PCR-confirmed malaria among children presenting with a decreased level of consciousness in Angola: a prospective, observational study

Background In malaria-endemic areas, children presenting to hospitals with a decreased level of consciousness remain a diagnostic dilemma. The definition of cerebral malaria in a comatose child demands exclusion of other possible reasons, which requires in-depth investigations that are not easily available. The aim of this study was to investigate the frequency and clinical characteristics of PCR-confirmed malaria in a cohort of children with a decreased level of consciousness, look for potential features that would aid in differentiating children with malaria from those without, and assess the performance of traditional thick film microscopy against the cytb-qPCR-method. Methods A total of 345 children aged 30 days–15 years old, presenting to Hospital Pediátrico David Bernardino in Luanda, Angola, with a decreased level of consciousness (Glasgow coma scale score < 15) were prospectively enrolled during 2014–2017. Malaria was defined as a positive cytb-qPCR result on any occasion in hospital. The clinical course and laboratory parameters were compared between children with malaria and those without. The performance of thick film microscopy was analysed against the PCR method. Results 161 of 345 children (46.7%) had a positive malaria PCR test result. All cases were Plasmodium falciparum species, and 82.6% (133/161) fulfilled the WHO criteria for severe malaria. Overall, children with malaria presented to hospital with a shorter duration of symptoms and less convulsions pre-admission compared to those without malaria. The median GCS score on admission was 8, which did not differ between children with or without malaria. Clinical findings on admission were mostly similar across the whole cohort, but an infection focus outside the central nervous system was more common in malaria-negative children. Moreover, severe anaemia, thrombocytopenia, and high CRP levels occurred more frequently in children with malaria. The case fatality ratio was 28.5% (91/319) and did not differ between parasitaemic children and those without malaria, although parasitaemic children died sooner after hospital admission. When neurological sequelae were also considered, a positive malaria test was associated with a better outcome. The performance of thick film microscopy against PCR yielded a sensitivity of 96.8% and a specificity of 82.7%. Conclusions In this cohort of children with a decreased consciousness, the frequent presence of a malarial infection could not be judged from the clinical findings on admission, but the combination of profound aneamia, thrombocytopenia, and a high CRP level increased the odds of a positive malaria test result. Mortality remained high regardless of etiology, but malaria infection associated with fewer neurological deficits at discharge. Thick film microscopy performed well compared to the cytb-qPCR method.