Some complications of oxygen therapy in patients with COPD are too well known to require elaboration. The precipitation of hypoventilation with high FIO2 is one example of a well established (and well known) fact. While researching oxygen administration recently, I came across a couple of interesting articles. One study (1) looked at the effect that breathing 100% O2 had on patients with COPD. They concluded that FEV1 was thereby decreased by about 6% and that the effect lasted at least 5 minutes. Comparing this result with a mixture of gases containing oxygen at room air concentration (0.21) but having a viscosity and density approximate to that of pure oxygen the authors concluded that it was this feature that caused a decrease in expiratory flow rates. I.e., the higher density and viscosity of 100% oxygen compared to room air caused increased airways resistance. Alternative explanations such as relief of hypoxic pulmonary vasoconstriction did not explain the comparable effects of the higher density (21% O2) gas mixture in comprimizing expiratory flow. Reflecting upon this and recalling my experience with the sudden deterioration that I had witnessed on several occasions in asthmatics given albuterol nebulizer treatments have made me wonder whether this effect was operative in those cases as well. The neb treatments are given at an oxygen flow rate of 15 L/min. I had written a post a few months ago questioning whether this sudden dterioration was caused by allergic reactions to albuterol and provided several refernces from the literature describing such occurences. A second study examined the effects of sudden discontinuation of O2 therapy on ABGs. Two groups of patients were studied. Group 1 had COPD with chronic CO2 retention and group 2 were asthmatics with hypoxemia and respiratory alkalosis. Both were given oxygen by mask (60%) for 60 minutes. The oxygen was then abruptly discontinued and ABGs measured periodically for 45 minutes. Mean baseline PO2, PCO2 in group 1 was 49 and 58 respectively. At the end of the the hour of oxygen administration they had increased to 184 and 68. Ten minutes after cessation of O2 both PCO2 and PO2 were no longer significantly different from baseline. However, although PCO2 remained at baseline levels the PO2 continued to fall and was significantly lower at 30 and 45 minutes (45 and 44). The reults for group 2 were nearly identical: Baseline PO2 = 63 PCO2 = 35; at the end of O2 administration PO2 = 223 PCO2 = 38; PO2 then fell back to the baseline level by ten minutes. At 45 minutes PO2 had fallen to 57. Thus both patients with and those without CO2 retention experienced statistically significant _undershoots_ 45 minutes after cessation of O2 therapy. The authors concluded that the most likely explantion for this phenomenon was the persistence of altered V/Q ratios caused by oxygen that persisted after it had beem discontinued. As a corallary to this study they concluded that optimally one should wait 15 mintues before obtaining ABGs in patients with chronic hypercapnia and 10 minutes in those without (asthma). Considering that O2 will fall further in both groups one has to wonder about this recommendation. Oxygen....you can't live with it (1), you can't live without it (2). H. Louzon MD (1) Johnson et. al. Forced Expiratory Flow is Reduced by 100% Oxygen in Patients With Chronic Obstructive Pulmonary Disease. South Med J 1995; 88(4):443-448 (2) Rudolf et. al. Changes in Arterial Blood Gases After a Period of Oxygen Breathing in Patients With Chronic Hypercapnic Respiratory Failure and in Patients With Asthma. Clin Science 1979;57:389-396 Date sent: Fri, 5 Jan 1996 16:02:27 -0600 (CST) From: Harvey Louzon To: emed-l@itsa.ucsf.edu Subject: emed-l oxygen delivery devices (VERY LONG) NOTE: The following is a very long post about oxygen delivery devices. If you are not interested, engage your e-mail delete services immediately. Recently I posted some comments about the respiratory response to hypercapnia. While looking through some physiology texts I noted that minute ventilation, in some cases, could exceed 100 L/min (LPM). I then recieved a private query as to how a non-rebreather mask (NRB) could provide 100% O2 when the patient's minute ventilation was greater than 15 LPM since that is the maximum flow rate generally available through a standard wall source. I surmised that the NRBs do not provide 100% in those circumstances. Although the reservoir may provide a temporary store of O2 when ventilation exceeds wall oxygen flow, it cannot do so indefinately, otherwise the bag will collapse. The resevoir bag will ultimately need to be replenished by oxygen from the wall source and a limited flow rate (15 LPM). Thus any sustained minute ventilation of greater than 15 LPM would have to be accomplished by entraining some room air and thus lowering the inspired oxygen concentration. Intrigued by this result I began asking our respiratory therapists questions. They tell me that only rarely is it not possible to keep the resevoir bag inflated. It is also possible to attach a Y-adapter to the patient's oxygen source and to attach a second high flow source of oxygen if needed. Out of exasperation one of our techs-in-training finally lent me their respiratory care