Infectious Disease Vexations — Recurrent Furunculosis and Cellulitis

Author: David R. Haburchak, M.D., FACP, Professor of Medicine, Program Director, Internal Medicine, Medical College of Georgia, Augusta Georgia.

Vexation: n. a vexing or being vexed // something which annoys one.

Editor’s Note—Recurrent bacterial skin infections produce significant morbidity among adults in both developed and developing countries. Most commonly manifested by the syndromes of recurrent furunculosis and recurrent cellulitis, these diseases attack patients with ostensibly normal immunity. As such they are "unexpected" by the patient and physician alike, particularly in their virulence and high recurrence rate. Furunculosis may also occur in families, leading to devastating psychological consequences.

Despite the relatively frequent and probably increasing prevalence of these "vexations," there are few established evidence-based guidelines for their therapy and prevention. This review provides a pathophysiological approach to diagnosis, therapy, and prevention of these infections for the practicing physician and his or her patients.

Patients with furunculosis or cellulitis should be rapidly assessed and triaged to outpatient, inpatient, or intensive care depending on the severity of the infection and systemic manifestations. Therapy must include appropriate antibiotics as well as meticulous supportive care and surgical intervention. Because of the likelihood of recurrence of disease, risk factors for infection should be assessed and corrected if possible, starting at the time of initial presentation. Elimination of patient organism reservoirs may be difficult but should be considered for high-risk patients and those prone to relapse. Long-term antibiotic prophylaxis has limited efficacy but might be considered in desperate cases.

It is possible that in the near future, new forms of prevention will become available. Most promising are new vaccines and drugs to prevent colonization of staphylococci and streptococci or to augment innate immunity against these organisms. Until then, clinicians will be challenged to expeditiously diagnose, efficiently treat, and efficaciously prevent recurrence of these infections.

Vexing Skin Infections

It is an understatement that recurrent cutaneous infections are vexing to both patient and physician. Frustrating, costly, and not infrequently dangerous, such illness constitutes a significant proportion of acute care. This report will present management strategies for 2 well-defined syndromes, recurrent furunculosis, usually caused by Staphylococcus aureus, and recurrent cellulitis, usually caused by various Streptococci. The approach in both instances is to define the risk factors for recurrent disease, and if possible ameliorate them. With early and effective intervention, a vexation may be lessened, even if not often cured.

Recurrent Furunculosis

This pesky illness manifests itself by recurrent folliculitis, furuncles, and other skin infections occurring in various locations on the body over months to years. The lesions sometimes erupt within families and may be seemingly spread from person to person.1 The prevalence and incidence are unknown but appear higher in warm, humid, and tropical environments. Even so, cases in US cities may account for up to 2% of emergency room visits.2 The painful and noxious lesions often necessitate incision and drainage, prolonged and repeated courses of antibiotics, and may leave disfiguring scars.

Etiology and Epidemiology
The cause of these infections, of course, is S aureus, and the most important risk factor (see Table 1, below) for first-episode and recurrent disease is carriage of a virulent strain3,4 This organism is persistently found in the anterior nose or skin in 20% of the population, transiently found in 60% of people, and never found in a fortunate 20%.5 Children become carriers early in life and tend to be persistently colonized until age 10, when most carriage becomes transitory.6 Under conditions of intense exposure, such as in hospitals and nursing homes, transient carriers will persist in their colonization.

Table 1. Risk Factors for Recurrent Furunculosis

Host Factors

Nasal carriage of Staphylococcus aureus

Chronic or recurrent skin disease

  • Atopic dermatitis
  • Recurrent infections or infusions
  • Occupational injuries

Comorbid disease

  • Diabetes mellitus, especially insulin requiring
  • Rheumatoid arthritis
  • Hemodialysis

Leukocyte defects (acquired and congenital)

Immunodeficiency diseases

  • Hyper IgE (Job) syndromes
  • HIV
  • Common variable immunodeficiency


  • Nifedipine
  • Corticosteroids


  • Nose picking
  • Scratching
  • Intravenous drug abuse

Organism Factors

  • Adherence factors
  • Virulence toxins
  • Small colony variants

Environmental factors

  • Tropical climates
  • Dusty conditions
  • Sharing towels, soap
  • Athletic teams

From the reservoir sites, organisms are spread by hand or by clothing to inoculate susceptible skin lesions such as abrasions and cuts, as shown through molecular strain analysis.4 Familial and community epidemics may be facilitated by close contact and sharing of soap and towels, as seen in an epidemic associated with a steam bath.7 Infection in abraded or injured skin may be facilitated by foreign material such as dirt or even blood.

Episodes of recurrent disease often follow closely upon initial infection, but the risk of recurrence in any given patient is multifactorial. In the absence of nasal carriage, the risk of recurrent furunculosis in any individual patient is low to non-existent. Over a period of 3-4 years, patients with recurrent furunculosis may clear their nasal colonization and subsequently have disappearance of their boils.8

Because of its importance as a reservoir for infection and subsequent transmission of furunculosis, the nose has received intense scrutiny. Propensity of nasal carriage is not related to either leukocyte or immunoglobulin deficiency, but it may be peculiar to characteristics of host glycoproteins or antibacterial characteristics of nasal fluid. Studies performed in human volunteers indicate the specific niche for S aureus is the moist squamous epithelium of the anterior septum and vestibule.9 This area is anterior to nasal hairs and ciliated epithelium sweep bacteria to the throat. Nasal secretions bathing this location appear defective in killing homologous strains carried by subjects despite high levels of neutrophil peptides 1 and 3 and Beta-defensin 2, components of the innate immune system. Interestingly, heterologous carrier strains were also impervious to killing by carrier nasal secretions, but such fluid could inhibit a laboratory strain of S aureus. The nasal fluid from noncarriers did inhibit the growth of strains from carriers.9

Locations other than the nose can be sites of chronic carriage.10 These include the vagina, particularly during menses, rectum, perineum, groin, and damaged skin.11 Some patients with persistent nasal colonization are colonized at multiple sites. While it is suggested that elimination of nasal carriage may allow disappearance of the organism from other areas of the body,5 it appears as likely that the nose may become re-colonized from peripheral sites.12

While human nasal and skin carriage is the major reservoir for family infection, there have been reports of both methicillin-sensitive and methicillin-resistant S aureus (MRSA) in noses of pet dogs.13 Additionally, small colony variant S aureus14 and high toxin or clumping factor strains15 seem to have advantage in persistence over other bacteria.

