Chronic (present > 48 hours) non-hypovolaemic hyponatraemia occurs frequently, can be caused by various conditions, and is associated with shorter survival and longer hospital stays. Many treatments, such as fluid restriction or vasopressin receptor antagonists can be used to improve the hyponatraemia, but whether that translates into improved patient-important outcomes is less certain.
This review aimed to 1) look at the benefits and harms of interventions for chronic non-hypovolaemic hypotonic hyponatraemia when compared with placebo, no treatment or head-to-head; and 2) determine if benefits and harms vary in absolute or relative terms dependent on the specific compound within a drug class, on the dosage used, or the underlying disorder causing the hyponatraemia.
We searched the Cochrane Kidney and Transplant Register of Studies up to 1 December 2017 through contact with the Information Specialist using search terms relevant to this review. Studies in the Register are identified through searches of CENTRAL, MEDLINE, and EMBASE, conference proceedings, the International Clinical Trials Register (ICTRP) Search Portal and ClinicalTrials.gov. We also screened the reference lists of potentially relevant studies, contacted authors, and screened the websites of regulatory agencies.
We included randomised controlled trials (RCTs) and quasi-RCTs that compared the effects of any intervention with placebo, no treatment, standard care, or any other intervention in patients with chronic non-hypovolaemic hypotonic hyponatraemia. We also included subgroups with hyponatraemia from studies with broader inclusion criteria (e.g. people with chronic heart failure or people with cirrhosis with or without hyponatraemia), provided we could obtain outcomes for participants with hyponatraemia from the report or the study authors.
Two authors independently extracted data and assessed risk of bias. We expressed treatment effects as mean difference (MD) for continuous outcomes (health-related quality of life, length of hospital stay, change from baseline in serum sodium concentration, cognitive function), and risk ratio (RR) for dichotomous outcomes (death, response and rapid increase in serum sodium concentration, hypernatraemia, polyuria, hypotension, acute kidney injury, liver function abnormalities) together with 95% confidence intervals (CI).
We identified 35 studies, enrolling 3429 participants. Twenty-eight studies (3189 participants) compared a vasopressin receptor antagonist versus placebo, usual care, no treatment, or fluid restriction. In adults with chronic, non-hypovolaemic hypotonic hyponatraemia, vasopressin receptor antagonists have uncertain effects on death at six months (15 studies, 2330 participants: RR 1.11, 95% CI 0.92 to 1.33) due to risk of selective reporting and serious imprecision; and on health-related quality of life because results are at serious risk of performance, selective reporting and attrition bias, and suffer from indirectness related to the validity of the Short Form Health Survey (SF-12) in the setting of hyponatraemia. Vasopressin receptor antagonists may reduce hospital stay (low certainty evidence due to risk of performance bias and imprecision) (3 studies, 610 participants: MD -1.63 days, 95% CI -2.96 to -0.30), and may make little or no difference to cognitive function (low certainty evidence due to indirectness and imprecision). Vasopressin receptor antagonists probably increase the intermediate outcome of serum sodium concentration (21 studies, 2641 participants: MD 4.17 mmol/L, 95% CI 3.18 to 5.16), corresponding to two and a half as many people having a 5 to 6 mmol/L increase in sodium concentration compared with placebo at 4 to 180 days (moderate certainty evidence due to risk of attrition bias) (18 studies, 2014 participants: RR 2.49, 95% CI 1.95 to 3.18). But they probably also increase the risk of rapid serum sodium correction - most commonly defined as > 12 mmol/L/d (moderate certainty evidence due to indirectness) (14 studies, 2058 participants: RR 1.67, 95% CI 1.16 to 2.40) and commonly cause side-effects such as thirst (13 studies, 1666 participants: OR 2.77, 95% CI 1.80 to 4.27) and polyuria (6 studies, 1272 participants): RR 4.69, 95% CI 1.59 to 13.85) (high certainty evidence). The potential for liver toxicity remains uncertain due to large imprecision. Effects were generally consistent across the different agents, suggesting class effect.Data for other interventions such as fluid restriction, urea, mannitol, loop diuretics, corticosteroids, demeclocycline, lithium and phenytoin were largely absent.
In people with chronic hyponatraemia, vasopressin receptor antagonists modestly raise serum sodium concentration at the cost of a 3% increased risk of it being rapid. To date there is very low certainty evidence for patient-important outcomes; the effects on mortality and health-related quality of life are unclear and do not rule out appreciable benefit or harm; there does not appear to be an important effect on cognitive function, but hospital stay may be slightly shorter, although available data are limited. Treatment decisions must weigh the value of an increase in serum sodium concentration against its short-term risks and unknown effects on patient-important outcomes. Evidence for other treatments is largely absent.Further studies assessing standard treatments such as fluid restriction or urea against placebo and one-another would inform practice and are warranted. Given the limited available evidence for patient-important outcomes, any study should include these outcomes in a standardised manner.