To evaluate all invasive treatments for suspected IPN.
The optimal invasive treatment for suspected IPN remains unclear.
A systematic search of randomized clinical trials comparing at least 2 invasive strategies for the treatment of suspected IPN was carried out. A frequentist random-effects network meta-analysis was made reporting the surface under the cumulative ranking (SUCRA). The primary endpoint regarded both the in-hospital mortality and major morbidity rates. The secondary endpoints were mortality, length of stay, intensive care unit stay, the pancreatic fistula rate, and exocrine and endocrine insufficiency.
Seven studies were included, involving 400 patients clustered as following: 64 (16%) in early surgical debridement (ED); 27 (6.7%) in peritoneal lavage (PL); 45 (11.3%) in delayed surgical debridement (DD), 169 (42.3%) in the step-up approach with minimally invasive debridement (SUA-DD) and 95 (23.7%) with endoscopic debridement (SUA-EnD). The step-up approach with endoscopic debridement had the highest probability of being the safest approach (SUCRA 87.1%), followed by SUA-DD (SUCRA 59.5%); DD, ED, and PL had the lowest probability of being safe (SUCRA values 27.6%, 31.4%, and 44.4%, respectively). Analysis of the secondary endpoints confirmed the superiority of SUA-EnD regarding length of stay, intensive care unit stay, pancreatic fistula rate, and new-onset diabetes. The SUA approaches are similar regarding exocrine function. Mortality was reduced by any delayed approaches (DD, SUA-DD, or SUA-EnD).
The first choice for suspected IPN seemed to be SUA-EnD. An alternative could be SUA-DD. PL, ED, and DD should be avoided.

Author