Specificity and sensitivity of automated external defibrillator rhythm analysis in infants and children

STUDY OBJECTIVES

The rhythm detection algorithms of automated external defibrillators have been derived from adult rhythms, and their ability to discriminate between shockable and nonshockable rhythms in children is largely unknown. This study evaluates the performance of 1 automated external defibrillator algorithm in infants and children and evaluates algorithm performance with anterior-posterior versus sternal-apex lead placement.

METHODS

We enrolled pediatric patients in a critical care unit, an electrophysiology laboratory, and a cardiac operating room. A monitor-defibrillator recorded ECGs by means of standard defibrillation-monitor pads. Selected 15-second rhythm samples were played into a LIFEPAK 500 automated external defibrillator, and the automated external defibrillator "shock/no shock" decision was documented. To determine sensitivity and specificity, the automated external defibrillator decision was compared with the "shockable" versus "nonshockable" rhythm classification provided by 3 expert clinicians who were blinded to the automated external defibrillator decision.

RESULTS

We recorded 1,561 rhythm samples from 203 pediatric patients (median age 11 months; range, day of birth to 7 years). The automated external defibrillator recommended a shock for 72 of 73 rhythm samples classified as coarse ventricular fibrillation by expert review (sensitivity 99%; 95% confidence interval [CI] 93% to 100%); and correctly reached a "no shock advised" decision for 1,465 of 1,472 rhythm samples classified as nonshockable by experts (specificity 99.5%). Specificity was 99.1% (95% CI 97.8% to 99.8%) with the sternal-apex lead and 99.4% (95% CI 98.1% to 99.9%) with the anterior-posterior lead.

CONCLUSION

This automated external defibrillator algorithm has high specificity and sensitivity when used in infants and children with either sternal-apex or anterior-posterior lead placement.

Optimal defibrillation response intervals for maximum out-of-hospital cardiac arrest survival rates

STUDY OBJECTIVE

Many centers optimize their emergency medical services (EMS) systems to achieve a target defibrillation response interval of "call received by dispatch" to "arrival at scene by responder with defibrillator" in 8 minutes or less for at least 90% of cardiac arrest cases. The objective of this study was to analyze survival as a function of time to test the evidence for this standard.

METHODS

This prospective cohort study included all adult, cardiac etiology, out-of-hospital cardiac arrest cases from phases I and II of the Ontario Prehospital Advanced Life Support (OPALS) study. Patients in the 21 Ontario study communities received a basic life support level of care with defibrillation by ambulance and firefighters but no advanced life support. Survival was plotted as a function of the defibrillation response interval. The equation of the curve, generated by means of logistic regression, was used to estimate survival at various defibrillation response interval cutoff points.

RESULTS

From January 1, 1991, to December 31, 1997, there were 392 (4.2%) survivors overall among the 9,273 patients treated. The defibrillation response interval mean was 6.2 minutes, and the 90th percentile was 9.3 minutes. There was a steep decrease in the first 5 minutes of the survival curve, beyond which the slope gradually leveled off. Controlling for known covariates, the decrement in the odds of survival with increasing response interval was 0.77 per minute (95% confidence interval 0.74 to 0.83). The survival function predicts, for successive 90th percentile cutoff points, both survival rates and additional lives saved per year in the OPALS communities compared with the 8-minute standard: 9 minutes (4.6%; -18 lives), 8 minutes (5.9%; 0 lives), 7 minutes (7.5%; 23 lives), 6 minutes (9.5%; 51 lives), and 5 minutes (12.0%; 86 lives).

CONCLUSION

The 8-minute target established in many communities is not supported by our data as the optimal EMS defibrillation response interval for cardiac arrest. EMS system leaders should consider the effect of decreasing the 90th percentile defibrillation response interval to less than 8 minutes.

Discrepancies between the upper limit of vulnerability and defibrillation threshold: prevalence and clinical predictors

INTRODUCTION

Upper limit of vulnerability (ULV) has a strong correlation with defibrillation threshold (DFT) in patients with implantable cardioverter defibrillators (ICDs). Significant discrepancies between ULV and DFT are infrequent. The aim of this study was to characterize patients with such discrepancies.

