Urine colour as an indicator for anastomotic leakage after robot-assisted radical prostatectomy

Objectives: To determine whether macroscopic haematuria predicts urethrovesical anastomotic leakage after robot-assisted laparoscopic radical prostatectomy (RALP) as well as a cystogram. Patients and methods: Participants were recruited before cystogram and catheter removal 5-14 days after RALP surgery. Urine colour in the collection bag was classified according to a three-step scale (clear, light red and dark red) and leakages in cystogram were graded with a four-step scale (Grade 0-3). Diagnostic accuracy parameters were calculated for urine colour. A multivariate logistic regression model was used to evaluate other leakage risk factors. Results: Of 214 patients, 201 (94%) had clear, six (3%) had light red and seven (3%) had dark red coloured urine. In the cystogram, 20 (9%) patients had leakage; 14 had Grade 1, five Grade 2 and one Grade 3 leakage. Overall, specificity and sensitivity of urine colour in predicting anastomotic leakage were 0.97 (95% CI = 0.95-100) and 0.38 (95% CI = 0.17-0.59), respectively. Negative and positive predictive values were 94% and 62%, respectively. Other significant risk factors for anastomotic leakage were previous transurethral resection or radiation therapy to the prostate, non-waterproof anastomosis at surgery, postoperative pelvic haematoma, long catheterization and surgeon's inexperience. In patients with no other risk factors, test sensitivity improved to 0.80 (95% CI = 0.45-1.15) and negative and positive predictive values to 99% and 44%, respectively. Conclusion: This prospective single-arm trial indicates that in patients with clear urine and no other risk factors for anastomotic leakage, a cystogram examination before urethral catheter removal can be safely omitted.