Skip main navigation

Temporal Trends in Sudden Cardiac Arrest

A 25-Year Emergency Medical Services Perspective
Originally publishedhttps://doi.org/10.1161/01.CIR.0000070950.17208.2ACirculation. 2003;107:2780–2785

Abstract

Background— Little is known about temporal trends in survival and prognostic characteristics of patients with out-of-hospital cardiac arrest treated by emergency medical services (EMS). We hypothesized that an evolving combination of beneficial and adverse factors may contribute to temporal patterns of survival.

Methods and Results— We evaluated a population-based cohort of EMS-treated adult patients with cardiac arrest (n=12 591) from 1977 to 2001 in King County, Washington. Time was grouped into an initial 5-year period and 5 successive 4-year periods. We sought to determine the potential impact of temporal changes in prognostic factors typically beyond EMS control termed “fate” factors (for example, patient age) and factors implemented by EMS termed “program” factors (programs of dispatcher-assisted cardiopulmonary resuscitation and basic life support defibrillation). Several characteristics associated with survival changed over time. Observed survival did not change over time among all patients with cardiac arrest (OR=0.98 [0.95, 1.01], trend for each successive time period) and improved over time among patients with witnessed ventricular fibrillation (OR=1.05 [1.01, 1.09]). In models that included all patients with cardiac arrest and controlled for fate factors, advancing time period was associated with an increase in survival (OR=1.08 [1.05, 1.11]). Conversely, in models that controlled for program factors, advancing time period was associated with a decrease in survival (OR=0.95 [0.93, 0.98]). Results were similar among patients with witnessed ventricular fibrillation.

Conclusions— The static temporal pattern of survival from cardiac arrest appeared to result from an evolving balance of prognostic factors. Programs implemented by EMS appeared to counter adverse temporal trends in prognostic factors typically beyond EMS control.

Over the last 3 decades, advances in the understanding of cardiac arrest and resuscitation have provided opportunities to strengthen the links in the chain of survival.1 Despite the apparent progress, survival has remained poor, and hundreds of thousands of persons die each year in the United States of out-of-hospital cardiac arrest.2 Paramedic services in King County, Washington, began 25 years ago. From the outset of paramedic services, information about each patient with cardiac arrest treated by emergency medical services (EMS) has been collected in a standard manner.3 During the 25-year period, the EMS system of King County implemented two programs designed to improve survival: (1) dispatcher-assisted telephone instruction in cardiopulmonary resuscitation (CPR) to increase the proportion of patients receiving prompt bystander CPR and (2) the use of manual and automated external defibrillators by first-tier responders to achieve earlier defibrillation in ventricular fibrillation arrest. The importance of these EMS-directed efforts with regard to temporal trends in survival relative to other potential prognostic characteristics is uncertain.4–8 We hypothesized that an evolving combination of beneficial and adverse factors may contribute to temporal patterns of survival. To address this hypothesis, we investigated temporal trends in both survival and prognostic characteristics of EMS-treated cardiac arrest caused by heart disease.

Methods

Study Design, Population, and Setting

The investigation was an observational study of all persons 18 years or older with out-of-hospital cardiac arrest caused by heart disease who were treated by EMS in King County, Washington (excluding the city of Seattle) from 1977 to 2001. Although some portions of King County had paramedic services beginning January 1, 1977, other areas did not initiate service until September 1, 1977. These areas were included only after the onset of paramedic services. King County, Washington (excluding Seattle), grew steadily over the past 25 years, with the population increasing from ≈667 500 in 1977 to 1 196 000 in 2000.9 The study was approved by the investigators’ institutional review board.

EMS System

The study area is served by a two-tiered EMS response system. Fire engines and/or basic life support (BLS) aid units staffed with emergency medical technician-trained fire fighters provide the first tier. A program of BLS defibrillation began in the late 1970s and became operational throughout the county by 1986. Paramedic-staffed advance life support (ALS) units provide the second tier. The BLS and ALS units are dispatched simultaneously by a standard protocol in the event of cardiac arrest. A program of dispatcher-assisted telephone CPR was initiated in 1982 whereby dispatchers offer CPR instruction to bystanders when a patient with cardiac arrest is not receiving CPR before EMS arrival.

