32 endonasal procedures). Endonasal DCR was four times more likely to fail compared to external DCR. This was statistically significant (95% confidence interval (CI) 1.25 to 12.84).
Authors' conclusions: The only trial included in the review provides evidence that endonasal DCR has statistically higher risk of failure compared to external DCR. However, this conclusion is limited by paucity of RCTs, small number of participants and lack of clarity of the methodological process. Well conducted RCTs with sufficient power are required to answer the research question.