Urinary tract infections in male spinal cord injured patients. Part one: Bacteriologic diagnosis.

Seventy-one paired samples of urine, obtained by collection of a clean voided specimen and by suprapubic aspiration of the bladder (SPA), were obtained from 53 male spinal cord injured patients who were free of indwelling catheters. A direct comparison of the number of colony forming units of bacteria within the paired specimens indicates that single voided specimen is adequate in 63 of 71 (88.7%) instances. While using a threshold of 10(4) organisms per cc of urine obtained by SPA, the presence of at least 10(4) organisms per cc of voided urine provided a sensitivity and specificity both of 100%. Such high degree of correlation between SPA and voided specimen is explained on the basis of altered urodynamics in these patients and repeated contamination and transmission of bacteria from anterior urethra into the bladder. In patients with colony counts of 10(2) to 10(4) per cc, in which the voided urine contains mixed flora or staphylococci, bladder urine is most likely to be sterile and accurate assessment may require SPA of the bladders. The lack of normal voluntary control of bladder emptying in these patients made collection of SPA samples easier than voided specimens. Conventionally accepted criteria for the diagnosis of urinary tract infection lack sensitivity when applied to this group of patients.