Assessment of the variation in American Society of Anaesthesiologists Physical Status Classification assignment in small animal anaesthesia



This article is corrected by:

  1. Errata: Erratum Volume 40, Issue 4, 389, Article first published online: 13 June 2013

Matthew McMillan, Department of Clinical Veterinary Medicine, University of Cambridge, Cambridge, UK. E-mail:



To evaluate the interobserver variability in the assignment of the American Society of Anesthesiologists Physical Status Classification (ASA-PSC) to compromised small animal patients amongst a group of veterinary anaesthetists.

Study design

Anonymous internet survey.


Hypothetical case presentations.


Sixteen hypothetical small animal cases with differing degrees of physiological or patho-physiological compromise were presented as part of an internet survey. Respondents were asked to assign a single ASA-PSC to each case and also to answer a number of demographic questions. ASA-PSC scores were considered separately and then grouped as scores of I–II and III–V. Agreement was analysed using the modified kappa statistic for multiple observers. Data were then sorted into various demographic groups for further analysis.


There were 144 respondents of which 60 (~42%) were anaesthesia diplomates, 24 (~17%) were post-residency (nondiploma holders), 24 (~17%) were current anaesthesia residents, 21 (~15%) were general practitioners, 12 (~8%) were veterinary nurses or technicians, and 3 (~2%) were interns. Although there was a majority agreement (>50% in a single category) in 15 of the 16 cases, ASA-PSC were spread over at least three ASA-PS classifications for every case. Overall agreement was considered only fair (κ = 0.24, mean ± SD agreement 46 ± 7%). When comparing grouped data (ASA-PSC I–II versus III-V) overall agreement remained fair (κ = 0.36, mean ± SD agreement 69 ± 19%). There was no difference in ASA-PSC assignment between any of the demographic groups investigated.

Conclusions and clinical relevance

This study suggests major discrepancies can occur between observers given identical information when using the ASA-PSC to categorise health status in compromised small animal patients. The significant potential for interobserver variability in classification allocation should be borne in mind when the ASA-PSC is used for clinical, scientific and statistical purposes.