Sodium restriction has long been an accepted non-pharmacological strategy for fluid and volume management in patients with heart failure. In the critical care setting, high sodium loads have been documented to be responsible for avoidable derailment of fluid balance, with fluid overload being the most serious adverse effect. This is also reported to be an independent risk factor for morbidity and mortality in critically ill and surgical patients [1].

The National Heart Foundation and Cardiac Society of Australia and New Zealand recommend restricting sodium intake to <2,000 mg/day or <87 mmol/day for patients with heart failure [2]. The same recommendations are made by the World Health Organisation, The Canadian Cardiovascular Society and The Heart Failure Society of America (for those patients with moderate-to-severe heart failure) [3-5]. In contrast, the European Society of Cardiology recommends <5,000 mg/day (<218 mmol/day) [3,6]. The SODIUM-HF trial undertaken in 2022 aimed to explore if a stricter sodium restriction to <100 mmol/day in patients with chronic heart failure reduced cardiovascular-related hospitalisation, cardiovascular-related emergency department visits and all-cause death within 12 months. Although this study referred to sodium <100 mmol/day, the participants were given a restriction of <1,500 mg/day, which equated to approximately <65 mmol/day. The results showed a stricter sodium restriction did not improve outcomes compared to usual care [7]. Despite this, the American Heart Association still recommends a sodium restriction of <1,500 mg/day [3]. In Australia, current national guidelines of <87 mmol/day are still accepted. A prior study including patients admitted to the cardiac intensive care unit extrapolated these dietary recommendations to the inpatient setting [8].

Sodium is often inadvertently administered as part of intravenous (IV) fluid resuscitation, medicine reconstitution, diluents, IV line flushes and as a component of medicine infusions. Several studies have investigated the amount and sources of sodium administration to critically ill patients or those in the intensive care (ICU) setting at a tertiary teaching hospital in South Australia. Each of these studies showed a high median daily sodium administration (approximately 225 mmol/day), primarily from sources such as IV infusions, IV medicines, and maintenance fluids [9,10]. These studies also identified that excessive sodium administration may be associated with poorer outcomes.

While research has been conducted into sodium administration in the intensive and critical care setting, limited literature was identified for heart failure-specific cohorts. It was hypothesised that inadvertent sodium would be administered to patients with heart failure, in excess of international guidelines.

Aim

This study aimed to determine the amount and sources of sodium administered to patients with heart failure with reduced ejection fraction (HFrEF) admitted to an academic teaching hospital network.

A retrospective cohort study was undertaken over a 3-month period across two hospitals within the Southern Adelaide Local Health Network (SALHN), Flinders Medical Centre (FMC, 756 beds) and Noarlunga Health Services (NHS, 104 beds). Patients were identified from the electronic medical record (Sunrise^©) using case-mix extraction and ICD-10 codes (International Classification of Diseases 10th Revision). Inclusion criteria were patients ≥18 years of age with heart failure with reduced ejection fraction (HFrEF or HFmrEF) admitted to FMC or NHS from October to December 2021 [11]. Data on patient’s left ventricular ejection fraction (LVEF) was recorded and categorised as mild dysfunction (LVEF 40 to 49%), moderate dysfunction (30 to 39%) and severe dysfunction (LVEF <30%) [12]. Where ejection fraction was not recorded in the medical record, from the current admission, this data was obtained from the patient’s prior admissions. Patients admitted with heart failure with preserved ejection fraction (HFpEF) were excluded due to inadequate evidence for the benefits of restricting sodium intake [13,14]. Additionally, due to research previously undertaken in the intensive care unit (ICU), days in which patients were in ICU were excluded. Those patients who were transitioned to palliative care during their admission were excluded from the main data set due to changing to end of life treatment goals.

Electronic medical records were used to extract data on the type and volume of IV fluids administered over each day of admission. Sodium (Na) content in medicines administered by infusion and boluses was calculated from both the sodium content of the medicine and the type and volume of crystalloid fluid or diluent. Data on sodium administered from oral and rectal medicines with known high sodium content was also collected. The amount of sodium administered each day was calculated based on published sodium concentrations, and guidelines for reconstituting and administering IV medicines (where local guidelines existed these were used in preference to data published in the Australian Injectables Drug Handbook) [15]. A volume of 10 mL Sodium Chloride 0.9% was assumed for IV line flushes (before and after administration as per local nursing guidelines). The sodium content of crystalloid fluids, medicines (IV and oral/rectal with known high sodium content) and diluents are presented in the Supplementary material.

Data was analysed using descriptive statistics to determine total sodium administration (mmol) per day of admission, sources of sodium administration and if there is any trend observed between sodium administration and the days or length of admission. Daily sodium administration was compared to current recommendations for sodium intake to determine if excessive sodium was being administered.

This project was reviewed by the Southern Adelaide Local Health Network (SALHN) Office for Research and deemed exempt from ethical approval (Reference Number: 2349).

A total of 123 patients were captured using inclusion and exclusion criteria. There were 116 eligible for inclusion in the study. Table 1 lists the characteristics of patients included in the study.

Table 1 . Demographic data.

