Manual Dilution of Continuous Renal Replacement Therapy Fluids in Anuric Patients with Severe Hyponatremia: A Case Report and Systematic Review

Patients with severe hyponatremia and the necessity of renal replacement therapy (RRT) are at risk for a too rapid correction of the sodium levels with neurological consequences. Various strategies for a slow correction of sodium levels are available, but dilution of sodium in dialysis and replacement fluids allows for a controlled increase in sodium, while allowing a sufficient renal replacement dose for clearance [1‒3]. Several papers demonstrated the theoretical basis of this strategy by means of calculation [4, 5]. Still, in clinical practice factors other than RRT dose and sodium concentration may influence the rise in sodium, such as fluid removal, body water content, and concomitant infusions [6]. Furthermore, the fluids may have to be diluted manually, which could lead to errors. It is therefore unsure if this dilution method leads to a controlled rise in sodium in clinical practice.

In this paper, we present a case in which dilution of dialysis and replacement fluids was used to allow RRT while slowly correcting the hyponatremia. We add a systematic review of individual patient data to investigate the rise in serum sodium levels during continuous renal replacement therapy (CRRT) with diluted replacement and dialysis fluids. We aim to describe the rate of sodium increase as a measure of safety and elucidate the factors that contribute to that rate. The CARE Checklist has been completed by the authors for this case report, attached as online supplementary material (for all online suppl. material, see https://doi.org/10.1159/000543718).

A 70-year-old, 70 kg and 1.62 m woman, with a history of hypertension and mammary carcinoma presented in the emergency department with complaints of diarrhoea, fatigue, and malaise since 2 weeks. On evaluation, she was somnolent, incoherent, and had a Glasgow Coma Scale (GCS) of E3M6V4. Her vitals were normal except for a blood pressure of 235/99 mm Hg. There was some oedema present in both legs, but no signs of pulmonary oedema on chest X-ray. Blood analysis showed a creatinine of 968 µmol/L, urea of 41.8 mmol/L, sodium of 100 mmol/L, potassium of 4.5 mmol/L, phosphate of 3.6 mmol/L, C-reactive protein of 203 mg/L, and a serum osmolality of 263 mOsmol/kg. Urine analysis was obtained from the only 10 mL of urine and demonstrated leukocytes, erythrocytes, and proteinuria. Urine sodium was 29 mmol/L and urine osmolarity was 340 mOsmol/kg. Abdominal ultrasound showed no hydronephrosis and a size of 12.8 cm and 12.5 cm of the right and left kidney, respectively. Further analysis demonstrated an anti-glomerular basement membrane (GBM) antibodies titre of 692 E/mL.

She was admitted to the internal medicine department with acute kidney injury and hypotonic hyponatremia. Correction of the hyponatremia was attempted with 300 mL 3% saline, 80 mg furosemide, 4 L 0.9% saline, and 500 mg oral sodium hydrogen carbonate over the first 24 h. The patient appeared to be anuric and the serum sodium increased to 107 mmol/L. Due to respiratory compromise as a result of fluid overload, the patient was transferred to the intensive care unit (ICU) to start with an intermittent session of ultrafiltration without dialysis of 800 mL in 2 h, which resulted in a sodium of 108 mmol/L. The next day continuous hemodiafiltration (CVVHDF) with regional citrate anticoagulation (Prismaflex with an ST 510 filter, Baxter, USA) was initiated for both clearance and a slow correction of sodium. The 5 L commercially available replacement and dialysate fluids (Regiocit^® and Biphozyl^®, Baxter, USA) were diluted with sterile water in order to clamp the sodium rise. Initially, 1 L of water was added to the 5 L bags of fluids, to obtain a sodium concentration of 117 mmol/L. This was determined in the laboratory on another occasion. CVVHDF was started with a dose of 37 mL/kg/h, a 50/50% division between dialysis/filtration and fluid removal was added. During therapy, sodium levels were monitored, and the dilution of dialysis and replacement fluids was adjusted accordingly, in order to obtain a slow increase of 5 mmol/L sodium per day. Figure 1 and online supplemental Data File A show the CVVHDF settings and sodium levels of dialysate and replacement fluids over time, with the resulting serum sodium levels.

