Article Text

Download PDFPDF

Dilemmas when managing cellulitis
  • Relevant BNF section: 5.1


Cellulitis is commonly seen in primary care. It is also a common reason for hospital admission accounting, for example, for around 3% of emergency medical consultations at one UK district general hospital.1 Here, we review the management of cellulitis, and try to resolve some common dilemmas: whether a patient with cellulitis should be treated in hospital; which antibacterial to use; whether it should be given orally or parenterally; and whether patients with recurrent cellulitis should take prophylactic antibacterials.

View Full Text

Statistics from

  • Relevant BNF section: 5.1

What is cellulitis?

Cellulitis, an acute non-contagious infection of the skin, is characterised by erythema, oedema and warmth accompanied by acute pain and tenderness. The condition is nearly always caused by an acute spreading bacterial infection of the dermis and subcutaneous tissues and may follow a skin abrasion or other similar trauma. At its mildest, the infection is relatively minor with local tenderness and inflammation and affects only a small area or a margin of a wound. If severe, it can affect a large area with spreading infection accompanied by chills, rigors, listlessness and lymphangitis or tender lymphadenopathy. In adults, cellulitis affecting the leg is the most common type necessitating hospitalisation, for example, accounting for 75% of admissions for cellulitis in a survey in Australia.2

Differential diagnosis

Cellulitis must be distinguished from deep vein thrombosis and from venous eczema.3 In cellulitis, the skin is usually smooth and shiny, in eczema it is crusting or scaling. Also, cellulitis is usually unilateral and is more likely to be associated with systemic symptoms.

Classically, cellulitis has been said to involve the subcutaneous connective tissue, and have an indistinct advancing border, compared to erysipelas, which is classically described as being superficial with a sharply demarcated border. However, in reality, cellulitis can also extend superficially and erysipelas deeply, so it may be difficult to distinguish between them. In practice, it is simplest to classify erysipelas as a form of cellulitis, rather than a distinct entity.4

Causes and risk factors

The commonest cause of cellulitis in adults is infection with streptococcal bacteria (especially Streptococcus pyogenes).5 The infection might also involve Staphylococcus aureus.6 Occasionally, other bacteria are the cause, especially if the cellulitis is related to a pre-existing skin problem (e.g. Pseudomonas in cellulitis arising from otitis externa); is acquired in hospital; occurs in a patient with diabetes or immunosuppression; or occurs in a child (e.g. Haemophilus influenzae was a recognised cause before routine HiB vaccination). In intravenous-drug misusers, a mixture of organisms may be involved, including Gram-negative and anaerobic bacteria.

Predisposing conditions particularly likely to be found in patients with leg cellulitis admitted to hospital are a site of entry for infection (e.g. leg ulcer, toe web intertrigo); lymphoedema; venous insufficiency; leg oedema; and being overweight.7 Other possible risk factors are diabetes mellitus, intravenous-drug misuse and previous surgery that has damaged blood or lymph vessels (e.g. saphenous venectomy, breast surgery).2,8,9

Recurrent cellulitis

In patients who have previously been admitted to hospital with cellulitis, around 30% develop a recurrence within a mean of 3 years, with venous insufficiency being the commonest predisposing factor.10


Cellulitis can lead to serious early complications, including septicaemia. Death reportedly occurs in up to 5% of patients hospitalised due to cellulitis and seems more likely in patients with multiple co-existing diseases, congestive heart failure or severe obesity.6 Necrotising fasciitis, usually due to streptococci but often associated with secondary anaerobic infection, is an uncommon, but life-threatening complication. This requires urgent surgical treatment.

Following cellulitis of the leg, around 7% of patients develop chronic oedema and a few patients develop persistent leg ulceration.11

Do laboratory tests help?

If a patient with cellulitis has a likely portal of entry or secondary blistering, bacteriology may help identify the pathogen. However, when there is no break in the skin, it is usually difficult to identify a specific bacteriological cause of cellulitis.12 Routine needle aspiration of the affected area rarely helps to determine the bacterial cause.12 Serological tests (e.g. antistreptolysin-O and anti-deoxyribonuclease B), done about a week after symptom onset, can be used to confirm streptococcal infection.

What is the best treatment?

Antibacterial therapy

There are too few published data to help determine the optimal antibacterial drug for treating cellulitis or optimal duration of treatment. Several published randomised controlled trials have compared different antibacterials,13 but these do not show a clear first choice of therapy.

