The study enrolled patients admitted to a geriatric acute care unit, with a mean age of 82.2 years and a standard deviation of 6.8 years; 70 percent of the participants were female. Within 24 hours of admission, the attending physician responsible for each patient recorded a clinical impression of dehydration, classifying the patient as either dehydrated or not dehydrated, with the latter category combining both normal hydration and hyperhydration. This clinical judgment served as the reference standard against which the laboratory measure was tested. In parallel, the researchers calculated serum osmolarity from routine laboratory values, including sodium, potassium, glucose, and urea, and classified patients as having high osmolarity if the calculated value reached or exceeded 295 mOsmol/l, or normal osmolarity if it fell below that threshold.

The results revealed a striking mismatch between the two approaches. Clinically, only 23 percent of the patients were judged to be dehydrated by their attending physicians. However, when the researchers applied the calculated osmolarity threshold, 38 percent of the patients were classified as having high osmolarity, meaning the laboratory measure flagged dehydration in a substantially larger proportion of patients than the clinical assessment did. This discrepancy immediately suggested that the two methods were not measuring the same thing, or at least not with the precision that clinicians might expect.

To quantify the diagnostic performance of calculated serum osmolarity, the researchers calculated sensitivity, specificity, positive and negative predictive values, and overall accuracy. Using the 295 mOsmol/l cut-off, the sensitivity of calculated osmolarity was 48.6 percent, meaning that less than half of the patients who were clinically judged to be dehydrated were correctly identified by the laboratory measure. The specificity was 65.1 percent, indicating that roughly one-third of patients who were not clinically dehydrated were incorrectly flagged as having high osmolarity. The positive predictive value was just 29.7 percent, meaning that fewer than one in three patients with high calculated osmolarity were actually clinically dehydrated. The negative predictive value was 80.7 percent, and the overall accuracy was 61.3 percent.

Perhaps most telling was the measure of agreement between the two methods beyond what would be expected by chance alone. Cohen’s kappa, a statistical measure of agreement that corrects for chance concordance, was just 0.112, a value conventionally interpreted as poor agreement. In practical terms, this means that knowing a patient’s calculated serum osmolarity tells a clinician almost nothing about whether that patient has been clinically judged to be dehydrated, and vice versa. The two approaches appear to be capturing fundamentally different aspects of the patient’s hydration status.

The researchers went one step further and examined whether blood glucose levels might explain some of the discordance between the two methods. They compared glucose concentrations across groups of patients who were concordantly classified by both methods and those who were discordantly classified. The results showed that mean glucose was significantly higher in all groups classified as osmolarity-positive, including those patients who were classified as osmolarity-positive but not clinically dehydrated, the so-called false positives. All planned pairwise comparisons reached statistical significance with p-values below 0.001. This finding suggests that elevated glucose, which is a major contributor to calculated serum osmolarity, may inflate the laboratory value in patients who are not actually dehydrated, creating a systematic source of misclassification.

The biological explanation for this phenomenon is straightforward. Calculated serum osmolarity is derived from a formula that incorporates the concentrations of osmotically active solutes in the blood, primarily sodium, glucose, and urea. When glucose levels are elevated, as they frequently are in older hospitalized patients, whether due to diabetes, stress hyperglycemia, or other factors, the calculated osmolarity rises even if the patient’s actual water balance is normal. This means that the laboratory measure can be elevated for reasons that have nothing to do with dehydration, leading to false positive classifications. In a population of older hospitalized patients, where elevated glucose is common, this effect can be substantial.

The implications of these findings for clinical practice are significant. Dehydration in older adults is associated with increased morbidity, mortality, and healthcare costs, and accurate identification is essential for appropriate treatment. However, the study suggests that calculated serum osmolarity, at least when using the conventional 295 mOsmol/l threshold, should be interpreted with considerable caution in the acute care setting. A high calculated osmolarity does not reliably indicate that a patient is dehydrated, particularly when glucose levels are elevated. Conversely, a normal calculated osmolarity does not rule out clinical dehydration, given the low sensitivity of the measure.

The researchers conclude that their findings support cautious clinical interpretation of calculated serum osmolarity in acute care, particularly when glucose is elevated. They suggest that clinical impression, while subjective, remains an important component of dehydration assessment in older hospitalized patients, and that calculated osmolarity should not be used as a standalone diagnostic tool. The study was approved by the ethical committee at the Medical Faculty of Ruhr University Bochum and registered in the German Clinical Trials Register. The work was supported by the Open Access Publication Fund of the Ruhr-Universität Bochum, with open access funding enabled and organized by Projekt DEAL.

As the population ages and the number of older hospitalized patients continues to grow, the need for accurate and reliable methods of assessing hydration status becomes increasingly urgent. This study adds to a growing body of evidence suggesting that no single laboratory measure can replace careful clinical evaluation in this vulnerable population. The challenge for future research will be to develop and validate assessment tools that combine clinical judgment with laboratory and other objective measures in a way that improves diagnostic accuracy without adding unnecessary complexity to the care of older patients. Until then, the humble clinical impression of the experienced geriatrician remains an essential and, according to this study, arguably more reliable indicator of dehydration than the laboratory values that clinicians have long relied upon.

Subject of Research: Diagnostic performance of calculated serum osmolarity versus clinical impression for identifying dehydration in older hospitalized patients

Article Title: Diagnostic performance of calculated serum osmolarity versus clinical impression for identifying dehydration in older hospitalized patients

Article References: Pourhassan, M., Funk, L., Schnepper, M., Levermann, I., Wang, B., Giehl, C., Trampisch, U., & Wirth, R. (2026). Diagnostic performance of calculated serum osmolarity versus clinical impression for identifying dehydration in older hospitalized patients. BMC Geriatrics. https://doi.org/10.1186/s12877-026-08503-6

Image Credits: AI Generated

DOI: 10.1186/s12877-026-08503-6

Keywords: dehydration, older adults, calculated osmolarity, serum osmolarity, glucose, geriatrics, diagnostic markers, hospitalized patients, clinical assessment, BMC Geriatrics, Diagnostic, performance