On the 6th day of CVVHDF, it was deemed safe to continue without diluted fluids for the serum sodium was 130 mmol/L. During 6 days of CVVHDF, 14 L of fluid was removed. CVVHDF was stopped and intermittent haemodialysis was initiated the day after. Clinically, the patient became more alert and her GCS normalized. Her mobility improved and she was discharged from the intensive care unit to the department of nephrology. She had a full neurologic recovery. She received no further treatment for the anti-GBM antibody glomerulonephritis since there was no pulmonary involvement and renal function was deemed unrecoverable, based on the severity of renal failure and anuria. Moreover, the possibility of side-effects and adverse events due to the immunosuppressant therapy was not deemed to outweigh the benefits of the treatment, if any.

The search with the terms “Renal Replacement Therapy OR Continuous Venovenous Hemofiltration OR Continuous Venovenous hemodialysis OR dialysis AND Hyponatremia OR hyponatremic OR low sodium” in the title or abstract, in Medline, Web of Science, and Google Scholar yielded 642 articles. Inclusion criteria were hyponatremia, CRRT with dilution of replacement or dialysis fluids and availability of individual patient data (i.e., at least one of the following: age, gender, course of hyponatremia, CRRT settings, and duration). There were no restrictions with regard to language or study design. Thirty-eight articles were deemed of interest for full article screening. Hereafter, 10 were excluded because no patient data were involved, 6 were excluded because no CRRT was used, 4 were excluded because they only described aggregate data, and 3 were excluded because the CRRT dialysis or replacement fluids were not diluted. Online supplementary Data File B presents the 15 remaining manuscripts in which replacement fluids or dialysate were diluted, consisting of 49 patients [7‒21]. The causes of renal failure varied and only 2 patients were reported to have sufficient diuresis. Continuous data of the systematic review were described as median (interquartile range).

In these 49 patients, all modalities of CRRT were used, most commonly CVVHDF (n = 33, 67%), but also CVVH (n = 6, 12%) and CVVHD (n = 10, 20%). Dilution of the dialysate and/or replacement fluids was accomplished using dextrose 5% (n = 2) or sterile water (n = 45) cases. Noteworthy was that there was one report of compounding by the pharmacy department [15] and one report used just two dilutions to increase safety [18]. The diluted dialysis or replacement fluids were adjusted over time to different sodium concentrations. If anticoagulation was mentioned, it was performed with citrate and heparin or without anticoagulation. During CRRT treatment, there were no reports of inadvertent increases in sodium or other adverse events.

To further investigate sodium increase over time, the data of all cases that reported a sodium difference over time per specific dilution were analysed, including the current case report [7‒13, 18, 21]. Of these 30 cases, 53 filter runs with specific dilutions were defined. Per filter run, the difference in serum sodium before and after the filter run, the difference between dialysate and replacement sodium versus serum sodium, the dose, and the fluid losses were collected. Table 1 demonstrates that with a median difference between serum/dialysate and replacement sodium of 8 mmol/L (5–11) (0–17) and a median CRRT dose of 27 mL/kg/h (22–30) (13–77.5), the hourly increase in serum sodium was 0.2 mmol/L/h (0.1–0.3) (0–0.7).

There were 10 filter runs in which the sodium increase exceeded the 0.4 mmol/L/h [7, 11‒15, 21]. The increase was intentional in 6 of these filter runs since there was a difference between serum and dialysate sodium of >10 mmol/L [7, 11, 15, 21]. In 2 filter runs, there were large volumes of infusion and transfusion during surgery with a large volume of blood loss and a CRRT dose >70 mL/kg/h [13] and in 2 filter runs there was no clear explanation, other than the small weight (6.6 kg) of the patient leading to a quicker equilibration between the CRRT fluids and serum [12]. The authors described in all these cases continuation of the increase in serum sodium, indicating that clinically the high increase in sodium was either intended or, at least, accepted.

We presented a case of a 70-year-old female with anti-GBM antibody glomerulonephritis, anuria, and severe hypotonic hyponatremia, who was treated with CRRT to gradually correct electrolyte disturbances and fluid overload. To accommodate this gradual increase in sodium, the commercially available dialysate and replacement fluids were diluted with sterile water to acquire a certain sodium concentration. A correction of the serum sodium of a maximum of 8 mmol/L per day was achieved with this treatment. A systematic review demonstrated that 49 patients were treated with CRRT with diluted commercial dialysis and replacement fluids within regular CRRT prescriptions, resulting in a median increase in sodium of 0.2 mmol/L/h without reports of adverse events. Dilution of commercially available dialysis and replacement fluids appears to be a viable option to slowly correct sodium levels during RRT.

