Hyponatremia is defined as plasma sodium concentration below 135 mmol/L. It is the most common electrolyte disorder, affecting 1–2 % of the general population, with an increased prevalence in older adults and people with hypertension, diabetes, cardiovascular disease, cancer, and psychiatric disorders [1]. Hyponatremia is present in 15–30 % of hospitalised patients [2], and is associated with prolonged hospital stay, increased readmission within 90 days, and increased mortality [3]. Hyponatremia can present with a broad spectrum of symptoms, from subtle or asymptomatic (e.g. nausea, gait instability) to severe or life-threatening (e.g. seizures, coma) [4].

Determining the cause of hyponatremia can be difficult. People with hyponatremia often have multiple comorbidities and take numerous medications, and their hyponatremia is often multifactorial. Traditional diagnostic algorithms can be challenging to implement and may not always lead to accurate diagnosis. In one survey, 46 physicians were given 4 clinical cases of hyponatremia and provided with 10 published diagnostic algorithms to guide them, yet only 10 % of physicians were able correctly diagnose the cause of hyponatremia [5]. Weaknesses identified in the algorithms included the belief that mild hypovolemia could be distinguished from euvolemia by clinical examination supported by routine laboratory data, a tendency to diagnose the syndrome of inappropriate antidiuresis (SIAD) prior to excluding other causes of hyponatremia, and reliance on generalisations rather than robust data (e.g. hyperkalemia must be present to diagnose primary adrenal insufficiency) [5].

This review provides a suggested approach for diagnosis of hyponatremia based on biochemistry and clinical assessment of volume status, according to best available evidence. Multiple diagnostic algorithms exist in the literature; however each has its limitations. Clinical history, medication history and examination are a critical part of diagnosis and provide important context for interpretation of biochemical assessments. However, fluid status assessment is often inaccurate, so it may not be appropriate to make treatment decisions based on fluid status assessment alone. Plasma and urine osmolality and urine sodium provide accurate information - but can have prolonged laboratory processing times and may not be immediately available to make preliminary management decisions. Furthermore, these measures are not static and can differ on serial measurements even if the pathophysiology is unchanged [6]. The algorithm we propose has a stronger emphasis on biochemistry supported by clinical assessment, which we believe provides the greatest diagnostic accuracy. Importantly, severely symptomatic hyponatremia is a medical emergency, and urgent treatment may be required prior to any assessment of its underlying cause. A list of suggested investigations to assist with determining the underlying cause of hyponatremia is provided in Table 1. Detailed discussion of the pathophysiology and treatment of hyponatremia is beyond the scope of this review.