Effect of HIV status on chest radiographic and CT findings in patients with tuberculosis

L. B. Haramati, E. R. Jenny-Avital, D. D. Alterman

Research output: Contribution to journalArticle

44 Scopus citations


Aim: To compare the chest radiographic and chest CT findings of tuberculosis according to HIV status. Patients and methods: Ninety-eight HIV-tested patients with cultures positive for Mycobacterium tuberculosis (Mtb) between January 1991 and December 1993 whose clinical charts and radiographic records were available for review formed the study population. There were 67 HIV-positive patients (51 men, 16 women) and 31 HIV-negative patients (23 men, 8 women). Chest CT scans were available for review in 15 HIV-positive and four HIV-negative patients. Results: On chest radiographs, HIV-positive patients had mediastinal lymphadenopathy (60% vs. 23%) and atypical infiltrates (55% vs. 10%) significantly more frequently than HIV-negative patients. Conversely, HIV-negative patients had infiltrates typical for reactivation tuberculosis (77% vs. 30%) and cavitation (52% vs. 18%) significantly more frequently than HIV-positive patients. The chest CT scans showed a similar trend, but significant differences were only seen regarding more frequent bilateral mediastinal lymphadenopathy in HIV-positive patients and more frequent cavitation in HIV-negative patients. Conclusion: This study demonstrates significant differences in chest radiographic and chest CT appearances of tuberculosis according to HIV status. HIV-positive patients have more frequent atypical infiltrates and mediastinal lymphadenopathy, and less frequent cavitation and infiltrates typical for reactivation tuberculosis than do HIV-negative patients.

Original languageEnglish (US)
Pages (from-to)31-35
Number of pages5
JournalBrain and Language
Issue number1
StatePublished - Dec 1 1997


ASJC Scopus subject areas

  • Language and Linguistics
  • Experimental and Cognitive Psychology
  • Linguistics and Language
  • Cognitive Neuroscience
  • Speech and Hearing

Cite this