Asymptomatic urinary tract infections are common in diabetic patients. The aim of this 10 months prospective study is to evaluate urinary dipstick tests versus cytobacteriological examination to conclude the absence of urinary tract infection in diabetic subjects. Each diabetic patient hospitalised for less than 8 hours and for whom it was decided a cytobacteriological examination was included in the study (141 samples). At the same time (and at patient's bedside) a dipstick urinalisys (glucose, leucocytes, nitrite, blood, protein, and ketone) was carried out. Sensitivity, specificity, negative predictive value, post test probability and negative likehood ratio were calculated.
Results: the combination of leucocyte zone with nitrite zone (both negative) has a 85,2% sensitivity, avoids 65% of cytobacteriology, but has an odd ratio at 0,20. Those results are improved when the glucose zone (negative test or less than 4 crosses) is taken into account, with a 96,3% sensitivity, 63,4% cytobacteriology avoided and a negative likehood ratio at 0,06.
Conclusion: The addition of the glucose test to the usual leucocytes and nitrite tests seems to allow one to conclude the absence of urinary tract infection in diabetic patients; this is worth studying with a more extensive sample.