	Cardiac unit	Non-cardiac unit*	Total cohort
Number of patients	57 (49.1%)	59 (50.9%)	116 (100.0%)
Study days	407	446	853
Length of stay (days), median (IQR)	6 (5–9)	6 (3–10)	6 (4–9)
Hospital
FMC	57 (100.0%)	49 (83.0%)	106 (91.4%)
NHS	0 (0.0%)	10 (17.0%)	10 (8.6%)
Age (years), median (IQR)	72 (59–76)	82 (78–90)	77 (68–85)
Gender
Male	39 (68.4%)	34 (57.6%)	73 (62.9%)
Female	18 (31.6%)	25 (42.4%)	43 (37.1%)
HFrEF ejection fraction (EF)
Not recorded	2 (3.5%)	10 (17.0%)	12 (10.4%)
Mild dysfunction (40–49%)	18 (31.6%)	14 (23.7%)	32 (27.6%)
Moderate dysfunction (30–39%)	10 (17.5%)	14 (23.7%)	24 (20.7%)
Severe dysfunction (<30%)	27 (47.4%)	21 (35.6%)	48 (41.3%)

IQR=interquartile range..

*Non-cardiac units included general medicine, acute medical unit, non-cardiac surgical wards, etc..

There were 853 days of hospital admission reviewed from the 116 patients eligible for inclusion. There was an even distribution of patients admitted under cardiac and non-cardiac units, at 57 (49.1%) and 59 (50.9%) respectively. The majority of patients were admitted to the principal hospital in the network FMC (91.4%) with a median length of stay across the total cohort of 6 days (IQR 4 to 9 days). The median age for all patients was 77 years (IQR 68 to 85 years), and 73 (62.9%) patients were male. Across the total cohort, most patients were reported to have severe HFrEF (41.3%), followed by mild HFrEF (27.6%), and moderate HFrEF (20.7%). There were 10.4% of patients who had their ejection fraction recorded in a previous admission. Table 2 reports the median sodium administration per day.

Table 2 . Daily sodium administration (mmol/day).

	Cardiac unit	Non-cardiac unit	Total
Median (IQR) (mmol/day)	3.74 (0.00–11.45)	7.00 (0.00–18.99)	5.20 (0.00–15.55)
Maximum (mmol/day)	181.61	456.05	456.05

The amount of sodium administered to patients was categorised by the different sources of sodium, including line flushes, IV fluids, IV medicines and diluents, IV continuous medicine infusions, oral/rectally administered medicines and subcutaneous medicines (Fig. 1).

Figure 1. Sources of sodium administration (total cohort).

In the total cohort, IV fluids contributed to 38.1% of daily sodium administration. This was followed by sodium from line flushes and sodium from oral/rectal sources, which accounted for 22.3% and 20.5% respectively. There was 18.2% of daily sodium administered from IV medicines (11.1%) and IV medicine diluents (7.1%).

The sources of sodium administration differed between patients admitted under cardiac and non-cardiac units (Fig. 2). In patients admitted under the cardiac unit, IV line flushes were the greatest source of administered sodium at 38.3%, followed by IV fluids at 28.4%, IV medicines at 17.2% and IV medicine diluents at 8.7%. Oral/rectally administered medicines contributed to 4.9% of sodium administration. In patients admitted under non-cardiac units, IV fluids were the greatest source of administered sodium at 42.1%, followed by oral/rectally administered medicines at 26.8%, IV line flushes at 15.8%, IV medicines at 8.6% and IV medicine diluents at 6.5%. Across the total cohort and for each subgroup, IV continuous infusions and subcutaneous medicines had negligible contribution to total sodium administration.

Figure 2. Sources of sodium administration (by unit).

Interquartile range (IQR) was used to graphically present sodium administration by day of admission. Any data points considered outliers (more than 1.5 times the IQR) were excluded from these figures. The results are presented as sodium administration (mmol) by day of admission for the total cohort, cardiac and non-cardiac units respectively (Fig. 3-5).

Figure 3. Sodium administration (mmol) by day of admission (total cohort).

Figure 4. Sodium administration (mmol) by day of admission (cardiac unit).

Figure 5. Sodium administration (mmol) by day of admission (non-cardiac units).

In this study the median daily amount of sodium administrated via medicines, crystalloid therapy, diluents and IV line flushes was 5.20 mmol/day, which is equivalent to 5.98% of Australian and international recommendations of <87 mmol/day [2-5]. As such, excessive sodium was not found to be administered to patients with HFrEF irrespective of patients being treated by clinicians from cardiac or non-cardiac units. These findings were more favourable than hypothesised based on results from previous studies in the ICU and critical care setting [8,16]. This may be due to a reduced need for IV medicines and intravenous fluid therapy in the non-ICU patient population, with many patients tolerating oral therapy and being more hemodynamically stable compared to patients with HFrEF being cared for in the ICU [17,18].

