Out-of-hospital cardiac arrest in patients aged 35 years and under: A 4-year study of frequency and survival in London☆

2.1. Study design 

The study is a retrospective analysis of data collected by the London Ambulance Service NHS Trust for every out-of-hospital cardiac arrest patient attended aged 35 years or younger. Data were collected for a total of 3084 patients between 1st April 2003 and 31st March 2007. Data are presented to allow the comparison of characteristics between different age groups within the paediatric and young adult cardiac arrest population.

2.2. System overview 

London represents the most populous urban centre in the European Union and is one of the largest cities in the developed world in terms of its built-up area. It has a population of 7.5 million, representing 12% of the UK total and equating to a density of 4699 people per km2 in a geographic area of 1579km2 (609 miles2). Approximately 4 million people are aged 35 years and under.9

London is served by the London Ambulance Service NHS Trust, the largest emergency ambulance service in the world to provide healthcare that is free at the point of delivery. It responds to over 1 million calls per year and attends over 900,000 emergency incidents.10 Cardiac arrest calls are designated ‘Category A’ by the UK's Department of Health and are assigned the highest priority level to ensure the most rapid response. Vehicles responding to cardiac arrest emergency calls are equipped with automated external defibrillators (AEDs) and all operational staff operate to the Resuscitation Council (UK) Guidelines.11

The London Ambulance Service NHS Trust oversees a programme of public access AED deployment and provides training in their use. In 2007 around 425 public access AED units were deployed across 150 sites in London.

Limited knowledge of the symptoms of myocardial infarction and signs of cardiac arrest, as previously identified in London,12 reduces the speed of access to emergency medical care in some cases. As in many major cities, knowledge of bystander CPR has also been identified as an area for improvement.12 During the study period, the London Ambulance Service NHS Trust undertook public awareness campaigns and provided training to over 30,000 Londoners in community resuscitation.

2.3. Data collection 

Data were extracted from the clinical records completed by London Ambulance Service NHS Trust crews attending each out-of-hospital cardiac arrest. These records capture information including patient demographic details, location, bystander interventions, witnessed information and clinical observations. Variables were supplemented with information from the emergency call record. For each patient taken to hospital with resuscitation ongoing, attempts were made to obtain survival outcome information from the 32 receiving London Accident & Emergency Hospitals. All data items were obtained in accordance with relevant legislations governing the collection and use of patient data.

Based on the information from the clinical record and emergency call record, patients were categorised by age, aetiology and whether or not they were viable for resuscitation attempts. Aetiology groups included cardiac, trauma or other non-cardiac cause. Patients with no reported cardiac disease history were presumed to have suffered a cardiac arrest owing to an underlying cardiac condition unless there was information to suggest otherwise. Trauma and non-cardiac cases were sub-divided into further categories to allow a more detailed aetiology analysis. Patients were categorised as ‘not viable for resuscitation’ if their clinical record reported obvious signs of death (e.g. rigor mortis, post-mortem staining) and the attending crews did not make any resuscitation attempt.

2.4. Data analysis 

Population parameter statistics were generated to summarise and compare patient characteristics that met the study inclusion criteria. Patients were split in to two groups for data analysis by age. These were children (17 years or less) and adults (18–35 years). Since different characteristics were exhibited by children of different ages, those aged 17 or less were further divided into three groups; infants (under 1 year old), younger children (1–8 years) and older children (9–17 years). Frequency data were analysed with non-parametric Chi-squared tests and the Marasculio procedure for multiple comparisons using SPSS 16.0 (Chicago, USA). Multivariate regression analyses were used to identify possible predictors of survival.

3.1. Frequency and characteristics 

Cardiac arrest occurred more frequently among adults (18–35 years) (n=2380, 77.2%) than among children (17 years or less) (n=703, 22.8%). Nearly half of the cardiac arrests among children occurred in the infant group (n=321, 45.7%; 10.4% of the complete dataset). Fig. 1 illustrates frequency of cardiac arrest per year of age.

View Large Image Download to PowerPoint

Fig. 1. Number of cardiac arrests per year of age.

Table 1 presents the patients categorised by aetiology, age, gender and whether they received resuscitation attempts from the ambulance crew.

