Abstract: |
BACKGROUND: Cellulitis has many potential mimickers, and its misdiagnosis often leads to unnecessary hospitalizations and higher health care costs. The ALT-70 predictive model offers an objective tool to help differentiate between cellulitis and other clinically similar conditions at the time of initial emergency department (ED) presentation. OBJECTIVE: To evaluate the performance of the ALT-70 predictive model at 24 and 48 hours following ED presentation. METHODS: We performed a retrospective review of our prior cohort and expanded our data collection to include data at 24 and 48 hours after initial ED presentation. We compared classification measures for the ALT-70 at the time of initial ED presentation, 24 hours after presentation, and 48 hours after presentation. RESULTS: There was a statistically significant difference in median ALT-70 score between patients with true cellulitis and those with mimickers of cellulitis at all time points. Sensitivity, specificity, positive predictive value, and negative predictive value of the ALT-70 score was similar across all 3 time points. LIMITATIONS: Single-center design may reduce generalizability. CONCLUSION: At 24 and 48 hours, the ALT-70 performed similarly to the way it performed at the time of initial ED presentation, allowing for its use in a wider array of clinical settings. |