3. Electronic Theses and Dissertations (ETDs) - All submissions
Permanent URI for this communityhttps://wiredspace.wits.ac.za/handle/10539/45
Browse
1 results
Search Results
Item Comparison of Chest X-ray findings in ambulatory and hospitalised children with pulmonary TB(2018) Baker, GregoryINTRODUCTION: Pulmonary TB is common in South Africa, with many children affected. Diagnosis can be challenging and chest x-ray remains fundamental for diagnosis. Interpretation is difficult and shown to have wide inter-reader variability. No study however has compared CXR findings and inter-reader agreement between ambulatory and hospitalised patients. AIM: This study compares the frequency of CXR changes, as well as inter-reader agreement in ambulatory compared to hospitalised children with suspected TB. METHOD: A pre-existing database containing CXR data of children worked up for PTB from 2008-2013 was used. Retrospective analysis of 69 ambulatory and 112 hospitalised patients, aged 0-12 years from Nolungile clinic and Red Cross Children’s Hospital respectively was done. Each sample contained 50% proven TB and 50% negative controls. Two paediatric radiologists and one paediatrician served as blinded, independent readers for the database using standardised tick sheets. RESULTS: Finding frequency; Fleiss/Free marginal Kappa (ambulatory and hospitalised respectively): Overall TB: 27.5% and 35.7%; -0.07 and 0.12 /-0.06 and 0.12 Parenchymal change: 34.8% and 67.9%; 0.24 and 0.49/0.25 and 0.60 Lymphadenopathy: 24.6% and 33.9%; 0.01 and 0.13/0.01 and 0.13 Pleural effusion: 7.3% and 18.8%; 0.27 and 0.61/0.84 and 0.80 Airway compression: 11.6 and 17.9%; 0.26 and 0.20/0.73 and 0.52 CONCLUSIONS: Our study demonstrated no significant difference in lymphadenopathy, but an increase in parenchymal change in the hospitalised group. We otherwise showed comparable results to literature regarding finding frequency, but poor inter-observer agreement. If the least expert reader were removed, results were comparable with available literature. This highlights the need for development and study of explicit CXR criteria for lymphadenopathy to improve the value of CXR for paediatric TB in all settings.