Published in the July 2011 issue of Today’s Hospitalist
Do you find yourself having to wait longer and longer for an ID consult? According to Shanta M. Zimmer, MD, associate professor of medicine and an infectious diseases specialist at the University of Pittsburgh, hospitalists can play a bigger role in working through many of the grey areas that typically surface in patients with infections.
Usually, said Dr. Zimmer, who spoke at this spring’s Society of Hospital Medicine meeting in Grapevine, Texas, many of the questions she is asked as an ID consultant revolve around a few basic areas. And many times, she added, hospitalists can find answers to those questions in clinical practice guidelines issued by the Infectious Diseases Society of America (IDSA).
"One of the most common questions I’m asked is ‘When can we treat patients for two weeks vs. a longer course?’ " she said. "The IDSA guidelines are really useful in answering that."
Here’s a look at four infectious disease cases that Dr. Zimmer presented.
CASE 1 Treating MRSA
A 52-year-old diabetic woman was admitted for nausea and vomiting due to hyperglycemia. She had a PICC line placed for IV fluids and IV insulin and responded to that treatment, but on the third day of hospitalization, she developed a fever. Blood cultures showed she had MRSA bacteremia.
How long should this patient be treated for MRSA? Because she is diabetic, Dr. Zimmer explained, a longer course of four weeks is warranted. But for most patients, the IDSA guidelines are fairly clear on how to treat patients with an uncomplicated bloodstream infection and fever that resolves within 72 hours “as long as there is no evidence of endocarditis, active malignancies or immunosuppression. Physicians should remove the catheter and treat those patients with a systemic antibiotic (in this case, vancomycin) for a short course of 14 days.
"There is growing evidence, but not great evidence, that short-term treatment may be acceptable for uncomplicated patients," Dr. Zimmer said. "If you ask 100 different ID physicians ‘When can I treat for two weeks?’ you will get 100 different answers, but there is growing evidence about the duration."
She did, however, add this caveat: "You definitely want to do a TEE [transesophageal echocardiography] if you are going to do the two-week course. A TEE is better than a transthoracic echocardiography [TTE] to rule out endocarditis."
When is an infection related to a catheter?
To find out if a patient’s bloodstream infection is related to a catheter, Dr. Zimmer said, hospitalists can use this fairly simple strategy: Ask that blood be drawn from both near the catheter tip and from the periphery. If cultures grow two hours earlier in the catheter-tip blood than in the peripheral blood, the infection is probably related to the PICC line.
"The differential time to positivity is something that we think about," Dr. Zimmer said. "This is something you can use on the spot when you are taking care of patients without being in a very sensitive lab. There are a lot more bacteria on the catheter if the infection is catheter-associated than if it is a peripheral blood culture."
The quick-and-easy way to do this, Dr. Zimmer said, is to look at when the blood cultures were drawn from the line and from the periphery, and note what time you get the call from the lab saying it was positive. "If it is earlier from the line," she explained, "it’s probably catheter-associated."
While the IDSA guidelines call for removing catheters that have been linked to a bloodstream infection, Dr. Zimmer acknowledged that doing so is often easier said than done. The line may have been inserted for the long run, or there may be specific reasons it can’t be removed. Such circumstances are likely to produce pushback from other physicians.
"I know there will be some catheters that you can’t remove, like in the case of dialysis patients," Dr. Zimmer said. But even when taking out a long-term line is an unpopular decision, "removing lines makes it a lot less difficult to treat an infection, and complications are going to be lower. In most cases, it’s the best thing for the patient."
While some patients need lines in while their infection is treated, Dr. Zimmer stressed that some bloodstream organisms are just so virulent and "sticky" that they will never be cleared as long as a catheter remains in place. Candida and Staphylococcus aureus, she said, are two good examples.
CASE 2 Candidemia
In Dr. Zimmer’s second case, a 35-year-old breast cancer patient between cycles of chemotherapy came to the hospital with fever and chills. Blood cultures from the port and the periphery produced C. albicans. The culture from the port also grew coagulase negative Staphylococcus. Dr. Zimmer was called in and asked what to do.
She immediately removed the port. Because the Candida species grew from both the periphery and the port cultures, Dr. Zimmer said, the infection was very likely related to the catheter.
"Candida is a sticky bug," she noted. "It creates a huge biofilm, and you can’t clear it without removing it. If you are going to leave a device in and somebody has candidemia, you really need to get an ID expert on board to decide why that is the case. Antibiotic lock therapy is not approved for use in fungal infections."
And while the patient had candidemia because of the catheter, she didn’t have bacteremia. Because the Staphylococcus grew only from the port, not the periphery, the patient’s bloodstream was not infected.
"There’s probably just a contaminant on the line," she said, "not necessarily something you need to treat."
For hospitalists who want to be sure about the presence of candidemia, Dr. Zimmer explained that you don’t need to send every patient with candidemia for echocardiography. Only if someone has metastatic complications or if a patient has a very hard time clearing the infection from the bloodstream, she said, would you consider doing an echo to look for an unrecognized source.
Dr. Zimmer also reminded hospitalists to not treat tip cultures, but to instead just pull the catheters out. "Only treat if the bloodstream is positive," she said. "If you find S. aureus on a catheter tip, but not in the bloodstream, don’t worry about it."
When a patient does have candidemia, how long should therapy continue? Dr. Zimmer again referred to the IDSA guidelines.
