Inductive Warming of Intravenous Fluids: Overheating of the Toroid Heating Element during Rapid Infusion

A 42-yr-old, 80-kg male with chondrosarcoma of the left pelvis was scheduled to undergo elective left internal hemipelvectomy. His medical history and physical examination results were unremarkable. After placing standard American Society of Anesthesiologists monitors, general anesthesia was induced and maintained without incident. Based on historic precedent, large fluid shifts were anticipated, which prompted placement of a right radial arterial catheter as well as 8.5-French and 14-gauge antecubital intravenous catheters. Central venous pressure was continuously monitored via  the distal port of a right internal jugular 9-French multilumen cordis. A Belmont FMS 2000 Rapid Infusion Device (Belmont Instrument Corporation, Billerica, MA) was connected to the 8.5-French antecubital catheter to deliver warmed fluids throughout the procedure.

The case proceeded uneventfully through the first 9.5 h of the operation. Despite receiving 12 units packed erythrocytes, 11 l crystalloid, 12 units fresh frozen plasma, 6 units platelets, and 2.5 l albumin, 5%, the patient remained warm (36.9°C at the conclusion of pelvic resection) and hemodynamically stable. During fascial closure of the wound, the surgeon observed a “dusky” left leg. Doppler examination disclosed absence of a femoral pulse, necessitating emergent femoral artery exploration and bypass.

Despite having just completed a highly invasive orthopedic surgical procedure, 5,000 units intravenous heparin was administered, as requested by the vascular surgeon, resulting in an activated clotting time of more than 300 s. Ten minutes after heparin administration, hemostasis became increasingly difficult, and hemorrhage was evident in the surgical field. Despite fulminant hemorrhage, central venous and systemic arterial blood pressures were maintained within 20% of baseline using the Belmont FMS 2000 to deliver an additional 8 units of packed erythrocytes and 3 l saline, 0.9% (flow rate ranged from 100 to 500 ml/min). However, near the conclusion of the resuscitative effort (and after the surgeons had regained hemostasis), we noted an acrid smell reminiscent of burning plastic that coincided with an abrupt profound hypotensive (blood pressure 110/50 → 50/30 mmHg) and tachycardic (heart rate 97 → 130 beats/min) episode that was readily treated with 200 μg intravenous phenylephrine. Both the blood pressure and the heart rate returned to preevent levels in less than 2 min. At the nadir of this brief but dramatic hemodynamic change, the overheat alarm sounded and automatically shut off the Belmont warming unit. The disposable tubing insert was removed, was noted to be disfigured, and seemed to have charred material in one half of the heating coil (figs. 1A and B). The warming unit itself appeared grossly undamaged, but it was promptly replaced with another Belmont FMS 2000 with a new disposable insert.

Subsequent fluid warming proceeded uneventfully, and the patient remained hemodynamically stable and warm (esophageal temperature 37.0°C at the time of discharge from the operating room). He was transferred to the intensive care unit, where serum electrolytes and urine output remained within normal limits. He was extubated on the second postoperative day and discharged to a rehabilitation facility on postoperative day 8, without adverse sequelae.

Rapid resuscitation with large volumes of unwarmed intravenous fluid and blood products may result in life-threatening hypothermia. In response to this concern, the medical equipment industry has developed several fluid warming devices that are capable of delivering large volumes of temperature-specific intravenous fluids at high rates of infusion. Unlike conductive warming, which requires physical contact between the intravenous fluid and the warming device, inductive warming occurs in the absence of physical contact between the intravenous fluid and warming element. The Belmont FMS2000 is an inductive fluid warmer capable of heating intravenous fluids to 37.5°C at flow rates ranging from 60 to 500 ml/min and to 39°C at rates of less than 60 ml/min.2With this device, intravenous fluid passes through a toroidal-shaped (donut-shaped) heat exchanger by a built-in peristaltic pump.2A magnetic field is created by a coil at the center of the heat exchanger.2Energy from the magnetic field is transferred to the steel rings of the heat exchanger and then to the intravenous fluid.2Infrared temperature sensors are located on both the input and output ports of the heat exchanger.2If the output sensor detects a fluid temperature of 42°C or greater, the system is designed to—and did so in our case—shut down and alarm, thereby preventing thermal-related injury.2 

