The Infectious Diseases Society of America and the European Society for Microbiology and Infectious Diseases recently updated its guidelines for treating acute uncomplicated cystitis in women, recommending only three common antibiotics and eschewing fluoroquinolones. (See FastLinks.) The same article also pontificated on the treatment for uncomplicated acute pyelonephritis in women.
The guidelines were limited to premenopausal non-pregnant women with no known urological abnormalities and no significant comorbidities. The population had so-called uncomplicated pyelonephritis, but many decidedly complicated individuals populate the ED. The resistance and prevalence of the local flora and the collateral damage of antimicrobial therapy were considered important factors in making optimal empirical treatment decisions, just as they were with the recommendations for acute cystitis.
International Clinical Practice Guidelines for the Treatment of Acute Uncomplicated Cystitis and Pyelonephritis in Women: A 2010 Update
Gupta K, Hooton TM
Clin Infect Dis
Unlike cystitis where treatment is similarly initially empirical and cultures are rarely obtained, the organizations recommend a urine culture and sensitivity always be performed on patients with suspected pyelonephritis. Antibiotic treatment choices are based on the need for hospitalization, although specific parameters mandating hospitalization were not forthcoming.
Only two outpatient oral antibiotics were deemed academically proper for pyelonephritis: two specific quinolones and TMP-SMZ. Acceptable empirical therapy for pyelonephritis include oral ciprofloxacin (500 mg BID for seven days), with or without an initial 400 mg IV loading dose of ciprofloxacin. This is considered acceptable therapy for outpatients only when the prevalence of resistance of community uropathogens to fluoroquinolones does not exceed 10 percent. If one is using the hospital antibiogram to compile sensitivities for all infections, the resistance rate will likely be well above 10 percent. No data look just at urine cultures of ED patients, so this recommendation may not work for the EP.
The initial treatment of a soon-to-be outpatient with an IV loading dose of an antibiotic is rather common, but in this document it seems to be somewhat of a soft recommendation rather than a standard or mandate. Many EPs have opted for this reasonable intervention for years because it seems so rational. Oral quinolones, however, produce serum levels equivalent to IV administration, so perhaps it makes sense to use a non-quinolone for the IV loading. Alternatives to IV ciprofloxacin include IV ceftriaxone (1 g) or a consolidated 24-hour dose of an aminoglycoside, such as gentamicin. A once-daily 1000 mg oral dose of ciprofloxacin for seven days and oral levofloxacin (750 mg for five days) are other quinolone alternatives. Note that other quinolones, such as moxifloxacin, an otherwise powerful antibiotic, should not be used to treat urinary tract infections. It does not concentrate in the urine.
The other acceptable oral therapy is oral trimethoprim-sulfamethoxazole (TMP-SMZ), with one double-strength tablet two times a day for 14 days. An initial IV loading dose of either ceftriaxone or an aminoglycoside was likewise suggested.
Oral beta-lactams are less effective than other agents available for pyelonephritis. If one chooses an oral beta-lactam, an initial IV dose of a long-acting parenteral antibiotic, such as 1 g ceftriaxone or a so-called consolidated 24-hour dose of an aminoglycoside is recommended. This “consolidated dose” of aminoglycosides may be unfamiliar to many.
A number of intravenous regimens are acceptable for patients requiring hospitalization for pyelonephritis. Unlike cystitis, the first recommended choice for inpatient pyelonephritis is an IV fluoroquinolone. Alternatively, an aminoglycoside with or without ampicillin is acceptable. An extended spectrum cephalosporin or an extended-spectrum penicillin, with or without an aminoglycoside, is acceptable, as is single therapy with an IV carbapenem.
The consolidated 24-hour dose of tobramycin or gentamicin is 7 mg/kg in someone with normal renal function. (Ann Intern Med 1996;124:717.) This order surely will be questioned by the pharmacy because it is so large compared with other aminoglycoside regimens that are spaced out over the day. Essentially, the total daily regimen is given as a single dose, an effective but rarely chosen option in the ED. The initial one-time loading dose of tobramycin or gentamicin is given to patients about to be discharged rather than spaced out during the day, as would occur in the hospital.
The guidelines curiously do not address the rather important issue of extended spectrum beta-lactamase production by E. coli and Klebsiella, common nasty uropathogens. Routine urine cultures are done because pyelonephritis is obviously a serious infection, resistance is common, and modifications of empirical therapy may be required based on culture results and patient response.
