Purpose: There is a need for a valid and reliable method to describe the severity of preseptal cellulitis.
Methods: Items of a scoring system were derived by an expert group and evaluated using a retrospective chart review. The results were used to construct the final Severity Index. Validity and reliability of the Severity Index was evaluated by prospective assessment of 17 children. The Severity Index was compared with a Global Score, a score based on clinical impression.
Results: The average Severity Index score was 2.0 for patients treated with oral antibiotics alone and 6.0 for patients treated with intravenous antibiotics. The Severity Index correlated well with the Global Score (Spearman rank correlation coefficient rS = 0.60, P = 0.01). Ranked clinical photographs of preseptal cellulitis correlated moderately to the Severity Index (rS = 0.66, P = 0.02). The Severity Index score after 24 hours of treatment was significantly lower than at presentation (P = 0.004). The agreement between paired Severity Index scores [intraclass correlation coefficient (ICC) = 0.80, P = 0.001] was better than the agreement between paired Global Scores (ICC = 0.45, P = 0.03).
Conclusions: The Severity Index is an objective clinical tool for evaluating severity of preseptal cellulitis in children. It correlates well with clinical constructs for severity and is sensitive to small changes in clinical status. It has better reliability than overall clinical impression. The Severity Index will also be valuable as an outcome measure for future therapeutic trials for preseptal cellulitis in children.