Secondary Logo

Journal Logo


O2, anybody?

Hendrickx, Jan F.A.; De Wolf, Andre M.; De Hert, Stefan

Author Information
European Journal of Anaesthesiology (EJA): June 2015 - Volume 32 - Issue 6 - p 371-373
doi: 10.1097/EJA.0000000000000089
  • Free

Hypoxic guard systems have been developed by anaesthesia machine manufacturers to help avoid the administration of hypoxic gas mixtures. The original hypoxic guard systems, developed in the 1970s and 1980s, used mechanical [Link-25 Proportion Limiting System; GE (formerly Datex-Ohmeda)] or pneumatic (ORMC and S-ORC; Dräger Medical, Lubeck, Germany) links between the oxygen (O2) and nitrous oxide (N2O) flow metres. These links apportion the fresh gas flows of O2 and N2O in such a manner that a hypoxic gas mixture cannot be delivered at the common gas outlet. It was only later, around 2000, that the Anesthesia Machine Standards were published, including the American Society of Testing Materials, Standard F1850-00, and the ISO and IEC Standard EN60601-2-13, clause 51.102.2. These standards require the presence of a system that prevents the unintentional selection of a mixture of O2 and N2O with an O2 concentration less than 21%. However, these standards have some serious flaws: they are limited to O2/N2O mixtures, and they focus on the O2 concentration delivered at the common gas outlet, not on the inspired oxygen concentration (FIO2). Nevertheless, these original (hereinafter ‘conventional’ for clarity) hypoxic guard systems use more stringent criteria than are required by the standards: Link-25, an O2 concentration delivered at the common gas outlet at least 25%; S-ORC, an O2 concentration delivered at the common gas outlet at least 25% and a minimal O2 fresh gas flow of 250 ml min−1. The latter results in a progressively increasing O2 concentration delivered at the common gas outlet with progressively lower fresh gas flows less than 1 l min−1 (Fig. 1a). More recently, the same S-ORC thresholds have been implemented through software in the electronically controlled anaesthesia workstations from Dräger (Zeus, Primus). Contrary to the mechanical/pneumatic version, this electronically controlled S-ORC is also active with O2/air mixtures. As well as the conventional hypoxic guard, currently, three modern anaesthesia machines have an automode that utilises a sensor to measure FIO2 and thus allows automatic control to ensure a nonhypoxic FIO2: the Aisys (GE, Madison, WI, USA), the Zeus (Dräger, Lübeck, Germany) and the FLOW-I (Maquet, Solna, Sweden).

Fig. 1
Fig. 1:
Conventional ‘delivered’ hypoxic guard limits (oxygen concentration as delivered at the common gas outlet) for three modern low flow anaesthesia machines for (a) O2/N2O and (b) O2/air mixtures. The conventional hypoxic guards make it impossible to select a combination of fresh gas flows (FGF, X-axis), which result in an O2 concentration (Y-axis) below the limits depicted by the lines in the graph: o Zeus (Dräger, Lübeck, Germany); • Aisys (GE, Datex Ohmeda, Madison, Wisconsin, USA); Δ FLOW-i (Maquet, Solna, Sweden). These lines, the so-called conventional hypoxic guard limits, differ not only for each machine but also between O2/N2O and O2/air mixtures. These limits are active; regardless of whether the machines are used in target control mode or not, the machines will not use fresh gas flow, O2 combinations that fall below these limits. When these machines are used in nontarget-controlled mode, the conventional ‘delivered O2’ hypoxic guards do not ensure FIO2 >21% across the entire fresh gas flow range; they are bound to fail in the 0.7 to 3 l min–1 fresh gas flow range (see text for details).

The fact that conventional hypoxic guard thresholds vary among anaesthesia machine manufacturers (Fig. 1a, b) suggests that these hypoxic thresholds have been chosen somewhat arbitrarily. Worse still, they have never been tested clinically, but they should have been: all the conventional hypoxic guard systems and the standards that describe them are to some degree flawed because they do not take into account the current trend towards lower fresh gas flows. What happens when fresh gas flows are lowered? More than 10 years ago, we showed that, when using O2/air mixtures in a circle system with low fresh gas flows, in order to prevent an FIO2 less than 21%, the O2 concentration delivered at the common gas outlet has to be increased significantly.1 When fresh gas flows decrease rebreathing of exhaled gases will cause the concentrations of the inspired gas mixture to be lower than the delivered concentrations. This is true for all gases and vapours including O2, N2O, N2, xenon, and the potent inhaled anaesthetics. Because conventional hypoxic guard systems do not sufficiently take this into account, they are bound to fail to prevent the creation of a hypoxic FIO2 at low flows. The performance of a modern electronic conventional hypoxic guard system was recently tested in the clinical setting (S-ORC, Dräger Zeus, set in the manual/conventional/nontarget-controlled mode, i.e. the anaesthesia provider selects the fresh gas flow and delivered concentrations) by some of the authors of this editorial.2 The results are startling: the S-ORC did not prevent the development of a hypoxic FIO2, and this at fresh gas flows that are only moderately reduced (0.7 to 3 l min−1). Calculation of mass balances easily predicts the development of a hypoxic FIO2. One could argue that a hypoxic FIO2 should immediately result in an alarm (as we all measure FIO2), but the anaesthesia provider may be confused as to its cause, because ‘we all know that the conventional hypoxic guard systems prevent the development of a hypoxic gas mixture’. In other words, the user has a false sense of security, potentially delaying proper intervention. It is worth emphasising that other conventional hypoxic guard systems are likely to be similarly flawed.