Comment: Acute pyelonephritis is less common than acute cystitis, although both are familiar denizens of any busy ED. A urinary tract infection in women begins with colonization of the vaginal introitus by uropathogens from fecal flora, usually E. coli. The bacteria ascend via the urethra into the bladder. Pyelonephritis then develops from pathogen migration to the kidneys via the ureters. It is not well appreciated that many patients with clinical symptoms of only cystitis also have a silent pyelonephritis, and may relapse if treated only for a bladder infection. Identifying that subgroup, however, is banter at the ID cocktail party, not part of ED sleuthing. A variety of factors, including immunocompromise and ureterolithiasis, dictate the progression from the bladder to kidneys. Unlike cystitis, pyelonephritis can be caused by a bacteremia seeding the kidneys, classic cases being caused IV drug abuse or endocarditis.
Pyelonephritis is usually associated with fever, chills, flank pain, CVA tenderness, and often nausea and vomiting. These patients are ill and feel miserable. Occasionally, patients rapidly progress to sepsis, renal failure, renal abscess multiorgan system dysfunction, and death. A preceding cystitis may or may not have been clinically obvious. A diagnosis of pyelonephritis may not be that straightforward for patients presenting with advanced sepsis.
Imaging studies are not usually required for most cases of pyelonephritis, although a CT scan will demonstrate abnormalities consistent with infection. Ordering a CT scan will rule in or out urolithiasis, the worst scenario being an obstructing stone. Pyelonephritis becomes a much more complicated issue with a stone in the GU tract. Complete obstruction of the ureter can occur secondary to a stone, producing a falsely benign-appearing urinalysis and rapid sepsis because it is essentially a closed-space infection with ready access to the blood. Patients with pyelonephritis can progress to a bacteremia. Simply stated, pyelonephritis is not a simple disease process even in the absence of urinary tract obstruction.
It's always a clinical call whether otherwise healthy women with suspected pyelonephritis should be admitted to the hospital. Pyelonephritis in the elderly, men, and children is by definition complicated, and is not an outpatient disease. My paranoia lends supports to a liberal admission policy for women with pyelonephritis, not a popular decision with CMS or insurance companies that seem to require the patient be on death's door to earn an inpatient bed. The general gestalt for me often hinges on “sick versus nonstick” and the ability to hold down oral medications, get to the drugstore, or have the wherewithal to get back to the ED if things worsen. Compliance with follow-up is always a criterion, but it is difficult to judge in someone you've never met before.
Initial treatment of pyelonephritis with an intravenous loading dose of ceftriaxone, aminoglycoside, or a fluoroquinolone is a common intervention and seems intuitively reasonable. I don't know why the recommendation on an IV loading dose is not emphasized more; it seems cavalier of the prestigious IDSA. I can see absolutely no downside to it, even if the patient does not appear that ill in the ED. Things go south quickly in the outpatient, and your prescription for well meaning and perfectly matched antibiotics does not fill itself.
Occasionally, it's a good call to administer the first dose of an IV antibiotic, and then observe a patient to decide whether outpatient therapy will be appropriate. There is also no reason not to start the first pill at that time, and maybe give one dose to go in case the patient cannot get to the drug store right away. Ward et al demonstrated that parenteral antibiotics given in a 12-hour observation unit followed by outpatient oral antibiotics is usually effective in mild to moderate pyelonephritis. (Ann Emerg Med 1991;20:258.) Obviously, any pregnant woman with pyelonephritis should be admitted. Drugs recommended for cystitis, particularly nitrofurantoin and fosfomycin, should be avoided in pyelonephritis because they do not achieve adequate renal tissue levels.
The emergence of ESBL-producing organisms creates risk for initially treating an infection caused by an ESBL organism and with standard empirical therapy and having it fail. ESBL-laden patients initially look the same to the clinician and to the urinalysis and CBC. ESBL infections are possible in a seemingly healthy individual, and ESBL bacteria do not seek out only diabetics, the elderly, or those with HIV. ESBL infections, especially of the urine, have become like C. difficile colitis — a community-acquired infection with no obvious reason to attack the normal citizen. The only treatment for pyelonephritis currently available for an ESBL infection is a carbapenem. Meropenem, ertapenem, and imipenem have been advocated, but there is no way to know if the dreaded ESBL Klebsiella is lurking in the urine of your seemingly well patient or if the culprit is merely a wimpy E. coli, readily annihilated by a few pills. Perhaps I am just feeling lucky when I send home anyone with pyelonephritis.
