Extracorporeal membrane oxygenation (ECMO) is an established treatment for patients with severe pulmonary and/or cardiac failure unresponsive to conventional treatment. Its goal is to provide time for treatment and recovery from the underlying disease process, as a bridge to decision-making1 or to transplantation.2 ECMO is complex both medically and technically, requires a specialised team and equipment and its complications can be devastating. Accordingly, extensive monitoring of patients undergoing ECMO is essential.
We report the simultaneous and recurring malfunction of two new miniaturised ECMO devices due to environmental influences.
During the 2010/2011 influenza A (H1N1) pandemic, 14 patients suffering from adult respiratory distress syndrome (ARDS) due to community-acquired H1N1 infection were treated in our ICU using ECMO. In February 2011, we introduced into clinical practice a new ECMO device (Cardiohelp; Maquet Cardiopulmonary AG, Hirrlingen, Germany). This device is exceptionally small (315 × 255 × 427 mm), lightweight (<10 kg) and can be carried by a single person. Its cube-shaped disposable contains a centrifugal pump, a gas-exchanger/oxygenator and also provides several integrated pressure sensors and a blood temperature probe. Furthermore, an ultrasound flow sensor, which also serves as a sensitive bubble detector, and a sensor for continuous photometric measurement of haemoglobin concentration, haematocrit and venous oxygen saturation can be attached to the venous drainage line (Fig. 1).
In March 2011, three patients in our ICU simultaneously underwent ECMO for H1N1-associated ARDS. On one day at about 12.00 h, we encountered simultaneous alarms of two of the three machines that were located in adjacent rooms. Both devices failed to measure haemoglobin concentration, haematocrit and venous oxygen saturation and no numbers were given for the respective measurements. One of the nurses in charge complained that both ECMO devices had given alarms for the same type of malfunction for a few days exactly 1 h before her shift changeover. The third ECMO device was in operation in a patient's room on the other side of the ICU and did not malfunction. We unsuccessfully attempted to alleviate the problem by recalibration of the probes according to the manufacturer's instructions. However, despite recalibration, the same malfunction occurred repeatedly around noon each day. Curiously, when a colleague by chance shaded the sensor of one device for a few seconds while trying to silence the alarm, this alone silenced the alarm and re-enabled measurements.
The same malfunction was encountered a few weeks later during transfer of a patient with severe ARDS on ECMO from a referring hospital. When sunlight fell on the probe through the back window of the ambulance, alarms went off and measurements were disturbed. The problem was resolved by covering the window.
The observed malfunction of the ECMO device's photometric sensors occurred repeatedly around noon and was alleviated by shading the sensor. Our ICU is located on the ground floor of the medical centre and most of the windows face south (Fig. 2a). Although the windows are slightly opaque in their lower part to prevent pedestrians from looking into the ICU rooms, bright sunlight shines into the rooms around noon. This is true particularly during winter months when the sun is low above the horizon (Fig. 2b). It became apparent that, when the sun shone on the sensor, malfunction ensued.
Our report on sunlight-associated ECMO malfunction demonstrates that the spectrophotometric measurement probe of the Maquet Cardio Help ECMO can easily be disturbed by direct sunlight. This probe has three light-emitting diodes (LEDs) that produce light at three different wavelengths in the visible and infrared spectra. We were unable to reproduce the sunlight effect by attempting to produce malfunction with an LED flashlight (InGaN blue LED with Ce3+:YAG coating) shining directly at the sensor; malfunction may be due to light of a specific wavelength. Several ambient light sources can generate the wavelengths sensed by pulse oximeter probes. However, it has been shown that different ambient lights (incandescent, quartz-halogen, infrared, fluorescent and bilirubin light) do not significantly affect pulse oximeter measurements.3 The spectrum of quartz-halogen and incandescent light ranges from 650 to 1000 nm, the spectrum of infrared lamps begins at 700 nm, bilirubin light peaks around 200 to 400 nm and fluorescent light at 405 to 579 nm.3 In contrast to ambient light, sunlight is composed of ultraviolet, visible and infrared light, with specific wavelengths of about 100 to 400 nm, 380 to 780 nm and more than 700 nm, respectively. Furthermore, the specific situation of the probe included in the ECMO circuit has to be considered. Whereas pulse oximeters measure ambient light during both static and pulsatile phases and use these measurements to null the incident energy from an ambient energy source,4 ECMO commonly uses continuous flow, making this method useless in abolishing errors caused by ambient light.
A representative of the company that manufactures the device was contacted and details of the phenomenon, as well as the possibility of solving the problem by covering the sensor, were described. The phenomenon had not been observed previously according to Maquet.
Although other possibilities to measure haemoglobin concentration, haematocrit and oxygen saturation exist in an ICU setting, no fallback methods are available during interhospital transfer. To solve this issue, the manufacturer should provide a modification of the probe. In other devices, similar phenomena could be obviated by mounting the sensor on top of the tubing, a measure impossible with the fixed sensor position on this particular device. However, covering the sensor might help to re-enable the sensor with this device. Although measurements were made completely inoperable by sunlight in our cases, one might speculate whether different light conditions might affect measured variables and lead to erroneous values displayed by the photometric sensors of this or other devices. Accordingly, intensivists should be aware of sun-evoked erroneous alarms, particularly at high noon.
Assistance with the study: none declared.
Financial support and sponsorship: none declared.
Conflicts of interest: FH received honoraria from Maquet for lectures.
1. John R, Liao K, Lietz K, et al. Experience with the Levitronix CentriMag circulatory support system as a bridge to decision in patients with refractory acute cardiogenic shock and multisystem organ failure. J Thorac Cardiovasc Surg
2. Marasco SF, Lukas G, McDonald M, et al. Review of ECMO (extra corporeal membrane oxygenation) support in critically ill adult patients. Heart Lung Circ
2008; 17 (Suppl 4):41–47.
3. Fluck RR, Schroeder C, Frani G, et al. Does ambient light affect the accuracy of pulse oximetry? Respir Care
4. American Association for Respiratory Care. AARC Clinical Practice Guideline: pulse oximetry. Respir Care