Study population
We ran a retrospective study including 45 patients from November 2018 to December 2019 in a screening protocol for early anastomotic leak detection developed in conjunction with the General Surgery Department in our hospital.
The protocol consisted of an abdominal CT scan performed on the 3rd and 4th postoperative days in rather symptomatic or asymptomatic patients who presented with CRP >15mg/dl or procalcitonin >0.65µg/l.
Study technique and acquisition:
A low dose non-enhanced abdominal CT scan of 80 kV was first obtained circumscribed to the surgical anastomosis, previously located on the scanogram, to differentiate the surgical hyperdense material from a possible contrast agent leak.
A portal venous phase abdominopelvic CT scan was then acquired after a diluted iodinated contrast agent in a saline solution was passed through an 18G rectal catheter and an IV contrast agent is injected in a dose of 1.5 ml/kg at a rate of 3 cc/s.
Volume and concentration of endoluminal contrast agent were chosen according to the distance to the anastomosis:
- Low anastomosis (rectum – sigmoid colon): 20 cc of iodinated contrast agent in 80 cc of saline solution.
- High anastomosis (next to the splenic flexure of the colon): 40-50 cc of iodinated contrast agent in 250 cc of saline solution. It might be necessary to double the volume at the same concentration so the contrast reaches the anastomosis.
Parameters studied:
Sex, age, type of surgery, anastomosis location (colorectal, ileocolic, etc.), time after surgery, CPR and procalcitonin serum levels, leukocyte count and presence of symptoms (pain, abdominal distention, fever, and vomiting) were collected.
CT scan findings recorded included the presence of intra-abdominal free fluid, perianastomotic fluid, pneumoperitoneum, perianastomotic air bubbles, perianastomotic collections, endorectal contrast agent leaks, and bowel mural defects. Percentage of anastomosis leak, false positives, and false negatives were registered, taking the posterior surgical scan, clinical outcome and need of reintervention as the gold standards, respectively.
Finally, we studied added complications, the need for reintervention and alternative diagnoses.
Statistical analysis:
Statistical data analysis was performed using IBM SPSS statistics program. The sensibility and specificity for each finding were obtained. The main characteristics of our population were calculated, including medium age, type of surgery and anastomosis. Prevalence of anastomotic leak in CT and in following surgery was also obtained.
Exclusion criteria:
The next exclusion criteria were defined:
- CT scans performed as routinary follow-up after surgery
- Right colon or small bowel surgeries (ileocaecal resection, ileal anastomosis, etc)
- CT scans performed with no endorectal contrast agent
- Patients after the 4th postoperative day that underwent CT scan due to complication suspicion