Background and Purpose- Rates of intracerebral hemorrhage (ICH) are estimated to be highest globally in sub-Saharan Africa. However, outcomes of ICH are poorly described and standard prognostic markers for ICH have not been validated in the region. Methods- We enrolled consecutive patients with computed tomography-confirmed ICH at a referral hospital in southwestern Uganda. We recorded demographic, clinical, and radiographic features of ICH, and calculated ICH scores. We fit Poisson regression models with robust variance estimation to determine predictors of case fatality at 30 days. Results- We enrolled 73 individuals presenting with computed tomography-confirmed ICH (mean age 60 years, 45% [33/73] female, and 14% [10/73] HIV-positive). The median ICH score was 2 (interquartile range, 1-3; range, 0-5). Case fatality at 30 days was 44% (32/73; 95% CI, 33%-57%). The 30-day case fatality increased with increasing ICH score of 0, 1, and 5 from 17%, 23%, to 100%, respectively. In multivariable-adjusted models, ICH score was associated with case fatality (adjusted relative risk, 1.48; 95% CI, 1.23-1.78), as were HIV infection (adjusted relative risk, 1.92; 95% CI, 1.07-3.43) and female sex (adjusted relative risk, 2.17; 95% CI, 1.32-3.59). The ICH score moderately improved with the addition of a point each for female sex and HIV serostatus (0.81 versus 0.73). Conclusions- ICH score at admission is a strong prognostic indicator of 30-day case fatality in Uganda. Our results support its role in guiding the care of patients presenting with ICH in the region.
Keywords: HIV infections; Uganda; cerebral hemorrhage; mortality; sex; stroke.