Published in the March 2016 issue of Today’s Hospitalist
HOSPITALISTS FREQUENTLY CALL on their infectious disease colleagues to help them work through challenging diagnoses and treatments. But their first order of business might be deciding whether a formal or a curbside consult is called for.
In a presentation at last fall’s management of the hospitalized patient at the University of California, San Francisco (UCSF), infectious disease specialist Jennifer Babik, MD, PhD, an assistant professor at UCSF, addressed that question with some eye-opening data. She cited a study published in the January 2013 Journal of Hospital Medicine that compared curbside and formal consults in terms of their effect on treatment and care planning.
The study found that curbside information was inaccurate or incomplete in half the cases (51%). Further, opting for a formal over a curbside consult changed management in 60% of cases, with 36% of those consults leading to major changes.
“Are curbsides OK? I think the answer is ‘yes,’ but we have to balance concerns about efficiency and patient safety as well as education,” Dr. Babik said. If a specialist can answer a basic question and the patient’s situation isn’t complicated, hospitalists may not need a formal consult. Appropriate curbsides include questions about the spectrum of a specific antibiotic or appropriate therapy duration for a certain type of infection.
“A formal consult for Staph aureus bacteremia can decrease mortality by anywhere from 20% to 30%.”
~ Jennifer Babik, MD, PhD, University of California, San Francisco
But hospitalists should request a formal consult when a diagnosis is unclear or a patient’s situation might be complicated. “We tell our ID fellows that if they have to look up the answer to the question—or if it’s early in the year—there probably should be a formal consult.”
Here’s a look at several common conditions that ID specialists are asked to consult on.
Treating Staph aureus
Dr. Babik frequently is asked how to treat Staph aureus bacteremia. She described a 68-year old patient with a prosthetic mitral valve and end-stage renal disease (ESRD) who’s on dialysis and is admitted with Staph aureus bacteremia. “That infection was thought to be due to his dialysis catheter because his blood culture cleared within two days of starting antibiotics,” she explained.
Here’s the question Dr. Babik posed to attendees: Can you treat this patient for only two weeks? If not, why not, and what information do hospitalists need to make an informed decision about treatment duration? First, said Dr. Babik, while questions about Staph aureus are probably the most common curbsides that ID specialists get, this case needs a formal consult. Several studies have found that formal consults are more likely to lead to better metastatic foci detection, the removal of affected prosthetic devices, more orders for repeat blood cultures and echoes, and a lower relapse risk.
“More importantly,” she said, “a formal consult in these situations can decrease mortality by anywhere from 20% to 30%. Consider getting one for all Staph aureus bacteremia cases if you have the resources.”
To treat, the first step is to ensure infection source control, which entails a thorough physical examination, paying particular attention to neurological, musculoskeletal and abdominal exams. Dr. Babik also always does a good spine and skin exam and looks at all joints. “I have a really low threshold for imaging,” she added.
Hospitalists should also test all patients for endocarditis by ordering either a transthoracic echocardiogram (TTE) or a transesophageal echo (TEE), depending on the situation. (More on that below.)
With those results in hand, hospitalists can proceed to choosing treatment, which include IV antibiotics for bacteremia and “always a beta-lactam for MSSA, if you can,” Dr. Babik advised.
How long to treat?
Duration of therapy depends on whether the bacteremia is complicated or uncomplicated. According to Infectious Diseases Society of America guidelines, bacteremia is likely to be uncomplicated if:
- patients don’t have endocarditis;
- there are no metastatic foci of infection;
- blood cultures are negative within two to four days;
- fever abates within three days of taking antibiotics; and
- patients have no implanted prostheses including valves, cardiac devices or joints.
Prosthetic heart valves and valve rings, Dr. Babik noted, run between a 20% and a 50% risk of being seeded hematogenously with unrelated Staph aureus bacteremia, while prosthetic joints and cardiac devices run a 30% risk.
For uncomplicated bacteremia, treatment duration is at least two weeks. But in complicated cases, four to six weeks of treatment is warranted. And Dr. Babik urged hospitalists to set a low threshold for considering an infection complicated.
