Sodium deficit in hyponatraemia

Clinical background

The sodium deficit in hyponatraemia is a calculation model that translates a desired rise in serum sodium into a concrete quantity of sodium in milliequivalents. It serves two purposes: to give a rough estimate of how much sodium is required to reach a target value, and to provide a starting point for the choice of fluid and infusion rate. The decision is difficult without a structured estimate because hyponatraemia may be acute or chronic, volume status varies, and the therapeutic margin between undertreatment and overcorrection is narrow. Too slow a rate of correction leaves the patient in symptomatic hyponatraemia; too fast a rate risks osmotic demyelination, an injury that may be permanent.

Calculating the sodium deficit

The formula rests on the principle that the serum sodium concentration is determined by the ratio of total body sodium to total body water (TBW). The sodium deficit is the amount of sodium that must be given to move the concentration from the current level to the target level, assuming an unchanged water volume:

$$\text{Sodium deficit (mEq)} = \text{TBW} \times (\text{target Na} - \text{serum sodium})$$

$$\text{TBW} = \text{body weight} \times \text{body water fraction}$$

The body water fraction varies with sex and age:

The model derives from Adrogué and Madias's review of the pathophysiology and treatment principles of hyponatraemia, published in the New England Journal of Medicine in 2000 [1]. It is not an empirically derived prediction model in the modern sense but a physiological derivation based on the established relationship between sodium, potassium and water in the body's fluid compartments. The article is a review, not a cohort study with discrimination or calibration in the traditional sense, and the formula has therefore never been "validated" in the sense of being tested against outcomes in a derivation cohort.

Interpretation in practice

The result gives the total amount of sodium theoretically required to reach the target level. It does not give the rate, and it is the rate that determines safety.

The clinical value lies in two steps:

1. Estimate the total requirement. In symptomatic hyponatraemia where hypertonic saline is being considered, the formula gives an idea of the sodium dose corresponding to a reasonable first rise, for example 4 to 6 mEq/L.
2. Choose the infusion and monitor. If the target is a rise of 4 to 6 mEq/L over the first few hours, one can calculate the volume of 3% saline (513 mEq Na/L) corresponding to this dose, and then check the serum sodium after 2 to 4 hours to confirm that the rise remains within safe limits.

The rate of correction must always be limited. European guidelines recommend a maximum rise of 10 mEq/L per 24 hours [2]. A systematic review of published cases of osmotic demyelination shows, however, that the injury can occur even at rates of correction within this range, particularly in patients with a starting serum sodium below 115 mEq/L. The authors therefore recommend that this limit be lowered to below 8 mEq/L per 24 hours for patients with severe hyponatraemia and additional risk factors [3].

The risk factors for osmotic demyelination identified in the same review were alcohol misuse, hypokalaemia, liver disease and malnutrition. Of 21 patients with osmotic demyelination despite a correction rate of up to 10 mEq/L/24 h, 12 had a starting level below 115 mEq/L, 7 of them below 105 mEq/L. Four patients died and nine had persistent neurological deficits [3].

Validation and performance

Since the formula is a physiological derivation and not an empirical prediction model, there is no traditional derivation cohort with a c-statistic or calibration curve to report. The ability of the formula to predict the actual rise in serum sodium after infusion of hypertonic saline has, however, been studied in clinical settings.

A prospective observational study in children with SIAD given 3% saline compared the Adrogué–Madias equation with the Voets equation. The study found that both equations could predict the direction of the rise in sodium, but that the Adrogué–Madias equation was less reliable than the Voets equation in this population [4]. This accords with earlier studies in adults showing that the Adrogué–Madias equation tends to overestimate the rise in serum sodium, particularly in euvolaemic hyponatraemia, where water and sodium balance is not static during treatment.

A separate question, which the formula itself does not address, is the risk of overcorrection. The SHOR score (Severe Hyponatremia Overcorrection Risk) was developed to predict which patients with severe hyponatraemia (serum Na below 116 mmol/L) are at risk of being overcorrected irrespective of the intended dose [5]. The score is based on factors such as reduced level of consciousness, vomiting, severe hypokalaemia, dilute urine and the baseline sodium, and it was derived in a cohort of 623 patients at a tertiary emergency department in Ottawa. The score had a dichotomised c-statistic of 0.77 (95 per cent CI 0.73 to 0.81) in the derivation cohort. On external validation in 95 patients, however, the association with overcorrection was not statistically significant (P = 0.39), which limits its generalisability [5].

Limitations

The formula rests on the assumption that total body water is known and unchanged during correction. In practice one neither knows the TBW exactly nor is it constant. In SIAD, free water retention continues alongside sodium administration, and in hypovolaemic hyponatraemia water shifts with volume expansion. As a result the actual rise in serum sodium often differs from the calculated one, usually because the formula overestimates the rise in euvolaemic hyponatraemia and underestimates it in hypovolaemic hyponatraemia, in which volume replacement itself raises the sodium.

The body water fraction is a rough estimate. In obesity, oedema, ascites or pregnancy the actual water fraction differs from the standard values the formula uses, which systematically distorts the calculated deficit.

The formula does not apply to acute hyponatraemia (duration under 48 hours), where the correction targets are different and a faster rise is acceptable, and it does not serve as decision support for whether correction should be undertaken at all. It gives a quantity, not an indication.

The commonest misuse is to treat the calculated value as a prescription rather than a starting point. The serum sodium must always be remeasured after treatment has begun, and the infusion rate and choice of fluid adjusted according to the actual rise, not the predicted one.

References

1. Adrogué HJ, Madias NE. Hyponatremia. N Engl J Med 2000;342(21):1581-9. PMID: 10824078
2. Spasovski G, Vanholder R, Allolio B et al. Clinical practice guideline on diagnosis and treatment of hyponatraemia. Eur J Endocrinol 2014;170(3):G1-47. PMID: 24569125
3. Tandukar S, Sterns RH, Rondon-Berrios H. Osmotic Demyelination Syndrome following Correction of Hyponatremia by ≤10 mEq/L per Day. Kidney360 2021;2(9):1415-1423. PMID: 35373113
4. Cherukuri VKR, Nair SG, Pavithran PV et al. Comparison of Voets and Adrogué-Madias equations in children with syndrome of inappropriate antidiuresis receiving 3% saline. Pediatr Nephrol 2026;41(10):3489-3496. PMID: 42065738
5. Woodfine JD, Sood MM, MacMillan TE et al. Derivation and Validation of a Novel Risk Score to Predict Overcorrection of Severe Hyponatremia: The Severe Hyponatremia Overcorrection Risk (SHOR) Score. Clin J Am Soc Nephrol 2019;14(7):975-982. PMID: 31189541