The rather effective and once popular but now obscure and largely forgotten initial IV therapy with the total 24-hour dose of an aminoglycoside is a strategy that may be unfamiliar to many younger EPs. Instead of giving tobramycin or gentamicin throughout the day, one simply administers a so-called 24-hour consolidated dose, usually 7 mg/kg intravenously. But any initial IV antibiotic tactic ensures compliance to at least the first dose, provides antibiotic coverage for 24 hours, and should relieve symptomatology enough that the patient is comfortable and can then be compliant with outpatient oral therapy. I prefer 1 g IV ceftriaxone and oral medication for those going home with pyelonephritis. If this patient is not markedly better in 24–36 hours, a recheck is in order.
The emergency physician should be very respectful of acute pyelonephritis. It should not be viewed as a simple extension of cystitis. Antibiotics concentrate in the urine, but it's a soft tissue problem when the kidney is infected. The threshold for admission should be low, although I have never seen strict criteria. All experts seem to agree that some infected kidneys can be totally treated at home. I guess they are feeling lucky with that approach. There is nothing wrong with keeping a patient for observation or in the ED for 12 hours to see if the IV loading dose will be effective or if oral therapy will be tolerated. Intravenous fluid, urine cultures, occasionally even blood cultures, and judicious narcotics make life easier. The propensity for pyelonephritis to turn into a bacteremia rather quickly should be appreciated.
One also has to consider the possibility of an infected stone. One helpful hint for the avid urolithiasis-seeker is always to look at urine pH. The prescient and sagacious clinician knows that if the pH of the urine in a patient with pyelonephritis is above 7, it signals infection with a urea-splitting organism, a setup for an infected stone. (See sidebar.)
Click and Connect! Access the links in EMN by reading this issue on our website or in our iPad app, both available on www.EM-News.com.
Why EPs Should be Concerned about Pyelonephritis if the Urine pH is 7.0 or Higher
Most clinicians look at the red and white blood cell counts, and consider the presence or absence of bacteria when cogitating on the urinalysis of a patient with suspected pyelonephritis. Normally, however, the urine pH is given a furtive glance with little or no analysis. The urine is somewhat acidotic, so distinctly alkaline urine is not only unusual but also a cause for concern.
The issue is one of struvite stones in the alkaline urinary tract. These stones are composed of magnesium ammonium phosphate (struvite) and calcium carbonate (apatite). Struvite stones only form when ammonium production is increased, and the urine pH is elevated enough to decrease the solubility of phosphate. The only situation producing this in the upper urinary tract is an infection with a urea-splitting organism, specifically Proteus or Klebsiella.
Struvite stones are called infection stones, and they are problematic. They may grow rapidly over a period of weeks to months. They often develop into the well known staghorn or branched calculus that fills the entire intrarenal collecting system. Alternatively, a more common calcium-based stone can be supplemented with struvite by repeated infections. Urea-splitting bacteria live in the interstices of a struvite stone where antibiotics may not penetrate. Here the bacteria create a persistently local alkaline environment promoting further stone growth.
Staghorn calculi are usually asymptomatic, but they do not dissolve and will not be spontaneously passed. A urinalysis may show ammonium phosphate crystals, but will otherwise not suggest the presence of a struvite calculus. Staghorn calculi need to be out of the kidney. If you see one on a radiograph taken for another reason, let the patient know he needs to see a urologist in the near future. Recommendations for definitive therapy include percutaneous nephrolithotomy or lithotripsy. Asymptomatic large struvite stones can clandestinely yet nefariously cause deterioration in kidney function and progress to end-stage renal disease. A staghorn calculus found on a radiograph of a patient with sepsis mandates consideration of pyelonephritis as the septic source.
Patients with flank pain, fever, and pyuria are considered likely to have pyelonephritis, and imaging of the urinary tract is not routine. Urine pH 7 or higher should prompt one to consider strongly not only the presence of a struvite stone, possibly an obstructing stone, but also a resistant infection due to ESBL-producing Klebsiella. The presence of ESBL-producing Klebsiella infections are on the rise, necessitating carbapenem treatment rather than the standard antibiotics recommended for uncomplicated pyelonephritis.
It is generally assumed that pyelonephritis in the presence of a urinary calculus is an indication for admission. Such patients are at higher risk for developing sepsis, and may not be adequately treated with an outpatient course of oral antibiotics. The urine pH may serve as a reasonable surrogate to decide on an imaging modality, which may be as simple as KUB, which will readily show a staghorn calculus.© 2012 Lippincott Williams & Wilkins, Inc.