Six lung zones carried most of the diagnostic signal for pulmonary tuberculosis triage
A compact point-of-care ultrasound strategy identified pulmonary tuberculosis with nearly the same discrimination as much larger scanning schemes in a Ugandan cohort, a finding that could ease deployment in busy clinics where time, training, and patient throughput all limit imaging. Investigators compared an 18-position examination with smaller symmetric lung-zone protocols and found that a fixed 6-zone approach preserved most of the diagnostic performance seen with 12-zone and 18-position scans. The practical appeal is immediate: a shorter exam is easier to teach, faster to perform, and better suited to triage settings where TB suspicion must be resolved quickly.
The study enrolled Ugandan participants age 13 years and older with confirmed pulmonary TB, household TB exposure, or a non-TB respiratory condition. Among 318 analyzed participants, 118 had pulmonary TB and 200 were TB-negative, including 102 with other respiratory conditions. The design reflects a real-world screening problem rather than a narrow laboratory sample, because the key challenge in TB triage is not simply identifying obvious disease but separating TB from other respiratory illness that can produce overlapping symptoms and similar chest findings.
A fixed 6-zone protocol held up against broader scans
The model-selected 6-zone protocol sampled bilateral anterior-superior, lateral-superior, and posterior-superior lung zones. For TB versus all TB-negative participants, the area under the curve was 0.82 for the 6-zone protocol, 0.81 for the 12-zone protocol, and 0.81 for the 18-position protocol. For TB versus other respiratory conditions, the corresponding values were 0.81, 0.85, and 0.85. Those differences did not reach statistical significance in the larger comparisons, while the 2-zone protocol performed worse than the 6-zone protocol when TB was compared with all TB-negative participants.
One sentence captures the central result: the shortest exam was not the weakest one, but the two-zone version was too sparse to carry the same diagnostic weight.
This pattern matters because ultrasound findings in TB tend to be patchy rather than uniformly distributed across the thorax. A protocol that deliberately samples symmetric upper lung regions can still capture the combination of peripheral inflammation, pleural irregularity, and subpleural pathology that often accompanies pulmonary infection. In low-resource settings, a shorter protocol can also reduce operator fatigue and improve reproducibility, both of which influence the reliability of point-of-care imaging far more than clinicians sometimes acknowledge.
Shred sign, consolidation, and irregular pleura anchored the TB signal
Across the prespecified findings, shred sign, consolidation, and B-lines with irregular pleura emerged as the strongest TB-associated abnormalities. Those features fit the biology of pulmonary TB. Subpleural inflammation and parenchymal destruction can create a fragmented interface between aerated and diseased lung, producing irregular pleural contours and the mixed echogenic patterns that ultrasound renders as B-lines, consolidation, and shred-like margins. In bedside terms, these are the sorts of abnormalities that move a lung scan from a vague screen toward a clinically actionable triage tool.
The strength of the signal also helps explain why a limited scan can work. TB does not need a full chest survey to announce itself when the disease involves the upper and peripheral lung zones that ultrasound accesses most efficiently. A broader sweep may add detail, but the core discriminatory findings were already present within the streamlined protocol. For clinicians working in high-burden environments, that could mean less time scanning and more time making treatment and referral decisions.
Why fewer zones may be enough in high-burden clinics
Point-of-care ultrasound has long had appeal in TB care because it avoids ionizing radiation, can be deployed at the bedside, and does not depend on fixed imaging infrastructure. Yet it has lacked a standardized acquisition framework for pulmonary TB, which has slowed adoption. A symmetric, fixed 6-zone protocol addresses one of the field’s most practical problems: when scan technique varies widely from one operator to another, diagnostic performance becomes difficult to compare across sites or scale across programs.
The present findings support a simpler acquisition model without sacrificing much discrimination. That makes the protocol attractive for triage, where the goal is not to replace microbiologic confirmation but to identify patients who need further testing, isolation, or expedited evaluation. In a setting where many patients present with cough and other respiratory symptoms, a tool that can quickly separate likely TB from alternative illness has value even when it is not definitive on its own.
What the cohort can and cannot prove
The analysis used a fixed 75/25 training-holdout split and tested the best protocol of each size in the holdout set, which strengthens the internal logic of the comparison. Even so, this remains a preprint and has not undergone peer review. The population also came from a high-burden setting, so the performance estimates may not translate directly to lower-prevalence environments or to sites with different operator experience, patient mix, or competing lung diseases.
The sample size was sufficient to compare protocols, but not large enough to settle every implementation question. The study does not show whether the 6-zone protocol would perform as well in the hands of minimally trained clinicians, whether it would retain accuracy across different ultrasound devices, or how it would behave in patients with advanced HIV-associated lung disease, prior TB, or other structural lung abnormalities. Those issues matter because TB ultrasound is ultimately an operational test, not just a statistical one.
Even with those caveats, the signal is strong enough to influence how future TB ultrasound work is designed. The field no longer needs to assume that more scanning automatically produces better triage. In this cohort, a carefully chosen 6-zone exam preserved the main diagnostic information, and the most clinically relevant question now is how easily that streamlined protocol can be taught, reproduced, and embedded into routine TB evaluation without losing the accuracy seen in this study.