Combined fractional excretion of sodium and urea better predicts response to saline in hyponatremia than do usual clinical and biochemical parameters.

BACKGROUND The treatment of hyponatremic patients requires physicians to make a therapeutic choice between saline infusion and water restriction. Therefore, they need readily available and reliable parameters to facilitate making that choice. This study was designed to determine whether the use of clearance ratios can help clinicians recognize saline-responding hyponatremic patients. PATIENTS AND METHODS Thirty-five nonedematous, hospitalized, hyponatremic patients were classified according to their history and saline response into four groups: diuretic-taking patients, polydipsic patients, saline responders, and saline nonresponders. Within these four groups, clinical and biochemical volume-related parameters, including clearance ratios, were prospectively evaluated before infusion of 2 L isotonic saline. Clearance ratios as well as usual clinical and biochemical parameters were tested for their accuracy in predicting saline responsiveness. RESULTS Both positive (70%) and negative (54.5%) predictive values for hypovolemia were unsatisfactory; clinical prediction of hypovolemia was also characterized by low sensitivity (41.1%), but acceptable specificity (80%). In the polydipsia and saline-nonresponder groups, plasma urea and uric acid values tended to be lower than in the diuretic and saline-responder groups. However, the usefulness of these parameters was limited by too large an overlap among the different groups. In both polydipsic patients and saline responders, urinary sodium concentration was low. The combined amount of urinary sodium and potassium in relation to plasma sodium did not discriminate among the different groups. Most helpful in distinguishing among the groups was a combination of several clearance ratios (fractional excretions of sodium, potassium, urea, and uric acid), since the predictive use of each parameter on its own was restricted. The best indicator of saline responsiveness was a low fractional excretion of filtered sodium (< 0.5%) combined with a low fractional excretion of urea (< 55%). CONCLUSION The accuracy of clinical evaluation for predicting the state of extracellular fluid volume in hyponatremia is low. The combination of low fractional sodium excretion (< 0.5%) and low fractional urea excretion (< 55%) is the best biochemical way to predict saline response, whereas high fractional potassium excretion (> 20%) indicates diuretic intake.