ABSTRACT
Background: Data collected in clinical trials have been used to develop scoring systems that identify adults with malaria at greatest risk of death. One of these, the RCAM score, can be simply determined by measuring a patient’s Glasgow Coma Score and respiratory rate on admission to hospital. However the safety of using the RCAM score to define high-risk patients has not been assessed outside of the clinical trial setting. Methods: A retrospective audit of medical records of all adults admitted with a diagnosis of malaria to two tertiary referral hospitals in Lower Myanmar in 2013 was undertaken. An RCAM score was calculated in all patients and related to their subsequent clinical course. Results: The recent decline in malaria hospitalizations at both sites continued in 2013. During the year 90 adults were hospitalized with malaria; 62 (69%) had Plasmodium falciparum monoinfection, 11 (12%) had Plasmodium vivax mono-infection, 17 (19%) had mixed infection. All seven (7.7%) deaths occurred in patients infected with P. falciparum. An admission RCAM score < 2 identified all the patients that would survive to discharge (positive predictive value (95% confidence interval (CI)) 100% (94.9-100%) and also predicted a requirement for less supportive care: 9/70 (13%) patients with an admission RCAM score < 2 required supportive care (blood transfusion, vasopressor support or oxygen supplementation) during their hospitalization compared with 12/20 (60%) patients with an admission RCAM score ≥ 2 (p < 0.0001). No patient with P. vivax monoinfectionrequired supportive care during their hospitalization. Patients with an oxygen saturation ≤ 95% on room air on admission were more likely to die before discharge (odds ratio 17.3 (95% CI: 2.9-101.2) than patients with a higher oxygen saturation (p = 0.002). Conclusions: Even outside a clinical trial setting the RCAM score reliably identifies adults with malaria who are at greatest risk of death and can be safely used in the initial triage and management of these patients.