Annual Scientific Meeting - 2001

The Confidential Enquiry into Perioperative Equine Fatalities (CEPEF)

 G.M.Johnston^1,2,4, J.Eastment^2,3, P.M.Taylor² and J.L.N.Wood³

¹Rossdale & Partners, Newmarket, UK ²Clinical Veterinary Medicine, University of Cambridge ³Epidemiology Unit, Animal Health Trust UK ⁴Vetstream plc, Langford Arch, Pampisford, Cambridge, UK

CEPEF1&2

Methods

Equine clinics were approached to participate in an observational study which would require them to record a brief one-line diary about every general anaesthetic carried out by their clinic as well as information about the surgeon and anaesthetist and type of clinic. Details about the study methods have been published (Johnston 1991).

Summary of results

Preliminary results of a smaller part of the database have been published (Johnston et al. 1995a; Johnston et al. 1995a; Johnston et al. 1996). Records of 41,787 operations under general anaesthesia were submitted by 149 clinics from 19 countries from 1991 to 1996. The total mortality from perioperative complications was 1.9%. The death rate in the 35,927 non-colic operations was 0.9% and was 7.9% for the 5,833 colics. Twenty two different causes of death were identified for non-colics. The most common causes of death were cardiac arrest (32.8%), fracture (23.3%) and myopathy (7.1%). The clinical presentation of the patients undergoing colic surgery were found to be very different from non-colics.

The most frequent cause of death all equidae in this study, intraoperative cardiac arrest (0.33%), was examined in greater detail in non-colics. Multivariable models were built using logistic regression for two groups of patients; all non colics and a second model for those who received only inhalation maintenance agents. This was because no deaths were observed in patients where anaesthesia was maintained using intravenous agents – a difference which was significant in the univariate analysis (p = 0.0045). The final model for both groups which included variables for age and gender demonstrated that sedative premedication, induction technique, type of clinic, starting time of anaesthesia, body position and day of the week were all associated with death from cardiac arrest. In particular, after adjusting for other variables, there was a marked decrease in risk when acepromazine alone (OR = 0.41, 95% CI = 0.20 - 0.77) or acepromazine and alpha-2 agonist (OR = 0.58, 95% CI = 0.35 - 0.94) were used for premedication and there was a marked increase in risk for patients which were anaesthetised using inhalational induction (OR = 3.26, 95% CI = 1.04 - 9.39). In addition there was a significant reduction in the number of patients that died from a cardiac arrest whose anaesthesia was maintained using intravenous maintenance compared to inhalational maintenance agents.

The data suggests that there are changes in pharmacological management of sedation, induction and maintenance of anaesthesia in equidae that could immediately reduce the risk of a cardiac arrest. Other studies may be necessary to confirm the benefits of other changes suggested by the results. 

Factors significantly associated with death in non colic horses were the age of horse (lowest risk in horses 2-7 years of age); duration of anaesthesia (increased risk with increased duration); choice of induction and premedicant drugs; time of day of surgery (evening and night time surgery associated with greater risk); operation type; and clinic type

CEPEF3: isoflurane versus halothane – results of a randomised trial

Isoflurane has recently been licensed for anaesthesia in the UK. Experimental studies suggest major benefits over halothane: particularly reduced depression of the cardiovascular system. We considered that this agent should impact, in particular, on frequency of cardiac arrest.

In the light of the previous phases of the study two hypothesis were formed: The primary hypothesis was that isoflurane would be associated with a 50% reduction in death rate compared to halothane and the secondary hypothesis was that isoflurane was associated with a 50% reduction in rate of non fatal complications compared to halothane.

The study design (Johnston 1996) was an international, multi-centre randomised prospective clinical trial. It was not blinded but this was not a significant requirement because of the objective outcome of the horse being alive or dead at seven days post surgery.

The study sample size was estimated on the basis of: 1.0% death rate in the halothane group; 90% power, α=0.05 and 10% failure of compliance. This showed that 7,000 horses were needed in each of the two treatment groups and that 94 deaths would be expected between the two groups.

Details of Methodology and results are available on line at www.sivauk.org

Summary of results

The total number of anaesthetic episodes recorded was 11,336.

There was no overall difference in the overall risk of a non fatal or fatal complication between the two treatment groups. The data were explored using stratified analytical procedures and then modelled using mixed effects logistic regression analysis. This analysis demonstrated an significant reduction in the death rate with isoflurane in horses aged 2 - 5 years. This was confirmed in the multivariable modelling. There was a similar reduction in risk in this age group observed in the model for non-fatal complications. However also in this model, there was a significantly increased risk associated with isoflurane usage in the oldest age group.

