PURPOSE: Videomicroscopy is very useful for burn depth assessment in an early phase; however, there is no practical classification that includes complicated anatomic, pathologic, and morphologic findings of burn wounds. The aim of this study was to propose a novel classification to assess burn depth in its early phase easily and reliably by videomicroscopy. METHODS: Forty-four patients with 56 intermediate-depth burn wounds were included. Burn depth was divided into each grade according to our proposed classification, which is composed of five categories based on dermal capillary integrity patterns. The intrarater and interrater reliabilities of the assessment by the second and third authors were evaluated by Cohen's unweighted κ-value. RESULTS: The results of the measurements according to the proposed classification showed an accuracy of 92.9%, sensitivity of 81.8%, and specificity of 100.0%. The intrarater reliability of the second and third authors showed substantial agreement (κ=0.719 and 0.729, respectively). The interrater reliability of the sum of each observer's variable also showed substantial agreement (κ=0.636). CONCLUSION: This pattern analysis system is easy to use even for inexperienced personnel, and is reliable with high accuracy and specificity. Intrarater and interrater statistics also support its reliability and reproducibility. Copyright Â
PURPOSE: Videomicroscopy is very useful for burn depth assessment in an early phase; however, there is no practical classification that includes complicated anatomic, pathologic, and morphologic findings of burn wounds. The aim of this study was to propose a novel classification to assess burn depth in its early phase easily and reliably by videomicroscopy. METHODS: Forty-four patients with 56 intermediate-depth burn wounds were included. Burn depth was divided into each grade according to our proposed classification, which is composed of five categories based on dermal capillary integrity patterns. The intrarater and interrater reliabilities of the assessment by the second and third authors were evaluated by Cohen's unweighted κ-value. RESULTS: The results of the measurements according to the proposed classification showed an accuracy of 92.9%, sensitivity of 81.8%, and specificity of 100.0%. The intrarater reliability of the second and third authors showed substantial agreement (κ=0.719 and 0.729, respectively). The interrater reliability of the sum of each observer's variable also showed substantial agreement (κ=0.636). CONCLUSION: This pattern analysis system is easy to use even for inexperienced personnel, and is reliable with high accuracy and specificity. Intrarater and interrater statistics also support its reliability and reproducibility. Copyright Â