The association of early postoperative desaturation with discharge to care facility

Rostin, B. J. Teja, S. Friedrich, S. Shaefi, K. R. Murugappan, S. K. Ramachandran, T. T. Houle and M. Eikermann

Anaesthesia 2019: 74(4); 457-467.

Aim of Study

  • does desaturation after tracheal extubation increases the likelihood of adverse hospital discharge (Nursing Home/Residential Home)?

Design and Location

  • Large, retrospective study based out of Massachusetts General Hospital
  • Jan 2007-Dec 2015


  • Retrospective database study using electronic notes
  • Recorded if sats <90% in 10 minutes following extubation
  • Inclusion C: Intubated and extubated in theatre, Sats monitoring documented every minute, Discharge destination known
  • Exclusion C: ASA>4, LTOT/home ventilated pts (except for OSA CPAP users), Died in hospital, missing data

Primary Outcome

Association between post-operative desaturation and discharge to residential/nursing home

Secondary Outcomes

Pulmonary complications (CAP, re-intubation, ITU admission), post-operative MI


  • Multivariable logistical regression
  • Complicated
  • Used pre-specified confounder model to attempt to control variables


  • 73% of patients included (71,000/ 96,808)
  • 6% had a desaturation
  • those with desaturations most likely to have adverse discharge (regression derived, OR 1.36)
  • more likely to have pulmonary complications and ITU stay
  • dose dependent, as those with sats < 80% higher risk
  • desaturation also associated with: MI and AKI
  • desaturation more common with high-long acting opioid dose (>34mg PO morphine equivalent), higher doses of neostigmine (> 4.2mg) and high FiO2 (> 60%)
  • the ‘worst’ anaesthetists had 34% more desaturation events


  • Major complications appeared dose-dependent
  • Rates of desaturation vary with anaesthetists (case adjusted)
  • Potential modifiable RFs (as above)

Stated Limitations from the Study

  • Huge reliance on electronic sats rate recording
  • No documentation of FiO2
  • Large numbers of patients excluded
  • Retrospective regression always inferior to randomisation

Summary by Dr O Tolson. Journal Club Meeting 05 September 2019.

Comments are closed.

Blog at

Up ↑

%d bloggers like this: