Managing urinary tract infections. (ID Consult).
Other important issues to consider include how your lab calculates colonies/mL, drug resistance rates in your community, and renal system imaging for first urinary tract infections (UTIs) in children.
Once you have ruled out acute otitis media or other causes of fever, you suspect UTI in girls, particularly those <2 years old, and in uncircumcised boys <6 months of age.
Although catheterization is invasive and takes 510 minutes more than bagging, in the long run it can save trouble.
A recent study of 7,584 outpatient samples in children 2 years old or younger had contamination of just 9.1% of catheterization-acquired urine (cath-urine) specimens, compared with 62.8% of carefully collected bag urine specimens. Of 3,440 contaminated urine specimens, risk of unnecessary interventions, such as antibiotics, imaging, or hospital admission, was more than four times higher (J. Pediatr. 139:168-69, 2001).
Therefore, bag urine specimens should not be the basis for antibiotics.
The next issue to consider is how your lab calculates colonies/mL. A thousand colonies/mL sounds like a lot. However, for every 1,000 colonies reported, there is only 1 colony on the agar plate. This results from 0.001 mL of urine being inoculated onto agar, so 1 colony on the plate calculates as 1,000/mL. Some labs inoculate 0.01mL, so each 10 colonies on the plate calculates to 1,000/mL and more reliably predicts contaminants at <50,000 colony-forming units/mL.
Find out which method your lab uses, so you know how few actual colonies were detected. In children, reports of <10,000 colonies/mL--even in cathurines--indicate a contaminant. For a dean-catch urine, <50,000 colonies /mL (particularly with more than one organism) is likely a contaminant.
Expect Escherichia coli to be the pathogen in >85% of UTIs. A non--E. coli pathogen or multiple organisms (>100,000 colony forming units/mL) can occur with abnormal urinary tract anatomy or in a patient on antibiotics.
It's important to know localE. ccli resistance rates to decide which of the three usual first-line drugs--trimethoprim/ sulfamethoxazole, nitrofurantoin (Macrodantin), or cephalexin (Keflex)--will work best in your practice. Where I practice, E. ccli are approximately 25% resistant in vitro to trimethoprim! sulfamethoxazole (TMP/ SMZ); about 15%-20% are resistant to cephalexin; and 5% to nitrofurantoin.
But, each of these drugs is concentrated from 10- to 1,000-fold in urine, compared with serum, so pathogens resistant in the blood can be susceptible in the urine, leading to good clinical responses despite resistance in the test tube. Nevertheless, if >30% of E. coli are resistant to a drug in vitro, I would nor have confidence in that drug as a first-line UTI drug.
If the first-line drug resistance is high in your area, consider an oral third- generation cephalosporin, such as ceftibuten (Cedax) or cefixime (Suprax). These drugs are approximately 100 times more active against E. coli than the standard UTI drugs.
If the urinalysis or clinical signs show UTI predictors in a toxic or highly febrile child, hospitalization for treating possible bacteremia and! or pyelonephritis is warranted. Even with a UTI-predicting urinalysis, toxic infants could have meningitis without meningeal signs, so their sepsis work-up should include a lumbar puncture, blood, and cath-urine culture.
Imaging, including ultrasound to look for abnormal anatomy and voiding cystourethrogram (VCUG) to assess reflux, is indicated in children with a first UTI. Ultrasound can be done immediately, but VCUG timing is controversial. Some experts order VCUG during initial treatment as long as treatment has sterilized the urine, while others wait 24 weeks because renal system inflammation during infection may make reflux appear one grade worse than it is.
If you wait 2-4 weeks for the VCUG, use prophylaxis--usually TMP/SMZ-- until results are available. Unfortunately, neither ultrasound nor VCUG definitively detect an upper tract infection that could injure parenchyma and physical indicators such as flank pain are not easily detected in young children.
I use C-reactive protein to help decide to order a dimercaptosuccinic acid renal scan to confirm parenchymal inflammation and/or injury. It doesn't add much cost and reflects systemic response to the UTI; the level should be higher with renal parenchymal inflammation. If the Creactive protein is high, greater than 10 (our normal is <0.8), upper tract infection is more likely and a renal scan 6 weeks later can detect long term parenchymal injury.
DR. CHRISTOPHER J. HARRISON is professor of pediatrics in the division cf pediatric infectious diseases at the University of Louisville and Kcsair Children's Hospital, Lcuisville, Ky.
Coming next month: Dr. Michael E. Pichichero shares his experience with PANDAS.
|Printer friendly Cite/link Email Feedback|
|Author:||Harrison, Christopher J.|
|Date:||May 1, 2002|
|Previous Article:||Overdiagnosis and overtreatment of Lyme disease a problem on Long Island. (Only 32% Correctly Diagnosed).|
|Next Article:||Pediatric studies of common cold drug continue. (Expert Panel Rejected Approval for Adults).|