• 
• 

Three Part Question

In [children suffering a cardiac arrest] does [the use of automatic external defibrillators] [improve outcome]?

Clinical Scenario

The major determinant of survival from cardiac arrest due to ventricular fibrillation (VF) is the time interval between collapse and defibrillation. For every minute that defibrillation is delayed, survival rates drop by as much as 7 – 10%. In an effort to hasten access to defibrillation, the automatic external defibrillator (AED) was developed.
AED's can now be found in a wide variety of public places from airports to supermarkets. Designed for use by people with little or no training, they are of proven benefit in improving the prognosis of adults who suffer out-of-hospital cardiac arrests.
Ventricular fibrillation was assumed to have an extremely low incidence in children, therefore AED were designed solely for the treatment of adults. In 1995, a study by Mogayzel et al. demonstrated that VF was the initial rhythm in as many as 19% of children suffering out-of-hospital cardiac arrest. Therefore the use of AED's may be beneficial to a significant number of children.
The efficacy of AED's in children is uncertain due to two major concerns. Firstly, there are concerns as to whether an AED can accurately assess ventricular arrythmia in children and respond appropriately. Secondly, AED's provide fixed-energy shocks that could potentially inflict myocardial damage in young children.
In the event of an out-of-hospital paediatric cardiac arrest, would it be safe to use an adult AED?

Search Strategy

Medline 1966 – 06/05 using the Ovid interface, Embase 1980 – 2005 Week 26 using the Ovid interface, The Cochrane Library, Issue 2, 2005
Medline: [exp Pediatrics/ OR exp Child/] AND [automatic external defibrillator$.mp OR automated external defibrillator$.mp OR automatic external defibrillation$.mp OR automated external defibrillation$.mp OR (AED adj20 defibrillator$).mp OR automatic defibrillator$.mp OR automated defibrillator$.mp OR automatic defibrillation$.mp OR automated defibrillation$.mp]
Embase: [exp Pediatrics/ OR exp Child/] AND [automatic external defibrillator$.mp OR automated external defibrillator$.mp OR automatic external defibrillation$.mp OR automated external defibrillation$.mp OR (AED adj20 defibrillator$).mp OR automatic defibrillator$.mp OR automated defibrillator$.mp OR automatic defibrillation$.mp OR automated defibrillation$.mp]
Cochrane Library: 'automatic external defibrillator child', 'automated external defibrillator child', 'automatic external defibrillator paediatric', 'automated external defibrillator paediatric'

Search Outcome

Medline yielded 45 references of which 18 were relevant. Embase yielded 22 references of which 11 were relevant. Excluding repeat results, between these two databases a total of 20 relevant papers were identified. Four papers provided evidence of insufficient quality for inclusion. The full texts of three articles were unavailable as one was in Spanish and two were from journals that could not be sourced.
The Cochrane Library yielded no papers relevant to the three-part question.

Relevant Paper(s)

