Rethinking cellulitis

Flaws in diagnosis, prescribing are common.


Busy hospitalists and emergency department physicians frequently diagnose their patients with cellulitis. But they are often mistaken, research shows. A recent study found that misdiagnoses accounted for 30% of cellulitis cases at a large urban hospital, leading to dozens of unnecessary admissions and antibiotic prescriptions.

“It's important to remind ourselves that not all that is red, hot, and tender on the lower extremities is cellulitis,” said Arash Mostaghimi, MD, ACP Member, director of the dermatology inpatient service at Brigham and Women's Hospital in Boston and senior author of the study, published by JAMA Dermatology in November 2016.

Photo by Thinkstock
Photo by Thinkstock

Cellulitis is not a straightforward diagnosis, Dr. Mostaghimi noted. It is nearly always unculturable and has enough mimics that physicians should expand and revisit their list of differentials if patients do not respond to treatment. Given the lack of quantitative tests for this condition, “all physicians should approach cellulitis with humility and have a low threshold to ask for a second opinion,” he said.

Allergic and irritant contact dermatitis, asteatotic eczema, and lichen simplex chronicus all can mimic cellulitis. But the most common type of pseudocellulitis is bilateral stasis dermatitis, said Misha Rosenbach, MD, assistant professor of dermatology and director of the dermatology consult service at the Hospital of the University of Pennsylvania in Philadelphia.

Affected patients have poor venous return secondary to conditions such as decompensated congestive heart failure, Dr. Rosenbach said. Venous insufficiency leads to “swollen legs, which turn red and weepy,” triggering a misdiagnosis of bilateral cellulitis and admission for unnecessary treatment with intravenous antibiotics.

Ironically, stasis dermatitis usually improves in the hospital because patients are lying flat and complying with other medications and dietary restrictions, Dr. Rosenbach said. This outcome just reinforces the pattern of misdiagnosis. After discharge, patients' legs often swell and turn red again, and they return to the hospital, only to be diagnosed with recurrent cellulitis.

A few clinical pearls can help physicians avoid this diagnostic blunder, according to experts. “True cellulitis comes on quickly and improves quickly with treatment,” Dr. Rosenbach stressed. “As an infection, it's usually accompanied by fever and an elevated white blood cell count, and it is nearly always unilateral.”

Overtreatment and unnecessary admissions

Even true cases of bacterial cellulitis usually do not merit broad-spectrum antibiotics, according to research and experts. The case-fatality rate among affected inpatients is less than 1%, and no clear evidence links nonpurulent cellulitis in typical hosts to either methicillin-resistant Staphylococcus aureus (MRSA) or gram-negative organisms, said Craig Gunderson, MD, director of hospital medicine at VA Connecticut Healthcare System and an assistant professor of medicine at Yale University School of Medicine in New Haven.

In one randomized trial, emergency department patients with cellulitis responded similarly to cephalexin therapy alone or with trimethoprim-sulfamethoxazole to cover for MRSA infection. Investigators reported the findings in the June 2013 Clinical Infectious Diseases.

The most recent guidelines on skin and soft-tissue infections from the Infectious Diseases Society of America reflect these observations, calling for narrower-spectrum anti-streptococcal agents such as beta-lactam antibiotics for typical cases and reserving broad-spectrum combinations, such as vancomycin with piperacillin-tazobactam, for situations in which patients have complicating factors such as penetrating trauma, nasal colonization with MRSA, or injection drug use.

But many hospital physicians ignore this guidance, according to Dr. Gunderson. “I think it's a relief for doctors not to have to worry about antibiotic choice. I also think the promotion of broad-spectrum antibiotics for septic shock has promoted the general philosophy of using broad-spectrum antibiotics early and aggressively,” he said.

The downsides of that approach include antibiotic resistance, drug reactions, and Clostridium difficile infections. It may also waste hospital resources by tying up beds unnecessarily. In one study of U.S. emergency departments, 42% of cellulitis admissions were solely for intravenous antibiotic therapy, and many patients did not need to be admitted at all, according to results published in the January 2015 Western Journal of Emergency Medicine.

“Many alternatives exist for providing antibiotics in an outpatient setting if a provider feels parenteral treatment is necessary,” said lead author David Talan, MD, who is chair emeritus of the department of emergency medicine at Olive-View UCLA Medical Center. For example, physicians can consider newer single-dose antibiotics, such as dalbavancin and oritavancin, or can place peripheral venous catheters or peripherally inserted central catheters for use at home or at infusion centers, he said.