METHODS AND RESULTS

The ULV and DFT were determined in 167 ICD patients. Univariate and multivariate analyses were used to evaluate clinical predictors of a significant difference (> or =10 J) between ULV and DFT. Only 8 patients (5%) had > or =10 J difference. ULV exceeded DFT in all of them. Absence of coronary artery disease (6/8 vs 48/159 patients; P = 0.05) and absence of documented ventricular arrhythmias (4/8 vs 12/159 patients; P = 0.01) were the only independent predictors of a significant ULV-DFT discrepancy.

CONCLUSION

Significant discrepancies between ULV and DFT occur in 5% of patients with ICDs. Absence of coronary disease and documented ventricular arrhythmias predict such a discrepancy. At ICD implant, DFT testing is recommended in these patients and in patients with a high (>20 J) ULV before first-shock energy and the need for lead repositioning are determined.

Overcoming ACLS dogma: how quickly should we change?

The ECC Guidelines 2000 considered interesting new evidence about a pre-defibrillation period of prescribed CPR to increase the probability that the postshock rhythm would be perfusing rather than asystole. If victims of out-of-hospital cardiac arrest have not received bystander CPR before the arrival of the defibrillator, a period of preshock CPR could enhance the value of the shocks. At the end of the year 2000 there was insufficient evidence to recommend any other approach than shock as soon as possible and perform CPR at all other times.

Determination of the upper limit of vulnerability using implantable cardioverter-defibrillator electrograms

BACKGROUND

The upper limit of vulnerability (ULV) correlates with the defibrillation threshold and can be determined with 1 episode of ventricular fibrillation (VF). To automate the ULV in an implantable cardioverter-defibrillator (ICD), the most vulnerable intervals must be identified from an ICD electrogram rather than the latest-peaking surface T wave (Tpeak). We hypothesized that the recovery time (TR), defined as the maximum derivative (dV/dt) of the T wave of the shock electrogram, correlates with the most vulnerable intervals.

METHODS AND RESULTS

We determined ULV, defibrillation threshold, and the most vulnerable intervals in 25 patients at ICD implantation. The ULV was the weakest T-wave shock that did not induce VF. The most vulnerable intervals were the ones associated with the strongest shocks that induced VF. Telemetered shock electrograms were stored on digital tape and differentiated offline to measure TR. Tpeak and TR were highly correlated (Tpeak-TR=-2+/-11 ms; rho=0.80, P<0.001). At least 1 most vulnerable interval timed between -20 ms and +20 ms relative to Tpeak in all patients and between -40 ms and +20 ms relative to TR in 96% of patients.

CONCLUSIONS

The recovery time of shock electrograms provides accurate information about global repolarization. TR closely approximates Tpeak. The ULV method may be automated in an ICD by timing T-wave shocks relative to TR.

Automated external defibrillators (AEDs)

Automated external defibrillators, or AEDs, will automatically analyze a patient's ECG and, if needed, deliver a defibrillating shock to the heart. We sometimes refer to these devices as AED-only devices or stand-alone AEDs. The basic function of AEDs is similar to that of defibrillator/monitors, but AEDs lack their advanced capabilities and generally don't allow manual defibrillation. A device that functions strictly as an AED is intended to be used by basic users only. Such devices are often referred to as public access defibrillators. In this Evaluation, we present our findings for a newly evaluated model, the Zoll AED Plus. We also summarize our findings for the previously evaluated model that is still on the market and describe other AEDs that are also available but that we haven't evaluated. We rate the models collectively for first-responder use and public access defibrillation (PAD) applications.

[Early defibrillation in the treatment of sudden cardiac arrest]

Recovery from nontraumatic cardiac arrest depends on the presence of all the elements in the chain of survival. "Early defibrillation" is critical because ventricular fibrillation is the most common initial dysrhythmia of sudden cardiac arrest. Defibrillation is the only treatment, and survival from ventricular fibrillation is determined by time. Out-of-hospital studies have demonstrated that defibrillation provided by first responders improves survival. Technologic advances have simplified defibrillation delivery through the development of automated external defibrillators (AEDs). Early defibrillation programs with AEDs are quickly becoming a standard of care for emergency medical service systems throughout the United States. Improvement in in-hospital survival rates from cardiac arrest is not as evident as in the emergency medical service community. Medical centers need to assess response times to cardiac arrest and implement AED programs. All the nurses should learn to use an AED as part of basic life support training.

