ADVERTISEMENT
Evidence Corner
Dear Readers:
Managing wound infection requires quick, accurate action. After a clinician diagnoses signs of infection in a wound,1 invading microorganisms are sampled, cultured and identified along with antibiotic(s) to which they are sensitive. Then an appropriate antibiotic is prescribed and administered to quell the infection. Research has identified valid, reliable diagnostic clinical signs and symptoms of infection in chronic wounds despite ambiguity of some of these signals.2 Techniques for sampling invading microorganisms are controversial. For example, quantitative biopsies with > 1 x 105 or > 1 x 106 organisms per gram of tissue, often regarded as the “gold standard” sampling technique, have limited specificity, frequently over-predicting infections.3,4 This can result in unnecessary antibiotic use, which has been associated with delayed chronic wound healing5 and may foster development of resistant organisms. Quantitative biopsies also have low sensitivity, missing invasive organisms in 25% of burn biopsies,6 and low reliability, varying in bioburden by more than 2- log10 colony forming units (CFU) in 27% of paired isolates.6 Quantitative swab samples may be a viable alternative.1,7 Recent research is deriving an operational definition of a valid, reliable, quantitative microbial sampling technique capable of selectively identifying invading microorganisms in chronic wounds clinically diagnosed with an infection.8 Research described in this edition of Evidence Corner explores this operational definition.
Laura Bolton, PhD, FAPWCA
Adjunct Associate Professor
Department of Surgery, UMDNJ
WOUNDS Editorial Advisory Board Member and Department Editor
Diagnostic Validity of Quantitative Microbial Swab Techniques
Reference: Gardner SE, Frantz RA, Saltzman CL, Hillis SL, Park H, Scherubel M. Diagnostic validity of three swab techniques for identifying chronic wound infection. Wound Repair Regen. 2006;14(5):548–557.
Rationale: Cultures harvested from wounds using quantitative swab techniques are commonly used to evaluate the bioburden of potentially infected wounds. However, information from these swabs is often unclear owing to wide or unreported variations in harvesting and culturing techniques.
Objective: Compare the diagnostic validity of 3 standardized quantitative swab techniques in identifying chronic wound infection and explore the association of these bioburden measures with risk factors for wound infection.
Methods: Concurrent quantitative wound surface swabs were harvested from 1 nonarterial full-thickness chronic wound on each of 83 patients in a Veterans Affairs Medical Center. Subjects were 18 years or older with white blood cell counts > 1500 cells/mm3, total lymphocyte counts > 800 cells/mm3, platelet counts > 125,000 cells/mm3 and were not receiving coagulopathy or anticoagulant therapy. Colony forming units (CFU) were quantified from quantitative, pre-weighed standardized swabs used to harvest, A) wound exudate obtained immediately after dressing removal from the wound before cleansing, B) a swab of the entire cleansed wound surface,using a close Z-stroke to harvest organisms while rotating the swab, or C) rotating the swab over 1 cm2 in the cleansed wound center with sufficient pressure on the wound surface to extract fluid from inside the wound (Levine’s7 technique). CFU/g were compared to those from a concurrent biopsy of viable tissue harvested beneath sample (c) using a 4 mm–6 mm dermal biopsy punch. Quantitative and qualitative CFU samples were similarly processed within 2 hours of harvesting using standardized culture media and isolation techniques blinded to sampling technique.
Results: “Infection” defined as > 1 x 106 CFU/g of biopsy tissue was noted in 30 (36%) of the 83 subjects predominantly with diabetes. Most subjects (77%) with > 1 x 106 CFU/g of biopsy CFU were not receiving systemic antibiotics at the time of biopsy. “Infected” wounds were longer in duration, but smaller in area than wounds with low bioburden. All 3 quantitative swab techniques had significant agreement (P < 0.011) with the biopsy standard. The Levine7 swab had highest agreement (P = 0.001) and a higher area under the Receiver Operating Curve, which plots true positive “infection” identifications as a function of false positives (P = 0.046). An optimal cut-off point of 3.7 x 104 CFU/g of swab content represented the criterionfor predicting > 1 x 106 CFU/g in the corresponding tissue biopsy. Sensitivity was maximal at 90% for all 3 techniques for critical cut-off points in the range of 2.36 x 105 CFU/g to 6.25 x 105 CFU/g. Specificity was maximal for a critical cut-off point of 1 x 106 CFU/g of swab content at 53% for Z-technique, 75% for wound exudate, and 90% for Levine’s technique, though sensitivity at this cut-off point was below 64% for all 3 techniques. Qualitatively, a range of 3–3.5 species were found per wound for all 3 techniques. Concordance values showed high agreement among all techniques for recovering all organisms.
