NEW ORLEANS – Key factors in distinguishing true cellulitis from its clinical mimics include the presence of predisposing coexisting medical conditions, unilateral symptoms, and warmth at the anatomic site.
These were among the main findings of a retrospective study of all patients admitted to Massachusetts General Hospital, Boston, for suspected cellulitis during a 6-month period in 2009, Elizabeth E. Bailey said at the annual meeting of the American Academy of Dermatology.
She reported on 390 adults admitted with suspected cellulitis after excluding everyone with osteomyelitis, abscess, or bites. Ultimately, 323 of the 390 patients were deemed to have true cellulitis by the physician having primary responsibility for their care.
Among the 17% of patients determined to have a condition other than cellulitis, the most common final diagnosis was stasis dermatitis/lymphedema, which accounted for 17 cases, said Ms. Bailey, a fourth-year medical student at Columbia University, New. York. The other common mimics of cellulitis included 6 cases of gout or pseudogout, 6 cases of hematoma, 4 cases of septic arthritis or bursitis, and 4 venous occlusions.
In reviewing the charts of all 390 patients, one of the significant differences that emerged between patients with true cellulitis, as compared to pseudocellulitis, was that the true cellulitis group had a higher prevalence of coexisting medical conditions considered by the investigators to be potential risk factors for cellulitis. One or more of these conditions, which included diabetes mellitus, malignancy, intravenous drug use, and history of organ transplantation, was present in 55% of patients with cellulitis and 39% of those with pseudocellulitis.
On physical examination, 53% of cellulitis patients presented with warmth at the anatomic site, compared with 34% of those with pseudocellulitis. Bilateral disease was present in 13.4% of patients with a condition mimicking cellulitis and 5.9% of those with true cellulitis.
There were no significant between-group differences in body temperature; heart rate; reports of pain, itching, or burning; or any other physical exam findings.
In a multivariate analysis, the strong, significant predictors of cellulitis included the presence of a predisposing coexisting medical condition, which was associated with a 90% increased likelihood of cellulitis rather than a mimic. Bilateral symptoms rendered a patient 60% less likely to have true cellulitis. Warmth at the anatomic site was associated with a 2.2-fold increased likelihood of true cellulitis.
Of note, although warmth at the anatomic site was a significant predictor of true cellulitis, the other physical findings classically attributed to cellulitis – erythema, tenderness, and edema – were not.
"The classic tetrad of calor, dolor, rubor, and tumor is truly a tetrad for inflammation, as originally described in the 1st century AD," Ms. Bailey observed.
One of the striking findings in her study was how seldom the dermatology service was consulted in cases of suspected cellulitis. Only five patients with pseudocellulitis saw a dermatologist while in the hospital, and in four of these cases the dermatologic consultation led to a change in diagnosis from cellulitis to one of its mimics. A dermatologic consult was obtained in 18 of 323 cases of what was ultimately deemed true cellulitis.
Laboratory findings did not prove useful in making the distinction between true and pseudocellulitis in this study. However, the final word on this score isn’t in, because several laboratory values showing some potential weren’t measured in a sufficient number of patients to allow for definitive conclusions; among these were C-reactive protein, erythrocyte sedimentation rate, lactate dehydrogenase, and creatine kinase.
This study limitation will be rectified in a planned prospective study to be carried out at the hospital. In that study, all patients who present to the emergency department with suspected cellulitis will undergo a standardized assessment, including a specified panel of laboratory tests, followed by evaluation by both an infectious disease specialist and a dermatologist, explained Ms. Bailey.
The impetus for this retrospective study, as well as for the planned prospective study, lies in the fact that even though cellulitis accounts for about 10% of all infectious disease–related hospitalizations in the United States, there is no established diagnostic gold standard. It is likely that many other dermatologic or systemic conditions are misdiagnosed as cellulitis, Ms. Bailey said. The goal of the project she is conducting with other investigators at Massachusetts General Hospital is to develop criteria that will increase the specificity of the diagnosis.
Ms. Bailey declared having no relevant financial interests.