Measures

For each patient with cardiac arrest, BLS and ALS personnel completed a medical incident report form that includes information about patient demographics (age and gender), event circumstances (witnessed, bystander CPR, arrest [victim collapse] before EMS arrival, presenting rhythm), and EMS characteristics (response intervals, treatment, and immediate outcome). Information about these variables has been collected as standard over time, although information about the location of the cardiac arrest was not recorded until the mid-1980s. In addition to EMS report forms, death certificates and/or hospital records for each patient were reviewed to determine a cardiac cause of the arrest and survival to hospital discharge (the study outcome).

Statistical Analysis

Two sets of analyses were conducted: one for all EMS-treated arrests and a second restricted to witnessed ventricular fibrillation arrests. Temporal patterns for various patients, events, and EMS characteristics were determined. Time was grouped into an initial 5-year period and 5 successive 4-year periods. The initial time period of 5 years was chosen a priori because paramedic services were instituted throughout 1977, resulting in fewer events during the initial year. Next, using data from the entire study period, the odds ratio of survival to hospital discharge associated with the various characteristics was computed by means of logistic regression.

Logistic regression was used to assess the observed temporal pattern of survival to hospital discharge as well as to investigate the potential explanations for these temporal patterns. We sought to determine the potential impact of temporal changes in “fate” factors and “program” factors on survival from out-of-hospital cardiac arrest.10 Fate factors were patient or event characteristics typically beyond EMS control; program factors consisted of changes in care that resulted from programs implemented by EMS, specifically dispatcher-assisted CPR and BLS defibrillation. The 1977 to 1981 period served as the referent time period.

The analyses contrasted the observed temporal pattern of survival with three models. The first model, termed “model 1: fate factors constant,” sought to determine potential temporal patterns in survival had fate factors remained constant over time, but the EMS programs of dispatcher-assisted CPR and BLS defibrillation occurred (as they did). Thus, in addition to time period, “model 1: fate-factors constant” adjusted for patient age and gender, witness status, presenting rhythm, and arrest before EMS arrival status in the analysis of all rhythms, and age, gender, and arrest before arrival status for the analysis restricted to witnessed ventricular fibrillation. The second model, termed “model 2: fate factors constant,” sought to determine potential temporal patterns in survival had fate factors including BLS and ALS response intervals remained constant over time, but the programs of dispatcher-assisted CPR and BLS defibrillation occurred. Although BLS and ALS response intervals may be due to a number of factors, EMS has maintained consistent response protocols for cardiac arrest victims throughout the study period. Consequently, for this analysis, BLS and ALS response intervals were considered to be fate factors. Thus “model 2: fate factors constant” included the variables of model 1 plus BLS and ALS response intervals. The third model, termed “model 3: programs constant,” sought to determine potential temporal patterns in survival had the programs of dispatcher-assisted CPR and BLS defibrillation not been implemented, but the observed changes in fate factors occurred. Thus, in addition to time period, “model 3: program constant” included citizen CPR status in the all-rhythms analysis and citizen CPR and defibrillation response interval in the witnessed ventricular fibrillation analysis. In sensitivity analyses assessing the temporal pattern of survival and the potential impact of fate and program factors, calendar time in years (ie, 1977, 1978, and so on) was modeled as a continuous variable. Analyses were performed with SPSS 11.0 software (SPSS Inc).

