How Can We Decrease Mortality Due to Anastomotic Fistula in Colorectal Surgery?

Dear Editor,

In his recent editorial, Parés described the measures that could reduce fistula-caused mortality in colorectal surgery (in the failure to rescue), insisting on the importance of “a good visit in the rounds and the human capacity to detect and respond early to clinical signs which patients display when they have a complication”.¹ This is a major problem in colorectal surgery, as is confirmed by the results of ANACO.² Without belittling the importance of doctors’ rounds and their clinical sensitivity, their value is quite limited in the early detection of a colorectal fistula. The literature shows that surgeons’ impressions are as unreliable as the clinical signs we were taught “when we were very young”.^3,4 Clinical signs appear at the end of the first week after surgery (on the 6th day in ANACO and on the 7th in the Danish national study and in IMACORS).^2,5,6 By that time there is already, in general, severe peritonitis and symptoms of sepsis. Due to this we cannot wait for these manifestations if we wish to improve the results of the failure to rescue; an early diagnosis is indispensible. C-reactive protein (CRP) is currently the best early marker of intra-abdominal infection, as it warns before a fistula and its disastrous consequences appear.^7,8 Its levels rise significantly, more from the first postoperative day in those patients who will go on to present the classic symptoms 5 days later.^8,9 Procalcitonin offers nothing in comparison with CRP, and it is 25 times more expensive.⁶ A recent meta-analysis confirmed that the best day to measure CRP is the 4th, as it is then that the markers attains its greatest discriminatory power.¹⁰ We have nothing else with so much scientific evidence in its favour and which is so cheap and easy to use in case of failure to rescue as CRP. Due to its high negative predictive value (>95%) it is included in fast-track and ERAS protocols. A patient with CRP <100 mg/l on the 4th day can be discharged with a minimum risk of complications. It remains to be decided what to do when patients have higher levels, because scanning on the 4th day seems to give a lot of false negatives. If we wish to improve failure to rescue, we should centre on managing those patients with a CRP >100 mg/l on the 4th day. Future studies of this subject will have to discuss the usefulness of early scanning (probably with water-soluble rectal contrast), together with the role of endoscopy (diagnostically and potentially therapeutically) and additional surgery.