Integrated Diagnostic Approaches and Clinical Predictors in Central Nervous System Infections: Evidence from Resource-Limited Settings in Vietnam

Abstract

This doctoral thesis focuses on improving the diagnosis and management of central nervous system (CNS) infections in resource-limited settings, using Vietnam as a case study. CNS infections are often life-threatening, and their diagnosis is challenging due to overlapping symptoms, prior antibiotic use, and limited access to advanced diagnostics. The thesis comprises three chapters, each addressing a key aspect of this clinical problem.

Chapter 1 evaluated the BioFire FilmArray Meningitis/Encephalitis (FAME) panel, a rapid molecular test that detects 14 common pathogens in cerebrospinal fluid (CSF). The study demonstrated that FAME significantly improved diagnostic speed and yield compared to conventional methods, especially in patients who had already received antibiotics. However, it also revealed limitations such as false positives/negatives and the panel’s inability to detect regionally important pathogens like Streptococcus suis and Mycobacterium tuberculosis. The findings highlight the need to adapt molecular panels to local epidemiology.

Chapter 2 explored the use of 16S Oxford Nanopore Technology (ONT) sequencing as a culture-independent diagnostic method for bacterial CNS infections. This approach successfully identified pathogens in a large number of culture-negative cases, including several rare and antibiotic-suppressed bacteria. It also demonstrated clinical impact by showing that many patients received inappropriate antibiotics initially, which could have been corrected earlier using ONT results. The chapter emphasizes ONT’s potential in guiding targeted therapy and improving antimicrobial stewardship in low-resource settings.

Chapter 3 analysed the impact of specific pathogens and clinical factors on patient outcomes. The most commonly identified organisms were Mycobacterium tuberculosis, Streptococcus suis, HSV-1, and Cryptococcus neoformans. Fungal and tuberculous meningitis were associated with the highest mortality. Two strong predictors of poor outcome were identified: delayed hospital presentation (≥5 days of symptoms) and altered mental status at admission. These findings can help clinicians recognize high-risk patients earlier and prioritize interventions accordingly.

Together, these chapters provide actionable evidence to improve diagnostic strategies, tailor treatment, and strengthen patient care for CNS infections in settings with limited resources.