Despite the finding that excessive sodium was not administered to patients in this study, it is important to note there was considerable interpatient variability in the amount of sodium administered. Across all cohorts, the IQR was below the recommended intake, however the maximum amount of sodium administered on any day was considerably higher than recommended amounts. These findings may be due to many patients having no sodium administered (i.e. no IV therapy or oral/rectal medicines with high sodium content) on several days, and some patients receiving large volumes of IV fluid with high sodium content on few days of the study (e.g. to acutely manage hypovolaemia or hypotension).

The differences in sodium administration between cardiac and non-cardiac cohorts in our study may be related to several factors. While the relative proportion of sodium from IV medicines and diluents to total sodium administration was less in the non-cardiac unit (15.1% compared to 25.9% in the cardiac unit group), the absolute amount of sodium from these sources was more (1,656.04 mmol compared to 1,163.73 mmol in the cardiac unit). This may reflect the fact that patients admitted to a non-cardiac unit are more likely to be admitted for non-heart failure specific causes such as infection, and hence more likely to receive medicines with higher sodium content such as IV antibiotics (e.g. piperacillin/tazobactam containing 11.36 mmol per standard 4.5 g dose). Other contributing factors include increased use of oral/rectal medicines with high sodium, medical and nursing staff from non-cardiac units potentially having less awareness of the sodium content of medicines, and less routine fluid restriction for patients with HFrEF admitted to non-cardiac wards (hence may be more likely to receive increased IV fluids containing sodium).

Major contributors to inadvertent sodium administration in this study included IV fluids (for example, Sodium Chloride 0.9% and Hartmann’s Solution), IV line flushes and oral/rectal medicines. The contribution from IV medicines and IV medicine diluents was less than 20% of the total sodium administration across the total cohort. This was less than anticipated based on previous studies in the ICU and critical care setting. Exploration into the benefit of practice changes to use of 5% Dextrose rather than Sodium Chloride 0.9%, both for fluid therapy and as a vehicle for drug infusions and boluses, may still be worthy of consideration to prevent potential inadvertent sodium administration [16].

Key differences were observed in the sources of sodium administration between patients admitted under cardiac and non-cardiac units. In the cardiac unit, IV line flushes were the greatest contributor to sodium administration. This may be related to patients admitted under a cardiac unit receiving multiple doses of IV furosemide each day to manage decompensation of HFrEF, subsequently increasing the number of IV line flushes (and thus sodium) administered.

In contrast to the low contribution of oral/rectal medicines in patients admitted in the cardiac unit, this was one of the major sources of sodium administration for patients admitted under a non-cardiac unit. In this study, sodium polystyrene sulfonate (Resonium A^® containing 61.5 mmol of sodium per standard 15 g dose), macrogol 1335 with electrolytes (Movicol^® containing 8.125 mmol of sodium per sachet), Fleet^® sodium phosphate enema (191.4 mmol of sodium per enema) and Microlax^® enemas (4.73 mmol of sodium per enema) were common contributors. Medicines such as Resonium A^® have limited alternatives with lower sodium content, and there is additional complexity in managing the comorbidities of these patients (e.g. HFrEF and hyperkalaemia either primary or secondary to kidney disease); however alternative laxative products exist for Movicol^®, Fleet^® enemas and Microlax^® enemas with lower sodium content. The amount of sodium administered to patients admitted under non-cardiac units may reflect a poorer general understanding of the sodium content in commonly used medicines such as aperients, and further education for hospital staff, including available alternatives (e.g. glycerol suppositories), may be beneficial.

No obvious trends could be observed between sodium administration and days of patient admission. There was comparatively more sodium administered on the first day of admission, which then reduces from Day 2. The higher sodium on Day 1 of admission may reflect the acute management of dehydration or hypotension with IV fluids on presentation. However, from Day 9 there is significant fluctuation in sodium administration. A similar trend was also observed when results were reviewed by cardiac and non-cardiac units. As most patients were admitted for four to nine days, the unpredictable trend in sodium administration may reflect a reduced sample size for longer admissions, and changes in patient’s clinical states over the admission.

There were some limitations to our work. The study was conducted in only two hospitals and conducted over relatively short duration. An extended data collection period may capture variations in sodium administration practices over time and provide a more robust representation of practices. Additionally, the study design was retrospective and involved manual data extraction from electronic medical records, creating the potential for inaccuracies and confounding factors to be introduced. The use of a prospective design with improved data collection methods may assist in providing more reliable results. Assumptions that Sodium Chloride 0.9% was used for line flushes were made which may not reflect circumstances where Heparin Sodium was used for this purpose. Lastly, the study did not capture sodium intake from dietary sources, which is controlled by the Dietetics Department within our institution.

This retrospective study suggests patients with HFrEF admitted to a tertiary hospital network, did not receive inadvertent sodium from medicines, crystalloid therapy or diluents greater than current national and international daily recommendations. Pharmacist vigilance reviewing the amount of sodium administered is a safety initiative to ensure patients with HFrEF do not receive excessive administration of sodium from inadvertent sources. Further study into health professionals understanding of the amount of sodium in commonly used medicines, formulations and intravenous fluid therapy, may inform translational strategies to ensure patients do not receive excess sodium inadvertently whilst admitted in hospital.

None.

Supplementary materials can be found via https://doi.org/10.59931/rcp.23.009.