Long-term determinants of death after stroke in Ghana: Analysis by stroke types & subtypes

Determinants of long-term mortality after stroke by mechanistic type and subtype are unknown in sub-Saharan Africa (SSA). Such data are crucial for targeting specific pathophysiologic pathways to improve stroke outcomes in the region.

To evaluate rates and predictors of mortality up to 8 years after stroke, by type and subtype, in Ghana.

We prospectively collected data on stroke patients presenting at a tertiary medical facility in Ghana between 2013 and 2018 who were followed up until October 31, 2021. Stroke diagnosis was confirmed using a Computerized Tomography scan; ischemic strokes were mechanistically typed using the TOAST classification while intracerebral hemorrhages were subtyped using the SMASH-U scheme. Demographic and clinical predictors of stroke mortality were evaluated using Cox proportional hazards regression modelling.

Of 564 patients encountered during the study period, data on vital status were available for 556 (98.6%) subjects at discharge and 442 (78.4%) on follow-up. Mean age was 61.1 ±15.1 years, and 223 (53.1%) were male. Mortality rates at 1, 3, 12, 36, 60 and 96 months were 37.5%, 43.2%, 49.7%, 57.4%, 62.9%, and 73.7% respectively. Three (3) factors remained significantly associated with risk of death namely age, adjusted hazard ratios (aHR) of 1.12 (95%CI: 1.04-1.20), no formal education 1.36 (95% CI: 1.02-1.81) and modified Rankin score 1.67 (95%CI: 1.42 - 1.98) for each unit rise. For ischemic stroke the four (4) factors associated with long-term mortality were low formal education, low monthly income, having diabetes mellitus and higher functional score on admission. For intracerebral hemorrhage, the two factors were increasing age and high functional score.

Stroke severity is the key predictor of long-term mortality after an index stroke in Ghana, regardless of the underlying pathophysiologic mechanism. Instituting acute stroke systems of care to facilitate timely reperfusion efforts may greatly improve long term survival outcomes after stroke in SSA.