Project Lead(s): Yoel Lubell
Issue
Every day, over 10,000 deaths occur in the tropics, due to bacterial, viral or malarial infection for which the commonest presentation is undifferentiated fever.
Healthcare systems in developing countries lack the diagnostic capacity to differentiate causes of fever, leading to over-treatment with antimicrobials, a probable driver of drug resistance.
While widespread deployment of rapid diagnostic tests for malaria in endemic areas has revolutionized the management of fever, it has also raised new challenges with the rising proportion of fevers with a negative malaria test result. In some instances, these patients require an antibiotic but they are not identified as such, while in many more cases antibiotics are given unnecessarily.
One approach to improve the management of fevers would be to use biomarkers of bacterial infection to identify patients who should be treated with an antibiotic. A combined malaria rapid test with a test for a biomarker of bacterial infection might offer health workers in the rural tropics with a simple, affordable and practical approach to guide antimicrobial treatment.
Solution
The aim of this project was to estimate the predictive value of procalcitonin (PCT) and C-Reactive Protein (CRP) levels in stored serum samples from patients with undifferentiated fever, to determine the need for antibiotic treatment. PCT and CRP are established biomarkers for bacterial infections in patients with sepsis and are already used in many high-income countries, although there is very limited data in the context of tropical infections. The findings of this study could help guide the development of a combined malaria test with a biomarker test for bacterial infection.
The project team obtained more than 2,000 serum samples from three fever studies in Cambodia, Laos and Thailand, where extensive laboratory investigations were carried out in febrile patients to assign a microbiologically confirmed cause of illness. A mini-VIDAS system and reagents were acquired to run the PCT assays and a NycoCard Axis Shield Reader was used for the CRP assays.
Outcome
The research showed that both PCT and CRP biomarkers are effective in discriminating between viral and bacterial infections in the tropics, and could therefore potentially inform the management of fever in these settings. The team published a paper in November 2015 reporting that, in contrast to many studies in high-income settings, CRP out-performed PCT in terms of accuracy, with an under-the-curve value of 0.83 for CRP compared to 0.74 for PCT. The accuracy and stability of existing CRP rapid tests were evaluated in a real-life setting, making use of a fever study in Laos, resulting in a second peer-reviewed paper in early 2016. A third paper using the Phase I data to feed into a cost-effectiveness model explored the potential impact of CRP testing if it were to be enrolled in routine care.
In order to assess the potential impact of CRP testing to improve health outcomes and safely reduce antibiotic consumption in a real-life setting, the team is now launching a clinical trial in health facilities in Thailand and Myanmar to evaluate the impact of CRP testing in febrile patients attending primary care facilities. The team will be applying for Phase II Transition To Scale funding to extend this study to Laos, and to develop a combined malaria/CRP test that can be used by village health workers in remote, malaria-endemic settings. If successful in these steps, the team will carry out a final evaluation of the use of the combined malaria/CRP test in the hands of village health workers in Cambodia, Laos and Myanmar.