Host risk factors for progression from nasal colonization to furuncle seem to cluster around defects of skin integrity and leukocyte function. Patients with frequent injuries, injections, atopic dermatitis and other dermatitides are prone to infection. Once infected, such skin may itself be colonized and promote chronic nasal carriage. Patients with illness requiring frequent skin breech such as insulin-dependent diabetes or renal patients requiring dialysis are infected often.

Occasionally, congenital leukocyte dysfunction such as in chronic granulomatous disease, Chediak-Higashi, and other more ill-defined syndromes may present in adolescence or adulthood.16 Hyper IgE17 and common variable immunodeficiency18 syndromes may manifest themselves as recurrent skin infection in adults, probably through defects in chemotaxis or opsonization. Both diabetes and rheumatoid arthritis19 are associated with leukocyte defects, as is chronic nifedipine administration,20,21 low serum iron,22 and vitamin A deficiency.23,24

Infections usually recur in locations specific to individual patients. This is probably related to the proximate site of colonization and means of transfer and inoculation. Habitual nose picking or nail chewing could thus transfer bacteria to face, neck, and back by scratching or soiling shirt collars. Similarly, shaving or application of antiperspirants might induce axillary disease. Lesions on the buttocks, waist, or perineum probably arise from contiguous colonization and are abetted by sweating and infrequent change of clothing. Finally, contaminated fingers or fomites could be responsible for spread of infection among households.

Clinical Features
Most patients present themselves acutely to the emergency room or their physician with a painful boil. Usually the diagnosis is obvious, but sometimes systemic illness is severe. Vomiting and diarrhea may provoke hypotension and should prompt aggressive management for possible toxic shock syndrome.25 Typically the skin is inflamed and may be exquisitely tender. Lesions in thick-skinned areas of the body, such as the upper back and neck, may be undermined with sinus tracts. Called carbuncles, these usually need surgical intervention for control.

Evaluation and Management (see Table 2, below)

Table 2. Diagnostic Principles for an Episode of Furunculosis


How long has patient been ill?

Establish risk factors for staphylococcal disease

Establish risk factors for MRSA

  • Known previous MRSA infection or carriage?
  • Potential health care exposure?
  • Potential community exposure?

Is this a recurrence?

Does patient have gastrointestinal symptoms suggestive of toxic shock?

Physical examination

Document number, size, location, fluctuance and spread of primary lesions

Does patient have suppurative regional lymphadenopathy?

Does patient have evidence of systemic toxicity?

  • Evidence of toxic shock syndrome?
  • Evidence of endocarditis?

Laboratory tests

Blood cultures

Complete blood count, differential and peripheral smear

Screening chemistries

Important historical aspects include predisposition to S aureus in general and MRSA in particular. Preexisting conditions such as diabetes, renal disease, drug abuse, wounds, skin disorders, and atherosclerosis all increase risk of staphylococcal disease.26 Of increasing concern is community-acquired MRSA. There have been focal MRSA outbreaks in some closed communities (Native Americans in the Midwest,27 high school wrestling and football teams,28 and drug addicts.29) It remains rare, however, that community-acquired MRSA is not traced back to a health care facility such as a nursing home, dialysis center, or previous hospitalization.30

Physical Examination
Key physical findings include evidence of toxic shock syndrome, bacteremia, endocarditis, or metastatic spread of infection. Particular attention should be made for evidence of hypotension, cardiac murmur, conjunctival and digital petechiae, and evidence of diffuse or exfoliative rash. Examine regional and distant lymph nodes for adenitis and suppurative lymphadenopathy that might benefit from aspiration.

Laboratory Tests
Patients who are systemically ill should have blood cultures prior to antibiotics, as well as complete blood count and screening metabolic profile, particularly looking for evidence of hyperglycemia or renal insufficiency. A peripheral smear may demonstrate leukocyte abnormalities such as toxic granulations, Dohle bodies, leukopenia, or abnormal morphology of leukocyte granules.

Therapeutic Management (see Table 3, below)
Emergency therapy should be stratified as to the severity of systemic signs. Patients with toxic shock syndrome need intensive supportive management and adjunctive therapy, preferably in an intensive care unit.25 Less systemically ill patients should be hospitalized and treated with intravenous antibiotics, with correction of underlying predisposing conditions. All patients, unless allergic, should get intravenous nafcillin or oxacillin in high dose. If there is suspicion of endocarditis, gentamicin should be added. Depending upon risk of MRSA exposure as noted, vancomycin might be considered in initial empiric therapy. Systemic antibiotics should be adjusted based on results of blood and other cultures. Linezolid for MRSA has the advantage of both parenteral and oral dosing but is very expensive. Gentamicin should be discontinued after 3 days even if blood cultures were positive. If blood cultures are negative, most patients may be switched to oral therapy after fever and tachycardia improve.

Table 3. Therapeutic Principles for an Episode of Furunculosis


Systemically ill patients

  • Toxic shock syndrome—intensive care unit
  • Signs of sepsis—hospitalize
  • High risk for MRSA—hospitalize

Limited infection—outpatient therapy

Is Lesion Amenable to Drainage?