“In some ways,” she said, “your default is that these patients need four to six weeks.”
As for whether to order a TTE or TEE, Dr. Babik noted that as many as 15% of patients with Staph aureus bacteremia have endocarditis. The presence of an implanted prosthetic is one important determinant.
Even though TEEs are more expensive and riskier, they’re more sensitive for vegetations, at about 85% to 90% vs. 75% for TTEs. And TEEs are preferred in patients with prosthetic valves, Dr. Babik said, because in many cases “you can’t see much on a TTE” due to shadowing from valves and device leads.
Hospitalists should start with an initial TTE, which Dr. Babik said “may have good negative predictive value in a very narrow subset of patients at low risk for endocarditis.” What’s considered low risk? Patients who have no nosocomial acquired bacteremia or hemodialysis, negative blood cultures, no prosthetic valve or cardiac device, and no signs of secondary infection.
But physicians should move on to a TEE, regardless of a TTE’s negative status, if the TTE is of poor quality or for high-risk patients. Those individuals include patients with a prosthetic valve or implantable cardiac device and those with congenital heart disease or a history of endocarditis. TEEs are also indicated when hospitalists suspect community-acquired bacteremia, a new murmur, multiple metastatic foci or embolic lesions, peripheral stigmata of infective endocarditis, prolonged bacteremia or fever, or new conduction abnormalities.
Another frequent curbside question: Should you treat E. coli (or any other bacteria) if it shows up in a urine culture on admission, even when the patient has no UTI symptoms? In general, said Dr. Babik, the answer is “no.”
She pointed to a study in the April 1, 2014, issue of Clinical Infectious Diseases, which showed that although up to 90% of inpatients with positive urine cultures are asymptomatic, about half those patients receive antibiotics—even though treating them does not decrease their risk of developing a UTI. “It’s just hard to ignore a positive culture,” she said.
Dr. Babik cautioned hospitalists to look at the big picture before treating asymptomatic bacteriuria. In patients with catheters, for instance, pyuria is very common, found in up to 75% of patients with short-term catheters and up to 100% of those with long-term ones.
“Urinalysis has low predictive value for catheterized patients,” she explained. “It’s helpful mostly if it’s negative because the absence of pyuria can suggest an alternative diagnosis.”
Knowing the microbiology isn’t any help either, she pointed out. “It’s the same for asymptomatic bacteriuria as for UTI: E. coli, Klebsiella and enterococcus” are all common causes of both asymptomatic bacteriuria and UTI. So how can you tell asymptomatic bacteriuria from a UTI? In the case of patients who are catheterized or who can’t report symptoms, look for “non-classic” symptoms that could indicate a UTI, including new or worsening fever, rigors, altered mental status and acute hematuria, as long as there is no other source of infection.
Dr. Babik did note three exceptions where asymptomatic bacteriuria should definitely be treated: pregnant women, those undergoing genitourinary procedures and some immunocompromised patients. “Although there are not a lot of data,” she said, “most people will treat in renal transplant patients for the first three months and for neutropenia.”
In patients with a Foley or altered mental status, the hardest category to decide whether to treat is when patients have both a positive urinalysis and a positive culture. For these patients, she said, hospitalists might proceed with UTI treatment “if you don’t identify an alternative diagnosis.”
Candida in urine, ESBL
How about treating Candida in patients’ urine? “Generally, you don’t,” said Dr. Babik. “In patients with catheters, changing the Foley can get rid of it in 20% to 40% of cases.” The exceptions for when you need to treat are the same as for asymptomatic bacteriuria: pregnancy, urologic procedure and neutropenia.
If, however, you find an actual Candida UTI, fluconazole is first-line treatment. “All other options have either poor efficacy or poor side effect profiles,” she said, singling out flucytosine and Amphotericin B bladder washes. Echinocandins are an option only if there’s systemic disease, but not to treat a UTI because they have poor urinary penetration.
Dr. Babik said she also fields a lot of questions related to extended spectrum beta-lactamase (ESBL) producing organisms and cystitis. Do doctors have good oral options?