Results

Over the entire study period, the EMS treated 12 591 persons with out-of-hospital cardiac arrest and 4775 persons with witnessed ventricular fibrillation arrest caused by heart disease (Figure). The distribution of several patient, event, and EMS characteristics changed over time in the analysis of all rhythms (Table 1). Specifically, average age and BLS and ALS response intervals and the proportions who were women, treated with citizen CPR, defibrillated by BLS, and arrested before EMS arrival increased over time, whereas the average defibrillation response interval and the proportions who were witnessed and presented in ventricular fibrillation decreased over time. Similar patterns were observed among witnessed ventricular fibrillation arrests except that the proportions who were women and arrested before arrival remained relatively stable over time (data not shown). Several patient, event, and EMS characteristics were associated with survival to hospital discharge in all rhythm and witnessed ventricular fibrillation arrests (Table 2).

EMS-treated cardiac arrest with an abridged Utstein template (1977 to 2001).

TABLE 1. Demographic and EMS Characteristics for All Cardiac Arrests According to Time Period*

CharacteristicTime Period
1977–1981 (n=1982)1982–1985 (n=2100)1986–1989 (n=2293)1990–1993 (n=2119)1994–1997 (n=2081)1998–2001 (n=2016)
*Continuous variables are presented as means and standard deviations. Location of arrest was not routinely recorded during the first two time periods. VF indicates ventricular fibrillation; NH, nursing home. For arrests after EMS arrival, BLS, ALS, and defibrillation response intervals are recorded as 0 minutes.
†Test for trend, P<0.05.
Age, y64.5 (12.6)66.6 (12.6)66.8 (13.5)67.0 (13.7)68.3 (13.8)68.3 (13.8)
Female, % (n)23.2 (459)29.1 (612)33.2 (762)31.4 (666)31.1 (647)33.6 (678)
Witnessed, % (n)70.3 (1393)64.0 (1343)61.1 (1400)61.8 (1310)61.0 (1268)58.4 (1178)
Bystander CPR, % (n)
    Bystander CPR without dispatcher assistance27.3 (542)31.9 (670)34.2 (784)31.6 (670)28.9 (601)29.6 (596)
    Dispatcher-assisted CPR0 (0)14.3 (300)16.0 (366)18.6 (395)21.0 (436)20.2 (407)
    No CPR before EMS72.7 (1440)53.8 (1130)49.8 (1143)49.7 (1054)50.2 (1044)50.2 (1013)
Location, % (n)
    Private residence72.1 (1654)72.4 (1534)72.2 (1503)65.2 (1316)
    Public17.0 (390)20.2 (428)18.9 (393)18.6 (374)
    NH/Medical10.9 (249)7.4 (157)8.9 (185)16.2 (326)
Arrest before arrival, % (n)86.7 (1719)88.9 (1866)90.5 (2075)90.8 (1923)90.0 (1873)90.0 (1815)
Presenting rhythm VF, % (n)58.8 (1166)49.8 (1045)44.1 (1011)46.7 (990)47.7 (993)43.8 (883)
Defibrillation by BLS unit, % (n)9.3 (108)37.6 (393)73.4 (742)77.1 (763)77.1 (766)75.0 (662)
BLS response interval, min3.8 (2.5)4.2 (2.7)4.6 (2.7)4.7 (2.5)4.8 (2.6)5.1 (2.7)
ALS response interval, min8.4 (4.0)8.6 (5.3)9.6 (5.5)9.9 (5.3)9.5 (5.2)9.0 (5.0)
Defibrillation response interval, min7.7 (3.8)5.8 (4.1)5.2 (3.1)5.1 (3.0)5.1 (2.7)5.3 (2.7)

TABLE 2. Univariate Odds Ratio of Survival Associated With Patient and EMS Characteristics for All Rhythms and Witnessed Ventricular Fibrillation for the Entire Study Period

CharacteristicAll Rhythms (n=12 591)Witnessed Ventricular Fibrillation (n=4546)
Odds Ratio95% CIOdds Ratio95% CI
*Odds ratio for 1-year increase in age.
†For arrests before arrival.
‡Referent location is private residence.
§Odds ratio for 1-minute increase in response interval.
Age*0.9730.969, 0.9770.9830.978, 0.988
Female0.700.63, 0.781.030.89, 1.19
Witnessed6.185.35, 7.14
Bystander CPR1.851.66, 2.051.431.24, 1.64
Location
Private residence1.01.0
Public2.782.46, 3.131.751.50, 2.04
Nursing home/medical office0.540.43, 0.691.020.72, 1.42
Arrest before arrival0.440.39, 0.510.490.41, 0.58
Presenting rhythm VF9.338.17, 10.65
BLS response interval§0.890.86, 0.910.890.86, 0.91
ALS response interval§0.950.94, 0.960.950.93, 0.97
Defibrillation response interval§0.880.86, 0.89

Observed survival did not change over time for all EMS-treated cardiac arrests (Table 3). However, advancing time period was associated with an increased odds ratio of survival when models were adjusted for factors typically beyond EMS control (models 1 and 2: fate factors constant). The results were similar when arrest before arrival status was not included in the model or when EMS response intervals were log-transformed. In contrast, advancing time period was associated with a decreased odds ratio of survival when the model included citizen CPR in addition to time period (model 3: program constant). Results were similar when an initial 4-year rather than 5-year time period was used as the referent group. Results were also comparable when calendar time in years was modeled as a continuous variable. There was a modest temporal increase in the proportion of patients who were admitted to the hospital; however, this increase was countered by a temporal increase in the proportion who died in-hospital (data not shown).

TABLE 3. Odds of Survival According to Time Period for All Rhythms*

Time Period
1977–1981 (n=1982)1982–1985 (n=2100)1986–1989 (n=2293)1990–1993 (n=2119)1994–1997 (n=2081)1998–2001 (n=2016)Trend (n=12 591)
*The 1977 to 1981 period serves as referent time period for each model.
†Model 1: Fate factors constant adjusts for age, gender, witness status, presenting rhythm, and arrest before arrival.
‡Model 2: Fate factors constant adjusts for the covariates of model 1 plus BLS and ALS response intervals.
§Model 3: Program constant adjusts for bystander CPR.
Survival, % (n)17.5 (346)17.3 (363)17.1 (393)16.2 (343)16.9 (352)15.7 (317)
Observed (unadjusted)1.00.99 (0.84, 1.16)0.98 (0.83, 1.15)0.91 (0.78, 1.08)0.96 (0.82, 1.13)0.88 (0.75, 1.04)0.98 (0.95, 1.01)
Model 1: Fate factors constant1.01.29 (1.08, 1.53)1.49 (1.25, 1.77)1.30 (1.08, 1.55)1.34 (1.12, 1.60)1.33 (1.11, 1.60)1.04 (1.01, 1.07)
Model 2: Fate factors constant1.01.36 (1.14, 1.62)1.72 (1.45, 2.06)1.51 (1.26, 1.81)1.58 (1.32, 1.91)1.59 (1.32, 1.92)1.08 (1.05, 1.11)
Model 3: Program constant§1.00.88 (0.74, 1.03)0.85 (0.73, 1.00)0.78 (0.66, 0.93)0.82 (0.69, 0.96)0.75 (0.63, 0.89)0.95 (0.93, 0.98)

Among witnessed ventricular fibrillation arrests, a temporal trend of improved survival was observed (Table 4). This pattern was accentuated when adjusted for fate factors (models 1 and 2: fate factors constant). However, when the model was adjusted for citizen CPR and defibrillation response interval (model 3: program constant), the odds ratios of survival were <1.0 for all subsequent time periods relative to the initial period, though no trend was apparent. Results were similar when all ventricular fibrillation arrests (unwitnessed and witnessed) were included in the analyses, when an initial 4-year rather than 5-year time period was used as the referent group, and when calendar time in years was modeled as a continuous variable.