Aspirate for culture and sensitivity

Simple incision and drainage

Surgical consultation

Empiric Antibiotics

Outpatient therapy: Cloxicillin, dicloxicillin, cephalexin, or cephradine

Clindamycin if beta-lactam allergic

Inpatient therapy:

  • Intravenous oxacillin or nafcillin
  • If beta-lactam allergic: Vancomycin
  • If bacteremia suspected: Add gentamicin for 3 days
  • If MRSA suspected: Add vancomycin

Antibiotic switch and duration

Change to best single agent after receipt of culture and sensitivity

Bacteremia: 2 weeks parenteral drug, with transesophageal echocardiogram

Non-bacteremia: switch to oral dicloxacillin when symptoms abate

Duration: Total 2 weeks therapy unless endocarditis established

Lesions amenable to incision and drainage should be aspirated for culture and sensitivity, particularly because of the rise of MRSA. Adequate drainage often requires surgical intervention, which may remarkably improve pain and systemic signs. If the patient is not to be admitted, cloxacillin, dicloxacillin, cephalexin, or cephradine is usually given in oral doses of 30-40 mg/kg/d in 4 divided doses. Duration of therapy should be no more than 7-10 days.11 Comparison of 10-14 days vs 2 months of therapy showed no benefit of longer therapy in prevention of recurrence.8

Prevention (see Table 4, below)
How to prevent recurrent staphylococcal infections is less well defined. As noted above, almost all patients who have significant disease are colonized in the nose and possibly other locations. Besides correcting underlying illness, such as uncontrolled diabetes, the most common approach has been to attempt decolonization of the patient with 2% mupirocin nasal ointment applied to each nostril twice daily for 5 days. This drug is the latest in a line of nasally applied or secreted drugs that have shown some promise in lessening nasal carriage.8 The Cochrane Review cites only 1 article supporting use of mupirocin nasal ointment.31 In this study, patients applied 7 days of mupirocin and also washed with chlorhexidine soap and used chlorhexidine powder to other possible carriage sites. Prior to treatment, colonization rates were nose 67%, axillae 22%, groin 23%, and perianal 19%. At 91 days after treatment, only 43% of those receiving mupirocin had positive nasal cultures compared to 89% of those receiving chlorhexidine/neomycin cream. A subsequent trial did show that monthly treatment with 5 days of mupirocin nasal ointment among patients with an average of 4 episodes of furunculosis per year significantly lowered the rate of recurrent infection.32 In this small Israeli investigation, the mean patient age was 25 years. Compared to placebo, the number of infections in the treatment group decreased by 50%, with 8 of 17 treated patients completely free of relapse over 1 year. There was good correlation between negative nasal culture and freedom from relapse, and a linear relationship between the number of monthly pretreatment cultures that were positive and subsequent number of infections.

Table 4. Prevention of Recurrent Furunculosis

Ameliorate risk factors

  • Diabetic control
  • Treat/prevent skin disease and injury
  • Remove or replace at risky drugs: nifedipine, steroids, immunosuppresives
  • Correct leukocytic defects: GM-CSF for neutropenia?
  • Infection control
  • Behavioral modification

Nasal de-colonization (Especially recommended for recurrent disease, MRSA, or toxic-shock syndrome associated disease)

  • Intranasal mupirocin 5-day course with chlorhexidine
  • Repeated (monthly to quarterly) mupirocin courses
  • Family and pet therapy
  • Rifampin

Experimental and anecdotal therapies

  • Vitamin C
  • Iron sulfate
  • Bacterial interference with less pathogenic strains
  • Pentoxifylline
  • Staphylococcal vaccines

Mupirocin trials performed among health care workers have also been encouraging, with 71% nasal eradication rates at 3 months after a single 5-day course.33 Unfortunately, the all-site eradication rate in hospitalized Swiss patients with MRSA was only 25% despite chlorhexidine washing, and there was only 44% nasal eradication at day 26.12 In a similar study, after 5 days of intranasal mupirocin AIDS patients had 63% nasal eradication at 2 weeks, but only 29% at 10 weeks.34 Still another group successfully decolonized 87% of liver transplant recipients, but 37% relapsed and 23% of patients developed staphylococcal infection.35 Even so, a much more vigorous thrice-daily nose and skin mupirocin/chlorhexidine wash protocol combined with quarantine successfully eradicated MRSA from a Finnish health center and nursing home but would be impractical in most other settings.36

Mupirocin appears well tolerated, with headache (9%), rhinitis (8%), and nasal congestion (5%) the side effects most reported. Unfortunately, rapid development of resistance can be a problem, particularly when used among large numbers of people or on the skin.37 Additionally, the drug is not inexpensive—a single 5-day course costing $39.52 in 1997.38

Five-day courses of rifampin every 3 months decreased unspecified staphylococcal infections by 50% in dialysis patients,39 but at the risk of development of rifampin resistance and potential drug-drug interactions. Other published, but uncontrolled, prophylactic treatments have included pentoxifylline,40 vitamin C,41 and iron.22,42

Probiotic therapy was attempted by nasal instillation of non-pathogenic S aureus strain 502A in the 1960s and 1970s in an effort to produce "bacterial interference" of virulent strains. A 50% decrease in recurrent infections was observed over a 6-month period compared to controls.1 These studies were later abandoned after reports of metastatic abscess associated with this organism.43,44

Finally, staphylococcal vaccines are beginning to reach clinical trials aimed first at dialysis patients.45-47 To be effective, these vaccines will have to target multiple virulence factors but could be of significant benefit.

A first episode of staphylococcal skin disease does not warrant subsequent mupirocin preventive therapy unless the infection has been with MRSA or has caused toxic shock syndrome. Any underlying predisposing condition such as diabetes, vitamin deficiency, or dermatitis should be addressed first. Families should also be taught infection control practices. Patients, however, with MRSA or toxic shock associated organisms should be de-colonized48 with combined nasal mupirocin and chlorhexidine washing beginning at the time of acute therapy. They should then undergo periodic surveillance for recolonization.8 Patient recolonization should prompt investigation of possible carriage by family members and probably even pets. On the other hand, because they are constantly re-exposed, chronically ill patients living in nursing homes or frequenting health care facilities will not benefit from eradication attempts.

