TITLE:
Profile and Outcome of Non-Traumatic Coma in a Limited-Resource Setting in the Southwest Region of Cameroon
AUTHORS:
Forchap Nkemanjong Milton, Nkouonlack Duquesne Cyrille, Denis Georges Teuwafeu, Vincent Verla Siysi, Yacouba Njankouo Mapoure
KEYWORDS:
Non-Traumatic Coma, Aetiology, Outcome, Focal Neurological Deficit
JOURNAL NAME:
Neuroscience and Medicine,
Vol.17 No.1,
March
5,
2026
ABSTRACT: Background: Non-Traumatic Coma (NTC) almost always poses a diagnostic and therapeutic challenge, even in tertiary hospitals worldwide. We therefore aimed to determine the prevalence, aetiologies, and outcome of non-traumatic coma in two secondary referral centers in Cameroon. Methods: A 5-year retrospective hospital-based study was carried out, where medical records of patients aged 18 and above who presented with coma, not associated with physical assault, and were admitted to the medical wards were included. Bivariate and multivariate analyses were employed to identify associations and independent predictors, respectively. Survival rates were analyzed using the Kaplan-Meier estimator. Results: A total of 191 patient files were recruited out of 9581, accounting for 2% of admissions. Diseases with No Focal Neurological Deficits (DNFND) were the leading clinical presentation (47.1%). Stroke was the most frequent aetiology of NTC at 31.7% (60/189). The in-hospital mortality of NTC was 84.8% (n = 162/191). 9.9% (n = 19/191) of patients recovered completely. Factors associated with death were: GCS ≤ 5, stroke, hypertension, a length of hospitalization of ≤3 days, and the female gender. A length of hospitalization of ≤3 days, a GCS of ≤5, and HIV seropositivity appeared to be predictors of death. Being HIV seronegative or presenting with a GCS ≥ 6 conferred a 1.8 times higher odds of survival during admission. Conclusion: NTC is associated with higher mortality rates in our setting. It is therefore paramount to adopt a more clinical approach to diagnose the etiology of NTC, given the limitations in establishing diagnoses in these settings. We have therefore proposed an algorithm which best suits these settings.