Patient-dependent current dosing for semi-automatic external defibrillators (AED)

The improvements in semiconductors and modern circuitry allow new waveforms to be created for treating life-threatening heart fibrillation. A comparison of common waveforms shows that there is no definite optimal waveform. Especially in the case of early defibrillation by novices, the question of dosage should be re-discussed. While a physician may be able to dose the intensity of the therapeutic electric shock, one can't expect that from someone having no medical training. Common AEDs have predefined energy levels, that are delivered to a patient regardless of the patient's size and weight, etc. Current-based defibrillation provides a therapy matched to patient parameters, keeping the myocard stress as low as possible so that the heart has better chances of resuming a normal rhythm.

[Out-of-hospital cardiac arrest. Mechanisms and treatment with automated external defibrillator]

In Denmark, approximately 4500 persons suffer yearly from an out-of-hospital cardiac arrest with mortality close to 100%. The principal arrhythmia is ventricular fibrillation, which can only be treated effectively with prompt external defibrillation. Automatic external defibrillators (AED) are small, portable, easily operated devices. They have documented high specificity and sensitivity. Moreover, biphasic automatic external defibrillators are at least as effective as traditional monophasic defibrillators. Survival rates with good neurological status as high as 60% have been reported. Better survival of out-of-hospital cardiac arrest victims requires, however, improvements throughout the chain of survival, not only more automatic external defibrillators. Therefore, the health care system has to discuss thoroughly and solve important questions regarding organisation, logistics, education, and legal aspects in order to improve survival for out-of-hospital cardiac arrest victims.

Time to first shock by emergency medical technicians with automated external defibrillators

The interval from collapse to electrical rescue shock is a critical determinant of successful defibrillation in cardiac arrest. In order to achieve the earliest possible defibrillation, many emergency medical services (EMS) systems equip first-responding units with an automated external defibrillator (AED).

OBJECTIVE

To measure the time from on-scene emergency medical technician (EMT) recognition of cardiac arrest to AED application and shock in ventricular fibrillation (VF) arrest. In addition, the authors sought to understand the reasons for delays.

METHODS

Using the AED recordings and written EMS reports, the authors conducted a retrospective cohort study of all persons who experienced an EMS-attended VF cardiac arrest in which an AED was applied and a shock delivered by an EMT, from January 1999 through December 2000 (n = 177). Based on the bimodal distribution of times, two groups were assembled: no delay (time to shock < or = 90 seconds) and delayed (time to shock > 90 seconds). Patient and event characteristics associated with delay status were determined using Mantel-Haenszel methods.

RESULTS

The median (25th, 75th percentile) time from cardiac arrest recognition to shock was 51 (43, 64) seconds. Ninety-four percent (n = 166) of the cohort received a shock within 90 seconds. Delayed shock was associated with unwitnessed arrest status (odds ratio = 9.3, 95% confidence interval = 2.3, 36.8) and nursing home location (odds ratio = 10.0, 95% confidence interval = 2.1, 47.5).

CONCLUSION

The findings suggest that a 1-minute goal and a 90-second minimum standard for time to first shock are appropriate for EMT AED defibrillation in the field.

Success of serial external electrical cardioversion of persistent atrial fibrillation in maintaining sinus rhythm. A randomized study

AIMS

The aim of this prospective, randomized study was to determine the efficacy of a serial external electrical cardioversion strategy in maintaining sinus rhythm after 12 months in patients with recurrent persistent atrial fibrillation.

METHODS AND RESULTS

Ninety patients with persistent atrial fibrillation lasting more than 72 h but less than 1 year were randomized in a one to one fashion to repetition of up to two electrical cardioversions in the event of relapse of atrial fibrillation detected within 1 month of the previous electrical cardioversion (Group AGG), or to non-treatment of atrial fibrillation relapse (Group CTL). ECGs were scheduled at 6 h, 7 days, and 1 month. Clinical examination and ECGs were repeated during the 6-month and 12-month follow-up examinations. Echocardiography was repeated during the 6-month follow-up examination. Clinical and echocardiographic characteristics were similar in the two groups. All patients were treated with antiarrhythmic drugs before electrical cardioversion and throughout follow-up. After 12 months, sinus rhythm was maintained in 53% of Group AGG patients and in 29% of Group CTL patients (P<0.03). After 6 months, left ventricular ejection fraction had recovered significantly only in Group AGG (56.8 +/- 9.0% at enrollment vs 60.4 +/- 9.4% at 6 months,P <0.001).

CONCLUSION

These results demonstrate that an aggressive policy towards persistent atrial fibrillation by means of repetition of electrical cardioversion after early atrial fibrillation recurrence is useful in maintaining sinus rhythm after 12 months.