The success and lack of adverse events depend on a correct dilution and safety measures to prevent and detect errors early. The majority of the studies reported a manual dilution scheme, from which health care providers dilute it themselves. Two safety measures were reported: dilution performed by the pharmacy department and the use of just two diluted concentrations. We believe that these are viable options, but they still do not guarantee a lack of error. We advocate to check for serum sodium levels at least every 4 h for sake of early error detection. In our case report, this safety measure was not followed 1 day and fortunately had no consequences. Still, we reinforced this safety measure after a local review of this case.

An alternative option for slowly correcting sodium during RRT is to perform RRT with a low dose [22]. Low-dose RRT may be an option if clearance is not acutely required. The safe dose is unknown but could be approximated using a calculation. For example, a 70 kg patient with a sodium of 100 mmol/L would have a body water content of 42 L (60% of body weight). With a RRT dose of 10 mL/kg/h, 700 mL of blood will be replaced with dialysis or replacement fluids of 140 mmol/L. Assuming equilibration between serum and the body, this would result in an increase of 8 mmol/L sodium in the first 12 h and 14 mmol/L in the first 24 h. Clearly, this is a logarithmic curve, leading to a larger increase in sodium initially. Furthermore, this calculation neglects changes in body weight, where fluid removal and diuresis would increase the change in sodium. It merely demonstrates that even with a highly reduced CRRT dose of 10 mL/kg/h, the increase in sodium would still be too much.

Another alternative is to administer concomitant hypotonic fluids during CRRT [2]. This would increase the water content, resulting in a gradual rise of sodium. However, it could also be regarded as a dilution of the CRRT sodium content in the venous system. Which would result in an uncertain sodium concentration when CRRT or infusion pumps are suddenly stopped for a longer period. Furthermore, it increases the volume status, unless fluid removal is added to the CRRT or diuresis is maintained. With the technique of diluting the sodium concentration in the dialysis and replacement fluids, these issues are corrected for. Adding hypotonic fluids may be an additional safety measure, but we believe it cannot safely replace dilution of the dialysis or replacement fluids.

We believe that several factors influence the rate of serum sodium increase, which are the difference between serum and CRRT-fluid sodium, CRRT dose, and concomitant fluid removal or infusion (i.e., changing the volume of distribution). This review over individual patient data with diluted CRRT fluids demonstrated a median increase of 0.2 mmol/L/h in sodium with a regular CRRT dose. This would result in a safe increase in sodium. There were 10 filter runs in which the sodium increase was over 0.4 mmol/L/h or 10 mmol/L/day, but this appeared expected for the majority of the cases and lead to no action which would decrease serum sodium. This indicates that there was no unacceptable high increase in serum sodium with this dilution strategy.

With this strategy, the concentration of molecules other than sodium is diluted as well. This is only important for molecules with high concentrations (i.e., sodium and chloride) since the diluted concentration of molecules with low concentrations (e.g., potassium, calcium, phosphate) would still remain in the clinical range. Regional citrate anticoagulation should be considered because if citrate is delivered as a predilution replacement fluid, it too is diluted. We recommend targeting circuit ionized calcium levels rather than using a standardized dose. Still, in our case the standard citrate-replacement rate obtained sufficiently low circuit ionized calcium levels, and it was not required to increase in rate to compensate for the lower citrate concentration. During the 6 days of treatment, 3 RRT-filters were used with a filter lifetime of 72, 44, and 21 h, indicating sufficient regional anticoagulation, since the last filter run was terminated based on the decision to stop CRRT.

There are several limitations of this paper one needs to consider. Mainly, all data are based on case reports, which are observational data with inherent limitations in reliability, selective reporting, and missing data. These data were extracted, pooled, and analysed to allow generalization for clinical care because prospective studies on this topic are unlikely to be feasible. Still, it should be interpreted with caution, even though we believe this to be the best level of evidence available at this point.

In conclusion, this case report and systematic review demonstrate that CRRT with manually diluted dialysate and replacement fluids in patients with severe hyponatremia and anuria can lead to a slow serum sodium increase. It allows a sufficient RRT dose and fluid removal while clamping the sodium level. Still, errors in dilution may occur and we recommend 4 hourly monitoring of serum sodium levels to early detect an inadvertent increase in sodium.

Ethical approval was not required in accordance with national regulation. Written informed consent was obtained from the patient for publication of this case report.

The authors have no conflicts of interest to disclose.

This study was not supported by any sponsor or funder.

M.V.K.: conceptualization, data curation, formal analysis, methodology, software, and writing the original draft. L.R.: data curation, formal analysis, and writing – review and editing. F.K. and E.S.R.: writing – review and editing.

All data generated or analysed during this study are included in this article and its supplementary material files. Further enquiries can be directed to the corresponding author.