In our observational studies, we identified factors associated with specific causes of death, particularly cardiac arrest. As the major difference between isoflurane and halothane is in their impacts on the cardiovascular system, it was clearly of interest to examine the effect of the drugs on the major individual causes of death. These analyses are not yet complete, but it is interesting to note that there was a statistically significant reduction in the rate of cardiovascular associated mortality in the isoflurane group. We are also now examining the data to see if there was any effect on the myopathy / fracture group.

Further research needed

Epidemiological studies of anaesthetic mortality is a very effective way of studying what goes wrong in anaesthesia in order to change clinical practice to reduce perioperative complications. CEPEF has already provided a wealth of information suggesting how anaesthetic practice might be change might be altered. However more analysis of the observational cohort study (CEPEF 1& 2) will reveal more detail particularly on other causes of death and on morbidity. Subgroups of deaths in CEPEF3 will also be examined. The observed difference between inhalational vs. intravenous maintenance agents requires further investigation; particularly because this finding is in accordance with (physiological experimental studies (Taylor 1989; Taylor 1990; Taylor 1992; Taylor et al. 1992a; Taylor et al. 1992b; Taylor et al. 1995; Taylor & Watkins 1992; Such a study should be a randomised control trial RCT

Lessons from equine epidemiological anaesthetic studies could be applied with good effect in small animal anaesthesia. It is clear that large multicentre prospective studies of equine perioperative fatalities have taught us a great deal, undoubtedly more can be learnt in other species.

Acknowledgements

The Authors are very grateful to The Home of Rest for Horses who provided the funds for CEPEF1, Horserace Betting Levy Board and The Elise Pilkington Trust plus a grant from the AVA Trust who funded CEPEF2. The Horserace Betting Levy Board funded the position of James Wood at the Animal Health Trust.

The Wellcome Trust supported funded CEPEF3

References

Johnston , G.M. (1991). Confidential Enquiry into Perioperative Equine Fatalities. Equine Veterinary Education 3, 5-6

Johnston, G.M., Taylor, P.M., Holmes, M.A. and Wood, J.L.N. (1995a)  Confidential Enquiry of Perioperative Equine Fatalities (CEPEF-1): Preliminary Results.  Equine Veterinary Journal  27, 193.

Johnston, G.M., Taylor, P.M., McGee, M.A., Holmes, M.A. and Wood, J.N.L. (1995b)  Confidential Enquiry into Perioperative Fatalities (CEPEF 1): Survival curves.  Veterinary Surgery. 182 (Abstract).

Johnston, G.M., Taylor, P.M., McGee, M.A., Holmes, M.A. and Wood, J.L.N. (1996)  Confidential Enquiry into Perioperative Fatalities: Effects of Halothane and Isoflurane on Mortality.  Veterinary Surgery. 160 (Abstract).

Johnston, G.M., Taylor, P.M., McGee, M.A., Holmes, M.A. and Wood, J.L.N. (1996)  Confidential Enquiry into Perioperative Fatalities: Effects of Halothane and Isoflurane on Mortality.  Veterinary Surgery. 160 (Abstract).

Johnston, G.M., (1996) Choosing an inhalational maintenance agent in equidae – the case for a randomised control trial.   Journal of Veterinary Anaesthesia. 23 (2) 44-46.

Taylor, P.M. (1989a)  Adrenocorticol response to propofol infusion in ponies: a preliminary report.  Journal of Veterinary Anaesthesia  17, 12.

Taylor, P.M. (1990)  The stress responses to anaesthesia in ponies: barbiturate anaesthesia.  Equine Veterinary Journal  22, 307.

Taylor, P.M. (1992b)  Total intravenous anaesthesia for major surgery in horses - the way forward? In: Dyke, T.M., (Ed.)  Equine anaesthesia, abdominal surgery and medicine of the foal,  pp. 67.  Artarmon:  Australian Veterinary Association.

Taylor, P.M., Brearley, J.C., Luna, S.P.L. and Johnson, C.B. (1992a)  Total intravenous anaesthesia in horses.  Animal Health Trust Annual Report 56.

Taylor, P.M., Luna, S.P.L., Brearley, J.C., Young, S.S. and Johnson, C.B. (1992b)  Physiological effects of total intravenous surgical anaesthesia using detomidine-guaiphenesin-ketamine in horses.  Journal of Veterinary Anaesthesia  19, 24.

Taylor, P.M., Luna, S.P.L., Sear, J.W. and Wheeler, M.J. (1995)  Total intravenous anaesthesia in ponies using detomidine, ketamine and guaiphenesin: pharmacokinetics, cardiopulmonary and endocrine effects.  Research in Veterinary Science  59 (1), 17.