Abstract: SA-PO735
Serum Osmolality Testing and Preserved GFR Are Associated with Improved 30-Day Mortality in Severely Hyponatremic Patients
Session Information
- Fluid, Electrolyte, Acid-Base Disorders: Clinical - II
November 04, 2023 | Location: Exhibit Hall, Pennsylvania Convention Center
Abstract Time: 10:00 AM - 12:00 PM
Category: Fluid, Electrolytes, and Acid-Base Disorders
- 1102 Fluid, Electrolyte, and Acid-Base Disorders: Clinical
Authors
- Fox, Oliver James Staley, University Hospital Southampton NHS Foundation Trust, Southampton, United Kingdom
- Davis, Cai, University Hospital Southampton NHS Foundation Trust, Southampton, United Kingdom
- Ash, Thomas, University Hospital Southampton NHS Foundation Trust, Southampton, United Kingdom
- Veighey, Kristin, University Hospital Southampton NHS Foundation Trust, Southampton, United Kingdom
Background
Despite UK guidelines advocating paired osmolalities and urinary sodium, inadequate investigation of hyponatremia is common. Although many hospitals suggest serum osmolality as an initial test, it is unclear whether this approach improves outcomes. We have built a logistic regression model to explore the association between serum osmolality (sOsm) testing status and mortality in severely hyponatremic patients. We also examined renal function to identify if this conferred greater risk.
Methods
We analyzed sodium values on admission from 87,174 patients in a large UK tertiary hospital. We extracted U/E, measured osmolalities, age, and blood glucose. Since we wanted to isolate the effect of a sOsm test, it was important to control for its value. In the absence of a measured osmolality, we therefore derived a calculated osmolality using the first glucose and urea available within 24h. We validated this by regressing 753 measured and estimated osmolalities (R2 = 0.6958, p <0.001). 30-day-mortality was modeled with covariates ‘Sodium Result’, ‘sOsm value (calculated if measured not available)’, ‘age’, ‘GFR’ and an interaction term identifying if testing was associated with improved mortality when sodium was <126meq/L.
Results
Testing for serum osmolality in patients with sodium values <126meq/L was associated with lower 30 day mortality (coefficient -1.527, p<0.001), as was preserved eGFR (p<0.01) and younger age (p<0.001). Of 1187 patients with Na+<126meq/L, just 39.4% received a sOsm test within 7 days of admission.
Conclusion
In this large hospital cohort, a low proportion of patients received a serum osmolality when severely hyponatremic. Performing serum osmolality testing is associated with lower 30 day mortality. Patients with impaired renal function were at increased risk. Not being able to identify patients who were admitted for palliation or identifying further clinical predictors of mortality were important limitations.
This is the first large study where sOsm has been calculated in place of a measured value to model the impact of the test itself on outcomes, a technique which could be used on other datasets. The results of this analysis have led to the adoption of an electronic alert to prompt sOsm testing, the impact of which will be measured in 3 monthly intervals.