During our first Journal Club of the year we discussed the retrospective study by El Masri et al.,1 evaluating (1) the outcomes of routine cystourethroscopic surveillance in patients with traumatic spinal cord injury and (2) relevant publications retrieved after a systematic search of the literature. The authors found no significant differences in findings between patients who were symptomatic and asymptomatic at the time of cystourethroscopy and concluded that routine cystourethroscopic surveillance ‘is essential’ in all patients with traumatic spinal cord injury. While these findings are certainly interesting, based on the information provided in the paper, the participants of our Journal Club did not completely agree with the authors’ conclusions.

Our principal concern is related to the study design that was applied in order to answer the raised question. The authors did not compare outcomes between those patients who did and those who did not undergo cystourethroscopy. Instead, findings of all patients who underwent cystourethroscopy were stratified by absence, or presence, of ‘symptoms’. Here also lies our second concern; the authors did not clearly define the criteria used to determine which group (that is, symptomatic or asymptomatic) a patient fell into. Without this information the distinction between the groups appears less clear, and thus the relevance of a lack of different outcomes between the two groups remains questionable. Furthermore, although the range of patients’ duration of catheterization was reported for the entire cohort, no adjustments for this putative confounder were made in the stratified comparison. It may well be that the incidence of findings such as metaplasia or malignancy increases with duration of catheterization.

While it was unclear why the results of the systematic literature review were presented in the discussion section of the paper, we were more concerned about the rigor of the authors’ search for relevant publications. Most of the discussed papers broadly supported the authors’ clinical findings. However, two key papers on the same subject that showed conflicting findings were not discussed. In 1999, Yang et al.2 published in this journal a retrospective review of 6 years of annual screening cystoscopies in spinal cord injured patients and concluded that there was no evidence to support the use of routine cystoscopy as screening for bladder malignancy. We were surprised not to see this reference in El Masri et al.’s paper, as it was included in the discussed NICE guideline.3 Similarly, in 2003 Hamid et al.4 published results from a very similar review of annual cystoscopies and concluded that routine screening of cystoscopy was not to be recommended.

As for defining the clinical relevance of ‘metaplastic changes’ in patients with neurogenic bladder dysfunction requiring chronic catheterisation, more work needs to be done on the diagnostic value of cystourethroscopic surveillance. While awaiting results of future methodologically sound research, we do not feel that El Masri et al.’s work provides compelling evidence yet to recommend endoscopic surveillance in this group of patients.