The optimal choice is usually fosfomycin, which has a 94% clinical cure rate and a 78% micro cure rate in either a single dose or three doses every other day (QOD). As for nitrofurantoin, one study found that a 14-day course had a 69% clinical cure rate. In another outpatient study, amoxicillin-clavulanate had a 93% clinical cure rate with a five to seven-day course.
“Fosfomycin does not achieve good renal levels, so you can’t use it for pyelonephritis,” said Dr. Babik.
Nitrofurantoin does not achieve good renal levels either, and it shouldn’t be prescribed in renal failure patients because of its toxicity profile.
As an oral treatment for E. coli ESBL cystitis when there are no other oral options, Dr. Babik said she recommends oral fosfomycin at 3 grams QOD for three doses or until clinical improvement occurs.
“Fosfomycin really is your best bet in terms of possible susceptibility,” she pointed out. “But I would not use an oral agent if the patient is ill, has bacteremia or can’t be followed closely.” In addition, there are very few data on using oral agents in the setting of ESBL E. coli pyelonephritis.
Switching to PO
Then there’s this perennial question: when to switch patients from IV to oral antibiotics. Dr. Babik cited the case of a 45-year-old patient with diabetes who is admitted with pyelonephritis. Both blood and urine cultures are positive for pan-sensitive Klebsiella. The patient was treated empirically with ceftriaxone, and her white blood count normalized.
How long should this patient receive IV antibiotics? A systematic review published online in November 2008 by Current Medical Research and Opinion looked at the efficacy of an early switch (during days 1-4) vs. a late switch (days 7-10) and found no difference.
“It’s OK to change to PO once the patient is improving clinically and if you know susceptibilities,” said Dr. Babik.
As for PO therapy for uncomplicated pyelonephritis, IDSA guidelines released in 2011 recommend 500 mg of cipro twice a day for seven days. The guidelines also suggest trimethoprim/sulfamethoxazole (TMP-SMX) for 14 days and beta-lactams for between 10 and 14 days. Pyelonephritis is considered uncomplicated if the patient is not pregnant, is premenopausal and doesn’t have urogenital abnormalities.
As for complicated pyelonephritis, no guidelines currently exist. “I think most people extrapolate from the uncomplicated guidelines and treat between seven and 14 days,” she said. “Urogenital abnormalities is the one situation where you may opt for the longer course.”
Hospitalists also struggle deciding when or if to replace an infected central line. Generally, when patients are bacteremic and the line is the presumed infection site, doctors’ inclination is to remove it. But that’s not always necessary or even advisable, Dr. Babik pointed out. Instead, hospitalists should first seek more information.
What’s often not helpful, she said, is evaluating inflammation at the line site, which is “incredibly insensitive.” Doctors can do a line tip culture, which has a 79% sensitivity and 92% specificity as long as patients have a positive blood culture and more than 15 colony-forming units per plate of the same organism from the catheter tip.
But what she and her colleagues at UCSF use is differential time to positivity. “It allows for diagnosis without removing the line,” she said. “You draw a culture from the line and peripheral blood at the same time, then look for the line culture to be positive first.” If the line culture is positive at least two hours before the peripheral culture, that’s 85% to 95% sensitive and almost 90% specific, Dr. Babik explained. If it’s positive in less than two hours, it’s likely a different source.
What do you do when the line culture is positive and the peripheral is negative? That’s a very common scenario, she noted, and is usually caused by either a line being colonized “or just a contaminant.” But among patients with urinary infections and low-grade bacteremia with only one out of two cultures positive, “you have a 50% chance that one positive culture will come out of the line,” she noted. “So you can’t just automatically assume it’s colonization or contaminant.”
When should the line definitely be removed? In the presence of virulent organisms such as Staph aureus, Pseudomonas and Candida, and for bacteria that are difficult to eradicate such as Micrococcus, Bacillus or propionobacteria.
Dr. Babik also singled out the following scenarios that warrant line removal: severe sepsis or bacteremia that doesn’t respond within 72 hours to antibiotics, septic thrombophlebitis, an exit site or tunnel infection or abscess, or evidence of metastatic infection.
Bonnie Darves is a freelance health care writer based in Seattle.