Background
Severe malaria has a case fatality rate of 10-20 %; however, few studies have addressed the quality of severe malaria case management. This study evaluated the diagnostic and treatment practices of malaria patients admitted to inpatient health facilities (HF) in Malawi.

Methods
In July–August 2012, a nationwide, cross-sectional survey of severe malaria management was conducted in 36 HFs selected with equal probability from all eligible public sector HFs in Malawi. Patient records from all admissions during October 2011 and April 2012 (low and high season, respectively) were screened for an admission diagnosis of malaria or prescription of any anti-malarial. Eligible records were stratified by age (< 5 or ≥ 5 years). A maximum of eight records was randomly selected within each age and month stratum. Severe malaria was defined by admission diagnosis or documentation of at least one sign or symptom of severe malaria. Treatment with intravenous (IV) quinine or artesunate was considered correct. Patients without documentation of severe malaria were analysed as uncomplicated malaria patients; treatment with an artemisinin-based combination therapy (ACT) or oral quinine based on malaria test results was considered correct. All analyses accounted for HF level clustering and sampling weights.

Results
The analysis included 906 records from 35 HFs. Among these, 42 % (95 % confidence interval [CI] 35–49) had a severe malaria admission diagnosis and 50 % (95 % CI 44–57) had at least one severe malaria sign or symptom documented. Severe malaria patients defined by admission diagnosis (93, 95 % CI 86–99) were more likely to be treated correctly compared to patients defined by a severe sign (82, 95 % CI 75–89) (p < 0.0001). Among uncomplicated malaria patients, 26 % (95 % CI 18–35) were correctly treated and 53 % (95 % CI 42–64) were adequately treated with IV quinine alone or in combination with an ACT or oral quinine.

Conclusions
A majority of patients diagnosed with severe malaria received the recommended IV therapy in accordance with national treatment guidelines. However, the inconsistencies between diagnosis of severe malaria and documentation of severe signs and symptoms highlight the need to improve healthcare worker recognition and documentation of severe signs and symptoms.