A healthy patient who suffers recurrent disease should initiate a trial of nasal mupirocin and chlorhexidine washing, particularly if the disease is on the upper body, since it is probably coming from his nose. A 5-day course of mupirocin with nasal swab confirmation of eradication 2 weeks later appears to be a reasonable investment. Later nasal surveillance for recolonization may be justified in some cases.

Recurrent Cellulitis

Cellulitis may be defined as a spreading infection of the dermis or epidermis. It is thereby distinguished from the more localized folliculitis and furunculosis. The most prevalent organisms associated with cellulitis are beta hemolytic streptococci. Impetigo is caused by either streptococci or staphylococci but may be distinguished by its appearance—either honey-crusts or small pustules.

Cellulitis is common. Along with abscess, its evaluation accounted for 158 visits per 10,000 patient risk-years in 1991 British medical practice.49 Starting in the 1970s, there was a shift in the disease from its classical appearance on the face to the lower extremities, now the location of more than 80% of cases.50 Remarkably, cellulitis recurs among 15-29% of patients within 3 years51,52 and 50% of patients sometime later in life.53 Recurrent disease is associated with specific host risk factors (see Table 5, below), among which are diabetes, alcoholism, trauma, dermatophyte infections (athlete’s foot), and lymphedema.53 Case control studies show highest risk exists with lymphedema, skin breaks such as tinea infection and wounds, venous insufficiency, leg edema, and being overweight.54 The current epidemic of obesity certainly seems to be facilitating cellulitis incidence.

Table 5. Risk Factors for Recurrent Cellulitis

Host Factors

Carriage of beta-hemolytic streptococci

  • Toe web carriage
  • Anal carriage
  • Pharyngeal carriage?

Chronic skin disease

  • Lymphedema
  • Tinea pedis
  • Venous insufficiency
  • Previous cellulitis

Comorbid diseases

  • Obesity
  • Diabetes mellitus
  • Alcoholism
  • Drug abuse
  • Postsurgical: saphenous venectomy, breast surgery

Environmental Factors

  • Homelessness
  • Insufficient hygiene
  • Exposure to children?

A special form of recurrent cellulitis is that due to "skin popping" of narcotics and clandestine drugs by addicts.55 Organisms may consist of various oral and skin bacteria that contaminate materials or devices used in this practice.

Two common surgical predispositions for recurrent streptococcal cellulitis are saphenous venectomy and breast cancer surgery. Infections tend to occur months to years after surgery, suggesting that these are not simple surgical site infections. Tinea pedis is a common predisposing condition for leg cellulitis. Risk factors for breast cellulitis include hematoma drainage, presence of lymphedema, the amount of breast tissue resected, previous number of biopsies, and number of breast seroma aspirations.53

Surprisingly little is known of the mechanisms by which cellulitis recurs. This stems from the difficulty in culturing bacteria from skin aspirates of cellulitis and the rarity of positive blood cultures.56 Presumably, lymphedema promotes growth of bacteria at the same time it inhibits phagocytosis, thereby allowing spread of infection. Nevertheless, using punch biopsy has been difficult to demonstrate either histological or immunological differences in infection-prone and resistant limbs.57 Poor hygiene, dermatophyte infection, and trauma presumably allow inoculation of skin colonizing streptococci or staphylococci into very fertile ground. In support of this concept, a study of postvenectomy patients showed strong correlation between toe web colonization with beta hemolytic streptococci and recurrent cellulitis.58 The perineum or anal canal may be another reservoir for Group B (59) and Group G beta hemolytic streptococci infecting the leg.60 Similar mechanisms may account for recurrent cellulitis in the breast arising from local trauma or migrating from hand infections.

Once infection has occurred, lymphatics are damaged in most patients as evidenced by lymphoscintigraphy.61 Many patients are left with clinically evident residual edema.62 This makes a second episode of cellulitis more likely than the first.

Group A and G streptococci, the most common organisms associated with cellulitis, resist complement-associated opsonization via their M proteins.63,64 Once infected, animals would be expected to develop antibodies against their specific colonizing M-type strain to protect against recurrent infection. Such antibodies, however, do not appear to protect,63 possibly due to subtle changes in bacterial M-protein structure.65 If so, this immune-evading mechanism is similar to that used by Borrelia recurrentis, the agent of relapsing fever.

Finally, immunosuppressive and anti-inflammatory drugs, including nonsteroidal anti-inflammatory agents (NSAIDs) delay patient recognition of infection. Patients taking such medicines may have greater risk of severe illness.66,67

Clinical Features
The clinical features of recurrent cellulitis differ little from the first episode except that the limb or breast may show scarring or other residue. The first symptom is usually pain, followed soon by redness and warmth in the affected skin. The inoculation site may be occult or obvious and is usually somewhat distal to the center of the spreading erythema or pain. Most patients seek medical attention quicker for recurrent disease than for a first episode. They therefore tend to have less defined erysipelas, lymphangitis, lymphadenopathy and systemic manifestations. Often the patient has initiated self-therapy with antibiotics. The relatively benign course of such patients and difficulty of recovering putative pathogens from skin aspiration or blood culture have prompted some to suggest allergic sensitization as the cause of recurrent disease. Differential diagnosis is usually limited to deep vein thrombosis, insect venom, trauma, and overuse.

