Incidence and risk factors of anaesthetic-related perioperative cardiac arrest. European Journal of Anaesthesiology 2017;34:1–7 doi:10.1097/EJA.0000000000000685
Presented by: Dr R Dean-Paccagnella
- Many studies have analysed perioperative mortality in speciality sub-groups, but few have looked at unselected patient populations. Many studies have excluded patients undergoing cardiac surgery.
- Previous papers have studied perioperative mortality but have not independently reviewed the incidence and risk factors of cardiac arrest.
- This study aims to measure the incidence of perioperative cardiac arrest in an unselected anaesthetic population and retrospectively identify significant risk factors.
Design & Setting
- Retrospective cohort study of non-ITU patients undergoing anaesthesia between January 2007 and December 2012 at a single tertiary hospital in Cologne, Germany.
- 169,000 adult and paediatric patients underwent anaesthetic procedures within the time period.
- Study population (n 318) was identified by the screening of critical incident report forms, performed by the authors.
- Cases were categorised into 1. “anaesthesia related” (directly caused by an anaesthetic procedure), 2. “anaesthesia contributory” (caused by both surgical and anaesthetic events) or 3. “anaesthesia contributory” (possibly caused by factors under the control of the anaesthetist).
- Undifferentiated anaesthetic procedures were analysed retrospectively.
- Incidence of pulselessness requiring chest compressions within 24hours after anaesthetic procedure.
- Incidence of perioperative cardiac arrest was 5.8/10,000 anaesthetic cases (95% CI 4.7-7.0).
- Significantly increased risk of perioperative cardiac arrest was associated with ASA grade or 3 or more, revised cardiac risk index of 3 or more, NYHA or 3 or more, out of hours procedures, emergency surgery and pre-existing cardiomyopathy.
- Multi-variate logistic regression identified 3 predictors of perioperative cardiac arrest. ASA grade of ≥3 (OR 2.59, p=0.007, 95% CI 1.29 to 5.19), emergency surgery (OR 4.00, p=0.001, 95% CI 2.15 to 7.54) and pre-existing cardiomyopathy (OR 17.48, p= <0.001, 95% CI 6.18 to 51.51).
- Age over 75 years or less than 3 years, Gender, BMI ≥30 kg m3 , and patients with known difficult airways were not identified to be at significantly altered risk of perioperative cardiac arrest.
- Patients with an ASA physical status grade of ≥3, undergoing emergency surgery or with pre-existing cardiomyopathy appear to be at an increased risk of perioperative cardiac arrest in this single centre European university hospital population.
- Incidence of paediatric cardiac arrest directly caused by anaesthesia was high (5 of 12 cardiac arrests directly related to anaesthetic procedure).
- Clinically relevant question addressing entire anaesthetic population.
- Findings are in-line with previous papers addressing ASA grade and risk of anaesthesia related cardiac arrest.
- Risk factors were identified retrospectively by reviewers. NYHA classification appears to have been categorised retrospectively by investigators.
- Strength of relationship between anaesthetic procedure and cardiac arrest categorised by authors (although independently).
- Single centre European study which may not provide generalisable results.
- Main outcome measure is an infrequent event, and as such small variation in number of events will greatly influence the frequency reported.
- ASA grading, urgency of surgery and pre-operative identification of cardiomyopathy may help identify high risk cases.
- Further studies of peri-operative cardiac arrest would be improved by establishing a consensus for the definition of anaesthesia-related and anaesthesia-contributory cardiac arrests.
- Incidence of anaesthesia-related cardiac arrest appears to remain relatively high in the paediatric population.
Potential for impact
- If felt to be generalisable, ASA grade ≥3, emergency surgery and cardiomyopathy may indicate patients at significantly increased risk of perioperative cardiac arrest, although this remains an infrequent event.