A Simple Approach for Gas Blender on Extracorporeal Membrane Oxygenation in Resource Shortage Context

Abstract

With the massive influx of patients during COVID-19 pandemic into intensive care unit, resources have quickly been stretched to the limit, including extracorporeal membrane oxygenation (ECMO). Gas blender attached to ECMO is used to allow precise adjustment of characteristics of fresh gas flow, that is, blood oxygen delivery and carbon dioxide removal. To cope with the gas blender shortage, we describe a back-up system set up in our French tertiary referral ECMO center using air and oxygen flowmeters. A table has been created to facilitate medical prescription but also nurse monitoring. This extraordinary situation forces physicians to adapt medical devices, and that could be useful in future viral pandemics.

How to Do It Article

Since the end of 2019, a new viral disease caused by the SARS-CoV-2 coronavirus has emerged.¹ COVID-19 can lead quickly to a severe acute hypoxemic respiratory failure. Selected patients may benefit from rescue therapies such as extracorporeal membrane oxygenation (ECMO).^2-4 To ensure accurate fraction of inspired oxygen in circuit (FsO₂, i.e., sweep gas inlet oxygen fraction) delivery and blood carbon dioxide removal, an air/oxygen blender is usually attached on usual ECMO devices (Figure 1A).⁵ Due to this unique crisis situation, intensive care unit departments had to quickly adapt with uncommon situations of equipments. To alleviate the shortage of devices, those from nonexpert ECMO centers, stored in case of local emergency and those initially used for transport, were transferred to ECMO centers. A significant proportion did not have gas blender. To avoid full continuous oxygen administration and lack of sweep, a back-up system had to be established. In Henri Mondor University Hospital, a French tertiary referral ECMO center, we had to face a blender shortage due to 53 ECMO implantations among 338 COVID-19 critically ill patients (at the date of March 28), with sometimes, as many as 24 patients on ECMO in the same day in our 120 COVID-19 intensive care unit beds departments. We present here a basic but effective installation to replace usual gas blender.

Figure 1.: A: Usual gas blender. B: Back-up system. C: Schematic diagram of back-up system. In this example, an oxygen flow of 3 L/min plus an airflow of 1 L/min delivers a total sweep flow of 4 L/min, and 80% FsO₂.

From wall medical gas outlets, air and oxygen flowmeters were connected together in a “Y” shape and derived to the ECMO membrane (Figure 1B). Tubing was designed to be as short as possible to ensure better gas delivery. All components met “CE” directives, thus complying with European Union safety standards. A schematic diagram summarizes the setup (Figure 1C). By adjusting the gas flow of each flowmeter, amount of oxygen delivered and appropriate sweep could be chosen precisely. These two parameters were calculated by following formulas:

To facilitate medical prescriptions and nurse monitoring and promote patient safety, a pragmatic and easy calculation table has been created (Table 1). For example, 3 L/min of oxygen flow at 1 L/min airflow equals to 80% FsO₂ and 4 L/min sweep. Conversely, to deliver approximatively 50% FsO₂ and 5 L/min sweep, 2 L/min of oxygen flow and 3 L/min of airflow should be administered.

Table 1. - Calculation Table for Fraction of Inspired Oxygen in Circuit (FsO₂) Without Air/Oxygen Blender

Oxygen Flow (L/min)	Airflow (L/min)	0	1	2	3	4	5	6	7	8	9
	FsO2 (%)
0		0	21	21	21	21	21	21	21	21	21
1		100	61	47	41	37	34	32	31	30	29
2		100	74	61	53	47	44	41	39	37	35
3		100	80	68	61	55	51	47	45	43	41
4		100	84	74	66	61	56	53	50	47	45
5		100	87	77	70	65	61	57	54	51	49
6		100	89	80	74	68	64	61	57	55	53
7		100	90	82	76	71	67	64	61	58	56
8		100	91	84	78	74	70	66	63	61	58
9		100	92	86	80	76	72	68	65	63	61

Up to now, no easy-to-use flow adaptation has ever been published. Twenty-nine patients benefited from this back-up system. Due to the severity of these patients, high sweep flow rates were required and we administered up to 15 L/min of sweep flow without adverse effects, even although an upper limit of flow rate was not possible to precise. System development showed disconnections in two patients with immediate reconnection with high gas flows (12 and 15 L/min, respectively), quickly resolved thanks to installation of pipes with diameters adaptive to the outflow of the flowmeters. As it is impossible to create some combinations, this back-up system is not optimal and should be restricted to resource limitation contexts, but it allows a better adaptation of characteristics of fresh gas flow. In addition, with this device, precise FsO₂ delivery measurement into oxygenator was not performed and was based on a calculation formula. Length of this device can be a limit to a precise oxygen delivery, notably in case of inequal gas flow between the two flows that may create Venturi effect. However, in the current context, we preferred to administer the most precise oxygen flow and sweep rather with a direct connection, that is, without sweep and with 100% FsO₂ as used in some ECMO transport situations, although being the safest for the patients. To largely validate the device, FsO₂ measurement would obviously be mandatory. However, our protocol included for a daily postoxygenator blood gas analysis. Partial pressures of arterial oxygen were within the expected ranges in all cases.

Despite France has a well-developed healthcare system, COVID-19 put some hospitals, especially in Paris area, in front of major shortage. We had to increase our 7 ECMO pumps stock to 24 pumps, thanks to other Parisian or German centers. Some of them were given without blenders, confronting us to practical problems that were also experienced by other French ECMO centers. Budgetary restrictions in recent years have not allowed for full stock’s renewal, and we had to adapt as quickly as possible. Our system could be an interesting adaptation for all healthcare systems if ECMO were to become more necessary in recurrent or new viral pandemic.

Even in a major crisis situation, and pending the publication of forthcoming studies, it is probably reasonable to adapt adequate gas flows to avoid hyperoxemia and to control capnia variations. Indeed, impact of hyperoxemia in patients on ECMO is still debated.⁶ COVID-19 pandemic was so overwhelming that even expert ECMO centers had to adapt in resources in such limiting crisis conditions, present or future.

Acknowledgment

We warmly thank Henri Mondor hospital perfusionists team: Françoise Arnoult, Nicolas Blochet, Audrey Caurant, and Jean-Luc Ermine, colleagues of Thomas Vincent, who all daily cared for these back-up devices as well as all cardiac surgeons and ICU teams.