The literature search identified 166 unique articles which were screened initially by title search, then by review of the abstract. After exclusion of unrelated and duplicate articles, five articles were included in this report. Table 1 provides a summary of the included articles from our literature search.
<Insert Table 1 here>
Table 1
Results of literature search strategy
Set# | Searched for | # Records |
S1 | aluminum or aluminium | 965740* |
S2 | intravenous fluid$ | 330283* |
S3 | warmer | 1404574* |
S4 | (s1 and s2 and s3) | 3636° |
S5 | ((s1 and s2 and s3)) and (dates(2010-2029)) | 1287° |
S6 | ((s1 and s2 and s3)) and (dates(2010-2029) and (searchtype("Scholarly Journals"))) | 205° |
S7 | ((s1 and s2 and s3)) and (dates(2010-2029) and (articletype("Article") AND searchtype("Scholarly Journals"))) | 169° |
S8 | ((s1 and s2 and s3)) and (dates(2011-2021) and (articletype("Article") AND searchtype("Scholarly Journals"))) | 166° |
* Duplicates are removed from the search but included in the result count. |
° Duplicates are removed from the search and from the result count.
A study, led by Perl et al in 2019[5], compared aluminium release between an uncoated enFlow device with the Fluido® Compact (The 37°Company, Amersfoort, the Netherlands), a device that uses a parylene coating between the aluminium blocks and the infusate. The study evaluated aluminium release into normal saline (0.9% NaCl solution) or a balanced electrolyte solution (Sterofundin 1/1E) at various infusion rates. Fluids were run through with the heating units activated, then again with the heating units off. Their results clearly showed that the uncoated device leached aluminium into the infusate at higher levels than did the coated device. The amount of leaching was higher at slow infusion rates and when the heating units were activated. Notably, there was virtually no leaching of aluminium by either device into saline, in striking contrast to the high levels seen in the balanced electrolyte solution. This suggests that the balanced electrolyte solution demonstrated a more corrosive effect on the aluminium blocks than did the saline.
In January 2019, the Medicines and Healthcare products Regulatory Agency (MHRA) learned of research indicating that one of the uncoated fluid warming devices (enFlow®, Vyaire Medical, Mettawa, IL) was releasing unusual levels of aluminium when used to warm certain infusates. This led to a formal investigation by MHRA into the medical device concerned. Following this report, the MHRA released a Medical Device Alert[14] warning clinicians of these findings as reported by PIke.[8] Vyaire initiated a voluntary recall of all enFlow uncoated devices in March 2013.[15]
Also in 2019, Taylor et al evaluated the uncoated enFlow heating cartridge using a broad variety of fluids. Plasma-Lyte 148, compound sodium lactate solution, 4% human albumin solution, expired resuspended packed red cells, and fresh frozen plasma were pumped through the warming system at 2 ml.min−1.[9] Infusates were infused over 60 minutes and samples for analysis were collected before heating, then every 10 minutes with heating for 60 minutes. The investigators found that balanced salt solutions were associated with potentially toxic levels of aluminium leaching while saline was not. Blood products were associated with lower levels of aluminium leaching, although levels still exceeded the US Food and Drug Administration (FDA) recommended level of 25 mcg.l−1 (0.9 lmol.l−1) for intravenous nutrition.[16]
The Level 1® Fast Flow Fluid Warmer (Smith Medical, Minneapolis, MN) is a rapid infusion device that utilizes a passivated anodized aluminium heat block to warm passing fluids. This device was evaluated in 2020 by Cabrera et al using normal saline, lactated Ringer’s, and heparinized whole blood at a constant flow rate of 30 ml.min− 1 over the course of 1 hour.[17] Samples were collected every 10 minutes and each fluid was evaluated in triplicate. The amount of aluminium released into saline was negligible, and that released into whole blood initially was high but dropped to undetectable levels after 20 minutes. Aluminium levels in lactated Ringer’s solution rose steadily over the one hour of infusion but did not reach toxic levels as defined by the FDA[18] and the Agency for Toxic Substances and Disease Registry.[19] In separate Letters to the Editor, both Perl[6] and Exley[20] disagreed with the assumptions of Cabrera’s study, and Exley asserted that the aluminium-based fluid warmers were not safe for use in humans. Cabrera responded by defending the assumptions and conclusions in his study.[21]
Perl et al compared aluminium elution using the uncoated enFlow warmer heating fluids including saline, with or without combinations of sodium acetate, sodium lactate, and with or without sodium hypochlorite or hydrochloric acid, infused at 4 ml.min− 1 over 60 minutes.[22] There was no elution of aluminium in pure normal saline, while the addition of either acetate or lactate significantly increased aluminium leaching at a level that increased over time. In their second experiment, the Level 1® H-1025 with DI-300 disposable tubing, and the ThermoSens® (Barkey GmbH & Co. KG, Leopoldshöhe, Germany), a low flow warmer that uses a parylene coated aluminium heating chamber as a heat exchanger, were compared using alkalinized lactate-spiked saline or Sterofundin®, balanced electrolyte solution. Results between the enFlow and the Level 1 device were nearly identical, while the parylene-coated ThermoSens resulted in undetectable levels of aluminium elution.
