InfectionsBMJ 2006; 332 doi: https://doi.org/10.1136/bmj.332.7545.838 (Published 06 April 2006) Cite this as: BMJ 2006;332:838
- Brendan Healy1,
- Andrew Freedman
- 1 Brendan Healy is specialist registrar in infection diseases and microbiology, and Andrew Freedman is honorary consultant in infectious diseases at University Hospital of Wales, Cardiff.
Despite optimal treatment some wounds are slow to heal. The challenge clinically and microbiologically is to identify those wounds in which healing is impaired as a result of infection or heavy bacterial burden and in which systemic or topical antimicrobial treatment will be of benefit.
Staphylococci and streptococci are the most commonly encountered pathogenic organisms in community acquired superficial wounds. More unusual organisms may be found in bite wounds, and these reflect the source of the bite. Pathogenic organisms causing surgical wound infections vary according to the anatomical site of surgery. Antibiotic resistant organisms, such as methicillin resistant Staphylococcus aureus (MRSA), are more commonly encountered, reflecting the hospital flora.
When to sample
It is inappropriate to swab all wounds: swabs should be taken only from overtly infected wounds and from wounds that are deteriorating, increasing in size, or failing to make satisfactory progress despite an optimal environment for wound healing. Indicators of wound infection include redness, swelling, purulent exudate, smell, pain, and systemic illness in the absence of other foci. Subtle signs of local wound infection include unhealthy “foamy” granulation tissue, contact bleeding, tissue breakdown, and epithelial bridging. Staphylococcus aureus
Types of sample
Superficial wound swabs—The ease of obtaining and processing superficial wound swabs, combined with their relatively low cost and non-invasive nature, make them in most instances the most appropriate method for wound sampling. Organisms cultured from a superficial swab may, however, simply reflect the colonising bacterial flora and are not always representative of the pathogenic organisms invading deeper tissue. This is particularly relevant to deep surgical and deep penetrating wounds in which infection from internal sources may occur.
Tissue and pus—Tissue or pus, or both, should be collected whenever possible, as growth from these samples is more representative of pathogenic flora. These are amenable to quantitative microbiological analysis and other techniques used to improve the diagnostic yield. Tissue biopsy should always be carried out when therapeutic debridement of the wound is done, in cases of osteomyelitis, and when superficial sampling methods have been ineffective.
Less invasive techniques—Less invasive sampling techniques—such as dermabrasion and various absorbent pads—have been developed. A wide range of products is available, but no single method is used routinely yet.
Most laboratories will perform a semiquantitative analysis on wound swabs. This entails grading bacterial growth as scanty, light, moderate, or heavy. Semiquantitative analysis introduces a bias towards motile and fast growing organisms.
This is the 10th in a series of 12 articles
Infection is a major source of failed wound healing
Fastidious organisms such as anaerobes may be under-represented. Semiquantitative counts have been shown to correlate with quantitative tissue counts in both burn wounds and diabetic foot ulcers.
Bacterial load greater than 100 000 organisms or colony forming units per gram of tissue or mm3 of pus is a predictor of wound infection.
However, some wounds that are more heavily colonised will heal spontaneously, and, conversely, some organisms are able to cause serious infection at much lower levels of colonisation. Infection depends on the pathogenicity of the organism, the type of wound, and the host response.
Interpretation of results
Most wound swabs will yield bacterial growth. Growth of bacteria from wounds is not synonymous with infection, and treatment based on microbiological results alone is not warranted.
Wound infections in association with systemic illness, deep invasion, or cellulitis require empirical systemic antibiotic treatment while culture results are awaited. Choice of treatment will depend on factors such as the type and site of wound; previous microbiological results; and host factors such as drug allergies. Clinicians must always be alert to the possibility of necrotising fasciitis. A high level of suspicion followed by prompt aggressive surgical debridement of devitalised necrotic tissue is essential if the patient is to survive. Important clinical markers include pain disproportionate to clinical signs, anaesthesia over the infected area, and systemic illness.
Superficial wound swabs are not always representative of the pathogenic organisms invading deeper tissue
Treatment of locally infected wounds with topical antiseptics such as silver compounds or iodine will be sufficient in most instances. Topical treatment avoids the potential side effects of systemic antibiotics, such as Clostridium difficile diarrhoea, anaphylaxis, gastrointestinal upset, and, perhaps most importantly, selection of resistant organisms. Systemic treatment may be indicated if topical medication is unsuccessful.
