Gastro-intestinal perforation is a condition that can become life-threatening in case of appearance of systemic symptoms, sepsis-related peripheral hypoperfusion and single or multiple organ failure needing a prompt intervention in Emergency Department (ED) setting. Literature reports disagreeing data about the effect of surgical timing on mortality and postoperative outcomes: Buck et al. described a 2.4 % of decreased survival every hour of surgical delay in case of perforated peptic ulcers. Other authors documented significantly longer postoperative hospital stay, greater health costs and a significant increase of postoperative complication and mortality rates when surgery is delayed in high-risk patients with comorbidities or age > 65 years. Azuhata described a highly significant relationship between delayed surgery and patients' survival: after 6 hours from admission to ED, patients with gastrointestinal perforation and associated septic shock don't survive to surgery. The aim of this study is to assess the impact of delay of time between patient admission to ED and surgery for source control on 30-d mortality and postoperative outcomes in patients with gastrointestinal perforation with or without septic shock. Furthermore, we want to define the time threshold within which surgery can affect patients' survival.

This is an Italian National multicenter study composed by a retrospective phase of data collection from patients of past ten years and a perspective one of next two years.

INCLUSION AND EXCLUSION CRITERIA We include in our analysis all patients with more than 18 years old undergoing emergency surgery for gastro-intestinal perforations, with or without signs of septic shock. All patients with covered perforations or abdominal perivisceral free air bubbles treated with non- operative management or delayed surgery will be excluded. An online Case Report Form (CRF) will be filled out by every participating Italian participating center.

ENDPOINTS Primary outcome analyzed is in-hospital Clavien-Dindo > 3 complication rate [12]. Secondary outcomes are 60-d mortality rate, total length of stay (LOS), LOS in Intensive Care Unit (ICU), days of mechanical ventilation, re-intervention rate and 30-d readmission rate.

Analyzed variables are:

• Hospital characteristics; teaching/non-teaching setting; I/II level ED; emergency surgery/volume x year; 24H/on-call radiology and surgery team; presence of a dedicated OR in ED;
• Patient demographics (gender, age, Body Mass Index (BMI) and American Society of Anaesthesiologists' (ASA) classification of Physical Health, medical history of past abdominal surgical operations and comorbidities according to Charlson Comorbidity Index (CCI), mainly related to corticosteroids therapy and immunosuppressive conditions),
• Patient management in ED: admission date and time; Priority code at admission; time of abdominal x-Ray; time of CT-scan; Time of patient evaluation by surgeon on- call; Preoperative resuscitation management (Y/N, type, time of beginning); presence of Sepsis (infection documented or suspected + SOFA >=2) and of septic shock (sepsis + persistent hypotension needing vasopressors to maintain a Mean Arterial Pressure (MAP) of 65 mmHg; serum values of lactates > 18 mg/dL (or 2 mmol/L) with an adequate volume resuscitation measures) at admission or during patient stay in ED; SOFA score at admission; Emergency surgery score; ACS NSQIP® Surgical Risk score.
• Radiological findings: CT-scan reports of patients selected will be analyzed by site data-collection centers with the aim to identify typical and atypical radiological findings (subdiaphragmatic free gas, free peritoneal fluid, bowel wall discontinuity, extraluminal oral contrast, extraluminal abscess, abdominal collections, fat stranding, portal venous gas, wall bowel thickening, pneumatosis and mucosal hyperenhancement).
• Surgery: Time of skin incision; Type and duration of surgery; source control surgery, Open Abdomen; surgical approach; site and cause of perforation; pathologic data; Emergency surgery score; ACS NSQIP® Surgical Risk score.
• Postoperative Course: Preoperative mortality (patients die before surgery); Medical and Surgical complications (Clavien-Dindo score); total LOS; LOS in ICU; days of Mechanical Ventilation; Re-intervention rate; 90-d mortality.

PLANNED SAMPLE SIZE With a binary response variable, β=0·95, α=0·05, an anticipated small effect size and an allocation ratio 1:10 (Early treatment Yes vs. No), it has been calculated that 3276 patients are required to detect an association between the variables and the endpoint.

PLANNED ANALYSES Time between patient admission to ED and surgery will be analyzed as a continuous variable with t-Student's tests, comparing means between the different outcomes (primary and secondary). Different cut-off will be tested to define a significant time threshold correlating with outcomes. If necessary, more time intervals will be evaluated to evaluate correlations between the variables collected and the time.

Inclusion Criteria:

• more than 18 years old
• emergency surgery for gastro-intestinal perforations
• with or without signs of septic shock

Exclusion Criteria:

• All patients with covered perforations or abdominal perivisceral free air bubbles treated with non- operative management
• delayed surgery

patients undergoing emergency surgery for gastro-intestinal perforations, with or without signs of septic shock