Authors’ Conclusions: Swab specimens collected from wounds using Levine’s technique were more accurate than those using either wound exudate or Z-technique and were acceptably accurate in representing quantitative cultures biopsied from wound tissue. Further research is needed to explore predictive value of bioburden sampling techniques for identifying wounds at risk of infection in order to plan preventive interventions.
Clinical Assessments and Biopsies of Chronic Wound Infection
Reference: Serena T, Robson MC, Cooper DM, Ignatius J. Lack of reliability of clinical/visual assessment of chronic wound infection: the incidence of biopsyproven infection in venous leg ulcers. WOUNDS. 2006;18(7):197–202.
Rationale: Classic signs of infection are unreliable or absent in venous ulcers. Infection signs with positive predictive value: delayed healing, friable and/or discolored granulation tissue, undermined wound edge, foul odor, wound deterioration, and increased pain2 remain to be correlated with biopsy CFUs.
Objective: Compare clinical examination with biopsy CFUs in making the diagnosis of infection in a large group of venous leg ulcers.
Methods: Patients with venous ulcers confirmed by duplex scan were enrolled in a 20-week study of effects of twice-weekly repifermin versus placebo on complete healing not reported in this publication. Enrollment criteria for this study included clinically screened noninfected venous ulcers 3 cm2–25 cm2 in area of at least 1- month duration.These were enrolled only if a standardized quantitative 6-mm punch biopsy of wound tissue confirmed absence of infection by containing < 105 CFU and no beta-hemolytic streptococci per gram of tissue (negative biopsy) in 1 of up to 3 biopsies. If the first biopsy contained > 105 CFU, the protocol called for 10 days of systemic or topical antibiotics followed by a second biopsy. If the second biopsy contained > 105 CFU another 10-day course of antibiotics was followed by a final biopsy. Quantitative and qualitative biopsy organisms and percent of subjects clinically diagnosed as infected were reported.
Results: The investigators’ clinical determination of venous ulcer infection status was confirmed by biopsy CFUs in 74% of subjects. Of 614 patients screened, 352 with a negative wound tissue biopsy were eventually enrolled in the study. Of 260 subjects enrolled after 1 biopsy, 60% healed in 20 weeks. Of 82 enrolled after requiring a second biopsy, 51% healed. Of 10 enrolled after 3 biopsies, 40% healed (P = 0.15 for the trend toward lower percents healed in subjects receiving more biopsies). Most cultures (95%) displayed a single organism, predominantly Staphylococcus aureus. Polymicrobial cultures were reported in 23% of patients experiencing more than 1 biopsy.
Authors’ Conclusions: Wound infection before enrollment was associated with a trend toward poorer healing even after high levels of tissue organisms had been treated. Patients with the most biopsies tended to heal more slowly.
Clinical Perspective
Neither article attempted to validate tissue CFUs against clinically diagnosed wound infection. Both acknowledge that it is controversial to use biopsy content as a definition of wound infection unless the wounds display prior clinical signs of infection.1 Gardner et al established validity of the Levine7 quantitative wound surface swab sampling technique compared to a standardized quantitative biopsy, but caution readers that defining infection as biopsy CFUs without reference to validated clinical criteria for infection gives clinicians no idea of the predictive value of the biopsy in accurately identifying wounds at risk of infection so they can plan preventive interventions. Serena et al showed that a quantitative biopsy frequently confirmed clinical diagnosis in venous ulcers. Ulcers with biopsy-confirmed low CFUs experienced healing delay associated with repeated biopsies, antibiotic use, or both.This raises questions about biopsy and/or antibiotic safety and biopsy predictive validity. Future research comparing healing after quantitative Levine swabs or biopsies may answer these safety questions. Both articles support using quantitative wound CFU sampling techniques to identify infecting organisms. More research is needed to determine whether and at what cut-off point wound CFUs are sensitive, specific, accurate, or predictive in diagnosing infected wounds according to clinically validated criteria2 for wound infection.
