Semler MW, Janz DR and Russell DW et al.
Chest 2017; 152(4): 712-722.
Doi: 10.1016/j.chest.2017.03.061
Background
Hypoxia complicates approximately 50% of intubations in ITU; these desaturations are associated with increased rate of cardiac arrest and death.
Question
Does patient intubation position alter time to intubation / incidence of hypoxia?
Design
- Multicentre, randomised control trial: Critically unwell patients in ICU
- Computer generated blocks of trial arms, opaque envelopes sealed until decision made to enrol patient.
- 248patients required for power of 80%
- Intention to treat analysis
Setting
- July 2015 – July 2016, 4 tertiary care centres (USA)
Population
- > 18 years old, intubated by critical care medical fellow, on ICU, with planned use of NMB / sedation.
- Most common reasons for intubation: sepsis, altered mental status, pneumonia
- Exclusion: Intubation required so urgently that unable to assign to study, and clinician decision re specific positioning required.
- 309 eligible, 260 randomised, 49 excluded (42 due to urgency, 6 due to specifically positioning required)
Intervention
- Intervention: Ramped position (15⁰ head up)
- Control: sniffing position (neck flexed, head extended)
Data Collection
- Independent observer (not involved in the procedure) (O2 saturations (SpO2), timing etc)
- Primary investigator: concurrent assessment of 10% of intubations
- Intubator: Cormack-Lehane grade – glottis view, subjective difficulty, airway complications
- Study personnel: From medical notes: baseline characteristics, pre/post intubation managements, and clinical outcome.
Outcomes
Primary outcome
Lowest SpO2 between induction and intubation: No statistically significant difference
- This hold true after multivariable linear regression analysis (adjusting for cofounders [SpO2 at induction, BMI, operator experience, video-laryngoscopy])
Secondary outcome
- View achieved AND difficulty of intubation was WORSE in the ramped ground
- Cormak-Lehane view grade III / IV: Sniffing (11.6%) vs ramped (25.4%) (p=0.01)
- Operator reported “difficult intubation”: Sniffing (4.6) vs Ramped (12.3%) (p=0.04)
- Rates of failed intubation attempts were higher in ramped group
- First attempt success: Sniffing 85.4% vs Ramped 76.2% (p=0.02)
- More than 3 attempts at laryngoscopy: Sniffing 2.3% ve ramped 7/7% (p-0.02)
- But… No difference in oxygenation or haemodynamic outcomes
Tertiary data, subgroup and post-hoc analysis
- No statistically significant in tertiary outcome of ventilator free days, ICU days, or in-hospital mortality
- NO effect of operators experience on primary outcome
Author’s Conclusions
Ramped position is not associated with increase in lowest oxygen saturations during intubation of critically unwell patients on ICU, but might increase the difficulty of intubation / likelihood of repeated attempts.
Strengths of Study
- Relevant to daily ICU practice
- Prospective, multicentre study, large numbers: generalisable
- Baseline characteristics well balanced
- No selection bias / observer bias
- Accuracy of data ensured.
Weaknesses of Study
- 49 excluded from trial: 42 too sick
- No blinding
- Only fellows / trainee clinical: what about senior clinicians? Did they deal with the sicker, more challenging cases.
- Blinding not possible
Discussion from Journal Club Meeting
- Positioning is important: often challenging in ICU – space, beds, less time to optimse
- Can combine sniffing and ramped in some cases
- Anatomical perfection vs physiological compromise
- Is this actually a study of best way to optimise tragus to sternal notch alignment, rather than ramping vs sniffing?
Summary by Dr S Kipling. Journal Club meeting 11 January 2018.
KingstonGasDocs 