"There’s no controversy about how long to treat," she said. "Treatment should continue until there is clearance from the bloodstream, and then 14 days from the date of a negative blood culture." Since Candida can take time to grow in cultures, Dr. Zimmer stressed the importance of waiting "four or five days before saying for sure that blood cultures are clear and it’s safe to determine the duration of therapy."
When treating fungal infections like Candida, she said, start with the echinocandin fungal drugs, such as caspofungin. Once cultures come back, and depending on each hospital’s own flora and resistance experience, you can switch patients to fluconazole, which has good bioavailability and is less expensive.
Dr. Zimmer emphasized that it’s important to treat until the bloodstream is clear of Candida because complications can be "pretty significant." Those include metastatic disease, endocarditis and endophthalmitis. And all patients need to have retinal exams because many patients “up to 10% of individuals who don’t have the infection for a prolonged period of time ” suffer ocular problems.
"A longer duration puts you a higher risk for ocular problems," Dr. Zimmer said, noting that some people losing their vision or their eye.
CASE 3 CNS Lyme disease
Another area that Dr. Zimmer regularly consults on is Lyme disease, particularly central nervous system (CNS) forms. She presented the case of a 25-year-old woman with no medical history who came to the hospital with bilateral facial palsy and low-grade fever.
According to Dr. Zimmer, bilateral facial palsies are rare; among possible noninfectious causes “from vasculitis to cancer “neurosarcoidosis is the most common. If you’re looking at potential infectious causes, however, Lyme disease is one of the more common, depending on where the patient lives. (Other differential diagnoses include the varicella-zoster, herpes simplex and Epstein-Barr, syphilis and bacterial meningitis.)
Diagnosis is difficult, Dr. Zimmer said, because the Lyme antibody is sensitive but not specific. "If someone comes to you saying they carry a diagnosis of chronic Lyme but their western blot is negative," she explained, "they probably don’t have it."
CNS Lyme disease is even more challenging to diagnose, she said, because "the symptoms are very nonspecific." As a result, it can present as a lot of things, including a symptom like bilateral facial palsy with a history that reveals that patients had an exposure.
In the case of this patient, she said, treatment is needed, but the exact details are controversial. While U.S. guidelines call for IV ceftriaxone or penicillin for 14 days, European studies have shown that doxycycline works as well. Studies have also shown that treating for a longer amount of time is no more effective than a shorter, 14-day course.
"Some people are going to have chronic symptoms," Dr. Zimmer said, "and treating them longer doesn’t change that."
CASE 4 Pneumonia in HIV patients
Treating pneumonia with respiratory failure in HIV-infected patients is another big source of calls to ID subspecialists. Although the list of differential diagnoses in HIV patients can be longer, Dr. Zimmer said, it’s important for hospitalists to remember that community-acquired pneumonia is still the most likely cause.
Consider the case of a 45-year-old man with HIV who had a three-week history of cough, fever and shortness of breath that initially improved but then worsened before he was admitted.
"Even though he is at risk for a lot of different diseases," Dr. Zimmer said, "community-acquired pneumonia is still the most common. Influenza pneumonia or post-influenza bacterial pneumonia is also a possibility. You are going to worry about what time of year it is and what the flu pattern is in your region."
If a patient has HIV, she explained, spend the time compiling a good history to find out if the patient was taking drugs like Bactrim for prophylaxis against pneumocystis (PCP) and whether there was exposure to tuberculosis. Also take a travel history and get information about other diseases in the community, such as H1N1 influenza or pertussis.
This patient was started on ceftriaxone and azithromycin to cover community-acquired pneumonia. He also received oseltamivir, and his blood cultures grew Streptococcus pneumoniae. Until tuberculosis is ruled out, Dr. Zimmer said, such patients should be kept in respiratory isolation. And if PCP is suspected, she recommends starting patients on steroids as well.
Deborah Gesensway is a freelance writer who covers U.S. health care from Toronto.
IN A PRESENTATION at this spring’s Society of Hospital Medicine meeting, Shanta M. Zimmer, MD, an infectious diseases specialist at the University of Pittsburgh, offered hospitalists general rules to help them with their infectious diseases patients:
- Remember that not all fevers are infectious, so don’t be shy about repeating blood cultures before starting patients on antibiotics. Having a second set of negative blood cultures can help you avoid "playing a guessing game" later when organisms are found growing in the bloodstream, but you don’t know if those were caused by the drugs you prescribed or if they are truly signs of an infection. Also, hold off on giving stable patients antibiotics when the infectious etiology is unknown.
"Antibiotics complicate things when a patient has a fever and it’s not clear if it has an infectious source," Dr. Zimmer said. "Take time. There are very few things in medicine that are an emergency, and we often have time to get more data before we make a decision."
- But don’t take too much time making a decision. Although your first choice of an antibiotic is important and you want to start broadly to cover all possibilities, narrowing the antibiotic as soon as possible is also key.
"Going from caspofungin to fluconazole for Candida infection is a good example," Dr. Zimmer said. "It is cheaper, it creates less resistance, and if you can use one drug instead of two, do so."
- Remove all lines when you can, and remember that in the case of some infections “particularly S. aureus and Candida “treatment probably won’t be effective if you don’t remove the infected catheter.
- Respect Staphylococcus aureus. "It never ceases to amaze me with how virulent it can be," said Dr. Zimmer. "A huge amount of my consult business is related to S. aureus. It always frightens me."