Evaluation of diagnostic criteria for bacteriuria in acutely spinal cord injured patients undergoing intermittent catheterization.

To evaluate diagnostic criteria for bacteriuria in acutely spinal cord injured patients undergoing intermittent catheterization, we studied paired urine specimens obtained by suprapubic aspiration and intermittent catheterization. Culture of suprapubic aspirate was used to define presence or absence of bacteriuria. Fifty patients were studied for an average of 5 consecutive days; bacteriuria occurred within the study period in 47 (94%). Low-level bacteriuria was frequent; thus, the traditional diagnostic criterion, greater than or equal to 10(5) cfu/ml of midcatheter urine, had unacceptably low sensitivity (gram-positive organisms 0.45; gram-negative organisms 0.65) for bacteriuria documented by suprapubic aspiration. The best diagnostic criterion for gram-positive bacteriuria was between greater than or equal to 10(1) cfu/ml (sensitivity 0.91, specificity 0.86) and greater than or equal to 10(2) cfu/ml (sensitivity 0.85, specificity 0.93). For gram-negative bacteriuria, greater than or equal to 10(1) cfu/ml was optimal (sensitivity 0.96, specificity 0.96); a more practical criterion, greater than or equal to 10(2) cfu/ml, retained excellent sensitivity (0.91). Suprapubic or flank pain and/or tenderness occurred in five of 47 bacteriuric subjects; nonspecific symptoms, possibly associated with bacteriuria, were seen in an additional 28 subjects. We conclude that, in this unique population, a criterion of greater than or equal to 10(2) cfu/ml of midcatheter urine should be used for diagnosis of bacteriuria.