Thread by @WuidQ

#IDgrandrounds 50/M with new HIV diagnosis (CD4 13), presents with miliary TB and pericarditis (moderate pericardial effusion, no evidence of constrictive physiology). Together with anti-TB medications, will you prescribe corticosteroids?

1/5) No strong evidence to support the use of steroids especially in people w/ HIV.

Guidelines differ however:

ATS/CDC/IDSA 2016 idsociety.org/practice-guide…:

-- initial steroids "not be routinely used"

WHO 2017 who.int/tb/publication…:

-- initial steroids "maybe used"

2/5) A recent meta-analysis (from our very own Dr. Ige George), showed no mortality difference w/ the use of steroids (Forest plot provided in the pic)

ingentaconnect.com/content/iuatld…

3/5) Largely driven by the largest RCT (Mayosi): nejm.org/doi/10.1056/NE…

-- 67% HIV (14% on ART)
-- no mortality benefit
-- reduction in constrictive pericarditis

WHO guideline: HIV in the Mayosi trial may have influenced result; people w/o HIV may still benefit from steroids.

4/5) In the same meta-analysis by George et al ingentaconnect.com/content/iuatld…

.. a sensitivity analysis was performed to assess the impact of hypothetical NEW studies to the meta-analysis (see Forest plot in the pic).

The results were the same.

5/5) On the side..

The great Sir William Osler was believed to have suffered from cutaneous TB (tuberculosis verrucosa cutis or prosecrutor's wart). Through out his career, he had performed >1,000 postmortem autopsies.

Read more here: sciencedirect.com/science/articl…