Introduction to The Wound Infection Evidence Matrix
The development of a structured survey of the evidence for wound infection was one of the principal outputs agreed by the International Wound Infection Institute at its inception. This process began with a simple listing of references (to be found in our “useful documents” section) and will conclude with a comprehensive list of reviewed, surveyed and abstracted papers on wound infection and its treatment. We see the provision of an evidence matrix as the next step along the route. The current version of the matrix is incomplete, but we believe that it is at an appropriate stage to be reviewed by our members. We welcome all of your comments.
One of the main features that we wanted to include was a rating system for the evidence. This is always problematic. The literature reveals multiple classifications and evidence hierarchies for ranking evidence 1,2,3,4,5. However, these evidence classifications and hierarchies contain many inconsistencies in interpretation and ranking. Although the gold standard for best evidence is generally considered to be a meta analysis or systematic review of double blinded randomised controlled trials, few, if any, such reviews are to be found in regards to the diagnosis or treatment of wound infection. Our rating system is focussed towards the levels of evidence as follows:
1. Meta analysis and systematic reviews
2. Randomised controlled trials
3. Non randomised controlled trials, case control trials, prospective cohort studies, animal studies, evidence summaries or evidence guidelines
4. Case reports, case series
5. Expert opinion, other literature reviews
Highly influential pieces have been published within all the above levels, some of which have contributed to changes in practice. Therefore we are considering the addition of an “impact” rating, separate to the evidence rating, based on the importance or significance of certain papers.
We hope that you find this draft of the evidence matrix to be useful. Equally importantly, please give us your opinion on how to improve and add to this document. In particular we would like your feedback on the following:
· How useful is the evidence matrix as provided?
· What other papers should be included?
· What impact score would you give to significant pieces?
Please let us know what you think. All comments will be gratefully received at
Keryln Carville, Chair, Evidence sub-committee, April 2009
References:
1. Upshur, R. (2003). “Are all evidence-based practices alike? Problems in the ranking of evidence”. CMAJ, 169(7), downloaded http://www.cmaj.ca/cgi/content/full/169/7/672
2. Brown JP, Josse RG; Scientific Advisory Council of the Osteoporosis Society of Canada. 2002 clinical practice guidelines for the diagnosis and management of osteoporosis in Canada. CMAJ 2002;167(Suppl 10):S1-34.
3. Centre for Evidence-Based Medicine. Levels of evidence and grades of recommendation. Oxford: The Centre. Available: www.cebm.net/levels_of_evidence.asp
4. Wright PJ, English PJ, Hungin AP, Marsden SN. Managing acute renal colic across the primary–secondary care interface: a pathway of care based on evidence and consensus. BMJ 2002;325:1408-12.
5. Evans, D. (2003). “Heirarchy of evidence: A framework for ranking evidence evaluating healthcare interventions.” Journal of clinical Nursing. 12(1), 77-84.
The Wound Infection Evidence Matrix – April 2009
Author, date / Comment / Title / Key points / Grade /Bacteria, biofilms and wound healing
(Bowler, Duerden et al. 2001) / Wound microbiology and associated approaches to wound management. / A thorough review of the literature on wound infection published up until 2001. Key points: all wounds are colonised and progression to infection is as much due to host factors as to the type and number of bacteria present; most open wounds are polymicrobial, with anaerobic bacteria constituting 50% of the species present in infected wounds; swab sampling is easy to carry out but results can be misleading and this should only be carried out if there are clinical signs of infection, if the wound fails to heal or is deteriorating; antibiotics induce bacterial resistance and antiseptics are preferred if topical treatment is required; debridement is an essential part of infection control. In my opinion a literature review as is stated in the conclusion / 5
(Ceri, Olson et al. 1999) / Outlines new technology for
selecting effective antibiotics in the treatment of biofilms / The Calgary biofilm device: new technology for rapid determination of antibiotic susceptibilities of bacterial biofilms. / Biofilms have an inherent lack of susceptibility to antibiotics. Ceri et al describe the Calgary Biofilm Device (CBD) which is a method for the rapid and reproducible assay of biofilm susceptibility to antibiotics.
The paper describes the formation of biofilms and confirmation of their presence using quantitative microbiology and SEM, followed by the rigorous testing and assessment of the CBD assay using NCCLS reference strains of E coli, P aeruginosa and S aureus. Growth curves demonstrated that biofilms grew uniformly in each of the 96 wells.