Malfunction of fluid warming devices has previously been reported.3–6However, regional overheating of the toroid is a unique observation. In hope of elucidating the etiology of malfunction, we sent the disposable warming element to the manufacturer for evaluation. They too noted charred material in one half of the heating coil (i.e. , in the vicinity of melted plastic; figs. 1A and B), whereas other areas were devoid of obvious abnormality. Uncertain as to the composition of the discolored material noted in the toroid, the manufacturer undertook studies to determine whether clot formation could have occurred. They reported such an occurrence was only possible when procoagulant solutions (e.g. , calcium, clotting factors, cryoprecipitate, or platelets) were administered through their device. However, being cognizant of this potential pitfall, procoagulant solutions were administered through an alternate fluid warming device (i.e. , Hotline fluid warmer, Model REFHL-90DC; Smiths Industries Medical Systems, Rockland, MA) during the above-mentioned operation. Such practice is in keeping with the clinical guidelines at our institution. Therefore, we doubt that the charred material resulted from clot formation or improper use. Also, under extreme testing conditions, the manufacturer determined that the melting point of the plastic housing was 100°C. Aside from providing this information, the manufacturer was unable to clearly identify the cause of overheating in our case.

Because the damaged portion of the toroid was likely exposed to temperatures of 100°C or greater (i.e. , at or beyond the melting point of the toroid), another consideration for the source of toroidal discoloration was thermal injury to formed elements in transfused blood. In regard to the latter, human erythrocytes are capable of withstanding temperatures near 45°C (i.e. , a value significantly greater than the maximal output temperature of commercially available fluid warmers) for prolonged periods of time with little effect on cellular integrity.7Denaturation of erythrocyte membrane proteins is responsible for increased osmotic fragility and hemolysis after exposure to temperatures exceeding 45°C.8–13The percentage of hemolyzed blood cells increases linearly as a function of exposure duration and temperature magnitude.8,10,12Of interest, on review of the patient’s laboratory studies, two arterial blood gas samples were noted to be hemolyzed during and soon after the period of hemodynamic instability. Ironically, all other samples drawn from the same arterial line either before or thereafter were not reported as hemolyzed specimens.

Unless leukocyte filtration is performed before or during transfusion, leukocytes are also present in units of packed erythrocytes. To our knowledge, the temperature of leukocyte lysis has not been reported. However, there is concern that thermal-mediated leukocyte free-radical production, complement activation, and release of vasoactive mediators (e.g. , prostaglandins, leukotrienes, interleukins, bradykinins, tumor necrosis factor, oxygen free radicals, and cytokines) from thermally lysed or degranulated leukocytes may contribute to physiologic perturbations such as hypotension.14 

As stated above, when the output sensor detects temperatures of 42°C or greater, the device shuts down, thereby preventing thermal injury. However, temperature within the toroid itself is not monitored. Therefore, it is plausible that formed elements in blood exposed to nonphysiologic extreme temperatures (i.e. , ≥ 100°C in malfunctioning half of the warming element) lysed, releasing vasoactive mediators. We speculate this may have been the cause of the brief, but profound hypotensive episode rather than an abrupt change in preload, as the central venous pressure was maintained within 20% of baseline and surgical control of hemorrhage had already been achieved. We envision that prevention of this vasoactive mediator exposure was not possible because the output sensor detects temperature of fluid mixed from both halves of the toroid. That is, admixture of fluid within the distal toroid resulted in effluent temperatures of less than 42°C until the temperature within the malfunctioning half of the toroid was so extreme that “cooler” fluid (from the normal functioning region of the toroid) was no longer able to moderate outgoing fluid temperature. Accordingly, up to this endpoint, flow through the toroid may have flushed vasoactive mediators into the patient’s circulation, resulting in hemodynamic instability.

In summary, we report a unique observation in which the heating element of an inductive fluid-warming device experienced overheating (i.e. , temperature ≥ 100°C) during rapid infusion. Without prompt intervention, such an occurrence could result in patient injury.