Author, date and country	Patient group	Study type (level of evidence)	Outcomes	Key results	Study Weaknesses
Atkinson, E. et al 2003 USA	ECG's from 203 children aged <8 years recorded by a monitor-defibrillator. The ECG's were subsequently analysed by an expert panel and a LIFEPAK 500 AED.	Prospective blinded cohort study (Level 3b)	Anterior-posterior/sternal-apex pad placement superior for specificity	Specificity was high when pads were in either the anterior-posterior position (99.4%) or the sternal-apex position (99.1)	The rhythms were not recorded directly by the AED being tested.
			Sensitivity (AED correctly recommends a shock for shockable rhythms)	Sensitivity was excellent at 99% (95% CI: 93%-100%)
			Specificity (AED does not recommend a shock for non-shockable rhythms)	Specificity was excellent at 99.5% (95% CI: 99.0%-99.8%)
Cecchin, F. et al 2001 USA	191 Children aged <12 years 134 children had rhythms recorded directly from a modified AED. 57 children¡¦s pre-recorded rhythms were digitised for later analysis by an AED. All rhythms were analysed by an expert panel and an Agilent Heartstream ForeRunner 2 AED.	Prospective and retrospective study (Levels 3b and 4)	Differences between paediatric and adult rhythm characteristics	Rate and conduction characteristics of paediatric VF/VT are significantly different to those of adults (P<0.001)	Analysis of digitised rhythms may differ from the analysis of a direct recording. Data for the adult database was collected in another study and needed to be normalised for comparison. More accurate comparison could be achieved by collecting adult data by the same method as the paediatric data in this study.
			Sensitivity (AED correctly recommends a shock for shockable rhythms)	Sensitivity was higher for VF than rapid VT (96% vs 71% respectively)
			Specificity (AED does not recommend a shock for non-shockable rhythms)	Specificity was excellent at 100%
Atkins, D.L. et al 1998 USA	18 children aged 5 – 15 (mean 12.1 ±3.7) years who suffered cardiac arrest between 1/7/1988 and 1/2/1997	Retrospective cohort study (Level 4)	Recognition of VF	VF accurately recognised 22/25 times. Overall sensitivity of the AED's was 88%.	Small sample size. The data was not independently reviewed and was not blinded. Many models of AED were included in the study. Each may have differing sensitivity and specificity. Data for two children who did not meet the inclusion criteria were included as a control for the group who received shocks – inaccurate comparison due to differing age groups. No statistical analysis.
			Recognition of other rhythms	Accurately recognised: Sinus bradycardia – 6/6 times, Sinus tachycardia – 4/4 times, Asystole/PEA – 32/32 times. Overall specificity of the AED's was 100%.
			Survival	9 patients suffered VF: 7 treated, 3 survived. 2 untreated, 0 survived. 9 patients suffered other arrhythmias:1 survived, 8 died.
Samson, R. et al 2003 USA	Relevant literature	Review/Practice guideline (Level 4)	Safety and efficacy of AED's in paediatric patients.	AED's may be used for children aged 1-8 years of age who show no signs of circulation. Ideally the AED should provide a paediatric dose. The arrhythmia detection algorithm should have high sensitivity and specificity for paediatric shocks. Lone rescuers should provide 1 minute of CPR before any attaching an AED or seeking help.	No indication as to how the papers were identified.
Atkins, D.L., Kenney, M.A. 2004 USA	Relevant literature	Review (Level 4)	Safety and efficacy of AED's in paediatric patients.	Review recognises that the technology now exists to extend AED use to children. Expected that as AED use on children gains acceptance, outcome of paediatric cardiac arrest will improve.	No indication as to how the papers were collected. No critical appraisals were included.
Rugolotto, S. 2004 Italy	Relevant literature	Review (Level 4)	Developments in the use of AED's in children.	Review finds current published evidence on paediatric AED use encouraging. Further research needed to determine if AED's do indeed improve the outcome of paediatric cardiac arrest.	No indication as to how the papers were collected. No critical appraisals were included.
Berg, R.A. 2004 USA	Relevant literature	Review (Level 4)	Safety and efficacy of attenuated adult biphasic shocks in children.	Review supports the recommendation that AED's can be used in children aged 1-8 years who show no signs of circulation.	No indication as to how the papers were collected. The reviewer referred heavily to his own past and present research. No critical appraisals were included.
Gurnett, C.A., Atkins, D.L. 2000 USA	39-month-old male child with a family history of hypertrophic cardiomyopathy	Case report (Level 5)	Survival	AED correctly identified VF and advised a shock. Shock converted VF to sinus bradycardia. Child discharged home following implantation of ICD.	Isolated case. No follow-up.
