34th Annual Scientific Meeting proceedings

The Confidential Enquiry into Perioperative Equine Fatalities 4 (CEPEF4), published online in June 2025, collected prospective data from 47,396 horses and donkeys undergoing general anaesthesia (GA) in 93 centres, 28 countries and four continents (Gozalo-Marcilla et al. 2025a). CEPEF4 aimed to report the current data on equine perioperative mortality occurring within 7-days of anaesthesia and to document anaesthetic practice, risks and potential means of prevention.  

Implementing technology (IT) improvements, CEPEF4 halved the data collection period compared to the similarly sized CEPEF2 over 20 years ago (Johnston et al. 2002). The key factor was developing a website which incorporated an error management system from an initial portable document format (PDF) (Domenech et al. 2024).

1.  Mortality data in horses and ponies undergoing GA

The mortality rates within the 7-day period, for CEPEF4 in 2025 versus CEPEF2 in 2002 were as follows: overall 1.2% versus 1.9%, COLIC death rate 4.2% versus 7.8% and NON-COLIC death rate 0.6% versus 0.9% (Johnston et al. 2002; Gozalo-Marcilla et al. 2025a). *(COLIC, variable with capital letters, includes emergency abdominal surgeries for colics, caesarean sections or bladder ruptures).

Looking at those numbers we can say that there has been a clear reduction in mortality which is reassuring; this suggests that the range of potential improvements (i.e. new drugs, anaesthetic protocols and surgical techniques, as well as developments in peri-operative management, monitoring and cardiopulmonary support) may be responsible.

In spite of these improvements, healthy horses still die, although the causes of death have changed in the intervening 20 years. In CEPEF2, NON-COLIC deaths were roughly divided into thirds (1/3 cardiac arrest or cardiovascular collapse, 1/3 fractures or myopathy and 1/3 other reasons) (Johnston et al. 2002). In CEPEF4 1/3 were due to fractures, about 1/5 were due to abdominal complications (i.e. postoperative colic or colitis), around 1/8 from CNS complications or neuropathy and approximately 1/10 from cardiac arrest or cardiovascular collapse (Gozalo-Marcilla et al. 2025a).

Using logistic regression we identified factors that may influence clinical decision-making and enhance patient safety. Pregnant mares, geriatrics, thin horses, those with a higher American Society of Anesthesiologists (ASA) grade, urgent procedures and those lasting < 1 hour or > 2 hours had increased odds of death. Ketamine infusions and plasma lactate measurement during maintenance were also associated with elevated risk. In contrast, monitoring arterial pressure (invasively), end-tidal CO₂, blood gases and body temperature were each associated with reduced odds of death. Premedication with agonist-antagonist/partial-agonist opioids with non-steroidal anti-inflammatory drugs and administration of α2-adrenoceptor agonists during recovery, alone or with acepromazine were associated with reduced odds of death (Gozalo-Marcilla et al. 2025a). 

Lessons learnt:

1. Healthy horses still die. The causes of death have changed in the last two decades and require in depth analysis to suggest how the mortality may be further reduced.
2. Mortality is still substantial in sick horses, although the proportion has also decreased. This could relate to factors such as improved peri-operative management, surgical techniques, monitoring and support of cardiopulmonary function.
3. Some of the risk/protective factors identified were consistent with expectation, whereas others were surprising and require further investigation.

2.  Standing sedation data

Standing sedation (SS) in horses is not risk free. Using the same system, we collected data from 12,307 cases undergoing SS from 61 centres in 23 countries (Gozalo-Marcilla et al. 2025b). Death rates within the 7-days period were 0.2% overall (19/12,307), 0.1% for cases classified as NON-COLIC (16/12,237) and 4.3% for COLIC (3/70). For the NON-COLICs (16 deaths from 12,237 NON-COLICs), the causes of death were abdominal in nine (56.25%), (re-)fractures in four (25.0%), two were “found dead” (12.5%) and one (6.25%) for “other reasons”.

Even without the complications associated with GA (which generally occur during recovery), horses still die unexpectedly within seven days of procedures carried out under sedation alone.

