Pancreaticoyeyunostomia vs Pancreaticogastrostomia: Meta-Ánalisis

Pancreaticojejunostomy versus pancreaticogastrostomy: systematic review and meta-analysis.
The American Journal of Surgery feb 2007; 193(2):171-83
Moritz N. Wente M.D., M.Sc.a, Shailesh V. Shrikhande M.D.a, b, Michael W. Müller M.D.a, Markus K. Diener M.D.a, Christoph M. Seiler M.D., M.Sc.a, Helmut Friess M.D.a and Markus W. Büchler M.D.a, , aDepartment of General, Visceral and Trauma Surgery, University of Heidelberg, Heidelberg, GermanybDepartment of Gastrointestinal Surgical Oncology, Tata Memorial Hospital, Mumbai, India

Objective
Pancreaticojejunostomy (PJ) and pancreaticogastrostomy (PG) are the commonly preferred methods of anastomosis after pancreaticoduodenectomy (PD). The ideal choice of anastomosis remains a matter of debate.

Data Sources
Articles published until end of March 2006 comparing PJ and PG after PD were searched.

Study Selection
Two reviewers independently assessed quality and eligibility of the studies and extracted data for further analysis. Meta-analysis was performed with a random-effects model by using weighted odds ratios.

Data Extraction and Synthesis
Sixteen articles were included; meta-analysis of 3 randomized controlled trials (RCT) revealed no significant difference between PJ and PG regarding overall postoperative complications, pancreatic fistula, intra-abdominal fluid collection, or mortality. On the contrary, analysis of 13 nonrandomized observational clinical studies (OCSs) showed significant results in favor of PG for the outcome parameters with a reduction of pancreatic fistula and mortality in favor of PG.

Conclusions
All OCSs reported superiority of PG over PJ, most likely influenced by publication bias. In contrast, all RCTs failed to show advantage of a particular technique, suggesting that both PJ and PG provide equally good results. This meta-analysis yet again highlights the singular importance of performing well-designed RCTs and the role of evidence-based medicine in guiding modern surgical practice.