			Cardiac damage	ECG demonstrated transient changes consistent with recent electric countershock. Creatine kinase and Troponin I normal. Cardiac catheterisation and echocardiography demonstrated good ventricular function.
König, B., Benger, J., Goldsworthy, L. 2005 UK	6 year old female child, previously asymptomatic	Case report (Level 5)	Survival	Initially asystolic, spontaneously converted to VF during CPR – recognised by AED as shockable. Reverted to asystole after 1st 150J shock. Rhythm changed to VF after third adrenalin dose. Reverted to asystole after 2nd 150J shock – converted to narrow-complex bradycardia after one-minute accompanied by spontaneous respiratory effort. Diagnosed with long QT syndrome. Discharged home following implantation of ICD.	Isolated case. No data on cardiac enzymes or ventricular function provided. No follow-up.
Berg, R.A. et al. 2005 USA	32 piglets 1 to 3 months of age as animal models for paediatric VF. VF induced (under isoflurane anaesthesia) by 100Hz AC current delivered to the right ventricle via a pacing catheter electrode. Randomised after 7 minutes in VF to receive either adult or paediatric defibrillation dose.	Animal study: Randomised trial (Level 5)	Termination of VF	Adult energy dosing was significantly better than paediatric energy at terminating VF on the 1st shock (P = 0.01)	The pacing electrode used to induce VF and the presence of catheters in the right atrium, left ventricle and aorta may have contributed to the myocardial damage detected. Post-resuscitation haemodynamic support was not provided thus likely reducing 24hr survival. The data was not blinded. Animal data may not be applicable to humans.
			Time until return of spontaneous circulation	There was no significant difference between adult and paediatric energy dose (P = 0.3)
			Left Ventricular Ejection Fraction (1 and 4 hours post resuscitation)	LVEF decreased less after paediatric energy doses (P < 0.05)
			Troponin T levels	Undetectable at baseline for all piglets. Adult energy dose: Raised in 6 out of 11 piglets
			24h survival with good neurological outcome	24h survival with good neurological outcome was significantly superior with paediatric energy dosing.
Berg, R.A. et al 2004 USA	48 female swine as animal models of paediatric VF. VF induced (under isoflurane anaesthesia) by 100Hz AC current delivered to the right ventricle via a pacing catheter electrode. Randomised after 7 minutes in VF to receive either monophasic weight-based shocks or attenuated adult biphasic shocks.	Animal study: Randomised trial (Level 5)	Number of shocks required to terminate VF	Attenuated adult biphasic dosage terminated VF with significantly fewer shocks than weight-based monophasic energy dosage across all weight categories. (P < 0.01)	Animal data may not be applicable to humans. The energy attenuators used were prototypes. The data was not blinded.
			Return of spontaneous circulation	Superior after attenuated adult biphasic shocks compared with weight-based monophasic shocks (23 out of 24 piglets vs 15 out of 24 piglets)
			Left ventricular ejection fraction 4 hours postresuscitation	LVEF was significantly higher after attenuated adult biphasic shocks in the 24kg group (P < 0.05)
			24h survival with good neurological outcome	Superior after attenuated adult biphasic shocks compared with weight-based monophasic shocks (20 out of 24 piglets)
Tang, W. et al 2002 USA	Phase 1: 20 male piglets weighing between 3.5 and 25kg. Resuscitated with a manual biphasic defibrillator set at 50J. Phase 2: 9 male piglets. Resuscitated with an AED equipped with energy attenuating paediatric pads. VF was induced by the delivery of AC current into the right ventricle (under pentobarbital anaesthesia).	Animal study (Level 5)	Success of defibrillation	Phase 1: All animals were successfully defibrillated Phase 2: All animals were successfully defibrillated	No comparison between manual defibrillator and AED. Animal data may not be applicable to humans.
			Duration of CPR	Phase 1: There were no statistically significant differences in the duration of CPR between different weight ranges (range: 84 – 150 seconds). Phase 2: There were no statistically significant differences in the duration of CPR between different weight ranges (range 86 – 153.7 seconds).
			Post-resuscitation haemodynamic and myocardial function.	Reduced in all animals immediately after resuscitation. Returned to baseline after 4 hours. No statistically significant differences between weight ranges. No evidence of myocardial damage in any animal at autopsy.
Jorgenson, D., et al. 2002 USA	Simulated paediatric thoracic impedances	Experimental study (Level 5)	Energy attenuation.	Attenuation circuitry reduced the AED's 150J output to 50J over a range of patient impedances.	Attenuation circuitry reduced the AED's 150J output to 50J over a range of patient impedances.

