COMPUTED TOMOGRAPHY PATTERNS OF SINONASAL DISEASES AND THEIR CORRELATION WITH HISTOPATHOLOGY FINDINGS AT SAINT PAUL HOSPITAL MILLENNIUM MEDICAL COLLEGE, ADDIS ABABA, ETHIOPIA

Abstract

ABSTRACT Background: Sinonasal diseases can arise from multiple etiologies. CT imaging has become the primary investigation tool for the assessment of sinonasal pathologies, providing critical information for surgical planning and guiding interventions. Objective: To investigate and characterize the CT patterns of sinonasal pathologies, patterns of histopathology, clinical-imaging and imaging-pathologic agreement at Saint Paul hospital millennium medical college Method: Hospital-based cross-sectional study was conducted at SPHMMC in June 2024 G.C. The study population comprised patients who were referred from the ENT department to the radiology department for CT imaging of the paranasal sinuses at SPHMMC during the study period and who meet the inclusion criteria. Simple random sampling technique was employed from a source population of 220 patients. Data was collected from the hospital's EMR & PACS to see the imaging reports of patients’ imaged from January 2023 to June 2024. The data was analyzed by SPSS version 27.0 statistical software. Descriptive statistics, Test statistics (sensitivity, specificity, PPV, NPV & accuracy) & Cohen’s Kappa were used to assess the agreement of CT imaging findings with clinical diagnosis and histopathology diagnosis. Result: A total of 135 patients were included of which 79 (58.5%) were females and 56 (41.5%) were males with age range from 5 weeks to 78 years. The median age was 22years. The most common presentation was nasal blockage accounting 54 (40%). Inflammatory etiology is the most common CT & histopathology finding accounting for 109 (80.7%) & 17 (54.8%) respectively. The concordance between clinical and CT diagnosis and CT & histopathology diagnosis is 93.33% and 77.42% with a kappa of 0.806 & 0.565 respectively. Conclusion: In this study inflammatory diseases were the most frequently detected disease entities both on CT and histopathology. The study found a high (93.33%) concordance rate between clinical and CT diagnoses, and a moderate agreement between imaging and the gold standard of histopathology. Further prospective study with large number of sample size is recommended as this hospital based in tertiary center may not represent the general population. Key word: Computed Tomography, Histopathology diagnosis, Sinonasal Pathologies