Peritonitis diverticular: ¿qué nos enseño DILALA? Análisis metodológico de su diseño, aplicación y resultados

Abstract

BACKGROUND: DILALA compared laparoscopic lavage (LL) with the Hartmann procedure (HP) in Hinchey 3. It was one of the few randomized trials carried out to date, so a thorough methodological review is necessary to validate u object to their conclusions.

AIM: To analyze the methodology used in the design, application, analysis of results, and conclusions of all DILALA publications. Secondly, to collaborate in the improvement of the research about diverticular peritonitis and to facilitate its analysis by the readers.

METHODS: The central parts of a research trial were analyzed, from the research question, hypothesis development, operationalization of variables and trial design, statistical analysis of results, to conclusions. Errors, biases, and weaknesses were searched for to try and challenge the trial's findings.

RESULTS: It was designed as a 2-branch, open-label, randomized, controlled trial with a 1: 1 distribution. This was a superiority study and its data were analyzed both by intention to treat and by protocol. Possible selection biases, weaknesses, and methodological errors were found in it. According to its authors, DILALA demonstrated that LL is safe and feasible, with less operative time and hospital stay, and with fewer reoperations (both in proportion of patients and in number of reoperations per patient) than HP.

CONCLUSIONS: The validity of the main variable (reoperations) was compromised by including the ostomies closures procedures and excluding percutaneous drains without anesthesia. In addition, 3 possible sample selection biases were found. The LL safety and feasibility inference was wrong because the sample was too small to show differences in morbidity and mortality. DILALA demonstrated the obvious, that HP implies the need for an ostomy and eventually surgery for its closure.