Comment(s)

Kerber et al. recommended that AED's should demonstrate sensitivity of >90% for VF, >75% for rapid VT and specificity of >99% for non-shockable rhythms. It has been shown by Atkinson et al and Cecchin et al that AED sensitivity for VF in children exceeds these recommendations. However, Cecchin et al noted that the sensitivity of the Agilent Heartstream ForeRunner AED for rapid VT in children fell short of these recommendations at 71%. Given the rarity of rapid VT in cardiac arrest, this lone observation should not preclude the use of AED's in children. Both studies have also demonstrated that AED's also exceed the recommended specificity for non-shockable rhythms in children. MacDonald et al showed that 53.85% of all AED errors that occurred in Boston between 1995 and 1997 were due to mistakes by the operator. Therefore the development of energy attenuating pads that do not complicate the operation of the AED is welcomed. Despite evidence that adult AED's can be successfully used for younger children, recent animal data (Berg et al 2005) suggests that paediatric energy dosing significantly reduces the risk of post-resuscitation myocardial damage and results in higher rates of survival. Therefore the use of energy attenuating paediatric pads in conjunction with adult AED's is highly desirable. The use of AED's in children younger than one year cannot be recommended with confidence due to limited data, small sample sizes and the greater potential for myocardial damage. Also, the usefulness of AED's in this very young age group may be limited due to their lower incidence of shockable rhythms.

Clinical Bottom Line

Based on the available evidence, it appears that automatic external defibrillators are capable of improving the outcome of VF in children older than one year of age. In addition energy attenuating paediatric pads should used if available for the one to eight year old age group.

References

1. Atkinson, E. et al. Specificity and sensitivity of automated external defibrillator rhythm analysis in infants and children. Ann Emerg Med. 2003; 42: 185-196
2. Cecchin, F. et al. Is arrhythmia detection by automatic external defibrillator accurate for children? Sensitivity and specificity of an automatic external defibrillator algorithm in 696 pediatric arrhythmias. Circulation. 2001; 103: 2483-2488
3. Atkins, D.L., Hartley, L.L., York, D.K. Accurate recognition and effective treatment of ventricular fibrillation by automated external defibrillators in adolescents. Pediatrics. 1998; 101: 393-397
4. Samson, R. et al. Use of automated external defibrillators for children: an update. An advisory statement from the Pediatric Advanced Life Support Task Force, International Liaison Committee on Resuscitation. Resuscitation. 2003; 57: 237-243
5. Mogayzel, C. et al. Out-of-hospital ventricular fibrillation in children and adolescents: causes and outcomes. Ann Emerg Med. 1995; 25: 484-491
6. Atkins, D.L., Kenney, M.A. Automated external defibrillators: safety and efficacy in children and adolescents. Pediatr Clin N Am. 2004; 51: 1443-1462
7. Rugolotto, S. Automated external defibrillators for children: what is new? Ital Heart J. 2004; 5: 6-10
8. Berg, R.A. Attenuated adult biphasic shocks for prolonged pediatric ventricular fibrillation: Support for pediatric automated defibrillators. Crit Care Med. 2004; 32(Suppl.): S352-S355
9. Gurnett, C.A., Atkins, D.L. Successful use of a biphasic waveform automated external defibrillator in a high-risk child. Am J Cardiol. 2000; 86: 1051-1053
10. König, B., Benger, J., Goldsworthy, L. Automatic external defibrillation in a 6 year old. Arch Dis Child. 2005; 90: 310-311
11. Berg, R.A., et al. Better outcome after pediatric defibrillation dosage than adult dosage in a swine model of pediatric ventricular fibrillation. J Am Coll Cardiol. 2005; 45: 786-9
12. Berg, R.A. et al. Attenuated adult biphasic shocks compared with weight-based monophasic shocks in a swine model of prolonged pediatric ventricular fibrillation. Resuscitation. 2004; 61: 189-197
13. Tang, W. et al. Fixed-energy biphasic waveform defibrillation in a pediatric model of cardiac arrest and resuscitation. Crit Care Med. 2002; 30: 2736-2741
14. Jorgenson, D., et al. Energy attenuator for pediatric application of an automated external defibrillator. Crit Care Med. 2002; 30(Suppl.): S145-S147.
15. Kerber, R.E., et al. Automatic external defibrillators for public access defibrillation. Circulation. 1997; 95: 1677-1682
16. MacDonald, R.D., et al. Performance and error analysis of automated external defibrilaltor use in the out-of-hospital setting. Ann Emerg Med. 2001; 38:262-267