Abstract
INTRODUCTION: Out-of-hospital cardiac arrest (OHCA) is a time-critical emergency in which ambulance clinicians must rapidly decide whether to initiate resuscitation. In the UK, up to 30% of OHCA patients also have a documented do not attempt cardiopulmonary resuscitation (DNACPR) decision. Despite these advance care plans, patients may still receive cardiopulmonary resuscitation (CPR), potentially causing distress to patients, relatives and clinicians, and increasing resource use. The frequency and predictors of such events in UK ambulance services remain poorly described.
METHODS: A retrospective service evaluation was undertaken using one regional ambulance service's OHCA registry (January-December 2024). Adults with a valid DNACPR or recommended summary plan for emergency care and treatment (ReSPECT) recommending against CPR, which was available to crews either electronically or physically at the time of the incident (or crews were informed by another HCP of the presence of a valid form), who suffered an OHCA and were attended by a regional emergency medical services (EMS) were included. The primary outcome was the commencement of any resuscitation attempt, defined as either over-the-phone CPR instructions or on-scene clinical intervention. Descriptive statistics compared cases with and without resuscitation. Univariable and multivariable logistic regression identified factors associated with a resuscitation attempt.
RESULTS: Between 1 January and 31 December 2024 there were 7809 OHCAs, of which 1827 patients had a DNACPR; resuscitation by the ambulance service or bystanders occurred in 377 (20%) cases. Call handlers initiated most attempts (87%), and 36% received on-scene interventions. Median age was 80 years (IQR 70-90), and 51% were female. Multivariable analysis showed that each additional clinician on scene was associated with increased odds of resuscitation by 69% (adjusted odds ratio [aOR] 1.69, 95% CI 1.55-1.84; p <0.001). A public location carried a six-fold higher likelihood of resuscitation compared with assisted-living or nursing homes (aOR 6.08, 95% CI 1.47-27.4; p = 0.013). Home OHCAs were common and associated with a resuscitation attempt (aOR 1.41, 95% CI 1.00-2.01). Where dedicated end-of-life care single points of access were available in an area, there were lower odds of a resuscitation attempt (aOR 0.32, 95% CI 0.12-0.81). Patient age, sex and response time were not independent predictors of the outcome.
CONCLUSION: One in five patients with a recorded DNACPR received a resuscitation attempt. This likely reflects the real-world complexities of DNACPR recognition in time-critical emergencies, including delayed access to documentation and potentially reversible presentations. Larger on-scene teams and public locations were associated with resuscitation, whereas a dedicated end-of-life care hub was associated with reduced attempts. Improving real-time visibility of advance care plans and strengthening end-of-life care pathways may help to ensure that care remains aligned with patient preferences.