Systematic review and meta-analysis of diagnostic test accuracy in chronic wound's microbiology

This study aims to assess the diagnostic accuracy of non-culture-based methodologies for detecting microorganisms in chronic wounds.

We systematically reviewed studies that evaluated the diagnostic accuracy of alternative tests in chronic wound samples, excluding studies focused on animal samples or unrelated conditions. The search encompassed PubMed, CINAHL, Scopus and Web of Science databases, employing the QUADAS-2 tool for risk of bias assessment. Our search included the PubMed, CINAHL, Scopus and Web of Science databases, and we assessed the risk of bias using the QUADAS-2 tool. A meta-analysis was conducted on polymerase chain reaction (PCR) and colorimetric methods to determine sensitivity, specificity, diagnostic odds ratio, and summary receiver-operating characteristic (sROC) curves using a random-effects model. For methods not suitable for quantitative synthesis, a narrative synthesis was performed.

Nineteen studies involving various types of chronic wounds were analysed, revealing diverse diagnostic methods including fluorescence, PCR, colorimetry, voltammetry, electronic nose, biosensors, enzymatic methods, staining and microscopy. Combining fluorescence with clinical signs and symptoms (CSS) versus culture showed significant accuracy. Colorimetry demonstrated low sensitivity but high specificity, with a diagnostic odds ratio of 6.3. PCR generally exhibited good accuracy, although significant heterogeneity was noted, even in subgroup analyses.

This study identified a broad spectrum of diagnostic approaches, highlighting the superior diagnostic accuracy achieved when microbiological analysis is combined with clinical assessments. However, the heterogeneity and methodological variations across studies present challenges in meta-analysis. Future research should aim for standardized and homogeneous study designs to enhance the assessment of diagnostic accuracy for alternative methods.