Hyperosmolar coma due to exclusive glucose accumulation: recognition and computations

Aim: To avoid electrolyte derangements during correction of hyperosmolar coma (HC), PNa_PREDICTED at the end of correction is presently estimated from plasma glucose (P_G, mM/L). When the rise in plasma osmolality (Posm) is entirely due to glucose addition (G_A, mM) to the extracellular volume (ECV), this PNa prediction can be improved by correctly estimating G_A and any associated water loss (δV), while excluding any concomitant Na loss (δNa). Methods: Indicating with ₀ the normal conditions, with ₁ the HC, establishes an exclusive G_A accumulation. We derived the equations for computing G_A, δV and PNa_PREDICTED. Computer simulations of HC were performed by adding the known G_A while subtracting the known δV and δNa in different combinations, obtaining exact values of PNa₁ and. Applying our formulas, we recognized and discarded all cases with concomitant δNa, and we computed G_A, δV and PNa_PREDICTED from PNa₁ and, as if they had been measured in patients. We extended these same calculation algorithms to 68 patients with HC. Results: In computer simulations, true and calculated G_A, δV and PNa_PREDICTED were identical, such that regression and correlation coefficients were 1 (P < 0.0001). Out of the 68 patients recruited, 13 fulfilled the boundary conditions of an exclusive G_A addition. The true values, obtained by balance studies performed on these patients, were not different from and significantly correlated with the calculated data (R² = 0.99, P < 0.001). Conclusion: Our new model system for HC and the new formulas improve to near exactness the accuracy in estimating PNa_PREDICTED, helping the physician to avoid unwanted electrolyte derangements during treatment.