Reliability of the use of photographs for the classification of dependence-related lesions

To examine the reliability of the classification of dependence-related lesions using photography.

A panel of 5 experts carried out a blind, anonymous and independent assessment of 24 photographs. Images were sent by electronic mail, together with an anonymous questionnaire that was sent back to a referee of the research group. We suggested 8 options for classification: pressure ulcer (categories I, II, III, IV, or 'suspect of deep-tissue damage'), moisture-associated damage, pressure-moisture combined aetiology and multifactorial lesion. Concordance index was calculated for every photography. Cohen's Kappa index with 'jackknife' technique were calculated for every photograph and category. We used the statistical programme Epidat 4.2.

Observed agreement was variable, between 4 and 1 depending on the photography. A consensus of 100% was observed in 6 images. Global Kappa resulted as moderate: 0.5202 (IC95%: 0.3850- 0.6542). Intra-group concordance resulted moderate in 4 categories, low in 3, and high only for pressure ulcer category II (Kappa 0.8924. IC 95%: 0.7388-1.0456). Inter-observer concordance was 0.6602 (IC 95%: 0.4969-0.8081).

The reliability of the use of photographs for the categorization of dependence-related lesions is moderate, being higher for category II pressure ulcers and low for more complex wounds. This method must be used with care, and education on the subject is required.