The most common cause of cardiac arrest was a presumed cardiac condition (n=1384; 44.9%). 30% (n=924) of arrests were due to trauma and 25.1% (n=775) were of another non-cardiac cause. Adults were significantly more likely to have a traumatic cardiac arrest than children (34.5% vs. 14.8%; χ2 (1, n=3084)=99.3, p<0.001).

Over two thirds of all patients in the study were male (n=2179; 70.0%). The adult group contained a significantly higher percentage of males in comparison to the group of children (75.0% vs. 57.4%, χ2=80.85, p<0.001).

Just over half of all patients were viable for resuscitation (n=1748; 56.7%). Of these, the largest aetiology group was presumed cardiac (41.3%; n=722). The proportion of patients for whom resuscitation was attempted was higher in the younger age groups. 83.5% (n=269) of infants (under 1 year old), 82.3% (n=153) of children (1–8 years) and 76.5% (n=150) of adolescents (9–17 years) received resuscitation attempts from the ambulance crew. In sharp contrast, only 49.4% (n=1176) of adult patients (18–35 years) received resuscitation attempts. In fact, 18–35 year olds accounted for 90.1% of all patients for whom resuscitation was not commenced.

There was a significantly higher proportion of presumed cardiac aetiology arrests among the group of patients who did not have resuscitation undertaken, compared to those who did (49.6% vs. 41.3%, χ2=7.82, p=0.005).

Patients who had a cardiac arrest of a non-cardiac (non-traumatic) cause were significantly more likely to receive resuscitative efforts than not (30.1% vs. 18.6%, χ2=53.4, p<0.001).

For all cardiac arrests of a traumatic or other non-cardiac origin, the specific cause of arrest was explored. A breakdown is presented in Table 2.

Table 2. Breakdown of the most common underlying causes cardiac arrests of a traumatic or other non-cardiac origin.

		Age group												Total, n
		Under 1 year			1–8 years			9–17 years			18–35 years
		Male,an (%)	Female,an (%)	Unknown,an (%)	Male,an (%)	Female,an (%)	Unknown,an (%)	Male,an (%)	Female,an (%)	Unknown,an (%)	Male,an (%)	Female,an (%)	Unknown,an (%)
	Non-cardiac aetiology
	Overdose	0	0	0	0	0	0	3 (1.1)	2 (0.7)	0	196 (73.4)	59 (22.1)	7 (2.6)	267
	Respiratory	6 (8)	5 (6.7)	0	6 (8)	2 (2.7)	0	6 (8)	7 (9.3)	1 (1.3)	22 (29.3)	20 (26.7)	0	75
	Seizure	1 (2)	2 (4)	0	1 (2)	1 (2)	0	2 (4)	0	0	32 (64)	10 (20)	1 (2)	50
	SIDS	29 (56.9)	22 (43.1)	0	0	0	0	0	0	0	0	0	0	51
	Terminal illness	3 (9.4)	1 (3.1)	0	2 (6.3)	3 (9.4)	0	0	1 (3.1)	0	14 (43.8)	8 (25)	0	32
	Miscellaneous	31 (10.3)	32 (10.7)	1 (0.3)	25 (8.3)	13 (4.3)	0	11 (3.7)	14 (4.7)	1 (0.3)	98 (32.7)	70 (23.3)	4 (1.3)	300
	Total	70 (9)	62 (8)	1 (0.1)	34 (4.4)	19 (2.5)	0	22 (2.8)	24 (3.1)	2 (0.3)	362 (46.7)	167 (21.5)	12 (1.5)	775
	Traumatic aetiology
	Assault	0	0	0	0	1 (3.1)	0	3 (9.4)	0	0	20 (62.5)	7 (21.9)	1 (3.1)	32
	Electrocution	0	0	0	0	0	0	0	0	0	9 (69.2)	1 (7.7)	3 (23.1)	13
	Fall	0	1 (1.4)	0	4 (5.6)	2 (2.8)	0	3 (4.2)	1 (1.4)	0	44 (62)	13 (18.3)	3 (4.2)	71
	Fire	0	0	0	2 (20)	2 (20)	0	0	1 (10)	0	3 (30)	1 (10)	1 (10)	10
	Hanging	0	0	0	0	0	0	5 (2.2)	7 (3)	0	163 (70.6)	49 (21.2)	7 (3)	231
	Road traffic incident	0	1 (0.4)	0	4 (1.7)	5 (2.1)	1 (0.4)	19 (8.1)	7 (3)	1 (0.4)	147 (62.3)	44 (18.6)	7 (3)	236
	Shooting	0	0	0	0	0	0	0	1 (2.9)	0	31 (91.2)	2 (5.9)	0	34
	Stabbing	0	0	0	0	0	0	8 (8.7)	1 (1.1)	1 (1.1)	69 (75)	11 (12)	2 (2.2)	92
	Drowning	0	0	0	1 (6.3)	0	0	1 (6.3)	0	0	8 (50)	5 (31.3)	1 (6.3)	16
	Suicideb	0	0	0	0	0	0	2 (2)	1 (1)	0	65 (65)	26 (26)	6 (6)	100
	Miscellaneous	3 (3.4)	1 (1.1)	0	6 (6.7)	2 (2.2)	0	4 (4.5)	2 (2.2)	0	50 (56.2)	16 (18)	5 (5.6)	89
	Total	3 (0.3)	3 (0.3)	0	17 (1.8)	12 (1.3)	1 (0.1)	45 (4.9)	21 (2.3)	2 (0.2)	609 (65.9)	175 (18.9)	36 (3.9)	924

