Sudden out-of-hospital cardiac arrest (OHCA) is a significant health burden in New Zealand. During the one-year period from 1 July 2014 to 30 June 2015, St John attended 4,298 OHCAs. The majority of patients (98%) attended were adults greater than or equal to 16 years of age, with an incidence of 132.8 per 100,000 person-years. Of the adult patients with OHCA attended by St John during this period, 1,996 (48%) had a resuscitation attempt initiated by the attending emergency ambulance crews.

St John operates a two-tiered response system for responding to OHCA in New Zealand. The two-tiered response system involves response of the fire service crewed with basic life support first responders, dispatch of the closest two ambulances and an intensive care paramedic (ICP) if possible. The usual standard for an ambulance crew is to be staffed with two trained responders. New Zealand is in a relatively unique situation in that some ambulances may be staffed with a single trained responder. If an ambulance is staffed with two trained responders, the crew composition may include a mix of ICP, paramedic (PARA) or emergency medical technician (EMT). However, the ideal skill composition and number of crew that attend an OHCA is unknown.1

There are various models of ambulance response to OHCA, ranging from one- or two-tier systems, different numbers of crew and different skill compositions.1–2 A limited number of studies have shown that a larger number of crew may not have an effect on patient outcomes.3–4 Primarily these studies make comparisons between a double crew (DC) to a crew of three or more. There are few ambulance services, other than in New Zealand, where an ambulance response can consist of a single crew (SC) outside of the context of a being a first tier or non-transporting response. As such, there are no studies that compare a DC ambulance response to a SC ambulance response.

As New Zealand is unique in that a portion of the ambulance crews may be staffed with a single responder, this study sought to investigate if there was any association between attendance by a single-crewed ambulance and cardiac arrest outcomes in New Zealand.

Method

Study design

This was a retrospective analysis using St John OHCA registry data for the period of 1 October 2013 to 30 June 2015. The data variables held within the registry are inclusive of all OHCA attended by St John in New Zealand, regardless of whether a resuscitation attempt was initiated or not. It includes dispatch data, on-scene data collected by the ambulance personnel in attendance and mortality data from district health boards or the New Zealand National Health Index. The dispatch data includes time of answering the emergency call in the ambulance clinical control centre, time of arrival of an ambulance at the scene, the level of qualification of the attending crew, the number of crew attending and whether the first locating vehicle was an emergency transporting ambulance or a non-transporting vehicle such as a car.

The St John OHCA registry covers the areas of New Zealand that are serviced by St John. This includes all of New Zealand with the exception of the Greater Wellington Region, which is serviced by Wellington Free Ambulance. The registry encompasses 90% of the New Zealand population (approximately 4.1 million people) and 97% of New Zealand.5–8

Variables

Variables included were ambulance crew number (SC or DC), gender, age, ethnicity, OHCA witnessed status, performance of bystander cardio-pulmonary resuscitation (CPR), defibrillation prior to ambulance arrival, shockable presenting rhythm, aetiology, location, incident region, service area (urban, rural or remote), minutes to scene, highest clinical level in attendance, ROSC on handover at hospital and discharged alive.

Ethics approval

Ethical approval for research using non-identifiable OHCA data was obtained from the Auckland University of Technology Ethics Committee (13/367) and the New Zealand Health and Disabilities Ethics Committee (13/STH/192/AM02).

Statistical methods

All data was entered into an IBM SPSS database (Version 23.0 IBM Corporation). Patient demographics, case characteristics and patient outcomes were compared between SC and DC groups using the Chi-squared test for categorical data and the Mann-Whitney-U test for numerical data. Binary logistic regression analysis of patient demographics and case characteristics were performed to select variables significant at the p<0.10 level for adjusted multivariate regression models.

Outcomes

Survival to hospital discharge was the primary outcome. Survival to hospital discharge was defined as having either a known date of being discharged alive from hospital or survival to 15 days post-event.

Results

Included cases

During the 21-month study period, St John attended 7,005 OHCAs of which 3,446 had resuscitation attempted. After excluding OHCA where age was less than 16 years, fire service or a non-transporting rapid response vehicle such as a car were the first locating vehicle and emergency response ambulances with greater than 2 crew; a total of 2,347 OHCA cases were included for analysis (Figure 1).

Figure 1: Flow diagram of included cases.

Study population

Of the 2,347 cases, 337 were attended by SC (14%) and 2,010 were attended by DC (86%). Gender and age were similar between SC and DC groups. A lower proportion of patients attended by SC were European, Pacific Peoples or Other ethnicity and a higher proportion were Māori (Table 1).

Table 1: Characteristics of OHCA by crew level.

‍

OHCA: out-of-hospital cardiac arrest, CPR: cardiopulmonary resuscitation, ICP: intensive care paramedic, EMT: emergency medical technician, PARA: paramedic, ROSC: return of spontaneous circulation.

OHCA characteristics

OHCA characteristics were similar between SC and DC groups (see Table 1). Approximately 50% were witnessed by a bystander and 19% were witnessed by ambulance staff, 58% had CPR performed by a bystander, less than 10% had defibrillation prior to ambulance arrival, nearly 40% had a shockable presenting rhythm and the majority had a presumed cardiac aetiology (approximately 80%).