textbooks and I looked into the matter..... There are two basic kinds of oxygen delivery systems: A variable performance system does not supply all of the patient's volume or flow needs. Some room air is thus entrained to supplement it and thus the percentage oxygen delivered will vary from patient to patient and from time to time in the same patient. A fixed performance system is designed to supply all of the patient's gas needs and thus supplies a fixed FIO2 irrespective of oxygen flow rate (within limits). There are 3 basic kinds of implementations: 1) A low flow system such as nasal canula or simple face mask is a variable performance system. A reservoir system (such as a 'venti' or non-rebreather mask) MAY be a fixed performance system but only if the following two conditions are satisfied: 1) volume must be sufficient to meet the patient's peak inspiratory flow rates and 2) No ambient air is entrained. We will see shortly that resevoir systems (in practice) are variable performance systems and thus do not provide a fixed percentage of oxygen. High flow systems provide ALL of the patient's inspiratory gas needs and a stable FIO2. They do this by providing flows of at least 50 - 60 LPM and in some cases up to 100 LPM. Note that, depending upon the design, they may or may not provide 100% O2. A venturi mask is such a system but can only reliably provide FIO2s up to about 40%. Now on to the specifics. A. Nasal Canulas: Typical of low flow systems they provide a variable percentage of oxygen. This may be approximated by the following equation: FIO2 = (FR + 0.21*(V-FR))/V Where FR is the flow rate of 100% O2 and V is the patient's inspiratory flow rate. The latter may be approximated by V = MV*(I:E) Where MV is the patient's minute ventilation and I:E is the sum of the patient's observed inspiratory, expiratory ratio. Example: Let's suppose TV = tidal volume = 600 ml, RR (respiratory rate) = 15. Then MV = TV*RR = 9 LPM. If the patient has an I:E ratio of 1:2 then inpsriation is occuring only 1/3 of the respiratiry cycle and the maximum flow rate would be (2+1)*9 L/min = 27 L/min. Let's plug in a flow rate of oxygen of 2 LPM above. FIO2 = (2 + 0.21*(27-2))/27 = 27% Note that the FIO2 will decrease if the patient's I:E ratio increases or, at a constant I:E ratio, if minute ventilation increases. This situation is complicated somewhat by the fact that an 'anatomic reservoir' exists in the upper airway which serves to increase the inspired oxygen content when the patient's ventilatory requirements exceed administered flow rates. Nevertheless, an aprroximation to the FIO2 administered as a function of flow rate is as follows: FIO2(%) = 21 + 3*FR The factor of the last term may be as high as 4 if the anatomic resevoir is providing signifcant O2 and if the patient's minute ventilation is not too high. Thus each increment in flow (LPM) will increase FIO2 by about 3 - 4% points. Note that the precise percentage that nasal canulas deliver will vary with the phase of repiration that the patient is in. I.e., very early or late in the inspiratory phase flows are very low and thus modest O2 flow rates (2 - 4 LPM) may be suffcient to provide 100% O2. In mid-inspiration peak flow rates are substantially greater and considerable room air is mixed in with the oxygen reducing O2 concentration. How does this compare to experience? A study whose results I will repeatedly refer to looked at delivered oxygen percentages for various devices (1). Nasal cannulas at 4 LPM delivered an _average_ of 30% O2. The range over several subjects ranged from a low of 27% to a high of 36%. The effect of nasal versus oral breathing patterns with nasal cannulas in a healthy volunteer was studied (3). With both normal respiratory rates and with hyperventilation, tracheal oxygen concentration was significantly greater with nasal as opposed to oral breathing. As expected hyperventilation at a fixed oxygen flow rate resulted in lower tracheal oxygen concentration than did normal vnetilation for both breathing patterns. B. Simple Face Masks. A simple face mask functions by effectively extending the size of the 'anatomic reservoir'. Oxygen flow rates of 5 - 6 LPM minimum are necessary to prevent the mask from becoming a CO2 reservoir, as well, due to expired gases. A study done on healthy volunteers looked at the effect of oxygen flow rate on CO2 rebreathing (2). It was found that, compared to breathing room air, flow rates of less than 5 LPM resulted in progressively increasing minute ventilations to maintain a normal CO2. Recall that, because CO2 is such a potent respiratory stimulant, minute ventilation increasees to prevent a rise in pCO2 in normal persons. Of course, in diseased states this would impose an inordinate amount of increased work upon already overtaxed respiratory muscles and CO2 retention would supervene. A simple mask at 6 LPM flow rate provides an average of 40% O2 with a range of 38% to 46% (1). C. Venturi Masks. The term 'venturi mask' is a misnomer. These are actually 'air entrainment' masks, the venturi principle based upon the Bernoulli equation has nothing to do with it's operation (4). The venturi principle relies upon a suction effect created by a moving stream of fluid to cause 'mixing'. An automobile carburetor