Presently, no formal practice guidelines or consensus statements exist for management of first or subsequent episodes of cellulitis, nor is there good evidence on prognosis.68 A pharma-sponsored expert panel, however, has recently published its deliberations.69

Pain and failure of self-therapy frequently brings the patient to the emergency room or clinic. At this time, the clinician should make a rapid assessment and clinically stage the patient according to the Eron Classification System70 (see Table 6, below). Key components of the history (see Table 7, below) are duration and intensity of both local and systemic symptoms, history of injury or exposure to bites or contaminated water, and previous attempted therapy. It is crucial to ascertain history of skin or bone infection and comorbid disease such as diabetes, cardiac, pulmonary, hepatic, hematologic, or renal impairment. As noted above, lymphedema and prior surgery are key risk factors.

Table 6. Eron Classification System for Patients with Skin and Soft-Tissue Infections


Patient Criteria Triage Decision
I Healthy, non-systemically ill Outpatient
II Toxic or febrile, stable comordid diseases Observation
III Toxic or febrile, unstable comordid diseases
Limb-threatening infection
Admit to Hospital
IV Deep seated infection, critical location, sepsis ICU, Surgery consult

Table 7. Diagnostic Principles for an Episode of Cellulitis


  • How long has patient been ill?
  • Exposure to injury, bite, water?
  • Prior skin or bone infection?
  • Comorbid disease?
  • Prior surgery?

Physical Examination

  • Signs of sepsis or systemic illness
  • Location of cellulitis: critical location?
  • Disproportionate pain
  • Measure and demarcate extent of cellulitis
  • Presence of discoloration, bullae, crepitus, flucuance, necrosis

Laboratory Tests

  • Complete blood count
  • Screening chemistries
  • Blood cultures if signs of sepsis
  • Imaging in suspected stage IV infection
  • Gram stain if unusual organism or anthrax suspected
  • Quantitative skin biopsy and culture for suspected stage III or IV infection

Physical examination and triage. Prime components of the physical examination include signs of sepsis, extent and severity of local inflammation, and stability of comorbid diseases. Early streptococcal toxic shock syndrome can result from proliferation of bacteria in bruised as well as punctured dermis and, unless the skin is closely examined, illness may be mistaken as gastroenteritis.67 Particularly worrisome are disproportionate pain, bullae, hemorrhage or violet discoloration, patchy anesthesia of the skin, crepitus, and fluctuance. These signs represent deep infections of the dermis, fascia, or even fat and muscle, with possible necrosis and gangrene. If so, the infection is Stage IV. Infections of the hands, genitalia, perineum, thighs, face, and head, as well as grossly contaminated wounds such as human bites are particularly dangerous. All such patients should rapidly be started on intravenous antibiotic therapy and admitted to the hospital with prompt surgical consultation.

Simple erysipelas or streaky lymphangitis of the lower leg is less worrisome, and might be considered Stage I infection in a healthy patient or patient with stable comorbidity. Most Stage I patients might be treated using closely observed outpatient oral antibiotic therapy and rest.

Unfortunately, up to 42% of patients present with fever, tachycardia, tachypnea, leukocytosis and other signs of sepsis.71 These Stage II and III patients should be admitted for observation and parenteral antibiotics. Once the systemic signs are improved, the patient might be discharged on oral or parenteral antibiotics to complete the course of therapy.

Laboratory Tests
Routine laboratory tests have not been shown to be useful for prognosis or management.68 Complete blood count and screening chemistry profile might be helpful to identify diabetes, renal failure, or thrombocytopenia and urinalysis to exclude glomerulonephritis. Stage IV patients warrant imaging studies to look for gas, osteomyelitis, venous thrombosis, or foreign bodies. Blood cultures should be performed in patients with sepsis, while wound biopsy, Gram stain and culture are indicated for severe stage III and stage IV disease. During periods of high bioterror alert status, bullous lesions should be examined for the possibility of anthrax. Anaerobic cultures should be performed on foul-smelling, necrotic, or gas-associated wounds.

Therapeutic Management (see Tables 8 and 9, below)

Table 8. Therapeutic Principles for an Episode of Cellulitis

Classify severity of infection

Triage patient

Empiric Antibiotics

  • Stage I: cloxacillin, dicloxacillin, cephalexin or cephardine Clindamycin if allergic to beta lactam drugs
  • Stage II-IV: Intravenous clindamycin until stable
  • Bites: (human and animal): Ampicillin/Sulbactam plus clindamycin
  • "Water bugs:" Intravenous ciprofloxacin
  • Diabetes: Penicillin G or clindamycin plus gentamicin
  • Anthrax: Intravenous ciprofloxacin plus rifampin

Surgical debridement for suspected fasciitis, myonecrosis

Switch to oral therapy when stable, able to take oral drug

Finish 14-day total course with oral therapy

Table 9. Initial Antibiotic Doses for an Adult with an Episode of Furunculosis or Cellulitis

(See text for indications and duration)

Cloxicillin 500 mg po q 6 hr
Dicloxicillin 500 mg po q 6 hr
Cephalexin 500 mg po q 6 hr
Cephardine 500 mg po q 6 hr
Nafcillin 2 gm iv q 4 hr
Oxacillin 2 gm iv q 4 hr
Clindamycin 750 mg iv q 8 hr or 450 mg po q 6 hr
Ampicllin/Sulbactam 3.1 gm iv q 8 hr
Ciprofloxacin 400 mg iv q 12 or 500 mg po q 12 hr
Vancomycin 1 gm iv q 12 hr
Rifampin 600 mg po q day
Gentamicin 2 mg/kg load, followed by 1.5 mg/kg iv q 8 hr


600 mg iv or po q 12 hr

Streptococci and staphylococci susceptible to beta lactam antibiotics cause most episodes of cellulitis without unusual historical features. An appropriate choice for early Class I infection might be cephalexin or dicloxacillin. These oral beta lactam drugs should be given in generous doses even for Stage I disease, at least 500 mg every 6 hours for adults. It is paramount that the affected limb or tissue be elevated and put at rest.