a Gender. b Not including hanging.

The single most common cause of non-cardiac (non-traumatic) cardiac arrest was overdose and these patients accounted for 34.5% (n=267) of this group. The vast majority (98.1%; n=262) were aged 18–35 years. Three times as many males as females suffered a cardiac arrest from overdosing (199 vs. 61) (Table 2).

Respiratory compromise was the second most common cause of non-cardiac (non-traumatic) arrest (Table 2). 88% received resuscitation attempts.

There were 51 cases of Sudden Infant Death Syndrome (SIDS) which accounted for 15.8% of all arrests in the infant group. Resuscitation was attempted in 68.6% of cases.

The single most common cause of traumatic cardiac arrest was road traffic incidents (25.5%), closely followed by hangings (25%) (Table 2). 65% of road traffic incident patients received resuscitative efforts. In contrast, resuscitation was only attempted in 34% of those whose arrest was due to hanging. Interestingly, the vast majority of hangings (70.6%; n=163) were adult males (Table 2).

10.8% (n=100) of traumatic cardiac arrests were attributed to suicide (e.g. jumping from a height; excluding hanging). Males aged 18–35 years accounted for the majority of traumatic suicide cases (65%; n=65).

Stabbings accounted for a total of 10% (n=92) of traumatic arrests. Most of these patients were aged 18–35 years (89.1%; n=82) with 10.9% (n=10) aged 9–17 years. 82% (n=75) of patients who were stabbed received resuscitative efforts. Those that were deemed not viable for resuscitation were exclusive to the 18–35 year age group (Table 2).

3.2. Location, witnessed and bystander CPR rates and presenting rhythm—patients who had resuscitation commenced 

Further information about location of cardiac arrest, witnessed status and bystander CPR was collected for the 1748 patients for whom resuscitation was started (Table 3).

Table 3. Bystander witnessed, crew witnessed and bystander CPR rates for patients for whom resuscitation was commenced.

		Age group				Total
		Under 1 year	1–8 years	9–17 years	18–35 years
	Number of patients (for whom information is known)	230	140	121	1020	1511
	Witnessed, n (%)	73 (31.7)	54 (38.6)	63 (52.1)	518 (50.8)	708 (46.9)
	Bystander witnessed, n (%)	59 (25.7)	47 (33.6)	51 (42.1)	366 (35.9)	523 (34.6)
	Crew witnessed, n (%)	14 (6.1)	7 (5.0)	12 (9.9)	152 (14.9)	185 (12.2)
	Bystander CPR, na (%)	105/268 (39.2)	59/153 (38.6)	49/150 (32.7)	288/1175 (33.0)	601/1746 (34.4)
	Bystander witnessed patients who received bystander CPR, nb (%)	36/59 (61)	24/47 (48.9)	27/51 (52.9)	201/366 (54.9)	288/523 (55.1)

a Denominator is the number of patients for whom this information was recorded. b Denominator is the number of bystander witnessed arrests.

Of the 952 patients for whom location information was available, 65.4% of cardiac arrests occurred in private locations (e.g. home or nursing home) and 34.6% occurred in public (e.g. street, work).

