- Email: [email protected]
Paediatric out-of-hospital cardiac arrests — epidemiology and outcome
RESUSCITATION
ELSEVIER
Resuscitation 30 (1995) 141-150
Paediatric out-of-hospital cardiac arrests outcome
epidemiology and
Markku Kuisma* a, Pertti Suominen b, Reijo KorpelaC aLuakariyksikko. Helsinki RescueDepartment, Department of Health, Agricolankatu 15, SF-00530Helsinki, Finland bDepartment of Anaesthesia,Jorvi Hospital, Turuntie 150. SF-02740Espoo. Finland ‘Departmeni of Anaesthesiaof Childrens’ Hospital, Helsinki University Hospital, Stenbiickinkatu 9, 00290 Helsinki, Finland Received 11 June 1995;revision received 7 August 1995;accepted 8 August 1995
Abstract Objective: To determine the epidemiology and aetiology of out-of-hospital paediatric cardiac arrest and the outcome of resuscitation and to apply the Utstein template for the paediatric cardiac arrest population. Design: Retrospective cohort study. Setting: A middle-sized urban city (population 516 000) served by a single emergencymedical services (EMS) system. Patients: 79 consecutive paediatric (age under 16 years) prehospital cardiac arrest patients between January 1, 1985and December 31, 1994.No patient was excluded. Intervention: Advanced paediatric life support according to the recommendations of American Heart Association. Main outcome measures: Survival from cardiac arrest to discharge and factors associatedwith favourable outcome defined as alive 1 year after discharge with Bloom category I or II. Results: 79 patients had cardiac arrest. The incidence of paediatric out-of-hospital cardiac arrest and sudden unexpected out-of-hospital death was 9.8 and 8.9/100OOOlinhabitantsaged under 16, respectively. The mean age was 2.9 years, 72.2%were under 18 months. SIDS was the leading causeof cardiac arrest followed by trauma, airway related cardiac arrest and (near)drowning. Fifty-two patients were considered for resuscitation in whom asystole was the most common initial rhythm (78.9%) followed by pulselesselectrical activity (13.5%) and ventricular fibrillation (3.8%). Resuscitation was attempted in 34 patients. The overall survival rate was 9.6%, for attempted resuscitation 14.7%,for attempted resuscitation when cardiac arrest was witnessed 25.0% and for attempted resuscitation with witnessedarrest of cardiac origin 0%. Favourable outcome was registered in four of five survivors. Factors associated with favourable outcome were collapse in a public place, the near-drowning aetiology of arrest, bystander initiated CPR and short duration of resuscitation. Multivariate regressionanalysis showed no factor related to favourable outcome, but MICU time interval < 10 min was related with.survival. Due to the retrospective nature of this study all core times could not be obtained. In spite of this, the Utstein template was applicable also in our paediatric cardiac arrest population. Conclusions: Survival from paediatric cardiac arrest has remained low. The overall survival rate was 9.6%, survival after attempted resuscitation 14.7%and 0% when resuscitation was attempted in witnessedarrest of cardiac origin. Asystole was the most common initial rhythm and the four leading causesfor cardiac arrest were SIDS, trauma, airway related arrest and (near)drowning. The Utstein template adopted for adult out-of-hospital cardiac arrests was was found applicable also in paediatric cardiac arrests. Abbreviations: AED, automated electrical defibrillator; ALS, advanced life support; EMS, emergencymedical system; EMT, emergency medical technicians; MICU, mobile intensive care unit; PEA, pulselesselectrical activity. * Corresponding author, Fax: 358 0 3936242. 0300-9572/95/X19.05 0 1995Elsevier Science Ireland Ltd. All rights reserved SSDI 0300-9572(95)00888-Z
hf. Kuisma et al. /Resuscitarion 30 (1995) 141-150
142
Keyworh:
Cardiacarrest;Resuscitation;Out-of-hospital;Paediatric;Epidemiology;Outcome;Utstein
1. Introduction
The number of paediatric cardiac arrest and resuscitation studies has remained low compared to those in adult populations [l-6]. Major reasons for the scarcenessof reports have probably been the rarity of paediatric cardiac arrest and the poor prognosis associated with resuscitation. Some of the previous studies have had inadequacies in case or survival definition - whether collapse was witnessed or not, by whom collapse was witnessed (bystander or EMS personnel), and exclusion of subgroups of trauma cases or those dead on arrival. Utstein recommendations were introduced in 1990 in order to standardize data reporting from out-of-hospital cardiac arrest [7]. The problem of nomenclature had been widespread and uniformly recognized. The survival rates had been shown to vary widely depending on the caseand survival definition selected [8, 91. However, no paediatric study according to the Utstein style has been published yet. The aim of this retrospective study was to report the incidence and aetiology of paediatric out-of-hospital cardiac arrests as well as outcome of resuscitation, without excluding any patient group.The Utstein style was applied whenever retrospectively possible. 2. Materials and methods
Helsinki EMS system Helsinki is the capital of Finland with a population of 516 000 inhabitants and a geographic area of 590 km*. In daytime during working days the number of people is approximately 10% higher than the number of inhabitants. The number of people under 16 years of age was 76 450 in 1985 and 85 400 in 1994 (15.7 and 16.6% of the total population, respectively). The number of children under 12 months was 7200 in 1994 (1.4% of the total population and 8.4% of children under 16 years).The number of deaths from all causesin the community was 982 per 100 000 inhabitants in
1994.During the study period the total number of paediatric deaths in Helsinki was 474; 58.6/100000 aged under 16 years/year. The Helsinki 112Dispatching Center dispatches annually 34 100 urgent medical missions in Helsinki. It is a combined center for medical, tire and rescueemergencies.In addition to Helsinki, it serves the province of Uusimaa which increases the population of the dispatching area up to 1 000 000. The dispatch is criteria based,computer aided, and the dispatchers are full-time employees who have passseda medical dispatching course. Most of the dispatchers give telephone assisted CPR instructions. Dispatching during event interrogation is routinely used. The first unit should be dispatched within one min and the unit should be mobile within two min from the receipt of the call. The control board shows the location of units with the accuracy of a city region. The EMS systemis three tiered and responsible for urgents calls. The non-urgent calls are taken care of by two private enterprises. The city is divided into eight regions each of which has an own rescuestation. The first tier consists of six ambulancesand eight tire engines(used as first reponding units) manned with emergency medical technicians-defibrillators (EMT-ds) capable of defibrillation, insertion of an intravenous line and adult intubation. The first responding unit turns the automated electrical defibrillator (AED) on when they arrive at the patient’s side in order to register that time. The clocks of AEDs are synchronized with the time in the dispatching center and checked daily. Three advanced life support (ALS) units make the second tier which is capable of admistering intravenous drugs and paediatric intubation. The third tier is the physician staffed mobile intensive care unit (MICU). MICU physicians are also responsible for education, surveillance and development programs in EMS. The calls are classified into four categories; D is nonurgent, C is urgent and the patient has to be reached in 20 min (neither blue lights nor siren), B is urgent with medium or unknown risk (blue lights and siren) and taken care of by basic and/or ALS-
M. Kuisma et al. /Resuscitation
30 (1995) 141-150
143
units and finally A in which the nearest unit and MICU are dispatched simultaneosly for high risk missions. During simultaneous A-calls, one of the ALS-units (medical supervisor) is dispatched and MICU is informed. Major changesin EMS during the study period have been the introduction of AEDs in 1988, the changefrom two to three tiered EMS in 1989(until then MICU was the only ALS provider), full-time dispatchers in 1990, the use of fire engines as first responding units in medical emergencies since 1990 and telephone assisted CPR in 1992. Cardiac arrest was defined as the absence of consciousness, spontaneous ventilation and palbable pulses in major arteries. The resuscitation of children has followed the latest version of guidelines from American Heart Association [lo]. MICU attends every resuscitation event even if an ALS-unit is present. Excluding hypothermic patients, efforts are ceasedin the field when there is no responseto resuscitation attempts. Children are transported to three receiving hospitals, however cardiac arrest patients are centered to the intensive care unit of Childrens’ Hospital which is the tertiary care center for children in Helsinki.
