Plasma sodium concentration in older patients as an indicator of severity in emergencies: Results from the Emergency Department and Elder Needs-16 study.

Abstract

Objectives: To study baseline factors associated with hypo- and hypernatremia in older patients attended in emergency departments (EDs) and explore the association between these dysnatremias and indicators of severity in an emergency.

Material and methods: We included patients attended in 52 Spanish hospital EDs aged 65 years or older during a designated week. All included patients had to have a plasma sodium concentration on record. Patients were distributed in 3 groups according to sodium levels: normal, 135-145 mmol/L; hyponatremia, 135 mmol/L; or hypernatremia > 145 mmol/L. We analyzed associations between sodium concentration and 24 variables (sociodemographic information, measures of comorbidity and baseline functional status, and ongoing treatment for hypo- or hypernatremia). Indicators of the severity in emergencies were need for hospitalization, in-hospital mortality, prolonged ED stay (> 12 hours) in discharged patients, and prolonged hospital stay (> 7 days) in admitted patients. We used restricted cubic spline curves to analyze the associations between sodium concentration and severity indicators, using 140 mmol/L as the reference.

Results: A total of 13 368 patients were included. Hyponatremia was diagnosed in 13.5% and hypernatremia in 2.9%. Hyponatremia was associated with age ($ 80 years), hypertension, diabetes mellitus, an active neoplasm, chronic liver disease, dementia, chemotherapy, and needing help to walk. Hypernatremia was associated with needing help to walk and dementia. The percentages of cases with severity indicators were as follows: hospital admission, 40.8%; in-hospital mortality, 4.3%; prolonged ED stay, 15.9%; and prolonged hospital stay, 49.8%. Odds ratios revealed associations between lower sodium concentration cut points in patients with hyponatremia and increasing need for hospitalization (130 mmol/L, 2.24 [IC 95%, 2.00-2.52]; 120 mmol/L, 4.13 [3.08-5.56]; and 110 mmol/L, 7.61 [4.53-12.8]); risk for in-hospital death (130 mmol/L, 3.07 [2.40-3.92]; 120 mmol/L, 6.34 [4.22- 9.53]; and 110 mmol/L, 13.1 [6.53-26.3]); and risk for prolonged ED stay (130 mmol/L, 1.59 [1.30-1.95]; 120 mmol/L, 2.77 [1.69-4.56]; and 110 mmol/L, 4.83 [2.03-11.5]). Higher sodium levels in patients with hypernatremia were associated with increasing need for hospitalization (150 mmol/L, 1.94 [1.61-2.34]; 160 mmol/L, 4.45 [2.88-6.87]; 170 mmol/L, 10.2 [5.1-20.3]; and 180 mmol/L, 23.3 [9.03-60.3]); risk for in-hospital death (150 mmol/L, 2.77 [2.16-3.55]; 160 mmol/L, 6.33 [4.11-9.75]; 170 mmol/L, 14.5 [7.45-28.1]; and 180 mmol/L, 33.1 [13.3-82.3]); and risk for prolonged ED stay (150 mmol/L, 2.03 [1.48-2.79]; 160 mmol/L, 4.23 [2.03-8.84]; 170 mmol/L, 8.83 [2.74-28.4]; and 180 mmol/L, 18.4 [3.69-91.7]). We found no association between either type of dysnatremia and prolonged hospital stay.

Conclusion: Measurement of sodium plasma concentration in older patients in the ED can identify hypo- and hypernatremia, which are associated with higher risk for hospitalization, death, and prolonged ED stays regardless of the condition that gave rise to the dysnatremia.