A pragmatic approach

For mild, uncomplicated cellulitis, streptococci are the most likely causative organisms, and so phenoxymethylpenicillin (or benzylpenicillin if parenteral therapy is needed) is the most appropriate 'blind' treatment. Some of our consultants recommend replacing this therapy with flucloxacillin if there is no improvement after a few days, as the infection may be due to S. aureus. An alternative first-line approach recommended by some experts is to use flucloxacillin at a high dose (e.g. for an adult, 1-2g intravenously four times daily), as this should be effective against both streptococci and staphylococci. These recommendations to use single-drug antibacterial therapy contrast with national prescribing guidelines: both the British National Formulary (BNF) and the Public Health Laboratory Service14 (now part of the Health Protection Agency) recommend a combination of phenoxymethylpenicillin and flucloxacillin for the treatment of cellulitis; however the BNF recommends phenoxymethylpenicillin alone for 'erysipelas' (unless staphylococcal infection is suspected). For a patient with severe or rapidly worsening cellulitis (who needs admission to hospital), it would seem prudent to use a combination of phenoxymethylpenicillin (or benzylpenicillin, if parenteral therapy is needed) plus flucloxacillin.

Knowledge of local resistance patterns,15 the results of bacterial culture tests, or certain associated factors (e.g. a diabetic foot ulcer,16 an animal bite, or i.v. drug misuse) might indicate that other, broader-spectrum antibacterial therapy is needed.

Oral vs. i.v. antibiotics

We could find no published randomised controlled trials indicating the optimal route of administration of antibacterials available in the UK, for treating patients with cellulitis. In a non-randomised study involving 73 adults admitted to hospital with erysipelas, i.v. penicillin was no more effective in reducing fever duration or hospital stay than was oral penicillin.17 Whether antibacterial therapy needs to be given intravenously should probably depend on the patient's condition (e.g. whether he or she is systemically ill or vomiting) and on the site of infection (e.g. for periorbital cellulitis, i.v. therapy might be considered necessary).

General measures

For a patient with cellulitis affecting an arm or leg, elevating the limb might help reduce oedema and pain. The patient might need an analgesic drug. There are several published case reports, but no proven link, between NSAID use in patients with cellulitis and the development or aggravation of necrotising fasciitis.18 Systemic corticosteroid therapy has no worthwhile benefit in either treating cellulitis19 or in preventing recurrences.20

There is no published evidence to support the practice of applying hot moist soaks or potassium permanganate solution to areas of skin inflamed due to cellulitis.

Where should patients be treated?

There are no published UK trials investigating whether patients with cellulitis are best treated in or out of hospital nor any guidelines advising on hospital admission. Our consultants suggest that admission would seem sensible at least for neonates; patients who lack home support; those who are unwell due to co-existing diseases or are immunocompromised; or those whose cellulitis is severe, affects the periorbital area, or is worsening despite antibacterial therapy.

Who should treat them?

Cellulitis is diagnosed and treated by GPs and by various hospital specialists. Hospitals should have clear, simple, protocols for treating cellulitis, drawn up in association with infectious diseases specialists. Dermatologists should be asked to see any patients with suspected cellulitis that is not responding to antibacterial treatment, to help exclude other skin conditions. Advice from an ophthalmologist is necessary in cases of periorbital cellulitis, or from an ENT surgeon if there is any suspicion of spread of infection from the sinuses.

Is it possible to prevent cellulitis?

There are no published randomised controlled studies on prevention of first episodes of cellulitis, nor on whether treating underlying conditions prevents recurrence. However, it makes sense to treat predisposing conditions such as athlete's foot,21 and venous leg ulcers22 where possible, and to try to reduce chronic oedema of the arm23 or leg.

In an open randomised study in 40 patients with erysipelas and lymphatic congestion or venous insufficiency and a history of at least two episodes of erysipelas during the previous 3 years, prophylaxis with phenoxymethylpenicillin (1-2g twice daily; or erythromycin 250-500mg twice daily for penicillin-allergic patients) did not reduce recurrence rates compared with no prophylaxis.24 By contrast, in a randomised trial in 32 patients with a history of two or more episodes of erysipelas or cellulitis, but no current infection, none of those given erythromycin 250mg twice daily for 18 months developed relapses (vs. 8 patients in the group receiving no prophylaxis, p<.001).25


Cellulitis is common, varies in severity and can occasionally be life-threatening. Athlete's foot, lymphoedema and vascular insufficiency of the leg appear to be important precipitants of the condition, and should be treated wherever possible. Patients who have undergone surgery that has caused damage to blood or lymph vessels (e.g. saphenous venectomy, breast surgery) should be told that they could be at increased risk for developing cellulitis, and so advised to take care to avoid injury to the skin. Most community-acquired cellulitis is caused by streptococci and should therefore be treatable 'blind' with phenoxymethylpenicillin or benzylpenicillin. Hospital admission is advisable for neonates, patients who lack home support, those who are immunocompromised or are unwell with a co-existing disease, or if cellulitis is severe, affects the periorbital area or is worsening despite antibacterial therapy. Although there is weak and inconclusive evidence on whether long-term antibacterial prophylactic therapy prevents recurrent cellulitis, it may be worth trying (for 1-2 years) in patients with predisposing conditions who have had at least two episodes of cellulitis.

For such a common, and potentially serious, problem, there is an alarming lack of evidence to guide major management decisions. This needs to be addressed urgently.


[M=meta-analysis; R=randomised controlled trial]

  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. R 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. R 24.
  25. R 25.
View Abstract

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.