Abstract
This protocol outlines a planned systematic review and meta-analysis designed to provide a head-to-head comparison of the diagnostic accuracy of Xpert MTB/RIF Ultra versus the original Xpert MTB/RIF for detecting tuberculous pleurisy.
Introduction: The Diagnostic Challenge of Tuberculous Pleurisy
Tuberculous pleurisy (TBP), an infection of the membrane surrounding the lungs, is the most common form of extrapulmonary tuberculosis. Despite its prevalence, it is notoriously difficult to diagnose early. The bacterial load in pleural fluid is often very low (paucibacillary), meaning traditional cultures and older molecular tests frequently return false negatives.
While the Xpert MTB/RIF Ultra (Ultra) test was designed with a lower limit of detection to catch these "hidden" cases, its actual performance compared to the original Xpert MTB/RIF specifically for TBP remains a subject of debate. This protocol establishes a rigorous "battle plan" to determine if the newer Ultra version offers the sensitivity boost needed to overcome the diagnostic challenges of this infection.
Methodology
As this is a protocol, the methods describe the statistical and procedural rigor the authors will apply to their upcoming review to prevent "cherry-picking" data once the analysis begins.
Search Strategy and Inclusion Criteria
The research team will perform a comprehensive search across five major databases: PubMed, Embase, the Cochrane Library, and two major Chinese databases (CNKI and Wanfang). The study is specifically looking for "head-to-head" studies—those that tested the same patient samples with both versions of the Xpert technology. This approach reduces the "noise" caused by comparing different patient populations across different studies.
To be included, studies must:
- Be published in English or Chinese.
- Have a sample size of at least 10 specimens.
- Provide enough raw data to reconstruct a "2x2 table" (True Positives, False Positives, False Negatives, True Negatives).
Statistical Modeling and Quality Control
The authors will use bivariate random-effects models to calculate pooled sensitivity and specificity. This is a standard but sophisticated approach in diagnostic meta-analyses that accounts for the inherent trade-off between sensitivity and specificity.
Heterogeneity will be assessed using $I^2$ statistics and subgroup analyses focusing on variables like sample preparation (e.g., homogenization or decontamination methods) and patient selection. Study quality will be evaluated using the QUADAS-2 tool, a specialized framework for evaluating the risk of bias in diagnostic accuracy studies.
Key Insights and Expected Impact
- The "Head-to-Head" Focus: By focusing on studies that ran both tests on the same samples, the authors aim to eliminate confounding variables like local TB prevalence or laboratory infrastructure.
- Addressing the "Paucibacillary" Challenge: Higher sensitivity in pulmonary TB (sputum) does not always translate perfectly to extrapulmonary fluids; this study aims to prove if that translation holds true for the pleural space.
- Standardizing the "Gold Standard": The study will measure the Xpert tests against a composite reference standard (CRS) or culture, acknowledging that no single test is currently perfect for TBP.
Limitations and Future Work
The authors acknowledge several constraints, including potential language bias (as the search is limited to English and Chinese) and data scarcity; the protocol requires at least four published studies to perform certain subgroup analyses. Once completed, this study will likely inform World Health Organization (WHO) guidelines and local clinical protocols for the management of extrapulmonary tuberculosis.
Summary updated