Forty-one studies were included in the review and 37 studies (2,868 patients) were included in the meta-analysis. Fifteen studies assessed short-term catheterisation, 10 assessed long-term catheterisation and 16 assessed complications in patients with spinal cord injury. Minor complications were common and there was a significant number of more serious complications.
Short-term catheterisation (15 studies). The most common complication was leakage of urine (10.6%, 95% CI 2.4 to 17.7). Gross haematuria (4.7%, 95% CI 0.0 to 10), accidental removal (4.0%, 95% CI 0 to 8.7) and catheter blockage (5%, 95% CI 1.7 to 10.4) were the next most common. The more serious complications of urethral stricture or erosion varied considerably between high quality studies (3.4%, 95% CI 1.0 to 7.0) and lower quality studies where it was much more frequent (16.7%, 95% CI 13 to 20.8).
Long-term catheterisation without spinal cord injury (10 studies, seven included in meta-analysis). There was variation in the results for many outcomes based on whether patients were in-patients or outpatients. Rates of accidental catheter removal were higher in in-patients than in outpatients (70.2% versus 7.3%), as was catheter blockage (44.4% versus 28.7%). Urinary leakage was higher in outpatients (52.1%) than in-patients (40%). The rate of gross haematuria in one outpatient study was 43.6% (95% CI 32.8% to 54.7%).
Long-term catheterisation with spinal cord injury (17 studies, 16 included in meta-analysis). These studies showed relatively consistent findings that indicated substantial levels of gross haematuria (13.5%, 95% CI 3.4 to 21.9), false passage (3.1%, 95% CI 0.0 to 6.3%) and bladder cancer (1%, 95% CI 0.0 to 5.0%). High levels of accidental removal were seen in a single study (46.7%, 95% CI 22.9 to 71.3). Urine leakage rates were low in studies that included mostly men (2.0%, 95% Ci 0.0 to 7.1) but too variable to allow pooling in studies that included mostly women (range 7.1% to 92.3%). Other outcomes that showed greater variability between men and women were bladder stones (pooled estimates for male-majority studies 27.2%, 95% CI 12.8 to 40.9%; range in women-majority studies 14.3% to 100%) and urethral stricture or erosion (pooled estimate for male-majority studies 8.7%, 95% CI 0.0 to 18.7; pooled estimate for women in higher quality studies were 37.2%, 95% CI 12.5 to 62.1). Rates of blockage showed great variation which could not be explained.