In total 46.9% of patients aged 35 years and under had their cardiac arrest witnessed by a bystander (34.6%) or ambulance crews (12.2%). The rates of bystander witnessed arrests were lowest in the infant group at 25.7% (Fig. 2 and Table 3). Fig. 2 shows the percentage of witnessed and unwitnessed cardiac arrests for all age groups.

View Large Image Download to PowerPoint

Fig. 2. Witnessed and unwitnessed arrests by age group.

There was a significant trend in the occurrence of bystander CPR by age group. 39.2% of infants, 38.6% of younger children, 32.7% of elder children and 33.0% of adults received bystander CPR (χ2 for trend=4.39, p=0.03). Prior to the arrival of ambulance crews, a total of 34.4% of patients viable for resuscitation received bystander CPR (Table 3). Of those that were bystander witnessed, bystander CPR was given in 55.1% of cases. The rate of bystander CPR remained constant over the 4-year study period.

The presenting rhythm was obtained for 1363 patients who had resuscitation commenced. Of these, 162 (11.9%) presented with ventricular fibrillation (VF) or ventricular tachycardia (VT). The remainder presented with asystole (68.6%), PEA (17.8%), or other rhythm (1.8%). The frequency of VF/VT was lowest for those aged under 1 year (no occurrences), 1.8% for those aged 1–8 years, 13.9% for those aged 18–35 years and highest for those aged 9–17 (16.1%). There were no differences in median response times for patients who presented in VF versus those who did not; emergency call to arrival at scene and emergency call to arrival at patient times were 6 and 7min, respectively for those aged 1–35 years. Median response times were slightly quicker for infants (under 1 year) at 5 and 6min, respectively.

3.3. Patient outcomes—patients who had resuscitation commenced 

The total number of patients aged 35 years and under achieving a return of spontaneous circulation (ROSC) following treatment for cardiac arrest and whilst still in the field was 126 (7.2%) (Table 4). There was a significant difference in the proportion of ROSC between the age groups (p=0.03). The highest frequency of ROSC was seen in those aged 9–17 years (10.0%) and the lowest in those aged 1–8 (2.0%). There was a significant difference between these two groups (p<0.05) and between those aged 1–8 years and 18–35 years (p<0.001 using the Marascuilo procedure for multiple comparisons) (Table 4).

Table 4. Rates of ROSC and survival for patients for whom resuscitation was commenced.

		Age group				Total
		Under 1 year	1–8 years	9–17 years	18–35 years
	Number of patients	269	153	150	1176	1748
	ROSC, n (%)	14 (5.2)	3 (2.0)	15 (10.0)	94 (8.0)	126 (7.2)
	Survival status unknown, n (%)	142 (52.8)	83 (54.2)	98 (65.3)	718 (61.1)	1041 (59.8)
	Survival status known, n (%)	127 (47.2)	70 (45.8)	52 (34.7)	458 (38.9)	707 (40.4)
	Survived to hospital discharge,an (%)	5 (3.9)	4 (5.7)	4 (7.7)	27 (5.9)	40 (5.7)
	Expired in field,an (%)	14 (11.0)	8 (11.4)	5 (9.6)	91 (19.9)	118 (16.7)
	Died in hospital,an (%)	108 (85.0)	58 (82.9)	43 (82.7)	340 (74.2)	549 (77.7)

a Denominator used for the calculation of percentages is the number of patients for whom the survival outcome was known.

As a result of missing documentation and difficulties in obtaining data from some hospitals, it was not possible to track survival data for over half of the patients for whom resuscitation was commenced (59.2%). The survival to hospital discharge rate for patients for whom data were available was 5.6% (n=40/713). Survival was not significantly different across age groups (p=0.90). It ranged from 3.9% for those aged under 1 year to 7.7% for those aged 18–35 years (Table 4). A multivariate analysis revealed that patients presenting with asystole were significantly less likely to survive (p=0.009, OR, 0.261; CI, 0.095 to 0.712) to hospital discharge than those presenting with VF/VT.

4.1. Limitations 

Although the sample size of patients aged 35 years and under was the largest known of its kind, survival outcome retrieval was low and limits the conclusions that can be drawn regarding factors associated with increased survival. Additionally, information from autopsy reports would have greatly increased the accuracy of aetiology allocation and reduced the risk of over reporting the number of cases assigned a cardiac origin.