Objective: The aim of this study was to determine the safety and efficacy of methods for intraoperative evaluation of urinary tract injury during pelvic surgery.
Methods: PubMed, EMBASE, CINAHL, Web of Science, Scopus, ProQuest, the Cochrane Library, and Clinicaltrials.gov were searched from 1947 to February 2018. Articles or abstracts describing the routine evaluation of urinary tract injuries during pelvic surgery in adults were included, surgical indications of urinary tract anomaly, stones, or malignancy were excluded. There were no restrictions on study design or language. Outcomes included injuries diagnosed intraoperatively, delayed diagnoses, adverse effects, subjective assessments, time to use, and cost. Data were extracted in duplicate at an individual-participant level. Prevalence of injuries, sensitivity, specificity, and predictive values of each diagnostic method were calculated. A combination of generalized linear models and a Bayesian approach were used to separately pool diagnostic accuracy measures.
Results: There were 5303 titles, 527 abstracts, and 164 full-text articles assessed for eligibility; 69 articles were analyzed. Diagnostic methods retrieved were cystoscopy using saline, dextrose or unspecified distention media, oral phenazopyridine and vitamin B, intravenous (IV) methylene blue, IV sodium fluorescein, IV indigo carmine, prophylactic retrograde ureteral stents, and transabdominal Doppler ultrasound. Prevalence of urinary tract injury ranged from 0.3% to 2.8%. Sensitivity ranged 63% to 91%, and specificity, 99.7% to 100%, with no significant differences suggested between methods.
Conclusions: All evaluable methods of intraoperative urinary tract assessment during pelvic surgery are safe and effective with specificity of greater than 99% and low rates of complications. Larger, more rigorous studies are required to evaluate the diagnostic accuracy of some newer methods.