Tuberculosis Diagnosis

This study aims to address critical diagnostic and data gaps in tuberculosis (TB) care among pregnant and postpartum women in Guinea-Bissau, a high-burden, resource-limited setting. Recognising that current TB screening during antenatal care (ANC) relies largely on unstructured symptom questions, the study will integrate more systematic and innovative approaches into routine maternal health services. The project will implement the Bandim TBscore II as a structured triage tool to classify symptom severity and guide referrals. To strengthen diagnostic capacity, two novel tools will be evaluated: stool-based GeneXpert testing (a method recommended in children and explored here as a feasible alternative for pregnant women) and artificial intelligence-powered chest X-ray interpretation software designed to enhance TB detection where radiological expertise is lacking. The study will also generate comprehensive, population-based data on TB and TB infection (TBI) among pregnant, postpartum women and women of reproductive age in Guinea-Bissau. The results are intended to inform health policy, both locally and in high-income countries, by providing evidence to improve TB screening protocols and care for this vulnerable group. Ultimately, the study seeks to develop scalable strategies that can be replicated across low- and middle-income countries to advance maternal and child health and support global TB eradication efforts.

Tuberculosis (TB) remains a major global health challenge, affecting over 10 million people annually. Hong Kong carries an intermediate TB burden, with \~3,200 new cases reported yearly. Pulmonary TB (PTB), the most common form, presents diagnostic difficulties. Traditional methods like sputum smear and culture often fail in patients unable to produce adequate samples, necessitating bronchoscopy to collect bronchoalveolar lavage (BAL) for mycobacterial testing. These limitations pose risks for patients and strain healthcare systems. Bronchoscopy is invasive, resource-intensive, and may delay treatment-especially for elderly patients with comorbidities. Blood-based inflammatory markers lack diagnostic specificity. A rapid, non-invasive alternative is urgently needed. The investigators developed a plasma-based assay that detects Mycobacterium tuberculosis cell-free DNA (MTB cfDNA) in blood. This liquid biopsy leverages metagenomic sequencing and computational analysis to identify TB-specific genetic material while minimizing contamination. Preliminary data show excellent diagnostic performance, with area under the receiver operating characteristic curve values \>0.94 for TB pleurisy. The investigators propose a prospective clinical validation study comparing plasma MTB cfDNA testing to bronchoscopy with BAL culture and molecular testing. The primary aim is to demonstrate non-inferiority of plasma cfDNA within a 10% sensitivity margin. Secondary aims include assessing how clinical and radiological features affect test performance and evaluating the assay's ability to detect drug resistance mutations for personalized therapy. Validation could transform TB diagnosis by offering a rapid, safe, and accurate blood test. Patients could avoid invasive procedures, receive faster diagnoses, and begin treatment sooner. Detecting resistance mutations directly from plasma would enable timely, targeted therapy-critical for addressing multidrug-resistant TB. This represents a paradigm shift toward precision medicine in TB care. Tailored to Hong Kong's epidemiological context, this study addresses a key diagnostic gap. The approach has global relevance, with potential to improve clinical outcomes, reduce costs, and accelerate progress toward WHO's TB elimination goals.

* Detection of sensitivity and specificity of microscopic examination and Molecular studies using polymerase chain reaction (PCR) Mycobacterium tuberculosis (MTB)/ non tuberculous mycobacterium (NTM) MDR-TB compared to culture which is the gold standard technique in the diagnosis of TB to reach the most reliable one for better health care service. * Assess the prevalence of tuberculosis cases among clinically suspected cases admitted to the hospital throughout the selected study period.