Stop Mucking Around and Throw the End Tidal On: A Safer Way to Perform Procedural Sedation and Analgesia!

This is part of a recurring series examining landmark articles in Emergency Medicine, in the style of ALiEM’s 52 Articles.

Discussing: “Does End-tidal Carbon Dioxide Monitoring Detect Respiratory Events Prior to Current Sedation Monitoring Practices?” Burton, J. Harrah, J. Germann, C. Et al. Acad Emerg Med, May 2006, 13(5): 500-4


Screen Shot 2015-09-13 at 5.42.38 PM


Main Points:

  1. This trial was stopped early after an interim safety analysis demonstrated that standard care was potentially placing patients at higher risk for delayed recognition of acute respiratory events. 17 out of 20 acute respiratory events recorded in the dataset demonstrated changes in the end tidal carbon dioxide (ETCO2) level indicative of hypoventilation or apnea.
  2. Twenty acute respiratory events were recorded in the study cohort of 60 procedural sedation and analgesia (PSA) encounters. In 14 of these episodes, the ETCO2 demonstrated changes prior to SpO2 change or clinically observed hypoventilation.


PSA is a routine practice that is critical to many tasks in the emergency department. There are many guidelines from various professional societies that discuss the components of PSA and associated requirements for safe patient monitoring. Many clinicians use a wide array of pharmacologic options based on patient characteristics when performing PSA, yet it is always challenging to predict how each individual will respond to the drugs used. Therefore, having an extra tool for patient monitoring that will improve safety is a great aid for quality patient care.

ACEP has issued multiple guidelines over the past decade to keep up with the evolution of clinical practice as well as technologic changes such as capnography. Currently ACEP has a level B recommendation for the use capnography as part of the PSA monitoring tools. The article by Burton and his colleagues is one of many studies that helped change practice patterns over the past ten years. The authors collected a prospective convenience sample of two 30 patient blocks undergoing PSA at a single tertiary ED. There was an attending and 3 residents collecting ETCO2 data during PSA events and the clinical team was blinded to this information. The purpose was to identify the ability of ETCO2 to detect acute respiratory events before conventional strategies.


This trial was a prospective cohort trial with a convenience sampling time that was vaguely identified. It appears that the 4 study investigators needed to be on shift for the data to be collected but the timeframe during which this occurred is unclear. Acute respiratory events were defined as clinically significant if: SpO2≤92%; increases in the amount of supplemental oxygen provided in response to apnea, hypoventilation, or desaturation (local guidelines required every patient to receive a minimum of 2L via NC during PSA); use of BVM or oral/nasal airway for ventilator assistance or apnea; repositioning or airway alignment maneuvers for ventilation or apnea; physical or verbal stimulation for depressed ventilation or apnea; and reversal agent administration. Study measurements included a Ramsay score clinical assessment and patient monitoring data (SpO2, HR, RR, and ETCO2). Surprisingly, in this cohort of 60 encounters 20 acute respiratory events were noted and each respective encounter is detailed in Table 2. These range from desaturation with response to stimulation to four patients requiring BVM and airway repositioning. This rate of complication is more than double the predicted rate the authors had identified based on their previous experience with PSA and their institutional database. This degree of complication may be attributable to patient selection and therefore represent sampling bias. During PSA encounters there was continuous capnography available to the research clinicians, but the study monitoring data only recorded ETCO2 at 30 second intervals. The time of ETCO2 change and clinical recognition of acute respiratory events ranged from 0 to 271 seconds. This data collection method may have failed to fully capture the precise time of patient hemodynamic changes and the value of ETCO2.

Level of Evidence:

According to the ACEP grading scheme this trial receives a grade two level of evidence for having a prospective cohort design with minor methodological flaws.


At this tertiary ED the physicians were performing PSA not only for the routine orthopedic complaints like those seen at RIH, but also for TEE and disimpaction! Propofol was the predominant agent used, 68% of the time, however, ketamine came in second at 20%.

 Relevant articles:

ACEP PSA Guidelines, October 2013


Ramsay Score


Source Articles:

Burton, J. Harrah, J. Germann, C. Et al. “Does End-tidal Carbon Dioxide Monitoring Detect Respiratory Events Prior to Current Sedation Monitoring Practices?” Acad Emerg Med, May 2006, 13(5): 500-4


Anatoly Kazakin MD

1 thought on “Stop Mucking Around and Throw the End Tidal On: A Safer Way to Perform Procedural Sedation and Analgesia!

  1. Important article to consider. Thanks for the excellent review, Anatole. You can also wonder about whether it would be better to routinely monitor anyone who requires oxygen monitoring (pulse ox) with ETCO2, or even INSTEAD of pulse ox. How many times do you go in to an ED room and find the pulse ox on the floor? The nasal placement of the ETCO2 might make it less likely to be pulled off or yanked off. Maybe if initial pulse ox is ok, just put pt on ETCO2 to make sure there is not hypoventilation/apnea. We would have fewer false alarm monitors going off, and it may improve safety, as this study with sedation suggests. Maybe one of you bright young residents should design a study to test this! BZ

Leave a Reply

Your email address will not be published. Required fields are marked *