Review of: Davis DP, Hwang JQ, Dunford JV: "Rate of Decline in Oxygen Saturation at Various Pulse Oximetry Values with Prehospital Rapid Sequence Intubation." Prehospital Emergency Care. 12(1):46-51, 2008.
This study comes to us from data collected during the San Diego Paramedic RSI Trial. A ground and an air medical agency were each equipped with pulse oximetry and capnography that recorded and stored their readings during the RSI procedure. This trial involved the administration of midazolam (ground service) or etomidate (air medical), both followed by succinylcholine for paralysis. The medics were encouraged not to perform bag-valve mask (BVM) ventilation following the administration of succinylcholine to avoid gastric insufflation.
The authors defined desaturation as SpO2 less than 90%. They attempted to determine whether there was an optimal oxygen saturation point at which if you began the RSI procedure you would avoid dropping below 90% during the period of apnea induced by the paralytic.
Of the patients with initial SpO2 greater than 90%, 83% were intubated. Of that 83%, 6% suffered desaturation events. The mean number of intubation attempts was 1.5, and 59% of patients requiring more than one attempt suffered desaturation events. A majority of desaturation events -- 86% -- occurred if their initial SpO2 was less than 93%.
The authors conclude "The rate of SpO2 decline increases as SpO2 decreases, with an inflection point occurring around 93%. Intubation attempts below this value are almost always associated with subsequent desaturation, suggesting that BVM should be used prior to laryngoscopy in these patients."
We do so much in EMS and medicine based purely on speculation. Much of our knowledge as it relates to hypoxemia and intubation comes from the anesthesia literature, which isn't applicable to our practice. The San Diego Paramedic RSI Trial continues to provide more insight into the physiology of out-of-hospital medicine.
We now recognize that "hyperventilation" is not appropriate but that "preoxygenation" is. But to what value? This study is the first to clearly show the results of performing RSI at various starting points on the desaturation curve. Does this data apply to your practice? In some ways it should apply to all. Simply put, perform BVM ventilation until oximetry is greater than 93% before administering the paralytic. However, unlike San Diego, not all programs prohibit BVM following paralysis. Instead, they rely on the Sellick maneuver and on being careful not to exceed 20cmH20 pressure.
But what if the initial saturation is less than 93%? How often have we boldly gone ahead with the procedure confident all would go well? The authors have set a benchmark of performance. Is it accurate? The only way to know is for more services to duplicate it, which shouldn't be difficult as we adopt the principle that all airway management should be guided by oximetry and capnography.
With this being said, there are difficulties incorporating this information into all cases. What if you're unable to obtain oximetry due to poor perfusion or a host of other conditions for which the device cannot operate accurately? Secondly, what were the end-tidal CO2 levels obtained during this study? I suspect this data will be forthcoming to further aid us in the critical decision making that is required. Until then, use the information you have, and give your patient every opportunity for a positive outcome from what is a very negative experience.