TABLE 4. Odds of Survival According to Time Period for Witnessed Ventricular Fibrillation*

Time Period
1977–1981 (n=962)1982–1985 (n=811)1986–1989 (n=780)1990–1993 (n=754)1994–1997 (n=787)1998–2001 (n=681)Trend (n=4775)
*The 1977 to 1981 period serves as referent time period for each model.
†Model 1: Fate factors constant adjusts for age, gender, and arrest before arrival.
‡Model 2: Fate factors constant adjusts for the covariates of model 1 plus BLS and ALS response intervals.
§Model 3: Program constant model adjusts for bystander CPR and defibrillation response interval.
Survival, % (n)31.4 (302)32.7 (265)34.4 (268)35.3 (266)36.5 (287)36.1 (246)
Observed (unadjusted)1.01.06 (0.87, 1.30)1.14 (0.94, 1.40)1.19 (0.97, 1.46)1.25 (1.03, 1.53)1.24 (1.00, 1.52)1.05 (1.01, 1.09)
Model 1: Fate factors constant1.01.10 (0.90, 1.35)1.20 (0.98, 1.47)1.26 (1.03, 1.55)1.34 (1.09, 1.64)1.31 (1.06, 1.62)1.06 (1.02, 1.10)
Model 2: Fate factors constant1.01.18 (0.96, 1.44)1.44 (1.17, 1.78)1.52 (1.23, 1.88)1.67 (1.35, 2.06)1.62 (1.30, 2.02)1.11 (1.07, 1.15)
Model 3: Program constant§1.00.73 (0.59, 0.90)0.72 (0.58, 0.90)0.74 (0.60, 0.93)0.79 (0.64, 0.98)0.78 (0.63, 0.99)0.98 (0.94, 1.02)

In analyses in which time was modeled as a grouped variable or as a continuous variable, the addition of covariates to the models investigating fate and program factors all significantly improved the fit of the model (as determined by the likelihood ratio test comparing nested models) with the exception of gender. No term was dropped due to collinearity.

Discussion

Observed survival to hospital discharge for all EMS-treated patients with cardiac arrest remained relatively unchanged over time in this community for the 25-year span from 1977 to 2001. However, the distribution of several characteristics associated with survival changed over time. Indeed, the static temporal pattern in survival appeared to result from a dynamic and evolving balance between beneficial and adverse prognostic characteristics. Specifically, the results suggest that the implementation of EMS-directed programs of dispatcher-assisted telephone CPR that increased the proportion of patients with cardiac arrest receiving citizen CPR and BLS defibrillation that decreased the defibrillation response interval may have in part balanced adverse temporal trends in several factors generally beyond EMS control.

Importantly, expected relationships for patient, circumstance, and EMS characteristics known to be associated with survival in cardiac arrest were evident in this study population. Although little is known about temporal patterns of demographic or circumstance characteristics (fate factors) or the causes of these temporal patterns, limited evidence from other communities suggests similar temporal trends of increasing age, greater proportion who were women, and the decreasing proportion with ventricular fibrillation that were observed in this study.11–13 In addition, several EMS factors demonstrated temporal changes. Basic and advanced life support response intervals increased over time despite an increase in the density (number of units per square mile) of EMS units and maintenance of standard protocols of EMS response for cardiac arrest victims. Although debatable, the increase may be due to population growth and urbanization, a factor that would be beyond EMS control. In contrast to these potentially adverse temporal trends, the proportion of patients who received citizen CPR before EMS arrival increased over time in large part as a consequence of dispatcher-assisted telephone CPR, an intervention that has been associated with improved survival.14,15 Likewise, despite temporal increases in BLS and to a lesser extent ALS response intervals, the defibrillation response interval declined over time as a result of training and equipping first-tier BLS personnel with defibrillators.

We explored the potential influence of temporal changes in prognostic factors on the temporal patterns of survival. In analyses of all rhythms, the results suggest that had fate factors (age, witness status, and so forth) remained stable over time and the EMS programs of dispatcher-assisted CPR and BLS defibrillation been implemented (as they were), a temporal trend of improving survival may have occurred. Conversely, the results suggest that given the observed adverse temporal patterns of fate factors, a temporal trend of decreasing survival may have occurred if the program of dispatcher CPR had not been implemented. Moreover, this latter model probably underestimates the potential decrease in survival had EMS not implemented the aforementioned programs because it cannot account for the temporal decrease in defibrillation response interval resulting from the BLS defibrillation program that would be pertinent only to patients with ventricular fibrillation.

Thus, analyses were also restricted to patients with witnessed ventricular fibrillation, a subset for whom both early defibrillation along with prompt CPR might be expected to affect survival. Indeed, a temporal trend of increasing survival was observed in this group of patients. Controlling for fate factors strengthened the temporal trend, whereas controlling for citizen CPR and defibrillation response interval nullified the observed temporal trend of increasing survival, suggesting that dispatcher-assisted CPR and BLS defibrillation were in part the means through which survival in ventricular fibrillation improved over time. Of note, although an overall temporal trend was observed, the trend of increasing survival in ventricular fibrillation was most apparent during the first four time periods and appears to subsequently plateau, a pattern that corresponds to the period of implementation of the EMS programs.

This study has limitations. The study investigated temporal trends in survival of out-of-hospital cardiac arrest and potential explanations for the observed trends. The observational cardiac arrest surveillance system of King County used in this investigation does not allow for definitive causal inference. Possible explanations outside the confluence of fate and program factors assessed in these analyses include hospital care, unmeasured patient characteristics, or other confounders. Information about hospital treatments as well as patients’ chronic clinical conditions was not routinely available. It is likely that in-hospital treatment did change over time.16 However, given the temporal patterns of hospital admission and hospital mortality rates (a temporal increase in the proportion admitted was countered by an increase in the proportion who died in-hospital), a clear trend for improved outcome as the result of in-hospital care was not obvious though still possible. Patients’ clinical conditions have been independently associated with survival in certain cardiac arrest settings4; the increase in average age over time would suggest that clinical morbidity would presumably have increased rather than decreased over time. Information was not available about functional status or quality of life among persons who survived to hospital discharge. These are important considerations and may have changed over time. Finally, the findings observed in this community may not be generalizable to other communities, though we believe the dynamic interplay of prognostic factors over time may be pertinent to other settings. These limitations should be balanced against the strengths of the study; the investigation reports the experience of a mature EMS system and used a standardized system of data collection incorporating information from EMS, death, and hospital records.

The static temporal pattern of survival in EMS-treated out-of-hospital cardiac arrest in this community over the past 25 years was not the result of a stable temporal pattern of patient, circumstance, and EMS characteristics but rather appeared to represent an evolving balance of prognostic factors. Programs of dispatcher-assisted telephone CPR and BLS defibrillation implemented by EMS appeared to counter adverse temporal trends in prognostic factors generally beyond EMS control and contribute to the temporal trend of improved survival among patients with ventricular fibrillation arrest. The results of the study suggest that ongoing innovations in resuscitation of patients with cardiac arrest that strengthen the chain of survival will be important if survival from out-of-hospital cardiac arrest is to be improved. Whether the dissemination of defibrillators, developments in defibrillation technology, variation in CPR, treatment with pharmaceutical agents, or other therapies will be important in the public health challenge of improving survival from cardiac arrest are unsettled but important questions that deserve more investigation.

The authors wish to acknowledge and thank the emergency medical technician fire fighters, paramedics, and emergency dispatchers of King County for their ongoing commitment to excellent patient care and the improvement of public health.

Footnotes

Correspondence to Thomas Rea, Public Health Seattle-King County, Emergency Medical Services Division, 999 Third Ave, Suite 700, Seattle, WA 98104-4039. E-mail

References

  • 1 Guidelines for cardiopulmonary resuscitation and emergency cardiovascular care: an international consensus of science. Circulation. 2000; 102 (suppl 8): 2000.Google Scholar
  • 2 Nichol G, Stiell IG, Laupacis A, et al. A cumulative meta-analysis of the effectiveness of defibrillator-capable emergency medical services for victims of out-of-hospital cardiac arrest. Ann Emerg Med. 1999; 34: 517–552.CrossrefGoogle Scholar
  • 3 Cummins RO, Chamberlain DA, Abramson NS, et al. Recommended guidelines for uniform reporting of data from out-of-hospital cardiac arrest: the Utstein Style. Ann Emerg Med. 1991; 20: 861–874.CrossrefMedlineGoogle Scholar
  • 4 Hallstrom AP, Cobb LA, Yu BH. Influence of comorbidity on the outcome of patients treated for out-of-hospital ventricular fibrillation. Circulation. 1996; 93: 2019–2022.CrossrefMedlineGoogle Scholar
  • 5 Kim C, Fahrenbruch CE, Cobb LA, et al. Out-of-hospital cardiac arrest in men and women. Circulation. 2001; 104: 2699–2703.CrossrefMedlineGoogle Scholar
  • 6 Swor RA, Jackson RE, Tintinalli JE, et al. Does advanced age matter in outcomes after out-of-hospital cardiac arrest in community-dwelling adults? Acad Emerg Med. 2000;762–768:.MedlineGoogle Scholar
  • 7 Holmberg M, Holmberg S, Herlitz J. Incidence, duration and survival of ventricular fibrillation in out-of-hospital cardiac arrest patients in Sweden. Resuscitation. 2000; 44: 7–17.CrossrefMedlineGoogle Scholar
  • 8 Valenzuela TD, Roe DJ, Cretin S, et al. Estimating effectiveness of cardiac arrest interventions: a logistic regression survival model. Circulation. 1997; 96: 3308–3313.CrossrefMedlineGoogle Scholar
  • 9 US Census Bureau. Available at http://www.census.gov. Accessed January 15, 2003.Google Scholar
  • 10 Eisenberg MS, Bergner L, Hallstrom AP. In Sudden Cardiac Death in the Community. New York, NY: Praeger Publishers; 1984:47–58.Google Scholar
  • 11 Kuisma M, Repo J, Alaspaa A. The incidence of out-of-hospital ventricular fibrillation in Helsinki, Finland, from 1994 to 1999. Lancet. 2001; 358: 473–474.CrossrefMedlineGoogle Scholar
  • 12 Herlitz J, Andersson E, Bang A, et al. Experiences from treatment of out-of-hospital cardiac arrest during 17 years in Goteborg. Eur Heart J. 2000; 21: 1251–1258.CrossrefMedlineGoogle Scholar
  • 13 Cobb LA, Fahrenbruch CE, Olsufka M, et al. Changing incidence of out-of-hospital ventricular fibrillation, 1980–2000. JAMA. 2002; 288: 3008–3013.CrossrefMedlineGoogle Scholar
  • 14 Carter WB, Eisenberg M, Hallstrom A, et al. Development and implementation of emergency CPR instruction via telephone. Ann Emerg Med. 1984; 13: 695–700.CrossrefMedlineGoogle Scholar
  • 15 Rea TD, Eisenberg MS, Culley LL, et al. Dispatcher-assisted CPR and survival in cardiac arrest. Circulation. 2001; 104: 2513–2516.CrossrefMedlineGoogle Scholar
  • 16 Spaulding CM, Joly LM, Rosenberg A, et al. Immediate coronary angiography in survivors of out-of-hospital cardiac arrest. N Engl J Med. 1997; 336: 1629–1633.CrossrefMedlineGoogle Scholar

eLetters(0)

eLetters should relate to an article recently published in the journal and are not a forum for providing unpublished data. Comments are reviewed for appropriate use of tone and language. Comments are not peer-reviewed. Acceptable comments are posted to the journal website only. Comments are not published in an issue and are not indexed in PubMed. Comments should be no longer than 500 words and will only be posted online. References are limited to 10. Authors of the article cited in the comment will be invited to reply, as appropriate.

Comments and feedback on AHA/ASA Scientific Statements and Guidelines should be directed to the AHA/ASA Manuscript Oversight Committee via its Correspondence page.