Otherwise Stage I infections associated with bites, feces, or water are likely due to polymicrobial oral, enteric, or environmental flora and need broader coverage, preferably in the inpatient setting. A parenteral third-generation antipseudomonal penicillin or cephalosporin plus a fluoroquinolone would be appropriate until the organism is identified. These drugs are particularly important for such "water bugs" as Vibrio vulnificus and Aeromonas hydrophila. Most episodes of recurrent cellulitis fortunately do not require such broad spectrum antibiotics.

Besides beta lactams, adjunctive intravenous clindamycin appears warranted for severe, well-established cellulitis with or without toxic shock syndrome.67 Clindamycin, as a protein-synthesis inhibiting antibiotic, has been demonstrated to enhance phagocytosis through alteration of the bacterial cell wall structure, inhibition bacterial toxin formation, and modulation of host cytokine release.72 Furthermore, clindamycin is not subject to high bacterial inoculum effect, the so-called Eagle phenomenon.72 This refers to the lack of bacterial proliferation and down regulation of penicillin-binding proteins when organisms have reached the maximum growth capacity of the infected tissue. Once the infection is under control, the patient may be switched to combined oral therapy for 3-5 days. The remainder of the 10-14 day antibiotic course is completed with the beta lactam. By limiting the duration of oral clindamycin, one hopes to reduce the risk of Clostridium difficile diarrhea.

Diabetic patients occasionally have positive blood cultures, often for Group B streptococcus. In such patients, synergistic combination penicillin and gentamicin is preferred. Again, once the infection is controlled, usually within 2-3 days, switch to an oral cephalosporin to complete the 14-day course of therapy.

All patients are instructed to elevate the affected limb or part, use compression stockings or bandage when up, and expect some pain for a few days to weeks. Nonsteroidal pain medications are usually sufficient. The patient should normally be seen in the follow-up clinic within a few weeks to ensure healing, to exclude abscess or venous thrombosis, and to initiate preventive therapy.

Prevention (see Table 10, below)
Prevention involves reducing risk factors such as edema and tinea pedis and educating the patient on pathogenesis. Patients might benefit from instruction to quickly self-treat minor skin trauma in edematous areas with antiseptics or topical bacitracin. Prophylactic long-term or cyclical antibiotic use has been attempted, but to date no clinical trials support such intervention.73 Well-fitting support hose extending above the level of lymphedema are helpful but can be very expensive. Moderate exercise with latter leg elevation may be useful in promoting lymphatic, venous and muscular function. Sequential pneumatic compression devices have facilitated impressive improvement of recurrence rate when used with cyclical 10-day courses of penicillin G in a recent report from Germany.74

Table 10. Prevention of Recurrent Cellulitis

Ameliorate Risk Factors

Control lymphedema by physical measures

  • Support hose
  • Pressure bandages
  • Elevation of limb

Control tinea pedis

Prevent or avoid skin trauma

Weight loss

Control diabetes

Plausible but untested

  • Chlorhexidine washing
  • Treatment of minor skin trauma with mupirocin or bacitracin ointment

Antibiotic Prophylaxis-reserve for after second episode

  • Penicillin VK 250 mg twice daily or
  • Clindamycin 150 mg once daily

Support for vigorous therapy of tinea pedis comes from a trial among patients with filariasis.75 Therapy of tinea pedis is best initiated with topical therapy such as tolnaftate or ketoconazole cream. Suppression of dermatophyte recurrence can usually be accomplished with foot powder and periodic foot inspection by the patient without resorting to systemic drugs such as itraconazole and terbinafine. These latter drugs may be necessary to eradicate toenail disease, but often pose more toxicity and cost than desired.76 Since the web space is the most implicated streptococcal entry point in recurrent cellulitis, this tissue rather than the nail should be the primary focus of attention.

Finally, low-dose antibiotic suppression such as twice-daily penicillin VK 250 mg or clindamycin 150 mg once daily has been useful for some patients,77 but monthly injectable penicillin was ineffective for high-risk patients in one study.78 Chronic antibiotic therapy should be reserved for patients after their second episode of cellulitis.

Recurrent cellulitis is common, with a lifetime risk of 50% after a first episode. Key management concerns are early recognition of surgical or life threatening disease and prompt administration of antiproliferative and antitoxin antibiotics, usually in the form of parenteral beta lactam and clindamycin therapy for stage III or IV disease. History is critical for identification of unusual organisms that may require broader antibiotic coverage. After therapy of the acute infection, patients should undergo vigorous therapy to limit lymphedema, and prevent tinea and other skin infections or trauma. Chronic suppressive antibiotic therapy may offer some benefit, but only after other prophylactic measures are instituted.


1. Steele RW. Recurrent staphylococcal infection in families. Arch Dermatol. 1980;116:189-190.

2. Meislin HW, et al. Cutaneous abscesses: Anaerobic and aerobic bacteriology and out-patient management. Ann Intern Med. 1977; 87:145-149.

3. Weems JJ, Beck LB. Nasal Carriage of Staphylococcus aureus as a risk factor for skin and soft tissue infections. Curr Infect Dis Rep. 2002;4:420-425.

4. Toshkova K, et al. The significance of nasal carriage of Staphylococcus aureus as risk factor for human skin infection. FEMS Microbiology Letters. 2001;202:17-24.

5. Kluytmans J, et al. Nasal carriage of Staphylococcus aureus: Epidemiology, underlying mechanisms, and associated risks. Clin Microbiol Rev. 1997;10:505-520.

6. Armstrong-Esther CA. Carriage patterns of Staphylococcus aureus in a healthy non-hospital population of adults and children. Ann Hum Biol. 1976;3:221-227.

