Diagnostic Utility of Xpert MTB/RIF Ultra in Abdominal Tuberculosis: A Prospective Observational Dataset
Description
This dataset provides diagnostic data collected during a prospective observational study aimed at evaluating the diagnostic accuracy of the Xpert MTB/RIF Ultra assay for abdominal tuberculosis (TB). The study was conducted at Sir Ganga Ram Hospital, Delhi, India, between September 2019 and March 2021. A total of 144 patients with clinico-radiological features of abdominal TB were enrolled. Key findings across demographic, clinical, radiological, endoscopic, microbiological, and histopathological domains were systematically recorded. Patient specimens (lymph node aspirates, tissues, pus, other aspirates, body fluids) were analysed using Xpert MTB/RIF Ultra, direct fluorescence microscopy, liquid culture (BD MGIT 960 system) with backup solid culture (Löwenstein-Jensen medium), and histopathology, as applicable. Key variables in the dataset include: • Patients’ age and gender distribution • Sites of sample collection • Xpert MTB/RIF Ultra results (including rifampicin resistance status) • Phenotypic and genotypic drug susceptibility testing (DST) results • Smear and culture results • Histopathology results • Results of other diagnostic modalities (i.e., radiological, endoscopic, biochemical) • Treatment response This dataset allows for the analysis of Xpert MTB/RIF Ultra's diagnostic performance using sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), positive likelihood ratio, negative likelihood ratio, and overall accuracy, in comparison with TB culture and a composite reference standard (CRS). CRS was defined as either a positive TB culture, or positive results in at least two of the following modalities: imaging or endoscopy suggestive of TB, elevated ascitic fluid adenosine deaminase (ADA) levels (>39 IU/mL), a positive acid-fast bacilli (AFB) smear, presence of granulomas on histopathology, or clinical with radiological and endoscopic improvement after 3 months of empirical anti-tubercular therapy. It is intended to facilitate secondary analysis, comparison with other diagnostic assays, and meta-analytical reviews. Applications: • Diagnostic performance assessment • Comparative studies of TB diagnostics • Educational resource for medical microbiology or infectious disease training All patient data were anonymised in compliance with ethical standards and institutional ethics committee.
Files
Steps to reproduce
1. Study setup and participant enrolment Conduct a prospective observational study at a tertiary care hospital with dedicated departments of Microbiology and Gastroenterology. Enrol adult patients with clinico-radiological features of abdominal TB. Obtain written informed consent and ensure approval from the institutional ethics committee 2. Sample collection Collect clinical samples such as: lymph node aspirates, tissues, pus, aspirates, and body fluids 3. Sample processing Add the sample reagent to the prepared sample in a 2:1 ratio as per Cepheid’s Xpert Ultra protocol 4. Xpert MTB/RIF Ultra testing: Use the GeneXpert IV system (Cepheid) and Xpert MTB/RIF Ultra cartridges Load the processed specimen into the cartridge and run the test Record automated output: Mycobacterium tuberculosis (MTB) Detected/Not Detected/Trace; Rifampicin (RIF) Resistance Detected/Not Detected/Indeterminate 5. Reference standard and additional tests Microbiology: Culture- inoculate samples into Mycobacteria Growth Indicator Tubes (MGIT) and place in the BD MGIT 960 system and monitor for up to 6 weeks; put backup solid cultures on LJ media AFB smear- perform auramine-rhodamine staining on all the samples and examine with fluorescence microscope Histopathology: Hematoxylin & Eosin (H&E) stain- to identify caseating granulomas or necrosis Ziehl-Neelsen (ZN) stain- to detect AFB Define CRS using: Culture, imaging, endoscopy, ascitic fluid ADA level, AFB smear, histopathology, and treatment response 6. Data management Enter data into Excel spreadsheet Variables should include: Patient identifier, age, gender, sample type, radiological/endoscopy/biochemistry results, Xpert result, DST results, AFB smear result, culture result, histopathology result, CRS diagnosis, treatment response 7. Statistical analysis Use IBM SPSS Statistics (version 28 or later) Calculate the following test evaluation indices: Sensitivity, specificity, PPV, NPV, positive likelihood ratio, negative likelihood ratio, and accuracy Generate Receiver Operating Characteristic (ROC) curves and calculate Area Under the Curves (AUCs) 8. Reagents and consumables Xpert Ultra cartridges (Cepheid, USA) Mycobacteria Growth Indicator Tubes and PANTA (Polymyxin B, Amphotericin B, Nalidixic acid, Trimethoprim, Azlocillin) supplement (BD, USA) GenoType MTBDRplus reagents and controls (Hain Lifescience, Germany) Auramine-rhodamine, H&E and ZN stain reagents Sterile containers, syringes, personal protective equipment 9. Software and tools Cepheid GeneXpert software (preinstalled) IBM SPSS Statistics for analysis Microsoft Excel for data entry and management 10. Quality control and ethics Calibrate and maintain equipment as per manufacturer recommendations Follow biosafety level 2 protocols for TB sample handling De-identify patient data before analysis or sharing
Institutions
- Sir Ganga Ram Hospital