OSI is the oxygenation saturation index. Is is normally defined as [Fio2 × mean airway pressure × 100)/oxygen saturation by pulse oximetry (Spo2)] and predicts outcomes of mechanically ventilated patients. In the case of HFNC patients, MAP may be estimated by the level of flow delivered, but no data is available about its utility.
There is probably no single variable that reflects the response to the treatment. I think that different things happen when the patient is doing well: oxygenation improvement, decrease in respiratory rate, relief in dyspnea feeling… Regarding the right flow, we know that most of the effects are flow-dependent and, therefore, when we start the treatment in paitents with acute hypoxemic respiratory failure, we try to use the highest tolerated flow. However, we can't start with 60Lpm as the patient does not tolerate it. So we start with 40Lpm and once the patient is used to receiving this amount of flow, we can progressively increase up to 60Lpm. This increase can usually be made in the first 30 minutes of treatment.
(Editor's note: "aspects" has been understood as "variables" for the purposes of this answer) Clinical examination, respiratory rate, use of accessory muscles, thoraco-abdominal asynchrony, SpO2, FiO2
There is no specific timeframe for expected improvement. However, it is true that some thresholds of different variables have been described as predictors of HFNC failure at different time-points.
The use of accessory muscles suggests that the inspiratory effort is excessive. Similarly, low PaCO2 or a negative swing in CVP could also suggest the same. (Please also see the second webinar for an answer to this question.)
I do not wait. If the patient is not responding to the treatment, I try to increase the flow up to the maximum tolerated. And if the patient is still not responding, one would need to escalate the treatment.
The evidence is sometimes controversial because the criteria for intubation may vary a lot between different countries, hospitals or even doctors in the same ICU. Thus, some studies compared early versus delayed intubation taking the time of ICU admission as moment 0. The majority of them have shown that earlier intubation is associated with better outcomes. In other words, delayed intubation may be associated with increased mortality.
My suggestion would be not to base the decision of intubating a patient based only on a number. The clinical examination of the patient is extremely important. The ROX may help you to decide if the patient is doing well or not, as you can repeat the measurement several times. The benefit of the ROX is that it is based on physiological variables that determine the outcome (need for intubation). We proposed an algorithm that may help in a review in ICM with Jean-Damien Ricard in 2020 that we are now testing in a RCT.
In our clinical practice we rarely do it. There is a good correlation between SpO2 and PaO2 if you keep the SpO2 < 98%.
These patients usually need lower flows and benefit more from active humidification that improves secretion clearance. But I would base my decisions in the same way as for HFNC patients.
As commented before, I would never make a decision based on just a number. I think that the ROX value should be combined with the clinical examination of the patient.
The contents of this page are for informational purposes only and are not intended to be a substitute for professional training or for standard treatment guidelines in your facility. The responses to the questions on this page were prepared by the respective webinar's speaker; any recommendations made here with respect to clinical practice or the use of specific products, technology, or therapies represent the personal opinion of the speaker only, and may not be considered as official recommendations made by Hamilton Medical AG. Hamilton Medical AG provides no warranty with respect to the information contained ion this page and reliance on any part of this information is solely at your own risk.
