Chest X-ray AI beats comorbidity scores for predicting ICU admission in COVID-19

Chest X-ray deep learning model predicted ICU admission within 24 h with AUC 0.78 (95% CI 0.74–0.81), outperforming Charlson (AUC improvement 0.16) and age-adjusted Charlson (0.12) comorbidity indices (p<0.001). Retrospective, single-site COVID-19 cohort; external validation nee…

• Retrospective study of 5,046 patients and 8,357 CXRs from a single institution during the COVID-19 pandemic.
• Model’s AUC of 0.78 significantly exceeded both Charlson Comorbidity Index (improvement +0.16) and age-adjusted Charlson Index (+0.12), p<0.001.
• Limitation: single-center, retrospective design confined to COVID-19 patients; no external or prospective validation performed.

RadPigeon summaries are original and for information only. They are not clinical advice.