sudden paediatric out-of-hospital deaths, a subgroup b was added into sections two and three in Fig. 1 for patients with irreversible signs of death on arrival. Subgroup b was not considered for resuscitation.
Data collection
Statistical methoa3
Data were retrospectively collected from MICU and EMS files, hospital records and autopsy reports. MICU files include every out-of-hospital cardiac arrest patient in Helsinki. None of the outof-hospital cardiac arrest patients was excluded. The arrival times of a first reponding unit and MICU at the patient’s side and return of spontaneus circulation (ROSC) were obtained. In 1985 1991only the dispatching time and the time when the vehicle stops were registered, and two min was added to estimate the time neededto processthe call before dispatching and to estimate the time neededto arrive at the patient’s side after the vehicle have stopped. Time zero was the time point when the call was first received. Cardiac arrests were placed into the Utstein template [7] (Fig. 1). Originally the Utstein template only applied to cardiac arrests with no irreversible signs of death (rigor, decapitation etc.) on arrival. In order to report the total number of
Statistical analysis which concentrated on linding factors related to survival or favourable outcome was performed by a statistician. Exact chi-square test and Student’s t-test (parametric data) were utilized in univariate analysis and logistic regression model in multivariate analysis.
Outcome tracking
The end points of the study were death or 1 year survival. Data were obtained from hospital records. The quality of secondary survival was measured by Bloom Classification on discharge and at 1 year if the category on discharge was other than I [11,12]: (I) No disability, active life. Children who on general examination are bright and appear intelligent. (II) Mild disability, active life, Ocular paresis, limited intention tremor, mild ataxia etc. (III) Partial disability such as severe ataxia or seriously reduced vision, but all are capable of selfcare. They may show definite impairment of intellect, but are capable of being taught a trade. (IV) Total disability. These cases are incapable of self-care. Favourable outcome was defined as alive at 1 year with Bloom category I or II.
3. Results Of the 79 consecutive study patients 52 were considered for resuscitation and 27 had irreversible signs of death. The total study population (79) was utilized to report incidence, aetiology, sex and age distributions and the scene of cardiac arrest. Initial rhythm, whether collapse was witnessed or not, the performance of bystander CPR, body temperature and cause of do-not-resuscitate (DNR) desicion were analyzed from the study population considered for resuscitation. Survival analysis was made from attempted resuscitations.
144
hf. Kuisma et al. /Resuscitation
I
30 (1995) 141-150
1. Population age under 16 yews served by EMS system N=80900
I
2. Confumcd cardiac wrests considered for wswitation
irreversible
death
+
wltnared
N-O
7. Arrest witnessed (bystanden) N=3 I
10. Initial rhythm
11. Initial rhythm
Fig. 1. Paediatric out-of-hospital cardiac arrests in Helsinki in 1985-1994 placed into the Utstein template. Subgroup b in sections 2 and 3 was added in order to report the total number of paediatric out-of-hospital deaths. Subgroup b was not considered for resuscitation. Further analysis for non-cardiac aetiology is presented in Figs. 2 and 3.
Cardiac arrests placed into an Utstein template are shown in Fig. 1, subgroup analyses are shown in Figs. 2 and 3. The incidence of paediatric out-of-hospital car-
disc arrest was 9.8/100 000 inhabitants aged under 16 years/ year for all causes, and 7.511000001 inhabitants agedunder 16 years for natural causes. The incidence of sudden unexpected out-of-
h4. Kuisma et al. /Resuscitation
145
30 (1995) 141-150
@iziy
I. Am& witnessed (laystandem) N=9
9.Anutrrlbsaed (EMS
penone
N-O
I + 10. Initial rhyam VF N==O
19. Expired in hospital
20. Discharged alive
+
t
11. Initial rhythm VTN=O
+
N=3
22. Alive at one yew N=3
Fig. 2. Subgroup analysis of non-cardiac aetiology from Fig. 1. Further analysis of unwitnessed cardiac arrests is presented in Fig. 3.
hospital death was 8.9/100 000 inhabitants age under 16 years/year forming 15.2% of all paediatric deaths in the community, and the incidenceof sudden out-of-hospital cardiac death was 0.4/100 000 inhabitants aged under 16 years/year. Forty-two patients were male (53.2%). The mean age was 2.9 ( f 4.8 SD) years, 57 (72.2%) of the patients were under 18 months (the age distribution is shown in Table 1). 63 (79.7%) collapsed at home or in another private apartment. The aeti-
ology was cardiac in three cases(3.8%), SIDS and trauma together were responsible for 72.2% of the cardiac arrests and 75.7% of the deaths (Table 2). Twenty-three of the 27 patients with irreversible signsof death had the diagnosis of SIDS. All SIDS caseswere unwitnessed. Seventy-three (98.6%) of those who died were autopsied. Autopsy was not performed in a patient with pulmonary hypertension who died waiting for a heart transplant. Of the 52 patients considered for resuscitation
U. Kuisma et al. /Resuscitation
30 (1995) 141-150
Fig. 3. Subgroup analysis of unwitnessed cardiac arrests from Fig. 2.
Table 1 Age distribution of the 79 cardiac arrests patients (those with irreversible signs of death included). Number of survivors in parenthesis Age (years)
No. patients
< 30 days
12 39 7 5 (1) 6 (2) 4 (1) 6 (1)
18 (34.6%) had a witnessed collapse. None of the collapses was EMS witnessed. Twelve patients (23.1%)received bystander initiated CPR. The initial rhythm was ventricular fibrillation in two (3.8%), asystole in 41 (78.9%), pulseless electrical activity (PEA) in seven (13.5%) cases and not available in two cases (3.8%). Resuscitation was attempted in 34 patients (65.4%). The mean time for the first responding unit and MICU to arrive at the patient’s side and to ROSC were 9.0, 12.4 and 24.1 min. For survivors and non-survivors the first responding unit and MICU mean time intervals were 8.V9.2 and 9.3/12.9 (P = 0.67 and 0.14).
M. K&ma et al. /Resuscitation 30 (1995) 141-150 Table 2 Aetiology of paediatric cardiac arrest (those with irreversible signs of death included) Aetiology
No. patients
No. survivors
SIDS Trauma Airway related infection mechanical obstruction pulmonary aspiration (Near) drowning Cardiac congenital primary arrythmia pulmonary hypertension Neurologic brain anomaly intracranial hemorrhage Intoxication Sepsis Hanging
47 (59.5%) lOa (12.7%) 8 (10.1%) 6 I I 5 (6.3%) 3 (3.8%) I I I 2 (2.5%) I I 2 (2.5%) I (1.3%) I (1.3%)
I -
3 -
I
aTwo of which by violence.
The mean resuscitation time for survivors and non-survivors was 11.4 and 30.6 min (P = 0.002). The causes for DNR decision were unwitnessed collapse (estimated long time-interval between the collapse and the call) in 12 and high energy trauma in six patients. The overall survival rate was 9.6% (5/52), for attempted resuscitation 14.7% (5/34), for attempted resuscitation when cardiac arrest was witnessed
147
25% (3/12) and for attempted resuscitation with witnessed arrest of cardiac origin 0% (O/3).When resuscitation was attempted in unwitnessed arrests the survival rate was 13.6% (3/22). Survival from resuscitation according to the initial rhythm was 18.5%(5127)in asystole and 0% both in ventricular fibrillation (O/2) and in PEA (O/5). The youngest child to survive was 2 years and 10 months old. When bystander CPR had been initiated and resuscitation continued by EMS personel, four of nine patients (44.4%) survived. None of the children who collapsed at home survived. The five children who survived to discharge, were alive also at 1 year and four of them had a favourable outcome (Table 3). The Bloom category on discharge was I in two, II in one and III in two patients, and at 1 year (for those three who were discharged with Bloom category other than 1) II in two patients and not available in one patient (no. 1). Patient 1 was discharged after 11 months of hospitalization at the age of 16 and her disability was estimated to be permanent defined as Bloom category III. In univariate analysis the scene of collapse in public place (P = O&l), the aetiology of arrest near drowning (P = 0.006) and bystander initiated CPR (P = 0.01) were related to favourable outcome, MICU time interval < 10 min was related to survival. The duration of resuscitation was significantly shorter in survivors with favourable outcome (P = 0.015). Multivariate analysis (logistic regression model) showed no factor related to favourable outcome, but MICU time interval < 10 min was related to survival.
Table 3 Characteristics of the live survivors. All were alive after I year of discharge. Time to ROSC was calculated from the time when the call was first received. Sex
Age
Aetiology
Bystander Witnessed Initial arrest rhythm CPR
Trauma Near drowning Near drowning Near drowning Hanging
No Yes Yes Yes Yes
years
I 2 3 4 5
F M M M F
15.7 5.0 2.9 6.9 14.6
‘See Material and methods section. n.a., not available.
Yes Yes No Yes No
Asystole Asystole Asystole n.a. n.a.
BAY temperature
Time to ROSC
Bloom categorya at discharge/ after I year
Normothermia Hypothermia Nonnothermia Normothermia Normothermia
I7 40 26 8 6.5
3lna 2l2 312 I/na I lna
148
M. Kuisma et al. /Resuscitation
4. Discussion
The etiologies of cardiac arrest and sudden unexpected death in children differ considerably from those in adults, cardiac origin being the minority [ 131.The aetiology seenin this study corresponds to that of previous out-of-hospital studies [1,6]. The two main categories were SIDS and trauma, which were responsible for more than two thirds of cardiac arrests. When airway related cardiac arrests and near-drowning patients are included these four major categories were responsible for 9 out of 10 cardiac arrests. The even higher proportion of SIDS (59.5%) in this study is explained by inclusion of patients with irreversible signs of death to aetiologic evaluations. The high autopsy frequency (98.6O/aof those who died) guaranteed that the aetiology could be verified in every arrest.The percentageof unknown etiologies has in previous cardiac arrest studies varied between 5 and 24 [1,3,6]. However, high autopsy frequency in paediatric sudden death population has been reported also by others [14]. The incidence of non-traumatic out-of-hospital cardiac arrest among individuals less than 18 has been reported to be 12.7/100OOO/year[l], and in the present study 8.51100OOO/year in children aged under 16.Neuspiel et al. [l4] found an annual incidence of 4.6/100 000 population for sudden natural death in persons aged l-2 1 years. The material also included patients developing cardiac arrest in hospital. The incidences are not directly comparable due to the different inclusion criteria for age and aetiology, but the incidence of out-of-hospital cardiac arrest in our study can be judged to be lower than previously reported. Marked variation in incidence rates has been shown to occur between different communities, and incidence rates have been found to be negatively related to survival rates [15]. Two thirds of paediatric cardiac arrests occur during the first 18 months of life [6]. In this study the proportion was 72.2%, SIDS being the most common aetiology. The incidence of cardiac arrest in children aged 2- 14 years was low but it increased again during the 15th year. At this age, the growing child begins to resembleyoung adults also with respect to cardiac arrest aetiology [16]. Compared to adult population, witnessed ar-
30 (1995) 141-150
rests in children are less common. Both Losek et al. [6] and Eisenberg et al. [l] reported a 29% frequency of witnessed arrests which is supported by the present study. Although survival from witnessed arrests was twice as likely as from unwitnessed arrests, almost half of the survivors came from the unwitnessed group. None of the study patients had an EMS witnessed arrest. Warning signs remain unrecognized or children may not have prodromal symptoms serious enough to activate EMS system before cardiac arrest. EMS witnessed arrests have not been specifically detined in previous paediatric studies. Every tenth ‘adult’ arrest, on the other hand, is EMS witnessed u71. Asystole, as the most common, and pulseless electrical activity are more common initial rhythms as ventricular fibrillation in paediatric cardiac arrests [4,6,18]. Asystole and PEA are typical initial rhythms in arrests of non-cardiac origin and unwitnessed arrest, and thus assumingly, of paediatric cardiac arrest. The development of VF has been assumedto be related to cardiac mass and the developing autonomic nervous system and therefore VF is not likely to occur in children with a small heart [ 181.The rhythm distribution in this study confirms previous results, although the proportion of ventricular fibrillation was even lower (3.8%). In adult populations, over 75% of the cardiac arrests has been found to occur in private residences [19]. 79.7%of the study patients collapsed at home or in a private apartment and none survived. On the other hand four of the five survivors in this study had a causeof cardiac arrest that is normally more likely to occur in public place (neardrowning, traffic accident). Cardiac arrests outside home are more often witnessed,more likely to have bystander CPR and CPR is initiated earlier which contribute to the more favourable outcome
PO1.
Reports of bystander initiated CPR are scarce. Losek et al. [6] reported a 48% frequency in witnessedpaediatric arrests and Lombardi et al. [21] 32% overall frequency in adult population. The overall bystander CPR frequency and the frequency in witnessed arrests were 23.1 and 22.2%, respectively, in this study. Survival to discharge in previous studies has
hf. Kuisma et al. /Resuscitation
been 6.7, 5.2, 0 and 7.9% [1,3,4,6]. Also in inhospital arrests the survival rate has been reported to be relatively low [2]. In contrast, in adult population survival rates up to 34% has been reported [8] when cardiac arrest is bystander witnessed, of cardiac origin and with the initial rhythm of ventricular fibrillation. In this study survival from attempted resuscitation was markedly higher (14.7%) than in previous studies. However, in 18 cases a DNR desicion was made by the MICU physician due to a hopeless situation. In other EMS systemsthese patients would probably have undergone unsuccessful resuscitation attempts. One-year survival incidence has not been reported previously in a paediatric resuscitation population. In adult populations those who survive to be discharged are likely to be alive at one year [22]. The same trend was also registered in this study - every survivor was alive after 1 year of discharge. However, neurologic sequelae have been reported to be frequent in survivors [3,6]. Especially when the arrest is airway related the total anoxia time, including the pre-cardiac arrest period, may be too long to allow total recovery [23]. In neurologically intact survivors resuscitation has been initiated and ROSC achieved promptly (31. In our study only two of the survivors were neurologically intact at discharge, but outcome was favourable in three patients, and after 1 year in four patients of five. It is concluded that most of the survivors are likely to lead a life of normal or near-normal quality. The Utstein style was found to be applicable to report retrospective data from paediatric out-ofhospital cardiac arrest. Owing to the aetiology and the rarity of witnessed arrest in paediatric population subgroup analyses of non-cardiac origin and unwitnessed arrests are essential in template presentation. Not all recommended core times could be obtained due to the retrospective nature of the study. There are 16 core and supplemental times in the Utstein recommendations [7]. Even in diligent record keeping it may be difficult to obtain all times reliably, and the number of recommended times to record should probably be reduced in order to allow uniform data reporting. The development of EMS systems and CPR have markedly increased the quantity and quality of survival from out-of-hospital cardiac arrest in
30 (1995) 141-150
149
adult population. The sametrend cannot be clearly seenin paediatric patients major reasons probably being long delay from the collapse to the initiation of CPR due to the high proportion of unwitnessed arrests and the different aetiology of cardiac arrest. Because these underlying prognostic factors cannot be modified no major improvement in survival is likely to occur in developed EMS systemsin the near future. However, based on the fact that injury is the leading cause of paediatric death above the age of 1 year, programmes aimed to prevent childhood injuries and drownings should receive vigorous attention [24,25]. As a medical system resource ALS has been shown to be associatedwith lower paediatric trauma death rates [26]. In addition to trauma prevention programmes, the teaching of bystander initiated CPR and the art of swimming should be further encouraged. conclusions Survival from paediatric cardiac arrest has remained low probably owing to the aetiology and high proportion of unwitnessed arrests. The overall survival rate in the present study was 9.6%, survival after attempted resuscitation 14.7% and 0% when resuscitation was attempted in witnessed arrest of cardiac origin. Asystole was the most common initial rhythm and the four leading causes for cardiac arrest were SIDS, trauma, airway related arrest and (near)drowning. Those five who survived to discharge were also alive at 1 year and four of them had favourable outcome. The Utstein style for recording adult out-of-hospital cardiac arrests was found applicable also in paediatric cases. Acknowledgements
The authors thank Per Rosenberg, MD, PhD, for valuable comments, Anneli OjajCrvi, MSc, for statistical review, Leila Saari, MICU nurse and Ville Voipio, assistant, for help in data collection. References [II
Eisenbexg M, Bergner L, Hallstrom A. Epidemiology of cardiac arrest and resuscitation in children. Ann Emerg Med 1983; 12: 612-674.
150
hf. Ku&ma et al. /Resuscitation
[2] Lewis JK, Minter MG, Eshelman SJ, Witte MK. Outcome of pediatric resuscitation. Ann Emerg Med 1983; 12: 291-299. [3] Torphy DE, Minter MG, Thompson BM. Cardiorespiratory arrest and resuscitation of children. AJDC 1984; 138: 1099-1102. [4] Applebaum D. Advanced prehospital care for pediatric emergencies.Ann Emerg Med 1985; 14: 656-659. [5] Zaritsky A, Nadkami V, Getson P, Kuehl K. CPR in children. Ann Emerg Med 1987; 16: 1107-l 11I. [6] Losek JD, HermesH, Glaeser P, Hendley G, Nelson DB. Prehospital care of the pulseless,nonbreathing pediatric patient. Am J Emerg Med 1987; 5: 370-374. [7] Cummins RO, Chamberlain DA, Abramson NS et al. Recommendedguidelines for uniform reporting of data from out-of-hospital cardiac arrest: the Utstein style. Ann Emerg Med 1991;20: 861-874. [S] Eisenberg MS, Cummins RO, Larsen MP. Numerators, denominators, and survival rates: reporting data from out-of-hospital cardiac arrest. Am J Emerg Med 1991;9: 544446. [9] Valenzuela TD, Spaite DW, Meislin HV, Clark LL, Ewy GA. Case and survival definitions in out-of-hospital cardiac arrest: effect on survival rate calculation. J Am Med Assoc 1992;261: 272-274. [IO] American Heart Association. Guidelines for cardiopulmonary resuscitation and emergency cardiac care. Recommendationsof the 1992national conference J Am Med Assoc 1992;268: 2251-2275. [ll] Bloom HJG, Wallace ENK, Henk JM. The treatment and prognosis of medulloblastoma in children: a study of 82 verified cases.Am J Roentgen01 1969; 105: 43-62. [12] Glauser TA, Packer RJ. Cognitive deficits in long-term survivors of childhood brain tumors. Child’s Nerv Syst 1991;7: 2-12. [13] Myerburg RJ. Sudden cardiac death: epidemiology, causesand mechanisms.Cardiology 1987;74: suppl. 2-9. [I41 Neuspiel DR, Kuller LH. Sudden and unexpected natural death in childhood and adolescence.J Am Med Assoc 1985;254: 1321-1325.
30 (1995) 141-150
I151 Becker LB, Smith DV, Rhodes KV. Incidence of cardiac arrest: a neglected factor in evaluating survival rates. Ann Emerg Med 1993;22: 86-91. [I61 Safranek DJ, Eisenberg MS, Larsen MP. The epidemiology of cardiac arrest in young adults. Ann Emerg Med 1992;21; 1102-l 106. [171 Eisenberg MS, Cummins RO, Litwin PE, Hallstrom AP. Out-of-hospital cardiac arrest: Significance of symptoms in patients collapsing before and after arrival of paramedics. Am J Emerg Med 1986;4: 116-120. Walsh CK, Krongrad E. Terminal cardiac eletrical activMl ity in pediatric patients. Am J Cardiol 1983;5I: 557-561. I191 Becker LB, Gstrander MP, Barrett J, Gondos GT. Outcome of CPR in a large metropolitan area - where are the survivors. Ann Emerg Med 1991;20: 355-361. PO1 Litwin PE, Eisenberg MS, Hallstrom AP, Cummins RO. The location of collapse and its effect on survival from cardiac arrest. Ann Emerg Med 1987; 16: 787-791. Lombardi G, Gallagher J, Gennis P. Outcome of out-ofWI hospital cardiac arrest in New York City. The prehospital arrest evaluation (PHASE) study. J Am Med Assoc 1994;271: 678-683. Kimman GP, Ivens EM, Hartman JA, Hart HN, SiWI moons ML. Long-term survival after successfulout-ofhospital resuscitation. Resuscitation 1994,28: 227-232. 1231 Orlowski JP. Effectivenessof pediatric cardiopulmonary resuscitation (editorial). AJDC 1984; 138: 1097. 1241 Hoekelman RA, Pless IB. Decline in mortality among young Americans during the 20th century: Prospects for reaching national mortality reduction goals for 1990. Pediatrics 1988;82: 582-595. 1251 Chameides L. CPR challenges in pediatrics. Ann Emerg Med 1993;22 (pt2): 388-392. WI Rutledge R, Smith CY, Axizkhank RG. A populationbased multivariate analysis of the association of country demographic and medical system factors with per capita pediatric trauma death rates in North Carolina. Ann Surg 1994,219: 205-210.
ELSEVIER
Resuscitation 30 (1995) 141-150
Paediatric out-of-hospital cardiac arrests outcome
epidemiology and
Markku Kuisma* a, Pertti Suominen b, Reijo KorpelaC aLuakariyksikko. Helsinki RescueDepartment, Department of Health, Agricolankatu 15, SF-00530Helsinki, Finland bDepartment of Anaesthesia,Jorvi Hospital, Turuntie 150. SF-02740Espoo. Finland ‘Departmeni of Anaesthesiaof Childrens’ Hospital, Helsinki University Hospital, Stenbiickinkatu 9, 00290 Helsinki, Finland Received 11 June 1995;revision received 7 August 1995;accepted 8 August 1995
Abstract Objective: To determine the epidemiology and aetiology of out-of-hospital paediatric cardiac arrest and the outcome of resuscitation and to apply the Utstein template for the paediatric cardiac arrest population. Design: Retrospective cohort study. Setting: A middle-sized urban city (population 516 000) served by a single emergencymedical services (EMS) system. Patients: 79 consecutive paediatric (age under 16 years) prehospital cardiac arrest patients between January 1, 1985and December 31, 1994.No patient was excluded. Intervention: Advanced paediatric life support according to the recommendations of American Heart Association. Main outcome measures: Survival from cardiac arrest to discharge and factors associatedwith favourable outcome defined as alive 1 year after discharge with Bloom category I or II. Results: 79 patients had cardiac arrest. The incidence of paediatric out-of-hospital cardiac arrest and sudden unexpected out-of-hospital death was 9.8 and 8.9/100OOOlinhabitantsaged under 16, respectively. The mean age was 2.9 years, 72.2%were under 18 months. SIDS was the leading causeof cardiac arrest followed by trauma, airway related cardiac arrest and (near)drowning. Fifty-two patients were considered for resuscitation in whom asystole was the most common initial rhythm (78.9%) followed by pulselesselectrical activity (13.5%) and ventricular fibrillation (3.8%). Resuscitation was attempted in 34 patients. The overall survival rate was 9.6%, for attempted resuscitation 14.7%,for attempted resuscitation when cardiac arrest was witnessed 25.0% and for attempted resuscitation with witnessedarrest of cardiac origin 0%. Favourable outcome was registered in four of five survivors. Factors associated with favourable outcome were collapse in a public place, the near-drowning aetiology of arrest, bystander initiated CPR and short duration of resuscitation. Multivariate regressionanalysis showed no factor related to favourable outcome, but MICU time interval < 10 min was related with.survival. Due to the retrospective nature of this study all core times could not be obtained. In spite of this, the Utstein template was applicable also in our paediatric cardiac arrest population. Conclusions: Survival from paediatric cardiac arrest has remained low. The overall survival rate was 9.6%, survival after attempted resuscitation 14.7%and 0% when resuscitation was attempted in witnessedarrest of cardiac origin. Asystole was the most common initial rhythm and the four leading causesfor cardiac arrest were SIDS, trauma, airway related arrest and (near)drowning. The Utstein template adopted for adult out-of-hospital cardiac arrests was was found applicable also in paediatric cardiac arrests. Abbreviations: AED, automated electrical defibrillator; ALS, advanced life support; EMS, emergencymedical system; EMT, emergency medical technicians; MICU, mobile intensive care unit; PEA, pulselesselectrical activity. * Corresponding author, Fax: 358 0 3936242. 0300-9572/95/X19.05 0 1995Elsevier Science Ireland Ltd. All rights reserved SSDI 0300-9572(95)00888-Z
hf. Kuisma et al. /Resuscitarion 30 (1995) 141-150
142
Keyworh:
Cardiacarrest;Resuscitation;Out-of-hospital;Paediatric;Epidemiology;Outcome;Utstein
1. Introduction
The number of paediatric cardiac arrest and resuscitation studies has remained low compared to those in adult populations [l-6]. Major reasons for the scarcenessof reports have probably been the rarity of paediatric cardiac arrest and the poor prognosis associated with resuscitation. Some of the previous studies have had inadequacies in case or survival definition - whether collapse was witnessed or not, by whom collapse was witnessed (bystander or EMS personnel), and exclusion of subgroups of trauma cases or those dead on arrival. Utstein recommendations were introduced in 1990 in order to standardize data reporting from out-of-hospital cardiac arrest [7]. The problem of nomenclature had been widespread and uniformly recognized. The survival rates had been shown to vary widely depending on the caseand survival definition selected [8, 91. However, no paediatric study according to the Utstein style has been published yet. The aim of this retrospective study was to report the incidence and aetiology of paediatric out-of-hospital cardiac arrests as well as outcome of resuscitation, without excluding any patient group.The Utstein style was applied whenever retrospectively possible. 2. Materials and methods
Helsinki EMS system Helsinki is the capital of Finland with a population of 516 000 inhabitants and a geographic area of 590 km*. In daytime during working days the number of people is approximately 10% higher than the number of inhabitants. The number of people under 16 years of age was 76 450 in 1985 and 85 400 in 1994 (15.7 and 16.6% of the total population, respectively). The number of children under 12 months was 7200 in 1994 (1.4% of the total population and 8.4% of children under 16 years).The number of deaths from all causesin the community was 982 per 100 000 inhabitants in
1994.During the study period the total number of paediatric deaths in Helsinki was 474; 58.6/100000 aged under 16 years/year. The Helsinki 112Dispatching Center dispatches annually 34 100 urgent medical missions in Helsinki. It is a combined center for medical, tire and rescueemergencies.In addition to Helsinki, it serves the province of Uusimaa which increases the population of the dispatching area up to 1 000 000. The dispatch is criteria based,computer aided, and the dispatchers are full-time employees who have passseda medical dispatching course. Most of the dispatchers give telephone assisted CPR instructions. Dispatching during event interrogation is routinely used. The first unit should be dispatched within one min and the unit should be mobile within two min from the receipt of the call. The control board shows the location of units with the accuracy of a city region. The EMS systemis three tiered and responsible for urgents calls. The non-urgent calls are taken care of by two private enterprises. The city is divided into eight regions each of which has an own rescuestation. The first tier consists of six ambulancesand eight tire engines(used as first reponding units) manned with emergency medical technicians-defibrillators (EMT-ds) capable of defibrillation, insertion of an intravenous line and adult intubation. The first responding unit turns the automated electrical defibrillator (AED) on when they arrive at the patient’s side in order to register that time. The clocks of AEDs are synchronized with the time in the dispatching center and checked daily. Three advanced life support (ALS) units make the second tier which is capable of admistering intravenous drugs and paediatric intubation. The third tier is the physician staffed mobile intensive care unit (MICU). MICU physicians are also responsible for education, surveillance and development programs in EMS. The calls are classified into four categories; D is nonurgent, C is urgent and the patient has to be reached in 20 min (neither blue lights nor siren), B is urgent with medium or unknown risk (blue lights and siren) and taken care of by basic and/or ALS-
M. Kuisma et al. /Resuscitation
30 (1995) 141-150
143
units and finally A in which the nearest unit and MICU are dispatched simultaneosly for high risk missions. During simultaneous A-calls, one of the ALS-units (medical supervisor) is dispatched and MICU is informed. Major changesin EMS during the study period have been the introduction of AEDs in 1988, the changefrom two to three tiered EMS in 1989(until then MICU was the only ALS provider), full-time dispatchers in 1990, the use of fire engines as first responding units in medical emergencies since 1990 and telephone assisted CPR in 1992. Cardiac arrest was defined as the absence of consciousness, spontaneous ventilation and palbable pulses in major arteries. The resuscitation of children has followed the latest version of guidelines from American Heart Association [lo]. MICU attends every resuscitation event even if an ALS-unit is present. Excluding hypothermic patients, efforts are ceasedin the field when there is no responseto resuscitation attempts. Children are transported to three receiving hospitals, however cardiac arrest patients are centered to the intensive care unit of Childrens’ Hospital which is the tertiary care center for children in Helsinki.
sudden paediatric out-of-hospital deaths, a subgroup b was added into sections two and three in Fig. 1 for patients with irreversible signs of death on arrival. Subgroup b was not considered for resuscitation.
Data collection
Statistical methoa3
Data were retrospectively collected from MICU and EMS files, hospital records and autopsy reports. MICU files include every out-of-hospital cardiac arrest patient in Helsinki. None of the outof-hospital cardiac arrest patients was excluded. The arrival times of a first reponding unit and MICU at the patient’s side and return of spontaneus circulation (ROSC) were obtained. In 1985 1991only the dispatching time and the time when the vehicle stops were registered, and two min was added to estimate the time neededto processthe call before dispatching and to estimate the time neededto arrive at the patient’s side after the vehicle have stopped. Time zero was the time point when the call was first received. Cardiac arrests were placed into the Utstein template [7] (Fig. 1). Originally the Utstein template only applied to cardiac arrests with no irreversible signs of death (rigor, decapitation etc.) on arrival. In order to report the total number of
Statistical analysis which concentrated on linding factors related to survival or favourable outcome was performed by a statistician. Exact chi-square test and Student’s t-test (parametric data) were utilized in univariate analysis and logistic regression model in multivariate analysis.
Outcome tracking
The end points of the study were death or 1 year survival. Data were obtained from hospital records. The quality of secondary survival was measured by Bloom Classification on discharge and at 1 year if the category on discharge was other than I [11,12]: (I) No disability, active life. Children who on general examination are bright and appear intelligent. (II) Mild disability, active life, Ocular paresis, limited intention tremor, mild ataxia etc. (III) Partial disability such as severe ataxia or seriously reduced vision, but all are capable of selfcare. They may show definite impairment of intellect, but are capable of being taught a trade. (IV) Total disability. These cases are incapable of self-care. Favourable outcome was defined as alive at 1 year with Bloom category I or II.
3. Results Of the 79 consecutive study patients 52 were considered for resuscitation and 27 had irreversible signs of death. The total study population (79) was utilized to report incidence, aetiology, sex and age distributions and the scene of cardiac arrest. Initial rhythm, whether collapse was witnessed or not, the performance of bystander CPR, body temperature and cause of do-not-resuscitate (DNR) desicion were analyzed from the study population considered for resuscitation. Survival analysis was made from attempted resuscitations.
144
hf. Kuisma et al. /Resuscitation
I
30 (1995) 141-150
1. Population age under 16 yews served by EMS system N=80900
I
2. Confumcd cardiac wrests considered for wswitation
irreversible
death
+
wltnared
N-O
7. Arrest witnessed (bystanden) N=3 I
10. Initial rhythm
11. Initial rhythm
Fig. 1. Paediatric out-of-hospital cardiac arrests in Helsinki in 1985-1994 placed into the Utstein template. Subgroup b in sections 2 and 3 was added in order to report the total number of paediatric out-of-hospital deaths. Subgroup b was not considered for resuscitation. Further analysis for non-cardiac aetiology is presented in Figs. 2 and 3.
Cardiac arrests placed into an Utstein template are shown in Fig. 1, subgroup analyses are shown in Figs. 2 and 3. The incidence of paediatric out-of-hospital car-
disc arrest was 9.8/100 000 inhabitants aged under 16 years/ year for all causes, and 7.511000001 inhabitants agedunder 16 years for natural causes. The incidence of sudden unexpected out-of-
h4. Kuisma et al. /Resuscitation
145
30 (1995) 141-150
@iziy
I. Am& witnessed (laystandem) N=9
9.Anutrrlbsaed (EMS
penone
N-O
I + 10. Initial rhyam VF N==O
19. Expired in hospital
20. Discharged alive
+
t
11. Initial rhythm VTN=O
+
N=3
22. Alive at one yew N=3
Fig. 2. Subgroup analysis of non-cardiac aetiology from Fig. 1. Further analysis of unwitnessed cardiac arrests is presented in Fig. 3.
hospital death was 8.9/100 000 inhabitants age under 16 years/year forming 15.2% of all paediatric deaths in the community, and the incidenceof sudden out-of-hospital cardiac death was 0.4/100 000 inhabitants aged under 16 years/year. Forty-two patients were male (53.2%). The mean age was 2.9 ( f 4.8 SD) years, 57 (72.2%) of the patients were under 18 months (the age distribution is shown in Table 1). 63 (79.7%) collapsed at home or in another private apartment. The aeti-
ology was cardiac in three cases(3.8%), SIDS and trauma together were responsible for 72.2% of the cardiac arrests and 75.7% of the deaths (Table 2). Twenty-three of the 27 patients with irreversible signsof death had the diagnosis of SIDS. All SIDS caseswere unwitnessed. Seventy-three (98.6%) of those who died were autopsied. Autopsy was not performed in a patient with pulmonary hypertension who died waiting for a heart transplant. Of the 52 patients considered for resuscitation
U. Kuisma et al. /Resuscitation
30 (1995) 141-150
Fig. 3. Subgroup analysis of unwitnessed cardiac arrests from Fig. 2.
Table 1 Age distribution of the 79 cardiac arrests patients (those with irreversible signs of death included). Number of survivors in parenthesis Age (years)
No. patients
< 30 days
12 39 7 5 (1) 6 (2) 4 (1) 6 (1)
18 (34.6%) had a witnessed collapse. None of the collapses was EMS witnessed. Twelve patients (23.1%)received bystander initiated CPR. The initial rhythm was ventricular fibrillation in two (3.8%), asystole in 41 (78.9%), pulseless electrical activity (PEA) in seven (13.5%) cases and not available in two cases (3.8%). Resuscitation was attempted in 34 patients (65.4%). The mean time for the first responding unit and MICU to arrive at the patient’s side and to ROSC were 9.0, 12.4 and 24.1 min. For survivors and non-survivors the first responding unit and MICU mean time intervals were 8.V9.2 and 9.3/12.9 (P = 0.67 and 0.14).
M. K&ma et al. /Resuscitation 30 (1995) 141-150 Table 2 Aetiology of paediatric cardiac arrest (those with irreversible signs of death included) Aetiology
No. patients
No. survivors
SIDS Trauma Airway related infection mechanical obstruction pulmonary aspiration (Near) drowning Cardiac congenital primary arrythmia pulmonary hypertension Neurologic brain anomaly intracranial hemorrhage Intoxication Sepsis Hanging
47 (59.5%) lOa (12.7%) 8 (10.1%) 6 I I 5 (6.3%) 3 (3.8%) I I I 2 (2.5%) I I 2 (2.5%) I (1.3%) I (1.3%)
I -
3 -
I
aTwo of which by violence.
The mean resuscitation time for survivors and non-survivors was 11.4 and 30.6 min (P = 0.002). The causes for DNR decision were unwitnessed collapse (estimated long time-interval between the collapse and the call) in 12 and high energy trauma in six patients. The overall survival rate was 9.6% (5/52), for attempted resuscitation 14.7% (5/34), for attempted resuscitation when cardiac arrest was witnessed
147
25% (3/12) and for attempted resuscitation with witnessed arrest of cardiac origin 0% (O/3).When resuscitation was attempted in unwitnessed arrests the survival rate was 13.6% (3/22). Survival from resuscitation according to the initial rhythm was 18.5%(5127)in asystole and 0% both in ventricular fibrillation (O/2) and in PEA (O/5). The youngest child to survive was 2 years and 10 months old. When bystander CPR had been initiated and resuscitation continued by EMS personel, four of nine patients (44.4%) survived. None of the children who collapsed at home survived. The five children who survived to discharge, were alive also at 1 year and four of them had a favourable outcome (Table 3). The Bloom category on discharge was I in two, II in one and III in two patients, and at 1 year (for those three who were discharged with Bloom category other than 1) II in two patients and not available in one patient (no. 1). Patient 1 was discharged after 11 months of hospitalization at the age of 16 and her disability was estimated to be permanent defined as Bloom category III. In univariate analysis the scene of collapse in public place (P = O&l), the aetiology of arrest near drowning (P = 0.006) and bystander initiated CPR (P = 0.01) were related to favourable outcome, MICU time interval < 10 min was related to survival. The duration of resuscitation was significantly shorter in survivors with favourable outcome (P = 0.015). Multivariate analysis (logistic regression model) showed no factor related to favourable outcome, but MICU time interval < 10 min was related to survival.
