Navarro, Miguel A. Mendoza MD; Lovato, Luis MD
Dr. Mendoza Navarro is a senior resident in the UCLA/Olive View Emergency Medicine Residency Program. Dr. Lovato is a clinical professor at the David Geffen School of Medicine at UCLA and the director of medical informatics for emergency medicine at Olive View-UCLA Medical Center. Read his past columns at http://bit.ly/Journalscan.
Not many studies have focused on the reliability of pyuria to predict simultaneous infection in patients with acute pain from nephrolithiasis, but a recent investigation found that about eight percent of those patients have urinary tract infections. Physicians, however, cannot always rely on the presence of pyuria to identify UTI.
Urinary tract infection is a common clinical condition, especially in women, accounting for more than one million U.S. ED visits every year. A good history and physical should guide initial medical decision-making, but incorporating a urinalysis is imperative, a strategy confirmed by a recent report that no single H&P finding can accurately diagnose a culture-proven UTI. (Acad Emerg Med 2013:20:631.)
Diagnosing UTI can become much more challenging in patients with nephrolithiasis because both can present with similar symptoms such as dysuria, frequency, urgency, and flank pain. Stones can develop as a result of recurrent UTIs caused by urease-producing organisms such as Proteus and Klebsiella (Kidney Int 2013;83:479), and can eventually lead to chronic kidney disease. (Curr Infect Dis Resp 2010;12:450.)
More importantly, patients with obstructive nephrolithiasis who develop infection have the potential to develop abscess or urosepsis, and may deteriorate rapidly. Successful therapy in these cases may hinge on emergent drainage and relief of obstruction. On the other hand, not every patient with nephrolithiasis and mild pyuria should receive a course of antibiotics given the risk of side effects and the propagation of antibiotic resistance.
Association of Pyuria and Clinical Characteristics with the Presence of Urinary Tract Infection among Patients with Acute Nephrolithiasis
Abrahamian FM, Krishnadasan A, et al
Ann Emerg Med
The authors performed a prospective observational study among a convenience sample of ED patients presenting from March 2003 to December 2006 at Olive View-UCLA Medical Center to determine whether pyuria and various demographic and clinical characteristics could diagnose UTI among those presenting with acute nephrolithiasis.
Patients included in the study had to be at least 18, show signs and symptoms of acute nephrolithiasis such as back pain, costovertebral angle tenderness, and hematuria, have evidence of at least one kidney stone on noncontrast CT of the kidneys, ureters, and bladder, according to a board-certified attending radiologist, and have a urine specimen sent for urinalysis and culture before antibiotic administration. Patients were excluded if they had received antimicrobials within 48 hours of presentation, had current urinary tract hardware such as an indwelling urinary catheter or stent, or if they had been enrolled in the study within the previous four weeks.
The primary data analysis included the proportions of patients with positive and negative culture results based on different demographics, such as sex, race, presenting symptoms, presenting signs, and relevant medical history. Other data analysis determined the sensitivity and specificity for urinalysis components. The performance characteristics for various WBC count thresholds among patients without objective or subjective fever was also examined.
The study enrolled 360 patients with CT-proven nephrolithiasis. UTI was defined as growth of a single pathogen at greater than or equal to 10^3 cfu/ml. Twenty-eight patients (7.8%) had UTI; 332 patients (92%) did not. Identified pathogens included Escherichia coli (23), Proteus mirabilis (4), and Staphylococcus aureus (1).
The median temperature of patients with acute nephrolithiasis and culture-proven UTI was 37.7°C compared with 36.7°C in patients without UTI (RR 9.9; 95% CI 4.6–17). Other factors with associated risk of UTI included female sex (RR 27; 95% CI 6.4–162), subjective fever (RR 6.6; 95% CI 3.1–13), history of UTI (RR 6.5; 95% 2.9–15), dysuria (RR 4.4; 95% CI 1.8–11), chills (RR 3.0; 95% CI 1.3–6.7), and urinary frequency (RR 2.5; 95% CI 1.1–5.5).
An elevated peripheral WBC count (>10.9 x 10^9/L) had a sensitivity of 68% and a specificity of 58% for detecting UTI. Pyuria (greater than 5 WBCs/hpf) had a sensitivity of 86% and specificity of 79% for detecting UTI.
Additional sensitivity analysis of performance characteristics of pyuria with higher colony count threshold revealed similar results. At 100,000 cfu/ml, the reported sensitivity was 90%, and specificity of pyuria was 78%. Of note, one patient with UTI had no pyuria (obstructive stone at UVJ), and three patients had only three WBCs/hpf.
The study was performed at a community hospital ED and may not be representative of populations in acute settings such as university hospitals or the urgent care setting. It also used a convenience sample of patients and was not blinded, which could potentially introduce an enrollment bias for patients with suspected infection. The authors also used a low CFU threshold for defining UTI, although they explained the desire to maximize sensitivity to detect infection as early as possible in these patients.
This study doesn't provide us an exact answer about whether a kidney stone is infected, but it does provide us with new information to consider. Fever, chills, prior UTI, dysuria, and frequency all marginally increased the relative risk of UTI. Female gender by far was the factor found to have the highest RR of UTI (RR 27; 95% CI 6.4–162). This finding suggests female patients with acute nephrolithiasis have a much higher risk of having concomitant UTI than their male counterparts. The factor with the next highest RR of UTI was elevated temperature (RR 9.9; 95% CI 4.6–17), although the median temperature elevation of 37.7°C is lower than most thresholds of clinically defined fever. This finding might suggest taking extra precautions in select patients such as checking temperature trends, checking a core temperature, and lowering your clinical threshold for antibiotic treatment even before patients meet established SIRS temperature criteria (>38°C).