is (was) an example of this. Bt contrast 'venturi' masks _do_not_ rely upon pressure differences to mix a constant proportion of air and oxygen. Rather this mixing occurs at constant pressure and is due to viscous forces acting upon the the ambient air induced by the high velocity oxygen flow. These devices consist of a jet (through which oxygen is piped), a mixing chamber and air entrainment ports. Percent oxygen delivered is inversely related to the size of the air entrainment ports. For example at a 24% O2 setting the manufacturer of our masks reccommend 3 LPM oxygen flow rate. This is achieved by mixing oxygen at 3 LPM with air at 25.3 LPM (i.e., (3*1+0.21*25.3)/(25.3+3) = 0.24. There is a total of 25.3 + 1 = 26.3 parts of gas. Thus the total flow rate is 26.3*3 LPM = 79 LPM. As long as the patient's minute ventilation is less than or equal to 79 LPM he will recieve 24% oxygen _irrespective_ of his respiratory rate or tidal volume. This provides a fairly comfortable margin. On the other hand consider what happens when the mask is set to deliver 50% O2. In this case 1.7 parts of air is entrained for each part pure oxygen. At 15 LPM recommended oxygen flow setting the total gas flow would thus be (1.7+1)*15 LPM = 41 LPM. This level is easily exceeded by a patient who is hyperventilating as I indicated in a previous post. How is the differnce made up? By entraining _more_ air then is required to sustain a 50% oxygen concentration. I.e., the actual percentage oxygen delivered would be LESS than 50%. Of course, if the patient's minute ventilation were to remain at or below 41 LPM then precisely 50% O2 would be delivered. Up to about 35% the delivered and predicted oxygen percentages coincide fairly well. Consider the actual delivered O2 as a function of that which is predicted (1): FIO2 O2 Flow Mean FIO2 Range 0.24 4 0.25 0.24 - 0.26 0.28 4 0.28 0.27 - 0.28 0.35 8 0.35 0.34 - 0.36 0.40 8 0.38 0.36 - 0.40 The following table lists predicted FIO2, air:oxygen ratio, recommended O2 flow and total gas flow for the type of ventimask we have available in our ED: FIO2 Air:O2 ratio Recommended O2 Total Gas Flow 24% 25.2:1 3 LPM 79 LPM 26% 14.8:1 3 LPM 47 LPM 28% 10.3:1 6 LPM 68 LPM 30% 7.8:1 6 LPM 53 LPM 35% 4.6:1 9 LPM 50 LPM 40% 3.2:1 12 LPM 50 LPM 50% 1.7:1 15 LPM 41 LPM The last column is, once again dervived by adding 1 + the air:O2 ratio and multiplying by the recommended O2 (in LPM). Any time the patient's minute ventilation exceeds the total gas flow, additional room air is entrained and the actual O2 delivered will be less than predicted. This is likely to occur above 35 - 40%. Thus the 'venturi' mask is a reliable fisxed performance device for delivering oxygen concentrations up to about 35 - 40%. Above this level increasing disparity will be seen between predicted and delivered oxygen concentration particularly in patients with high minute ventilations. D. Partial Rebreathing Masks. The rationale behind the rebreather is, once again, to provide a reservoir of highly concentrated oxygen to augemnt the anatomic reservoir whenever minute ventilation temporarily exceeds oxygen flow. Oxygen flow rate should be adjusted to keep the reservoir about 2/3 full at all times. Some of the expired air is diverted into the bag. Theoretically, this (dead space) air is high in O2 and very low in CO2. It is said that for oxygen flow rates of greater than 5 - 6 LPM CO2 contamination of the reservoir is minimal. About 1/3 of the exhaled volume is delivered to the bag and, with subsequent increase in pressure in the bag and mask, the remaining 2/3 is vented to the atmosphere. FIO2s range from about 35 - 60% at 6 - 10 LPM oxygen flow rate. E. Non-rebreathing masks (NRB). It was to (finally) answer the question of the actual percentage of oxygen delivered by NRBs that I went on this expedition in the first place. It did not seem possible, theoretically, to deliver 100% O2 using an oxygen source at 15 LPM, even with a reservoir, to sustain minute ventilations that might exceed 100 LPM. This is borne out in practice. Non-rebreather masks deliver an average of 63% O2 with a range of 57% to 70% (1). Finally, how accurate do you think the rotameters are in delivering the indicated O2 flow rate? One study (5) found that at 2 LPM flow they actually delivered 1.6 to 2.4 LPM (a 40% range). At 8 LPM they were accurate to within about a 15% range. All of the above studies were done in healthy volunteers. I am not sure how this could be extrapolated to patients with diseased lungs, however. Would their minute ventilations exceed the upper limit of gas flow of these devices? Probably. H. Louzon MD (1) Redding et. al. Oxygen Concentrations From Commonly Used Delivery Systems. South Med J. 1978;71(2):169-172 (2) Jensen et. al. Rebreathing During Oxygen Treatment With Face Mask: The Effect of Oxygen Flow Rates on Ventilation. Acta Anes Scan. 1991;35:289-292 (3) Poulton et. al. Tracheal Oxygen Concentrations With a Nasal Cannula During Oral and Nasal Breathing. Resp Care 1980;25(7):739-741 (4) Scacci R. Air Entrainment Masks: Jet Mixing is How They Work; The Bernoulli and Venturi Principles Are How They Don't. Resp Care 1979;24(10):928-931 (5) Sray et. al. Accuracy of Inpatient Oxygen Administration. Thorax 1989;44:1036-1037