Antibiotic susceptibility testing demonstrated that, compared to planktonic forms of the same bacteria, 100 to 1000 times the concentration of an antibiotic was required to eradicate the biofilm. The authors claim that the CBD provides a method for the rational selection of antibiotics effective against microbial biofilms. / 3
(Costerton, Stewart et al. 1999) / A good review of biofilms / Bacterial biofilms: a common cause of persistent infections. / A good and well-referenced review of biofilms. The review explains how biofilms form and develop, how they differ from planktonic bacteria, the importance of quorum sensing as a possible target for interfering with their development. / 5
(Davies, Parsek et al. 1998) / The importance of signalling in biofilms is described and it is suggested this could be a way to control biofilms. / The involvement of cell-to-cell signals in the development of a bacterial biofilm. / This study demonstrates that a cell-to-cell signal (quorum sensing) is required for the differentiation of individual cells of P aeruginosa into complex biofilms. When differentiation is hindered by a mutation, the biofilm becomes abnormal and is sensitive to a detergent biocide (SDS). Without the signalling device, the biofilms were not able to grow with the proper architecture and did not leave sufficient space between colonies.
The authors suggest that inhibition of the quorum sensing signals could be possible ways to control biofilms, given their resistance to most antibiotics. / 3
(Dow, Browne et al. 1999) / Review / Infection in chronic wounds: controversies in diagnosis and treatment. / A thorough review that includes: definitions of contamination, colonization and infection; the pathogenesis of wound infection and how the inflammatory response can delay wound healing; diagnosis of wound infection; quantification of bacterial burden.
Regarding bacterial burden, the authors caution that there is no hard cut-off figure above which colonization turns to infection and that the level of microbial burden alone can not be used to define infection. They stress the importance of bacterial species and variety, and the capacity of the host to tackle infection. They warn of the difficulties of quantitative biopsy and argue the case for semi-quantitative assessment.
The paper also provides a detailed critique of specimen collection and culture techniques and concludes with a thorough overview of treatment options including mechanical (debridement), antiseptics and the role of antibiotics. A useful table summarises the antimicrobial options for empiric therapy.
A small number of the concepts have been challenged over the years since the publication of this review, but on the whole it is a thorough and valuable resource on the subject of wound infection. / 5
(Dowd, Sun et al. 2008) / Highly significant study / Survey of bacterial diversity in chronic wounds using Pyrosequencing, DGGE and full ribosome shotgun sequencing. / This study used specific techniques to identify the major populations of bacteria that occur in the biofilms found in three types of chronic wound: diabetic foot ulcer, venous leg ulcer and pressure ulcer. The techniques were: three separate I 6S-based molecular amplifications, followed by pyrosequencing, shotgun Sanger sequencing and denaturing gradient gel electrophoresis.
All chronic wound types contained certain specific major populations of bacteria: Staphylococcus, Pseudomonas, Peptoniphilus, Enterbacter, Stenotrophomonas, Finegoldia and Serratia species. However, each of the wound types showed marked differences in their bacterial populations.
For example, in venous ulcers over 80% of the bacteria were facultative anaerobes, compared with 62% in diabetic foot ulcers, and just over 20% in pressure ulcers. Pressure ulcers on the other hand comprised over 60% of strict anaerobes, compared with nearly 30% in diabetic foot ulcers and virtually none in venous ulcers.
Different wound types also showed a different level of oxygen tolerance amongst their bacterial populations. The authors suggest that this may imply each wound type has a distinct pathophysiology that affects the ecology of the wound environment determiging which bacteria can develop.
Results were compared with those from traditional culture-based analyses. In only one wound type did the culture methods correctly identify the primary bacterial population. Standard culturing techniques are inherently flawed as they only examine the 1% of microorganisms that are able to grow rapidly in pure culture. Also, certain populations may never be cultured in the laboratory due to reduced metabolic activity, obligate cooperation with other bacteria, need for specialized nutrients or environmental conditions.
The paper gives full details of the bacteria identified. / 3
(Hill, Davies et al. 2003) / Molecular analysis reveals a much greater diversity of microflora in chronic wounds than do culture techniques. / Molecular analysis of the microflora in chronic venous leg ulceration. / Culture analyses of CVLU generally reveal staphylococci, streptococci, enterococci and facultative Gram-negative bacilli. However, anaerobic isolation techniques and prolonged incubation reveal the presence of fastidious and slow-growing anaerobic species such as Fusobacterium and peptostreptococci. Cultivation-dependent methods for characterising the microflora of chronic wounds are limited.