The Food and Drug Administration (FDA) subsequently issued a letter to health care providers warning that multiple fluid warmers have been restricted or recalled.[23] Included among the recalls was the uncoated enFlow® intravenous fluid warmer which was subsequently redesigned to include a parylene coating over the heating block.[7, 24] According to the FDA, Vyaire performed a complete product recall in 2019.[23] Two years later, the other companies had changed their Instructions For Use (IFU).
In an analysis by Waldmann et al, the team measured the concentration of aluminium that leached into three solutions (Sterofundin ISO, Plasma-Lyte 148, and whole blood) which were continuously pumped (0.2 and 5.5 mL.min−1) and warmed to 40°C for 5 h using the parylene-coated enFlow cartridge. In addition, prolonged quasi-static bench tests were performed that measured aluminium concentration in 16 different clinically relevant solutions which were gently rocked within the enFlow cartridge (parylene-coated) for 72 h at 40°C. It was determined that there was virtually no elution of aluminium resulting from the parylene coated enFlow, even in fluids high in lactate.[7] Despite the additional layer between the actual heating surface and the fluid, studies have found that there is no negative effect on the performance characteristics of the enFlow device.[25]
A summary of the literature search is presented in Table 2.
Table 2
Summary of literature search on aluminium elution from aluminium block intravenous fluid warmers
Author, year | Device | Fluid | Infusion rate | Duration | Findings |
Perl, 2019[5] | enFlow (uncoated) | Saline | 2, 4, and 8 ml.min−1 | 60 min | Very low aluminium elution |
Sterofundin | 2, 4, and 8 ml.min−1 | 60 min | Very high elution of aluminium, higher at low flow, higher when warming vs. room temperature |
Fluido Compact (coated) | Saline | 2, 4, and 8 ml.min−1 | 60 min | Very low aluminium elution |
Sterofundin | 2, 4, and 8 ml.min−1 | 60 min | Very low aluminium elution |
Taylor, 2019[9] | enFlow (uncoated) | Saline | 2 ml.min−1, heated and unheated | 60 min | Very low aluminium elution |
Distilled water | 2 ml.min−1, heated and unheated | 60 min | Very low aluminium elution |
Distilled water with acetic acid | 2 ml.min−1, heated and unheated | 60 min | High elution of aluminium, higher when warming vs. room temperature |
Plasma-Lyte 148 | 2 and 16.6 ml.min−1, heated and unheated | 60 min | Very high elution of aluminium, higher at low flow, higher when warming vs. room temperature |
Compound sodium lactate | 2 ml.min−1, heated and unheated | 60 min | Very high elution of aluminium, higher when warming vs. room temperature |
Expired packed red blood cells | 2 ml.min−1, heated and unheated | 60 min | High elution of aluminium, higher when warming vs. room temperature |
Fresh frozen plasma | 2 ml.min−1, heated and unheated | 60 min | High elution of aluminium, higher when warming vs. room temperature |
4% human albumin solution | 2 ml.min−1, heated and unheated | 60 min | High elution of aluminium, higher when warming vs. room temperature |
Cabrera, 2020[17] | Level 1 Fast Flow Fluid Warmer | Saline | 30 ml.min−1 | 60 min | Undetectable |
Lactated Ringer’s | 30 ml.min−1 | 60 min | Did not reach toxic levels as defined by FDA |
Heparinized whole blood | 30 ml.min−1 | 60 min | Undetectable |
Perl, 2021[22] | enFlow (uncoated) | Saline | 4 ml.min−1 | 60 min | Very low aluminium elution |
Saline with sodium acetate | 4 ml.min−1 | 60 min | High aluminium elution |
Saline with sodium lactate | 4 ml.min−1 | 60 min | Very high aluminium elution |
Saline with sodium lactate, acidified | 4 ml.min−1 | 60 min | High aluminium elution |
Saline with sodium lactate, alkalinized | 4 ml.min−1 | 60 min | Very high aluminium elution |
Saline, acidified | 4 ml.min−1 | 60 min | High aluminium elution |
Sterofundin | 4 ml.min−1 | 60 min | Very high aluminium elution |
Level 1 | Saline with sodium lactate, alkalinized | 4 ml.min−1 | 60 min | Very high aluminium elution |
Sterofundin | 4 ml.min−1 | 60 min | Very high aluminium elution |
ThermoSens | Saline with sodium lactate, alkalinized | 4 ml.min−1 | 60 min | Very low aluminium elution |
Sterofundin | 4 ml.min−1 | 60 min | Very low aluminium elution |
Waldmann[7] | enFlow (parylene coated) | Plasma-lyte 148 | 0.2 and 5.5 ml.min−1 | 5 hours | Undetectable |
| | Sterofundin ISO | 0.2 and 5.5 ml.min−1 | 5 hours | Undetectable |
| | Whole blood | 0.2 and 5.5 ml.min−1 | 5 hours | Undetectable |
| | Sixteen challenge solutions* | Quasi-static soak | 72 hours | All below toxic exposure levels |
* Sterofundin ISO; Plasma-Lyte 148; single donor human whole blood; human packed cells; Ringer’s lactate in 5% dextrose; human platelet lysate; human buffy coat; human plasma diabetic type 2; 5% dextrose solution; 3% sodium chloride injection; human serum albumin 25%; normal human serum off-the-clot charcoal-dextran 1; human cord blood; leukocytes; potassium chloride in 5% dextrose and 0.9% sodium chloride; and 10% dextrose and 0.45% sodium chloride. |
<Insert Table 2 here>