Each year, there are more than 1 million ED visits for urinary tract infection (UTI) in the United States.¹ Antibiotic treatment for UTIs is avoidable in a significant proportion of patients. Emergency physicians tend to overdiagnose and overtreat because asymptomatic bacteriuria is very common in all age groups, urine cultures are frequently ordered without an appropriate indication, and urinalysis results are often misinterpreted. In this column, my aim is for you to reflect on your practice when it comes to UTI diagnosis and treatment.

Which Historical Features Are Useful?

No single clinical symptom, sign, or lab test is accurate enough to rule in or out a UTI.² Only about half of patients who present with dysuria and frequency will have a UTI. However, for those patients who present with dysuria and frequency and do not have any symptoms suggesting vaginitis or cervicitis (vaginal irritation, bleeding, and discharge), the likelihood of a UTI increases to more than 90 percent, with a positive likelihood ratio (+LR) of 24.6.2 Other helpful historical features include a self-diagnosed UTI, which has a +LR of 4 (as good as a positive dipstick), and cloudy urine appearance in a noncatheterized patient, which has a specificity of 96 percent for UTI.²

Which Patients Require a Workup?

Urine tests are not required for the majority of patients with suspected lower UTI, as it is a clinical diagnosis and the urine tests can be misleading, resulting in over- and undertreatment. Urine tests are usually not necessary for very-low-risk patients or for patients with a very convincing clinical presentation, as your posttest probability will not be changed significantly.

Indications for urine tests for suspected lower UTI include those patients with an intermediate pretest probability for UTI, immunocompromised patients, a history of multiple courses of antimicrobial therapy, a history of antibiotic resistance, or a history of multiple drug allergies. Remember that even a urine culture has a 5 percent false-positive rate due, in large part, to asymptomatic bacteriuria and a 25 percent false-negative rate due to antibiotic use and sample overdilution.²

Think About Differential Diagnosis of Pyuria

Not all pyuria is caused by UTI. Pyuria can result from a wide array of conditions. Dehydration, acute renal failure, sexually transmitted infections, appendicitis, diverticulitis, and the presence of a bladder catheter can all cause white blood cell counts of more than 5 cells/mL on microscopy. While pyuria on its own has a sensitivity as high as 94 percent, its specificity is poor unless combined with positive nitrite. Pyuria and positive nitrite in the setting of a clinical history for UTI has a specificity of 100 percent. However, if microscopy in isolation is used, the result is overtreating 44 percent and undertreating 11 percent of UTIs.⁴ The urine dipstick is even worse. If the dipstick is used in isolation, UTIs are overtreated 47 percent and undertreated 13 percent.

Interpreting Epithelial Cells and Bacteria on Urine Microscopy

The classic teaching is that more than five epithelial cells per high-powered field represents an uncontaminated sample.⁴ However, while a “contaminated” sample may negatively affect the ability to obtain a reliable culture, it does not affect the accuracy of the dipstick or microscopy to the same degree. While bacteria seen on microscopy is predictive of a positive culture, it is not necessarily diagnostic of a UTI, as the positive culture could represent a contaminant or asymptomatic bacteriuria. Just as bladder catheters can cause pyuria, they can cause bacteriuria without UTI as well.⁵

Which Patients Require Antibiotics for Asymptomatic Bacteriuria?

According to the Infectious Diseases Society of America, asymptomatic bacteriuria is not linked to long-term adverse outcomes, and treatment does not decrease the rate of symptomatic UTI.⁶ Asymptomatic bacteriuria is very common in all age groups and is often misdiagnosed as a UTI. Antibiotics for asymptomatic bacteriuria should only be considered in patients undergoing an invasive urologic procedure and in pregnant patients.