An improved method to compute the solute and water derangements of hyperglycaemia.

Evaluation and treatment of hyperglycaemic hyponatremia, being quantitatively inaccurate, is open to new advancements. We herein describe the improvement of previous calculations of glucose appearance (G(A)), solute and solvent changes. From G(A) we derive the predicted plasma sodium concentration (PNa(G)), assuming no change in total body water (TBW), but only water shift from cells to the extracellular space (ECV). This assumption is validated by the respective solute ratios (PCl/PNa) unchanged from normal values, as well as the ratios between actual and normal solute concentrations (PNa(1)/PNa(0), PCl(1)/PCl(0)), identical for all solutes. When the assumption is met, G(A) can be exactly calculated. When the ratios are different from normal, they indicate the presence of a mixed abnormality due to a loss either of sodium, or sodium and water. These are estimated by computing the difference between PNa(G) and the actual PNa measured (PNa(1)). PNa(1) approximately equal PNa(G) if TBW and Na are unchanged, PNa(1) < PNa(G) in the presence of prevalent Na depletion, PNa(1) > PNa(G )when volume depletion prevails. In the first circumstance the ECV expansion is exactly established by appropriate mathematical formulas, in the latter conditions either Na or volume depletion are empirically estimated with algebric expressions. These equations were validated on computer-simulated models, and applied to 49 subjects with plasma glucose concentration >15 mM/L. G(A) and PNa(G) were computed, and, with the same formulas used in computer-simulated experiments, we calculated water and Na deficits. The PNa measured after correction of hyperglycaemia was correctly predicted (R(2) = 0.63, P < 0.0001). This method provides a firm ground to select the correct equation to accurately estimate the initial conditions of hyperosmolar hyperglycaemia, significantly improving its quantitative correction.