7. Landen MG, et al. Outbreak of boils in an Alaskan village: A case-control study. West J Med. 2000;172:235-239.

8. Hedstrom SA. Treatment and prevention of recurrent staphylococcal furunculosis: clinical and bacterological follow-up. Scand J Infect Dis. 1985;17:55-58.

9. Cole AM, et al. Determinants of Staphylococcus aureus nasal carriage. Clin Diagn Lab Immunol. 2001;8:1064-1069.

10. Lowy FD. Staphylococcus aureus infections. N Engl J Med. 1998;339:520-532.

11. Waldvogel FA. Staphylococcus aureus (Including Staphylococcal Toxic Shock). In: Mandell GL, et al, eds. Mandell, Douglas, and Bennett’s Principles and Practice of Infectious Diseases. 5th ed. Philadelphia, Penn: Churchill Livingstone; 2001:2069-2083.

12. Harbarth S, et al. Randomized, placebo-controlled, double-blind trial to evaluate the efficacy of mupirocin for eradicating carriage of methicillin-resistant Staphylococcus aureus. Antimicrob Agents Chemother. 1999;43:1412-1416.

13. Manian FA. Asymptomatic nasal carriage of mupirocin-resistant, meticillin resistant Staphylococcus aureus (MRSA) in a pet dog associated with MRSA infection in household contacts. Clin Infect Dis. 2003;36:326-328.

14. Von Eiff C, et al. Small colony variants of Staphylococci: A link to persistent infections. Berl Munch Tierarztl Wochenschr. 2000; 113:321-325.

15. Slobodnikova L, et al. Staphylococcus aureus in chronic and recurrent infections. Folia Microbiol (Praha). 1995;40:655-668.

16. Forte WC, et al. Repeated furunculosis in adult male with abnormal neutrophil activity. Allergol Immunopathol (Madr). 2000; 28:328-331.

17. Shemer A, et al. The hyper-IgE syndrome. Two cases and review of the literature. Int J Dermatol. 2001;40:622-628.

18. Sidwell RU, et al. A case of common variable immunodeficiency presenting with furunculosis. Br J Dermatol. 2002;147:364-367.

19. Nada R, et al. Neutrophil functions in rheumatoid arthritis. Indian J Pathol Microbiol. 1999;42:283-289.

20. Kaul N, et al. Effect of neifedipine administration the functional capacity of neutrophils. Mol Cell Biochem. 1993;120:81-85.

21. Levy R, et al. Inhibited neutrophil functions in patients treated with nifedipine but not with verapamil or diltiazem. Eur J Clinic Invest. 1996;26:376-381.

22. Demircay Z, et al. Phagocytosis and oxidative burst by neutrophils in patients with recurrent furunculosis. Brit J Dermatol. 1998;138:1036-1038.

23. Twining SS, et al. Vitamin A deficiency alters rat neutrophil function. J Nutr. 1997;127:558-565.

24. Aukrust P, et al. Decreased vitamin A levels in common variable immunodeficiency; vitamin A supplementation in vivo enhances immunoglobulin production and down regulates inflammatory responses. Eur J Clin Invest. 2000;30:252-259.

25. Issa NC. Thompson RL. Staphylococcal toxic shock syndrome. Suspicion and prevention are keys to control. Postgrad Med. 2001;110:55-56, 59-62.

26. Trilla A, Miro JM. Identifying high-risk patients for Staphylococcus aureus infections: Skin and soft tissue infections. J Chemother. 1995;7 Suppl 3:37-43.

27. Groom AV, et al. Community-acquired methicillin-resistant Staphylococcus aureus in a rural American Indian community. JAMA. 2001;286:1201-1205.

28. Lindenmayer JM, et al. Methicillin-resistant Staphylococcus aureus in a high school wrestling team and the surrounding community. Arch Intern Med. 1998;158:895-899.

29. Charlebois ED, et al. Population-based community prevalence of methicillin-resistant Staphylococcus aureus in the urban poor of San Francisco. Clin Infect Dis. 2002;34:425-433.

30. Salgado CD, et al. Community-acquired methicillin-resistant Staphylococcus aureus: A meta-analysis of prevalence and risk factors. Clin Infect Dis. 2003;36:131-139.

31. Leigh DA, Joy G. Treatment of familial staphylococcal infection-comparison of mupirocin nasal ointment and chlorhexidine/neomycin (Naseptin) cream in eradication of nasal carriage. J Antimicrobial Chemother. 1993;31:909-917.

32. Raz R, et al. A 1-year trial of nasal mupirocin in the prevention of recurrent staphylococcal nasal colonization and skin infection. Arch Intern Med. 1996;156:1109-1112.

33. Doebbeling BN, et al. Long-term efficacy of intranasal mupirocin ointment. A prospective cohort study of Staphylococcus aureus carriage. Arch Intern Med. 1994;154:1505-1508.

34. Martin JN, et al. A randomized clinical trial of mupirocin in the eradication of Staphylococcus aureus nasal carriage in human immunodeficiency virus disease. J Infect Dis. 1999;180:896-899.

35. Paterson DL, et al. Lack of efficacy of mupirocin in the prevention of infections with Staphylococcus aureus in liver transplant recipients and candidates. Transplantation. 2003;75:194-198.

36. Kotilainen P, et al. Eradication of methicillin-resistant Staphylococcus aureus from a health center ward and associated nursing home. Arch Intern Med. 2001;161:859-863.

37. Farr BM. Mupirocin to prevent S aureus infections. N Engl J Med. 2002;346:1905-1906.

38. SmithKline Beecham Pharmaceuticals advertising brochure. January 1997.

39. Yu VL, et al. Staphylococcus aureus nasal carriage and infection in patients on hemodialysis: Efficacy of antibiotic prophylaxis. N Engl J Med. 1986;315:91-96.

40. Wahba-YahavAV. Intractable chronic furunculosis: Prevention of recurrences with pentoxifylline. Acta Derm Venereol. 1992; 72:464-462.