The authors describe the analysis using 16S rDNA sequences of tissue from a CVLU which demonstrated significantly greater bacterial diversity than culture methods. Sequences even suggested novel species of bacteria. This technique can clearly not be used routinely so the clinical application is limited but may inform treatment in deteriorating or long-lasting wounds.
The study states that this was one patient and one wound that was analysed? Would that make it a 4? / 3
(James, Swogger et al. 2008) / Well-designed and highly significant study revealing that biofilms may be present in at least 60% of chronic non-healing wounds / Biofilms in chronic wounds. / Chronic wound specimens were taken from 77 subjects and acute wound specimens from 16. All specimens were cultured using standard techniques and in addition, light and scanning electron microscopy were used to analyse 50 chronic and 17 acute specimens. Molecular analyses were performed on the other 27 chronic specimens.
There was a statistically significant difference between the chronic and acute specimens in terms of presence of biofilm: 60% of the chronic but only 6% of the acute (p<0.001).
Molecular analysis showed that there were polymicrobial communities and bacteria, including strictly anaerobic, that were not revealed by culture.
The study shows that biofilms are not necessarily capable of detection using standard clinical techniques. / 3
(Laato, Niinikoski et al. 1998) / This paper suggests a mechanism to explain why S aureus sometimes appears to accelerate wound healing. / Inflammatory reaction and blood flow in experimental wounds inoculated with Staphylococcus aureus / This paper is often quoted as it demonstrated that certain low levels of bacteria in a wound could actually enhance healing, through stimulating inflammation which in turn would enhance local blood flow.
In an animal experiment, wounds were created and sponge implants were used as a matrix to encourage growth of granulation tissue. The implants were injected either with saline (control), or S aureus at concentrations of 102 or 105 microorganisms/ml. Implants inoculated with 105 organisms developed infection with pus formation, while implants inoculated with 102 showed no signs of infection but had an enhanced local blood flow. / 3
(Madsen, Westh et al. 1996) / P aeruginosa can enlarge ulcers and delay healing. S aureus and haemolytic strep also delay healing / Bacterial colonisation and healing of venous leg ulcers. / Fifty-nine patients with VLUs were followed with frequent semi-quantitative culture of bacteria from the ulcer surface for 180 days. The ulcer area was also measured. At 90 and 180 days the authors found that ulcers colonised with P aeruginosa were significantly larger than those without; and significantly fewer of them healed completely during the observation period.
Ulcers with S aureus or haemolytic streptococci healed significantly more slowly than those without. / 3
(Murphy, Robson et al. 1986) / Suggested that 105 bacteria per gram of tissue was a critical level for infection. / The effect of microbial contamination on musculocutaneous and random flaps. / Granulating wounds were inoculated with varying levels of bacteria per gram of tissue: 104, 105 or 106 and were then covered with musculocutaneous or random flaps or left uncovered. Bacterial proliferation was evident in all the heavily contaminated wounds (106) while in the minimally contaminated wounds (104) both types of flap achieved wound healing and decreased the bacterial level in the wound. In the intermediate group (105) musculocutaneous flaps lowered the bacterial count and allowed wound closure, whereas random flaps failed.
This study is relevant to clinicians beyond the field of surgery as it established that the 105 level of bacteria was an important tipping point in the development of infection – all other things being equal.
Significant impact / 3
(Ovington 2003) / Low evidence but useful piece / Bacterial toxins and wound healing. / An educational piece that describes critical colonisation as a stage of colonisation that occurs before invasive infection. The current view is that critical colonisation is actually the presence of biofilm.
The rest of the paper deals with bacterial toxins, describing the nature and effects of exo- and endo-toxins. While many educational pieces describe the effects of bacteria in wounds, this is one of the very few that deals specifically with the actions of bacterial toxins.
The author explains that while antimicrobials may reduce the amount of exotoxin produced by bacteria, they have no effect on exotoxin that is already in the wound. Also, when Gram-negative bacteria are destroyed, they release endotoxins from their cell wall, so topical antimicrobials may contribute temporarily to an increase in endotoxin levels. The author advocates the use of absorbent dressings and activated charcoal to remove toxins from the wound bed and describes a silver dressing that is based on an activated charcoal cloth (Actisorb). / 5