Dermatology consults

Inpatient dermatology consults are ideal for many cases of suspected cellulitis, especially those described as “atypical, recurrent, bilateral, or nonresponding”—all of which can point to pseudocellulitis, according to Dr. Mostaghimi.

Research supports the value of consults. One study of academic medical centers found that only 5% of inpatient dermatology consults were for cellulitis, but 74% of consults revealed pseudocellulitis. Unnecessary antibiotics and hospitalizations for these cases caused about 9,000 nosocomial infections each year in the United States and up to 5,000 Clostridium difficile infections and six cases of anaphylaxis, the authors estimated. Expanding inpatient dermatology consults could help correct this problem, they wrote in the July 2015 Journal of the American Academy of Dermatology.

Dermatology consults can also help hospitalists manage true cases of cellulitis—for example, by covering both gram-negative and gram-positive pathogens in patients with bullous cellulitis and liver disease or by rapidly identifying and treating Pseudomonas infections in patients with acute myeloid leukemia and cellulitis of the toe web space, Dr. Rosenbach said. Dermatologists also have the expertise to suspect atypical organisms, such as Candida, in immunocompromised patients with symptoms such as dull, red-brown cellulitis with deep induration, he noted.

Finally, consulting dermatologists can help identify and treat cutaneous risk factors for cellulitis, such as tinea pedis or untreated eczema, Dr. Rosenbach said. But it can be hard to get dermatologists into the hospital in the first place. Many cite lack of time, reimbursement problems, medicolegal concerns, and inconsistencies with electronic medical records as obstacles, he said. “Efforts to smooth these barriers could help encourage dermatologists to engage in the inpatient setting more easily,” he added. Hospitals also could consider expanding access to teledermatology, which has proven accurate for triaging dermatologic conditions in small studies of inpatients, he said.

But consults are only part of the answer, Dr. Mostaghimi noted. “Cellulitis is a common diagnosis [that is] seen mostly by general practitioners and emergency department physicians,” he said. “Specialty care such as dermatology is not available in all places and certainly does not scale to the extent we would need to systematically evaluate all cases. We need interdisciplinary collaboration to help identify cases and consider alternative diagnoses without impairing workflow and hospital throughput.”

Systemic changes

The involved disciplines could include hospital medicine, ED physicians, and infectious disease specialists. At the University of Utah Medical Center in Salt Lake City, these groups collaborated after surveillance data showed hospitalists were routinely prescribing empiric vancomycin for admissions with the diagnosis of cellulitis, said Peter Yarbrough, MD, FACP, a hospitalist and assistant professor of internal medicine at the University of Utah School of Medicine.

He and his colleagues developed distinct care pathways for purulent and nonpurulent cellulitis. The pathway for nonpurulent cellulitis prioritized a narrow-spectrum beta-lactam antibiotic (cefazolin) and recommended reserving vancomycin for treatment-refractory cases.

Broad-spectrum antibiotic prescriptions for cellulitis fell by about 25% when physicians used the care pathway, and pharmacy and total health care costs also decreased, as did average days in the hospital, according to his study in the December 2015 Journal of Hospital Medicine. Strong relationships among hospitalists, infectious disease specialists, and ED physicians helped the collaboration succeed, he added.

This work highlighted the power of collecting clinician-level data on prescribing practices, Dr. Yarbrough said. After identifying baseline problems, project leaders worked with prescribers to “nonjudgmentally” involve them in the intervention, he said. “Reliable and timely data [also] was essential to understanding if [it] was successful.”

But there was a caveat. “Unfortunately, physicians in that study still commonly opted to not use the pathway, showing that there needs to be strong institutional buy-in and leadership,” said Dr. Gunderson, who was not involved in the study. The field also needs more research, he added. “Physicians are always going to err on the side of caution, and there are almost no randomized controlled studies of cellulitis.”

Dr. Talan agreed. Despite overwhelming evidence to the contrary, physicians often fear that an otherwise stable patient with fever or a large area of cellulitis “might suddenly deteriorate, or might not be responsible for their own care and follow-up if they are discharged,” he said.

Another block on the road to cellulitis care is “a culture in which we do not question each other's diagnoses,” Dr. Mostaghimi said. Regardless of specialty, physicians are highly unlikely to challenge an existing diagnosis of cellulitis, he added. Shifting the status quo means understanding that diagnosing cellulitis can be hard even for experienced dermatologists, he emphasized. “We shouldn't have to wait for treatment failure to reconsider a diagnosis. If you doubt the diagnosis, speak up. And if you made the original diagnosis, don't take it personally if somebody considers an alternative.”