Hypotension during induction of anaesthesia is neither a reliable nor a useful quality measure for comparison of anaesthetists' performance

R. H. Epstein, F. Dexter, E. S. Schwenk

Research output: Contribution to journalArticlepeer-review

13 Scopus citations


Background: Identification of statistically reliable outcomes for comparison among anaesthetists is challenging. Time-weighted intraoperative mean arterial pressure <65 mm Hg (AUC65) is associated with increased odds for myocardial damage. We explored retrospectively whether such hypotension before incision was statistically reliable for peer comparison. Methods: We retrieved electronic data between 2006 and 2015 at a tertiary care, academic hospital in the USA for patients at risk for myocardial damage (inpatient after surgery, ASA physical status ≥III, ≥50 yr of age, and case duration ≥60 min). We determined the percentage of anaesthetists comparable based on caseload and case-mix. The AUC65 was compared amongst anaesthetists supervising ≥100 cases involving at-risk patients during the last 12 months. Results: Only 14.1% [95% confidence interval (CI) 13.6-14.5%] of cases involved patients who were 'at risk' during the 10 yr study period. A yearly average of 49 (sd 6) anaesthetists supervised ≥100 cases of any type, of whom only 52% (95% CI 47.1-56.0%) supervised ≥100 cases involving at-risk patients. Thus, nearly half the anaesthetists would have been excluded from peer comparison. During the last 12 months, there were two outliers among 34 evaluable anaesthetists (P<0.05, controlling for false discovery). However, their contribution to total hypotension amongst cases for all patients was small, because hypotension was widely distributed (e.g. 80% of hypotension attributable to 61.8% of anaesthetists, 95% CI 59.8-63.7%). There was no relationship between the AUC65 and propofol induction dose. Conclusions: The AUC65 of time-weighted pre-incision hypotension is not a suitable metric for comparing anaesthetists. There were few at-risk patients, half the anaesthetists were not evaluable because of their case-mix and caseload, and hypotension was widely distributed.

Original languageEnglish (US)
Pages (from-to)106-114
Number of pages9
JournalBritish Journal of Anaesthesia
Issue number1
StatePublished - Jul 1 2017


  • Healthcare
  • Hypotension
  • Medical informatics computing
  • Peer review

ASJC Scopus subject areas

  • Anesthesiology and Pain Medicine


Dive into the research topics of 'Hypotension during induction of anaesthesia is neither a reliable nor a useful quality measure for comparison of anaesthetists' performance'. Together they form a unique fingerprint.

Cite this