Ambulance response

A higher proportion of events attended by SC ambulances were in the Central Region (73%), a higher proportion of events were in rural areas (61%), the median response time for the SC group was longer and they were less likely to have the highest qualified responder in attendance being an ICP (67%) (Table 1).

Outcomes

The univariate analysis of ROSC sustained to hospital handover showed there was no difference in the rates of ROSC sustained to hospital handover in patients attended by either SC ambulances or DC ambulances (OR 0.779, 95% CI 0.602–1.010, p=0.059). However, patients were significantly less likely to survive to hospital discharge when attended by SC ambulances compared to those attended by DC ambulances (Table 2, Figure 2). After adjustment for confounders there remained a significant difference between SC and DC groups with patients attended by SC ambulances having a lower odds of survival (Figure 2).

Table 2: Univariate logistic regression model for survival to hospital discharge.

‍

CPR: cardiopulmonary resuscitation, EMT: emergency medical technician, ICP: intensive care paramedic.

Figure 2: Odds ratio (OR) of survival to hospital discharge for patients attended by SC ambulances vs DC ambulances.

Univariate unadjusted (OR 0.642, 95% CI 0.451–0.914, p=0.014).
Multivariate regression with adjustment for confounders (OR 0.533, 95%CI 0.320–0.888, p=0.016).

Discussion

This retrospective analysis of data from the St John New Zealand OHCA registry has clearly shown that attendance by a SC ambulance was associated with a reduction in the odds of survival to hospital discharge when compared to attendance by a DC ambulance. There was no difference in rates of ROSC on hospital handover between patients attended by a SC or DC ambulance.

Due to the retrospective observational nature of this study, potential causes for the associated reduction in survival to hospital discharge with attendance by a SC ambulance cannot be drawn. It is possible that factors such as patient characteristics, CPR quality, pre-shock and post-shock pauses, defibrillation mode and responder fatigue may be implicated.

During the first few minutes of a cardiac arrest, the quality of CPR is critical to achieving survival.9 The chest compression fraction, or the percentage of time that chest compressions are performed during resuscitation, has been previously identified as an independent predictor of survival.10–11 It may be that in the setting of attendance by a SC ambulance the chest compression fraction is less than that with attendance by a DC ambulance. With a SC ambulance a lone responder would need to perform multiple tasks such as performing chest compressions, attaching defibrillation pads and using the defibrillator, whereas in the case of a DC ambulance, one of the two responders would perform continuous compressions while the other responder completed all of the other tasks.

Chest compression rate and depth have also been identified as elements of high-quality CPR. Both have been shown to deteriorate over time with responder fatigue.12–13 To ameliorate the observed deterioration in chest compression quality over time, the resuscitation guidelines recommend that the responder performing chest compressions should change every two minutes.14 In the case of attendance by a SC ambulance the lone responder may either continue compressions for longer than the recommended two-minute period or may potentially recruit an untrained lay-responder to continue the compressions. Both situations have the potential to compromise of the quality of compressions.

Although it was not able to be investigated in this study, it is possible that in the context of attendance by a SC ambulance that the defibrillator may be used more frequently in an advisory mode. The use of a defibrillator in the advisory mode has been shown to be associated with longer interruptions in CPR and lower rates of ROSC compared to using a defibrillator in manual mode.15 In addition, Cheskes et al (2011) previously demonstrated that the pre-shock and post-shock pauses in compressions, which are typically longer in the advisory mode, were associated with significant reductions in survival.16

With a SC ambulance, even with the use of manual defibrillation, there may be an increased length of pre- and post-shock pauses in chest compressions. With a DC ambulance in order to minimise the pre-shock pause it is usual practice for one responder to continue compressions while the other charges the defibrillator, pausing compressions only to confirm a shockable rhythm exists and to deliver the defibrillator shock. However, with a SC ambulance this is not possible as the lone responder would need to cease compressions while they charge and use the defibrillator, potentially resulting in prolonged pre- and post-shock pauses in compressions. In order to further limit reductions in chest compression quality it may also be that advanced life support skills, such as intubation and administration of drugs are not performed in patients attended by a SC ambulance as these skills would require prolonged pauses in chest compressions.

This study represents the first comparative analysis of patient outcomes according to attendance by a SC or DC ambulance in response to OHCA in the New Zealand context. The strength of this study is that the registry held equivalent information on patients attended by both SC and DC ambulances that enabled sufficient numbers for statistical analyses. In addition, the responders were all from a single ambulance service utilising the same clinical procedures and guidelines. However, there are several limitations of the study in that the study was unable to investigate relationships between electronic defibrillator data, the characteristics of the receiving hospitals or patient comorbidities, all factors which may also influence survival outcomes following OHCA.

Conclusion

This study examined survival differences in patients where the first locating ambulance was single crewed or double crewed in the setting of OHCA in New Zealand. Patients had lower survival to hospital discharge outcomes when the first locating ambulance was single crewed than those where the first locating ambulance was double crewed. Although further investigation into relationships is required, the number of crew in the first locating ambulance impacts OHCA survival rates. This research supports the elimination of single-crewed ambulances in New Zealand.

‍