Where do we go from here? Do we require even more stringent conventional hypoxic guard criteria? This is likely not a very good option, because mass balances predict that even more stringent criteria may not prevent a hypoxic FIO2 when, for example, patient oxygen consumption is higher than normal. Also, the configuration of the anaesthesia circle system may influence the effects of rebreathing. Therefore, we propose that all manufacturers develop ‘smart’ hypoxic guard systems on the basis of software and measured FIO2 (or possibly the end-expired O2 concentration). With current anaesthesia machines, there are four different circumstances to be considered. First, the modern anaesthesia workstation with an electronic gas mixer and in automatic (target control) mode allows us to set a target FIO2 (or end-expired O2 concentration), and its algorithms will set the required individual fresh gas flows using feedback control from measured FIO2. This is the ultimate solution because in this mode, there is no need for a conventional hypoxic guard system. Currently, only three such workstations are available: the Aisys, the Zeus and the FLOW-i. Second, such a workstation, when used in a manual (nontarget-controlled/conventional) mode, should have an electronic (‘smart’) hypoxic guard system that will, when necessary, adjust fresh gas flows of the individual carrier gases based on measured FIO2 or end-expired O2 concentration, thereby overruling the anaesthesia provider. Currently, only one such workstation is available, the FLOW-i. If FIO2 decreases below 21%, within 20 s, the system will increase the fresh gas flows and the O2 concentration delivered at the common gas outlet, restoring FIO2 to at least 25% within 73 s after its activation.3 Third, if anaesthesia machines with electronic conventional hypoxic guard systems, such as the Dräger Primus, cannot be modified to incorporate an active inspired oxygen (‘smart’) hypoxic guard, then the conventional hypoxic guard thresholds should be increased, especially in the 0.7 to 3 l fresh gas flow range, and this for all carrier gas combinations. Finally, anaesthesia machines with classic glass rotameters constitute the Achilles, heel because these cannot be retrofitted with an electronic hypoxic guard system that can adjust individual fresh gas flows. There are few options in this case. Teach! Make people aware that inspiratory hypoxic mixtures can still occur despite the presence of a hypoxic guard system, and make alarms smarter. Still, it is clear that future anaesthesia workstations should take advantage of the added safety that modern technology has to offer by using target-controlled FIO2 (or target-controlled end-expired O2 concentration), or by utilizing smart, electronic hypoxic guard systems, based on measured FIO2 or end-expired O2 concentration.

Even more disconcerting is how it is possible that machine standards dealing with hypoxic guard systems have been written the way they are? Even the latest standards (ISO are no improvement because no clear-cut distinction is made between delivered and inspired concentrations. How is it possible that we, as a specialty, have missed the fact that these standards are bound to fail? Although these are relevant questions, what is important to our patients now is that the anaesthesia community be educated, and that the current standards be rewritten to meet the highest possible safety standards. Machine manufacturers and anaesthesia providers share a common responsibility here to improve patient safety: the manufacturers should include modern safety systems, and anaesthetists should insist that such systems be acquired. O2, anybody?

Acknowledgements relating to this article

Assistance with the editorial: none.

Financial support and sponsorship: none.

Conflicts of interest: JH has received lecture support, travel reimbursements, equipment loans, consulting fees and/or meeting organizational support of basically all companies involved with inhaled agent delivery (alphabetically): AbbVie, Acertys, Air Liquide, Allied Healthcare, Armstrong Medical, Baxter, Dräger, GE, Hospithera, Heinen und Lowensein, Intersurgical, Maquet, MDMS, MEDEC, Micropore. Molecular, NWS, Philips, Quantum Medical.

Comment from the editor: this editorial was checked and accepted by the editors but was not sent for external peer review. SDH is an associate editor of the Eur J Anaesthesiol.


1. Hendrickx JF, De Cooman S, Vandeput DM, et al. Air-oxygen mixtures in circle systems. J Clin Anesth 2001; 13:461–464.
2. De Cooman S, Schollaert C, Hendrickx JF, et al. Hypoxic guard systems do not prevent rapid hypoxic inspired mixture formation. J Clin Monit Comput 2014; [Epub ahead of print].
3. Ghijselings I, Hendrickx J, De Wolf AM, et al.. Performance of the inspired hypoxic O2 guard. Posters 2014 NAVAt meeting. - accessed 8/1/2015.
© 2015 European Society of Anaesthesiology