While World Health Organization (WHO) guidelines recommend empirical antibacterial therapy as the standard of care in all African children with severe falciparum malaria, there are fewer data to guide the management of adults with the disease in low transmission settings.

Presently WHO guidelines do not recommend empirical antibacterial therapy in adults with malaria in low transmission settings, instead antibacterial therapy is only clearly recommended in those patients in whom a serious bacterial co-infection is clinically suspected. However, in a pilot study in Myanmar (High Frequency of Clinically Significant Bacteremia in Adults Hospitalized With Falciparum Malaria PMID: 26989752) we found that 13% of adults hospitalized with falciparum malaria were bacteremic, with bacterial co-infection suspected by clinicians in the minority. Patients with serious bacterial infection are commonly not bacteraemic and so this probably underestimates the frequency of significant bacterial co-infection.

In that pilot study, over 75% of patients received empirical antibacterial therapy on admission to hospital, which would not accord with published WHO guidelines as clinicians suspected bacterial co-infection in only 17%. However, the study's 100% survival rate - when over half of the patients were at high risk of death - suggests that the administration of antibacterial therapy may be appropriate until bacterial co-infection is excluded.

There is also academic debate about the role of co-morbidities in the presentation of patients severely ill with vivax malaria. Bacterial co-infection has been reported in some - but by no means all - studies of severe vivax infection. It would be useful to determine the relative contribution of bacterial co-infection to the clinical presentation of patients with vivax malaria.

By systematically seeking evidence of bacterial co-infection in all patients hospitalized at the study sites, this study aims to determine if the bacterial infection is really as prevalent as was the case in the pilot study. Accordingly it aims to determine the utility of a strategy that includes empirical antibacterial therapy in adults hospitalized with malaria in low transmission settings, until significant bactrila infection has been excluded.

The study's participants will be adults hospitalized with malaria at four tertiary referral hospitals in Myanmar. The study will be a collaboration between clinicians at Insein General Hospital North Okkalapa General Hospital and Thingangyun General Hospital in Yangon and Naypyidaw General Hospital in Naypyidaw, Myanmar and the Department of Global Health at Menzies School of Health Research.

The study's participants will be cared for by local doctors and nurses according to current WHO guidelines for the management of severe malaria. Management will include a systematic screen for bacterial infection with a detailed history, examination, radiological and laboratory studies, including the collection of blood cultures in all patients (this is recommended in WHO guidelines, but rarely performed in the resource poor setting). The only difference to the current WHO guidelines for the management of severe malaria will be the administration of empirical antibacterial therapy to all participants with falciparum malaria on admission. Participants with vivax malaria - who have a much lower incidence of complicated disease - would be managed according to the existing WHO guidelines, which suggest that antibacterial therapy should only be commenced in the presence of convincing evidence of severe bacterial co-infection.

After informed consent is obtained, artesunate is administered and a detailed work up for bacterial co-infection is performed, all participants with falciparum malaria will receive empiric levofloxacin. This will continue until 48 hours when, if the patient is improving and the work up for infection is negative, antibiotics will be ceased. Conversely if cultures are positive, targeted antibacterial therapy will be commenced as soon as sensitivity data are available. In the setting of convincing clinical evidence of bacterial infection when cultures are negative (a patient with a chest x-ray consistent with pneumonia but in whom cultures are negative for example), appropriate antibacterial will be continued. Participants who deteriorate in the first 48 hours will have their antibacterial therapy escalated to vancomycin and meropenem (based on microbiological findings in our pilot study).

Participants hospitalized with vivax malaria will also receive parenteral antimalarial therapy and have a detailed work up for infection, but empirical antibacterial therapy will not be commenced. Study clinicians may start antibacterial therapy in the setting of positive cultures or unequivocal evidence of bacterial infection if cultures are negative.

The subsequent clinical course of the participants and the frequency of confirmed bacterial co-infection would be compared to historical controls at the study sites to determine the utility of empirical antibacterial therapy in adults hospitalized with malaria in Myanmar. These data would be generalizable to other low transmission settings potentially.

The plan is to continue the study for 3 years; we expect that this will allow sufficient time to enrol enough patients to answer the study questions. The data will be analysed annually by independent safety monitors (from the University of Medicine 2, Myanmar and The Kirby Institute, Australia) and the study terminated if clear benefit or harm is identified. Independent safety monitors will also follow the study locally, review all deaths and be available to address any concerns of the participants and their families.

Inclusion Criteria:

• Admitted to hospital
• Positive film for asexual forms Plasmodium falciparum or Plasmodium vivax or positive Rapid Diagnostic Test for either pathogen if blood film not immediately available (with confirmation on blood film as soon as possible).
• Informed consent

Exclusion Criteria:

• Age less than 16
• Pregnancy
• Patient's family declines consent

Adults hospitalized with malaria at the four study sites

The data will be stored for a period of five years after any resulting publication. Researchers who apply and receive approval from the Human Research Ethics Committees of the Menzies School of Health Research and the University of Medicine 2, Yangon will be able to review the data.

The data will be available from October 2016 for 5 years.

Researchers who apply and receive approval from the Human Research Ethics Committees of the Menzies School of Health Research and the University of Medicine 2, Yangon will be able to review the data.