Nitric oxide supply systems

Nitric oxide is officially a member pf the "Pulmonary vasodilators", a fascicle bundled with other pharmacology topics from Section 2.1.21 in the second edition of the CICM Syllabus for the Second Part Examination.  However, this chapter is all about the nitric oxide cylinder, which is the form this gas is usually found in, and to enhance the relevance of it to the modern reader, the large proprietary delivery apparatus attached to it. Before you ask, yes you do need to know this. This has come up at least once, in  Question 28.3 from the first paper of 2011. Moreover, one does not wish to ever be in the position of a grown-ass intensivist so confused by an unfamiliar technology that they are forced to call on NICU nurse educators to help coordinate the process of establishing an adult patient on this therapy. One would prefer to confidently swagger through that scenario, like it's something done every day.

It is indeed done every day in some places. Specifically, one author (Rogerson et al, 2021) drove a local quality improvement project to reduce their institutional use of the gas, and celebrated a decrease from 1515 hours month to 930 hours per month (across three busy paediatric ICUs). 

The gas tank itself

Question 28.3 from the first paper of 2011 presented the candidates with an image depicting a characteristically coloured branded tank. 

This is the product offered by Ikaria, a subsidiary of Mallinckrodt Pharmaceuticals (the pharma giant also historically involved in the processing of uranium and the state-authorised sale of cocaine). 

The gas comes in a pressurised cylinder, under a brand name "INOmax". It contains 99.92% nitrogen and only 0.08% nitric oxide, or 800 parts per million (ppm).  The tank is usually coupled to a proprietary delivery system that charges the hospital by the hour (in NSW, the cost seems to be fixed at $70 per hour in 2023 dollars, irrespective of the dose being delivered). And if the reader is at this point wondering whether they are being overcharged for basically compressed nitrogen with impurities:

You can't, like, own a gas, man.

Except with enough helpful regulation, you absolutely can. Consider that in the distant past, which is to say the 1990s, Ikaria did not exist, and hospitals acquired their nitric oxide much in the same way they acquired their other medical gases, i.e directly from industrial vendors. Tibballs et al (1993) recall buying it from Commonwealth Industrial Gases Limited, which rebranded to BOC a year later. Those guys also supplied it as a NO-N₂ mixture at either 220 or ~1000 ppm. The gas was then blended with the oxygen at the ventilator low pressure inlet using a Y-piece connector and an analogue flowmeter, such that the final desired  gas concentration was achieved (eg. for a 20 ppm end mixture,  Tibballs et al suggested blending NO and O₂ at a roughly 1:9 ratio).

This was, of course, inelegant. For a speciality that prides itself on carefully deliberate treatment decisions, any administration directions that in their precision resemble the recipe for a cake could not be acceptable. It should be noted that nitric oxide is a microdose agent;  the highest concentration you would ever use is about 80ppm, or 0.008% by volume, which corresponds to a gas mixture of 90% whatever, and 10% Inomax. A higher dose brings all sorts of problems (for example, it transforms to nitric oxide in combination with oxygen and water), and a lower dose leaves one wondering whether the agent is having any effect whatsoever. Given the precarious flowmeter math used to calculate the rate of administration and the lack of measurement options for NO or N₂O this status quo could not carry on for very long. By the mid-2000s, when Ikaria incorporated and gained approval for INOmax, most regulatory authorities had lost patience with the risk profile of the older delivery methods. One should also point out that buying random tanks of the gas from the back of a truck was not significantly cheaper: Petros et al (1995) noted a $4,722.85 cost per six months of use (about 234 hours), which works out about $10,362.61 after adjustment for inflation (about 148 hours of use at modern rates). Interestingly, the gas only accounted for about 10% of that cost- the rest was the cost of the monitoring sensors and other equipment, much as it is today. 

The nitric oxide delivery system

The system is essentially a narrow stalk on wheels, with a monitor and gas blender at the top, and gas tanks as ballast at the bottom, preventing the tall narrow device from tipping over. 

The on switch is at the back:

Also at the back are the two hose adaptors for the two cylinders at the bottom of the device (there is wisdom in having both attached) and the outlet for free gas flow. The latter is not going to the INOmax device - instead, it goes to this thing:

INOblender. An unpretentious object clamped under the main device. This is the thing the 1990s intensivists wished they had: a regulator device which, presented with an unregulated oxygen flow, will titrate the flow of nitric oxide to achieve the specified dose in ppm, as specified on the control knob. The INOblender manual suggests that this is, if anything, less accurate than the old methods, as they only guarantee +/-20% accuracy in the dose of NO it delivers. This, of course, matters very little, because if you are at the point where you are manually ventilating the patient with problems that require this product, then the times are bad indeed, and a small difference in dose may not play much of a role.

The gas tubes connect to two large cylinders at the back. The tubing is thinner than what you will be used to from working with medical air and oxygen supply because those gases are under much greater pressure when they leave their regulator, whereas the nitric oxide regulator assembly both looks and behaves differently:

Downstream from the regulator, the DSIR Plus manual specifies that the gas outlet pressure is around 170-240 kPa, as opposed to the 13,000 kPa at which the gas is stored. These gas cylinders at the back will usually be the "88" type, a 15.4L cylinder that contains 1918 usable litres of compressed NO. At 800 ppm concentration, a peak dose rate of 80ppm and a ventilator minute volume of 10L/min  gives you a cylinder gas delivery rate of 1L per minute, which means on a full tank you get about 1918 minutes, or 32 hours, of continuous gas supply.

The rest of the gas delivery system can be crudely summarised in this diagram:

The nitric oxide/nitrogen mixture is piped into the ventilator tubing before the humidifier, via a gas injection module that looks like this:

The gas injection module has such a beefy power supply because it is in fact a hot wire anemometer in the path of the ventilator gas flow, for measuring the rate of fresh gas delivery. This is the reason it needs to be upstream of the humidifier: the humid gas mixture could interfere with the measurement. The rate of delivered fresh gas mixture is then factored into the rate of nitric oxide delivery, thereby preventing the accumulation of nitric oxide between breaths or the drop in concentration with tachypnoea. Prior to this innovation, experimental evidence suggests that the tidal variation in NO delivery was something in the order of 5 to 90 ppm.

Things to worry about while using nitric oxide

The following adverse effects have been reported with its use:

• Methemoglobinaemia, as abundantly discussed already
• Hypotension (maybe some of it does leak into the systemic circulation, or maybe this the effect of depressed LV function
• Rebound hypoxia after abrupt withdrawal
• Thrombocytopenia (in as many as 10% of patients)
• Increased susceptibility to pulmonary infections probably due to NO₂ formation and associated lung injury

There are also several contraindications.

• Left ventricular failure: NO seems to cause lots of adverse effects in this group of patients- particularly, pulmonary oedema. In fact, halfway through one study, the investigators had to start excluding CCF patients from the trial because of these effects.
• Left to right shunting: NO will decrease the pulmonary (and thus the RV) pressure, increasing the amount of blood shunted via a septal defect.
• Uncontrolled haemorrhage: Though there is no human evidence, in animal studies NO had been shown to increase bleeding times.
• Existing methaemoglobinaemia: obviously, it will get worse.