The research, conducted at Hospital Alberto Barton Thompson in Callao and published in Frontiers in Nephrology, set out to answer a deceptively simple question: when a bedside exam, a bioimpedance scan, and a lung ultrasound are all performed on the same patient on the same morning, do they agree? The answer, quantified with statistical rigor, was largely no. The three methods captured overlapping but partially distinct slices of a patient’s true fluid status, and the disagreements were not random noise but appeared to reflect genuinely different physiology.

The technical centerpiece of the study is the B-line, a vertical, laser-like reverberation artifact that streaks down an ultrasound image from the pleural line and moves in synchrony with lung sliding. B-lines increase when fluid accumulates in the interlobular septa and alveoli of the lung, making their count a practical surrogate for extravascular lung water. The team used an abbreviated six-zone protocol, scanning two intercostal spaces along the midclavicular line and one along the midaxillary line on each side, a compromise that previous validation work has shown performs comparably to the traditional 28-zone survey while cutting acquisition time dramatically. Two formally trained nephrologists performed all scans with a handheld Butterfly iQ+ device, blinded to every other clinical and bioimpedance finding, and adjudicated disagreements by consensus.

Alongside the ultrasound, each patient underwent whole-body bioimpedance spectroscopy using a Fresenius Body Composition Monitor, which estimates how many liters of excess extracellular fluid a person carries relative to their normohydrated state. The investigators also performed a standardized clinical examination, recording dyspnea, crackles, peripheral edema, blood pressure, and interdialytic weight gain, and defined clinical euvolemia as the absence of all congestion signs. All measurements were taken immediately before the same midweek dialysis session, eliminating the temporal drift that has muddied earlier comparisons.

The headline result is striking. Of the 49 patients judged clinically euvolemic, 27, or 55.1 percent, had three or more B-lines on lung ultrasound, meeting the prespecified definition of subclinical pulmonary congestion. That subgroup alone constituted 23.3 percent of the entire cohort. In other words, nearly one in four dialysis patients in this study was walking around with objectively wet lungs despite a clean physical examination, no edema, and no crackles. Overall, 70.7 percent of participants had ultrasound-defined pulmonary congestion, and 41.4 percent had moderate-to-severe congestion with ten or more B-lines.

Agreement statistics drove the point home. Clinical examination showed only fair agreement with lung ultrasound at the three-B-line threshold, with a Cohen’s kappa of 0.28, improving to a still-modest 0.41 at the ten-B-line cutoff. Agreement between lung ultrasound and bioimpedance-defined overhydration was slight, at kappa 0.16, and the ultrasound agreed barely at all with elevated interdialytic weight gain, at kappa 0.08. Even the two objective technologies, which might naively be expected to converge, classified the same patients differently. The authors argue this physiological discordance is the story itself: fluid overload describes expansion of the body’s extracellular compartment, whereas congestion describes where that fluid actually accumulates and redistributes, and the two do not move in lockstep.

By combining clinical and ultrasound findings, the team carved the cohort into four volume phenotypes. Twenty-two patients were concordantly euvolemic, 55 were concordantly congested, 12 showed clinical signs without ultrasound congestion, and 27 occupied the discordant zone of subclinical pulmonary congestion. The phenotypes sorted cleanly along physiological gradients: concordantly congested patients had the highest weight gain, the greatest bioimpedance overhydration, and the heaviest B-line burden, while concordantly euvolemic patients sat at the bottom of every measure. The subclinical group fell in between, with intermediate overhydration that a stethoscope simply could not see.

Regression analysis sharpened the clinical message. Among euvolemic patients, each additional liter of bioimpedance-derived overhydration was associated with a 24 percent higher prevalence of subclinical pulmonary congestion, while interdialytic weight gain, the traditional dialysis-unit benchmark, lost all statistical significance after adjustment. Male sex was independently associated with a 45 percent lower prevalence of occult congestion. Blood pressure told a similarly bifurcated tale: systolic pressure correlated moderately with systemic overhydration but not with B-lines, whereas diastolic pressure showed only a weak link to lung water. These patterns reinforce the idea that whole-body fluid volume, weight fluctuation, and pulmonary congestion are separable physiological dimensions rather than interchangeable labels.

The findings resonate with a growing international literature. Meta-analyses report pulmonary congestion in roughly one half to two thirds of maintenance hemodialysis patients worldwide, and the landmark LUST study found that pulmonary crackles, the classic auscultatory sign, detected significant congestion with a sensitivity of only about 22 percent. Studies from Asia and Europe have likewise shown that patients deemed normohydrated by bioimpedance frequently carry ultrasound-visible lung water. The Peruvian cohort, notable as one of the first Latin American populations examined with this triple assessment, adds weight to the argument that clinically silent congestion is a reproducible global phenomenon rather than a local artifact.

What the study cannot yet say is whether acting on this hidden information saves lives. Its cross-sectional design precludes causal inference, it came from a single center where most patients dialyzed through central venous catheters, and echocardiography was available for only about half the cohort. The three-B-line threshold for subclinical congestion was prespecified but not universally validated, though sensitivity analyses at five and ten B-lines produced similar patterns. Still, the clinical direction is clear. The authors do not propose replacing the hands and stethoscope of conventional assessment, but they make a strong case for layering lung ultrasound and bioimpedance on top of it when fluid status is uncertain. With randomized trials of ultrasound-guided fluid management already underway, the era in which a wet lung could hide in plain sight on a dialysis chair may soon be ending.

Subject of Research: Discordance among clinical examination, bioimpedance spectroscopy, and lung ultrasound for assessing fluid status and pulmonary congestion in maintenance hemodialysis patients

Article Title: Physiological discordance between clinical examination, bioimpedance spectroscopy, and lung ultrasound in maintenance hemodialysis patients without heart failure

Article References: Benites-Flores, I. R., Rodríguez Revilla, J., Taype-Rondan, A., Fernández-Anticona, L. C., Prado Ramírez, P. A., Seminario Alegre, N., Rodríguez Tudero, C., Moral Berrio, E., Jimenez Mayor, E., Chandu Nanwani, A., & De la Flor, J. C. (2026). Physiological discordance between clinical examination, bioimpedance spectroscopy, and lung ultrasound in maintenance hemodialysis patients without heart failure. Frontiers in Nephrology, 6, Article 2000700. https://doi.org/10.3389/fneph.2026.2000700

Image Credits: AI Generated

DOI: 10.3389/fneph.2026.2000700

Keywords: hemodialysis, lung ultrasound, bioimpedance spectroscopy, pulmonary congestion, B-lines, volume assessment, fluid overload, point-of-care ultrasound, nephrology, subclinical congestion, dry weight, cardiovascular risk