Write a comment

Summary

This study evaluates the diagnostic accuracy and resource efficiency of pooled sputum testing using the Xpert MTB/RIF Ultra assay for tuberculosis detection among adults in Vietnam. The authors use a cross-sectional, multi-site design with 2,396 participants and compare individual Xpert Ultra testing, pooled testing, and liquid culture as the reference standard. The inclusion of both facility-based and community-based case finding allows for comparison across different epidemiologic contexts.

The results show that pooled testing reduces cartridge use by 46.5 percent and lowers costs while maintaining high specificity. However, sensitivity is lower than individual testing overall, 82.4 percent compared to 86.5 percent, and drops substantially in community-based settings to 59.1 percent. The authors conclude that pooled testing may be useful in certain contexts but should be implemented carefully due to the risk of missed cases.

Strengths

This study has several important strengths. The large sample size and multi-site design improve the reliability of the findings and make them more generalizable across different screening settings. Including both facility-based and community-based case finding is especially valuable, since it allows the authors to show how test performance changes depending on TB prevalence and population characteristics, which is highly relevant for real-world implementation.

Another major strength is the use of three testing approaches, individual Xpert Ultra, pooled Xpert Ultra, and liquid culture as the reference standard. This design provides a strong framework for evaluating diagnostic accuracy and allows for a clear, direct comparison between strategies. It also strengthens confidence in the results, since culture serves as a well-established reference standard.

The study also stands out in how it captures variation across key factors such as setting, geography, and bacillary load. This level of detail makes the findings more informative and useful for policy decisions, rather than just presenting overall averages.

Finally, the authors present a balanced and careful interpretation of their results. They clearly acknowledge the tradeoff between efficiency and sensitivity and avoid overstating the benefits of pooled testing. The discussion reflects a good understanding of the practical implications and highlights the importance of context-specific implementation, which strengthens the overall credibility of the study.

Major Concern

The main limitation of this study is that the economic analysis focuses on cartridge savings and direct cost reductions without considering the broader clinical and public health consequences of reduced sensitivity.

Although pooled testing reduces resource use, it also misses a meaningful proportion of true TB cases. The study shows that approximately 11.4 percent of TB cases would be missed under a pooled testing strategy. These missed diagnoses are more common in community-based settings and among individuals with low bacterial load, where sensitivity drops substantially.

These missed cases could lead to delayed treatment, continued transmission, and worse health outcomes. Because these downstream effects are not included in the economic analysis, the reported cost savings may not reflect the true impact at the program level.

In addition, the study shows clear geographic variation in sensitivity. Pooled testing sensitivity is 70.1 percent in Hanoi compared to 85.7 percent in Ho Chi Minh City. This suggests that a strategy that appears effective in one location may perform much worse in another, even within the same country. This point is important but is not fully explored in the discussion, particularly in terms of how it should influence implementation decisions.

Minor Concern

Another limitation is that the study evaluates a simulated pooled testing strategy rather than implementing it in real clinical settings.

While the simulation is carefully done, it may not capture practical challenges such as sample handling, delays in processing, or how clinicians would use the results. The results section is also quite dense, with many comparisons across different groups, which makes it difficult at times to clearly follow the main findings.

Suggestions for Improvement

• Expand the economic analysis to include the impact of missed TB cases, including delayed diagnosis and transmission

• More clearly distinguish between resource savings and overall cost effectiveness

• Discuss geographic variation in sensitivity in more detail and explain how programs should adapt strategies across different settings

• Clarify how the pooled testing strategy would function in real clinical practice and what challenges might arise

Overall Assessment

This is a well-conducted and policy-relevant study that provides useful evidence on the tradeoff between diagnostic accuracy and efficiency in TB testing. I especially appreciated the use of three testing approaches, individual Xpert Ultra, pooled testing, and liquid culture as the reference, which strengthens the comparisons and makes the findings more convincing. The results are particularly relevant for low- and middle-income settings where resources are limited. Addressing the limitations around cost interpretation and real-world implementation would further strengthen the study and improve its usefulness for decision-making.

Competing interests

The author declares that they have no competing interests.

Use of Artificial Intelligence (AI)

The author declares that they did not use generative AI to come up with new ideas for their review.