41. Levy R, et al. Vitamin C for the treatment of recurrent furunculosis in patients with impaired neutrophil functions. J Infect Dis. 1996;173:1502-1505.

42. Weijmer MC, et al. Preliminary report: Furunculosis and hypoferremia. Lancet. 1990;336:464-466.

43. Drutz DJ, et al. Bacterial interference in the therapy of recurrent staphylococcal infections: Multiple abscesses due to the implantation of the 502A strain of Staphylococcus. N Engl J Med. 1966;275:1161-1165.

44. Houck PW, et al. Fatal septicemia due to S aureus 502A. Report of a case and review of the infectious complications of bacterial interference programs. Am J Dis Child. 1972;123:45-48.

45. Shinefield H, et al. Use of a Staphylococcus aureus conjugate vaccine in patients receiving hemodialysis. N Engl J Med. 2002; 346:491-496.

46. Michie CA. Staphylococcal vaccines. Trends Immunol. 2002; 23:461-463.

47. Lee JC. Development of antistaphylococcal vaccines. Curr Infect Dis Rep. 2001;3:517-524.

48. Andrews M-M, et al. Recurrent nonmenstrual toxic shock syndrome: clinical manifestations, diagnosis, and treatment. Clin Infect Dis. 2001;32:1470-1479.

49. Office of Population Censuses and Survey. Morbidity statistics from general practice. Fourth National Study. Her Majesty’s Stationery Office. 1992 (Series MB5): 272.

50. Ronnen M, et al. Erysipelas. Changing faces. Int J Dermatol. 1985;24:169-172.

51. Jorup-Ronstrom C, Britton S. Recurrent erysipelas: Predisposing factors and costs of prophylaxis. Infection. 1987;15:105-106.

52. Crickx B. Erysipelas: Evolution under treatment, complications. Annales de Dermatologie et de Venereologie. 2001;128:358-362.

53. Baddour LM. Cellulitis syndromes: An update. International Journal of Antimicrobial Agents. 2000;14:113-116.

54. Baddour LM. Recent considerations in recurrent cellulitis. Curr Infect Dis Reports. 2001;3:461-465.

55. Binswanger IA, et al. High prevalence of abscesses and cellulitis among community-recruited injection drug users in San Francisco. Clin Infect Dis. 2000;30:579-581.

56. Perl B, et al. Cost-effectiveness of blood cultures for adult patients with cellulitis. Clin Infect Dis. 1999;29:1483-1488.

57. Baddour LM, et al. Biopsy specimen findings in patients with previous lower extremity cellulitis after saphenous venectomy for coronary artery bypass graft surgery. J Am Acad Dermatol. 1997;37:246-249.

58. Semel JD, Goldin H. Association of athlete’s foot with bacterial cultures of ipsilateral space samples. Clin Infect Dis. 1996; 23: 1162-1164.

59. Binnick AN, et al. Recurrent erysipelas caused by grup B streptococcus organisms. Arch Dermatol. 1980;116:798-799.

60. Eriksson BKG. Anal colonization of group G beta-hemolytic streptococci in relapsing erysipelas of the lower extremity. Clin Infect Dis. 1999;29:1319-1320.

61. de Godoy JM, et al. Lymphoscintigraphic evaluation in patients after erysipelas. Lymphology. 2000;33:177-180.

62. Cox NH, et al. Management and morbidity of cellulitis of the leg. J Roy Soc Med. 1998;91:634-637.

63. Bisno AL, Gaviria JM. Murine model of recurrent group G streptococcal cellulitis: No evidence of protective immunity. Infect Immun. 1997;65:4926-4930.

64. Bisno AL, et al. Molecular basis of group A streptococcal virulence. Lancet Infect Dis. 2003;3:191-200.

65. Villasenor SA, et al. Variable susceptibility to opsonophagocytosis of group A streptococcus M-1 strains by human immune sera. J Infect Dis. 1999;180:1921-1928.

66. Stevens DL. Could nonsteroidal anti-inflammatory drugs (NSAIDS) enhance the progression of bacterial infections to toxic shock syndrome? Clin Infect Dis. 1995;21:977-980.

67. Stevens DL. Streptococcal toxic shock syndrome. Clin Microbiol Infect. 2002;8:133-136.

68. Morris A. Cellulitis and erysipelas. Clinical Evidence. 2001; 6:1280-1283.

69. Eron LJ, et al. Key Decision Points in Evaluating and Treating Cellulitis: Expert panel recommendations on managing cellulitis from hospital to home. Supplement to Clinical Geriatrics. 2002.

70. Eron LJ. Infections of skin and soft tissue infection. Outcomes of a classification scheme. Clin Infect Dis. 2003;31:287-

71. Aly AA, et al. Case survey of management of cellulitis in a tertiary teaching hospital. Med J Australia. 1996;165:553-556.

72. Russell NE, Pachorek RE. Clindamycin in the treatment of streptococcal and staphylococcal toxic shock syndromes. Ann Pharmacother. 2000;34:936-939.

73. Morris A. Cellulitis and erysipelas. Clinical Evidence. 2002;8: 323-329.

74. Allard P, et al. Cyclical intravenous antibiosis as an effective therapy concept in chronic recurrent erysipelas. Hautarzt. 1999; 50:34-38.

75. Santos I, et al. Predisposing and triggering factors in erysipelas. Med Cutan Ibero Lat Am. 1984;12:399-402.

76. Del Rosso JQ. Current management of onychomycosis and dermatomycoses. Curr Infect Dis Rep. 2000;2:438-445.

77. Sjoblom AC, et al. Antibiotic prophylaxis in recurrent erysipelas. Infection. 1993;21:390-393.

78. Wang JH, et al. Role of benzathine penicillin G in prophylaxis of the lower legs. Clin Infect Dis. 1997;25:685-689.