Lessons learnt:

1. Increased use of SS rather than GA reduces the risk of mortality but does not avoid them altogether.
2. Standing sedation also requires close monitoring.

3.  Morbidity data: a parallel study

Eight of the 93 collaborating centres collected data from cases which did not culminate in death or euthanasia. The data are currently under analysis and will be published soon.

In setting up this morbidity investigation, Loomes et al. (2024) undertook a systematic review which included 67 published articles from 2000 to 2023. They reported that the prevalence of post-operative colic, surgical site complications, laminitis, diarrhoea/colitis, fever/pyrexia, jugular thrombophlebitis/thrombosis and respiratory complications were significantly higher after colic surgery compared with elective/non-abdominal surgery under GA. Myopathy/neuropathy was the only outcome with similar prevalence in the two groups (OR 1.86; 95% CI: 0.86-4.16; p = 0.16).

Unfortunately, most of the studies were retrospective and the definition of morbidity, the data collection periods, follow-up time and methods varied between studies.

Lessons learnt:

1. There is a clear need for our prospective morbidity study which aims to overcome the prime limitations of a systematic review.

4.  The importance of the term “perioperative” in the CEPEF studies. Reducing equine mortality further requires teamwork

The term “perioperative” in the acronym CEPEF emphasizes the importance of working together in a clinical setting. Certainly, few deaths could be attributable purely to anaesthesia (a certain anaesthetic technique), and a few could be entirely attributable to some aspect of the surgery (e.g. failed haemostasis). However, our data reflect the importance of the term perioperative as many of the fatalities are inherent to the nature of the horse: a prey animal with a marked flight response (fractures in recovery) and an anatomical tendency to gastrointestinal disorders (prone to colic). Fluent, easy communication and a good teamwork is mandatory for working together to reduce the death rate.

Lessons learnt:

1. We can only reduce perioperative mortality almost to zero if all the specialities work together with the same aim. Achieving absolute zero is, probably, impossible due to the nature of the horse itself and its physiology.

References

1. Domenech L, Redondo JI, Taylor PM, Bettschart-Wolfensberger R, Johnston GM, Doménech J, Gozalo-Marcilla M. Error management in large data sets: A critical component of data collection in veterinary medicine. Vet Anaesth Analg. 2024 Nov 17:S1467-2987(24)00342-8. doi: 10.1016/j.vaa.2024.11.005.
2. Gozalo-Marcilla M, Redondo JI, Bettschart-Wolfensberger R, Domenech L, Doménech J, Johnston GM, Taylor PM. The Confidential Enquiry into Perioperative Equine Fatalities: phase 4 (CEPEF4). A worldwide observational, prospective, multicentre cohort study in 2025. Vet Anaesth Analg. (in press).
3. Gozalo-Marcilla M, Redondo JI, Taylor PM, Domenech L, Doménech J, Johnston GM, Bettschart-Wolfensberger R. Standing sedation in horses and ponies: perioperative fatalities in a worldwide observational, prospective, multicentre cohort study. Abstract sent to the World Congress of Veterinary Anaesthesia and Analgesia, September, Paris (France).
4. Johnston GM, Eastment JK, Wood J, Taylor PM. The confidential enquiry into perioperative equine fatalities (CEPEF): mortality results of Phases 1 and 2. Vet Anaesth Analg. 2002 Oct;29(4):159-170. doi: 10.1046/j.1467-2995.2002.00106.x.
5. Loomes K, de Grauw J, Gozalo-Marcilla M, Redondo JI, Bettschart-Wolfensberger R. A systematic review of the prevalence of post-operative complications after general anaesthesia in adult horses (2000-2023). Equine Vet J. 2025 Jul;57(4):827-861. doi: 10.1111/evj.14416.

Acknowledgements
The study was approved by the International Ethical Review Committee of the Association of Veterinary Anaesthetists (AVA), under protocol 2020-009.

The authors would like to thank the Kate Borer-Weir Fund of the AVA, the AVA Trust, and the Horse Trust for supporting us with this study.

The project would have been impossible without the open mindedness, hard work, honesty and diligence of the personnel of every single collaborating centre who submitted the cases for these preliminary results.

The authors declare no conflict of interest.