TYPE: Late Breaking Abstract
TOPIC: Imaging
PURPOSE: To investigate the association of chest radiographic severity Brixia score and in-hospital mortality among admitted Filipino adult COVID-19 confirmed patients.
METHODS: We retrospectively reviewed 125 baseline chest x-rays of adult Filipino COVID-19 confirmed patients admitted at a private tertiary hospital in Batangas between March 2020 to March 2021. Patients’ demographic data and clinical outcome were obtained. The Brixia score for each baseline CXR was designated independently by two radiologists who were blinded from the patient’s outcome. Logistic regression was done with Brixia score to determine association with in-hospital mortality.
RESULTS: A total of 99 patients (79.2%) recovered while 26 patients died (20.8%) during this period. The results showed that there was a statistically significant association of Brixia scores (p value = 2.68 x 10-5 , OR = 1.28) with in-hospital mortality. Likewise, age (p value = 0.000636, OR = 1.05), hypertension (p value= 0.01579, OR = 3.00) , immunocompromised state (p value = 0.0042926, OR = 11.23) and smoking history (p value = 0.01579, OR = 2.16) were also significantly associated with in-hospital mortality. The average Brixia score from both radiologists yielded an optimal cut-point of 8 for in-hospital mortality with an AUC of 0.7706.
CONCLUSIONS: The findings of this study suggest that Brixia scores obtained from baseline CXR have a statistically significant association with in-hospital mortality among COVID-19 confirmed Filipino patients.
CLINICAL IMPLICATIONS: The use of Brixia scoring system from Italy may help Filipino clinicians systematically report CXR of COVID-19 patients and may help identify patients at higher risk for in-hospital mortality upon admission.
DISCLOSURE: No significant relationships.
KEYWORD: Brixia score