Jonathan Sheele1 and Ali S. Raja2 1 Department of Emergency Medicine, Mayo Clinic, Jacksonville, FL, USA 2 Department of Emergency Medicine, Massachusetts General Hospital, Harvard Medical School, Boston, MA, USA Urinary tract infections (UTIs) are common complaints in emergency medicine practice, affecting both women of all ages and older men. The lifetime incidence of UTIs in women is between 50% and 60%,1 and approximately 5% of women presenting to U.S. EDs have genitourinary complaints.2 UTIs are relatively uncommon in young men but do affect older men and are often associated with disorders of the prostate. The most common method for diagnosing UTI in the ED is either a urine dip or laboratory urinalysis. The dipstick urine test measures leukocytes, nitrite, blood, protein, and pH, while a urinalysis quantifies cell counts such as white blood cells, red blood cells, and squamous cells. The diagnosis of UTI can be difficult in the ED because of the inconsistent relationship between the clinical symptoms, bacteriuria, and pyuria. In addition, because the criterion standard test – a clean‐catch or catheterized specimen urine culture – cannot be completed in the ED (because it can take 2–3 days to grow), and the reality that some patients cannot obtain follow‐up within that timeframe to act upon the culture result, emergency physicians must often diagnose and treat UTIs without criterion standard testing. The consequence of these uncertainties is tension between emergency medicine and other specialties regarding the overdiagnosis and overtreatment of UTI.3–5 These differences in approach have led some specialists to analyze the root cause of ED physician’s propensity to misdiagnose UTI that identify factors such as time constraints, knowledge deficits, automatic testing, and malpractice fears as likely precipitants.6 Can history or physical exam accurately rule in or rule out urinary tract infection? A 2013 diagnostic systematic review addressed this question for women and reported positive likelihood ratio (LR+s) and negative likelihood ratio (LR−s) for several symptoms and signs and combinations using data from published literature, focusing on four studies that included ED patients.7 These are listed in Tables 30.1 and 30.2. Overall, neither history and symptoms nor physical examination findings demonstrated large positive or negative LRs to predict the presence of UTI. Table 30.1 History and symptoms predicting urinary tract infections (UTIs) Source: Data from [7]. Table 30.2 Physical examination findings predicting urinary tract infections (UTIs) Source: Data from [7].
Chapter 30
Urinary Tract Infection
Background
Clinical question
LR+ (95% confidence interval)
LR− (95% confidence interval)
Previous UTI
1.4 (0.9–2.0)
0.9 (0.7–1.0)
Dysuria
1.0 (0.8–1.4)
1.0 (0.7–1.2)
Urgency
1.3 (0.8–2.1)
0.9 (0.8–1.1)
Frequency
2.3 (1.4–3.6)
0.2 (0.0–0.6)
Back pain
1.6 (1.0–2.4)
0.8 (0.7–1.0)
Abdominal pain
0.8 (0.5–1.1)
2.0 (0.8–5.0)
Fever
2.2 (1.0–4.6)
0.9 (0.8–1.0)
Hematuria
1.4 (0.6–3.4)
1.0 (0.9–1.0)
LR+ (95% confidence interval)
LR− (95% confidence interval)
Temperature > 37.2 C
1.9 (1.2–3.0)
0.8 (0.6–0.9)
Costovertebral (CVA) tenderness
1.4 (0.8–2.4)
0.9 (0.8–1.0)
Vaginal discharge
0.4 (0.2–1.0)
1.1 (1.0–1.1)
Full access? Get Clinical Tree