In general, topical antibiotics are not recommended. Reasons for this include inadequate penetration for deep skin infections, development of antibiotic resistance, hypersensitivity reactions, systemic absorption when applied to large wounds, and local irritant effects leading to further delay in wound healing. Short courses of silver sulfadiazine or topical metronidazole can be useful, however, in certain circumstances—for example, with burns and chronic ulcers.
Osteomyelitis associated with wound infection
Osteomyelitis may develop after direct inoculation of bone from a contiguous focus of infection. This can be a devastating complication of wound infection, requiring specialist intervention and management.
The diagnosis of osteomyelitis should be considered in any chronic wound that does not heal despite optimal treatment or in any wound (especially in those with diabetes) that can be probed to bone. Plain x rays of the affected area should be the first line of investigation.
Radiographic changes, however, can lag behind the evolution of infection by at least two weeks; a single, negative plain x ray film does not, therefore, exclude osteomyelitis. Magnetic resonance imaging is more sensitive than plain radiography. Nuclear scintigraphy—either a technetium bone scan or a labelled white cell scan—may also be helpful but requires careful interpretation. It can be difficult to differentiate osteomyelitis from chronic soft tissue infection.
Antibiotics penetrate poorly into devitalised bone, and long courses of antibiotics may be required. It is therefore important to define the infecting organism(s) from the outset so that antibiotic treatment can be targeted. Ideally, in the absence of systemic illness, antibiotics should not be started before microbiological sampling of the infected bone.
Surgery followed by prolonged intravenous antibiotic treatment (generally a minimum of six weeks), is indicated in selected patients. Periodic antibiotic treatment at times of wound deterioration or of systemic illness may be appropriate if cure is unachievable.
Surgery enables debridement of all necrotic bone and tissue and provides deep samples for microbiological analysis. In some patients, surgery is not possible either because of the site of the wound or because of the patient's debility. Under these circumstances, a prolonged course of antibiotics may be warranted.
Choice of treatment is dependent on the antibiotic sensitivity pattern of the infecting organism(s) along with antibiotic properties, such as bone penetration, and host factors, such as drug allergy. Combination therapy is often used to gain maximal effect. Inflammatory markers (including C reactive protein and erythrocyte sedimentation rate) and radiological images can be used to monitor response.
Methicillin resistant Staphylococcus aureus
The incidence of MRSA wound infection and osteomyelitis is increasing. Isolation of MRSA from a wound, however, does not require treatment in the absence of clinical signs of infection. Topical antimicrobial agents, such as iodine and silver compounds, have activity against MRSA and may be used in localised wound infection when there is no evidence of invasion, cellulitis, or systemic upset.
In a systemically unwell individual, a glycopeptide (vancomycin or teicoplanin) should be administered. In all cases of MRSA osteomyelitis and in some MRSA wound infections a second antistaphylococcal agent with good penetration to bone and superficial skin sites should be added—for example, fusidic acid or rifampicin. Both rifampicin and fusidic acid can cause hepatitis and require regular monitoring of liver function tests.
With the exception of linezolid, evidence for the use of oral antibiotics in MRSA infections is lacking. However, when oral antibiotics are used, combinations are recommended to protect against the development of resistance. Combinations of rifampicin or fusidic acid with either trimethoprim or minocycline have been used with some success. The combination of rifampicin with fusidic acid is not advisable because of the increased risk of hepatotoxicity.
Linezolid, an oxazolidinone, is a new agent active against MRSA. It has excellent bioavailability, can be administered orally, and has good skin and bone penetration. Linezolid is generally well tolerated, but can cause bone marrow suppression, and regular haematological monitoring is therefore required. Linezolid use is currently limited by its high cost.
Agents that may be available in the near future include daptomycin, tigecycline, and dalbavancin.
Radiological improvement will often lag behind clinical improvement by up to six weeks
The ABC of wound healing is edited by Joseph E Grey (), consultant physician, University Hospital of Wales, Cardiff and Vale NHS Trust, Cardiff, and honorary consultant in wound healing at the Wound Healing Research Unit, Cardiff University, and by Keith G Harding, director of the Wound Healing Research Unit, Cardiff University, and professor of rehabilitation medicine (wound healing) at Cardiff and Vale NHS Trust. The series will be published as a book in summer 2006.
Competing interests For series editors'competing interests, see the first article in this series.
The figure showing the spectrum of interaction between bacteria and host was supplied by J E Grey and Stuart Enoch.