C-reactive protein testing to discriminate bacterial from viral pathogens among febrile patients in Southeast Asia and its impact on antibiotic prescribing in primary care

Thesis

In an era marked by the worldwide decline of malaria, fever remains the most frequent reason for seeking health care in the tropics. Non-malarial pathogens are mainly studied among hospitalised patients, while evidence in primary care is scarce. In addition to the limited evidence on the local epidemiology, prescribers routinely face difficulty in identifying patients who really need an antibiotic. This translates into high antibiotic prescription, fuelling the worldwide spread of antimicrobial resistance. C-reactive protein (CRP) is one of the most studied host-response biomarkers for guiding the management of acutely febrile patients, although evidence mainly originates from hospitals in high-income countries, implying that findings may not be applicable in primary care in low-to-middle-income countries (LMICs). My thesis aims to explore the bacteria and viruses driving acute fever in primary care in Southeast Asia, as well as the potential utility of CRP in identifying patients who need an antibiotic. To this end, we designed an individually randomised controlled trial of CRP-guided treatment in all febrile children and adults, ‘the CRP Study’. We also collected blood and respiratory specimens to ascertain the causes of fever in these patients. We found a significant antibiotic reduction of 7.8 percentage-points using CRP at a 40 mg/L threshold compared to routine practice, without differences in clinical outcome. Routine prescription levels were lower than prior to study implementation, suggesting that future studies should adopt a cluster rather than individual randomised design. Influenza virus type A, dengue virus and respiratory syncytial virus were the key pathogens among our primary care patients, although ascertaining causality was challenging due to lack of acontrol group, especially in nasopharyngeal swabs. Most bacterial infections showed low inflammatory levels and recovered without antibiotics, questioning the interpretation of microbiological findings among non-severe patients. CRP performance in distinguishing bacterial from viral pathogens was limited, suggesting that CRP may rather discriminate self-limiting from serious infections. We also evaluated the performance of CRP in a cohort of febrile patients in Tanzania. Here CRP showed a high sensitivity for detecting bacterial bloodstream infection, indicating that prospective evaluations on its effect on antibiotic prescriptions could be envisaged. The detection of bacterial zoonotic pathogens was of lower accuracy, warranting further evaluations for various levels of healthcare settings and age categories.

Authors: Althaus, T

• DOI: 10.5287/ora-zbajdg2yq
• Type of award: DPhil
• Level of award: Doctoral
• Awarding institution: University of Oxford
• Language: English
• Keywords: Antimicrobial resistance; Primary care; Southeast Asia; C-reactive protein; Causes of fever; Antibiotic prescription; sub-Saharan Africa; Fever
• Subjects: Antimicrobial resistance; Point-of-care diagnostics; Primary care; Infectious diseases; Lo-to-middle-income countries