Table 3 Characteristics of the live survivors. All were alive after I year of discharge. Time to ROSC was calculated from the time when the call was first received. Sex
Age
Aetiology
Bystander Witnessed Initial arrest rhythm CPR
Trauma Near drowning Near drowning Near drowning Hanging
No Yes Yes Yes Yes
years
I 2 3 4 5
F M M M F
15.7 5.0 2.9 6.9 14.6
‘See Material and methods section. n.a., not available.
Yes Yes No Yes No
Asystole Asystole Asystole n.a. n.a.
BAY temperature
Time to ROSC
Bloom categorya at discharge/ after I year
Normothermia Hypothermia Nonnothermia Normothermia Normothermia
I7 40 26 8 6.5
3lna 2l2 312 I/na I lna
148
M. Kuisma et al. /Resuscitation
4. Discussion
The etiologies of cardiac arrest and sudden unexpected death in children differ considerably from those in adults, cardiac origin being the minority [ 131.The aetiology seenin this study corresponds to that of previous out-of-hospital studies [1,6]. The two main categories were SIDS and trauma, which were responsible for more than two thirds of cardiac arrests. When airway related cardiac arrests and near-drowning patients are included these four major categories were responsible for 9 out of 10 cardiac arrests. The even higher proportion of SIDS (59.5%) in this study is explained by inclusion of patients with irreversible signs of death to aetiologic evaluations. The high autopsy frequency (98.6O/aof those who died) guaranteed that the aetiology could be verified in every arrest.The percentageof unknown etiologies has in previous cardiac arrest studies varied between 5 and 24 [1,3,6]. However, high autopsy frequency in paediatric sudden death population has been reported also by others [14]. The incidence of non-traumatic out-of-hospital cardiac arrest among individuals less than 18 has been reported to be 12.7/100OOO/year[l], and in the present study 8.51100OOO/year in children aged under 16.Neuspiel et al. [l4] found an annual incidence of 4.6/100 000 population for sudden natural death in persons aged l-2 1 years. The material also included patients developing cardiac arrest in hospital. The incidences are not directly comparable due to the different inclusion criteria for age and aetiology, but the incidence of out-of-hospital cardiac arrest in our study can be judged to be lower than previously reported. Marked variation in incidence rates has been shown to occur between different communities, and incidence rates have been found to be negatively related to survival rates [15]. Two thirds of paediatric cardiac arrests occur during the first 18 months of life [6]. In this study the proportion was 72.2%, SIDS being the most common aetiology. The incidence of cardiac arrest in children aged 2- 14 years was low but it increased again during the 15th year. At this age, the growing child begins to resembleyoung adults also with respect to cardiac arrest aetiology [16]. Compared to adult population, witnessed ar-
30 (1995) 141-150
rests in children are less common. Both Losek et al. [6] and Eisenberg et al. [l] reported a 29% frequency of witnessed arrests which is supported by the present study. Although survival from witnessed arrests was twice as likely as from unwitnessed arrests, almost half of the survivors came from the unwitnessed group. None of the study patients had an EMS witnessed arrest. Warning signs remain unrecognized or children may not have prodromal symptoms serious enough to activate EMS system before cardiac arrest. EMS witnessed arrests have not been specifically detined in previous paediatric studies. Every tenth ‘adult’ arrest, on the other hand, is EMS witnessed u71. Asystole, as the most common, and pulseless electrical activity are more common initial rhythms as ventricular fibrillation in paediatric cardiac arrests [4,6,18]. Asystole and PEA are typical initial rhythms in arrests of non-cardiac origin and unwitnessed arrest, and thus assumingly, of paediatric cardiac arrest. The development of VF has been assumedto be related to cardiac mass and the developing autonomic nervous system and therefore VF is not likely to occur in children with a small heart [ 181.The rhythm distribution in this study confirms previous results, although the proportion of ventricular fibrillation was even lower (3.8%). In adult populations, over 75% of the cardiac arrests has been found to occur in private residences [19]. 79.7%of the study patients collapsed at home or in a private apartment and none survived. On the other hand four of the five survivors in this study had a causeof cardiac arrest that is normally more likely to occur in public place (neardrowning, traffic accident). Cardiac arrests outside home are more often witnessed,more likely to have bystander CPR and CPR is initiated earlier which contribute to the more favourable outcome
PO1.
Reports of bystander initiated CPR are scarce. Losek et al. [6] reported a 48% frequency in witnessedpaediatric arrests and Lombardi et al. [21] 32% overall frequency in adult population. The overall bystander CPR frequency and the frequency in witnessed arrests were 23.1 and 22.2%, respectively, in this study. Survival to discharge in previous studies has
hf. Kuisma et al. /Resuscitation
been 6.7, 5.2, 0 and 7.9% [1,3,4,6]. Also in inhospital arrests the survival rate has been reported to be relatively low [2]. In contrast, in adult population survival rates up to 34% has been reported [8] when cardiac arrest is bystander witnessed, of cardiac origin and with the initial rhythm of ventricular fibrillation. In this study survival from attempted resuscitation was markedly higher (14.7%) than in previous studies. However, in 18 cases a DNR desicion was made by the MICU physician due to a hopeless situation. In other EMS systemsthese patients would probably have undergone unsuccessful resuscitation attempts. One-year survival incidence has not been reported previously in a paediatric resuscitation population. In adult populations those who survive to be discharged are likely to be alive at one year [22]. The same trend was also registered in this study - every survivor was alive after 1 year of discharge. However, neurologic sequelae have been reported to be frequent in survivors [3,6]. Especially when the arrest is airway related the total anoxia time, including the pre-cardiac arrest period, may be too long to allow total recovery [23]. In neurologically intact survivors resuscitation has been initiated and ROSC achieved promptly (31. In our study only two of the survivors were neurologically intact at discharge, but outcome was favourable in three patients, and after 1 year in four patients of five. It is concluded that most of the survivors are likely to lead a life of normal or near-normal quality. The Utstein style was found to be applicable to report retrospective data from paediatric out-ofhospital cardiac arrest. Owing to the aetiology and the rarity of witnessed arrest in paediatric population subgroup analyses of non-cardiac origin and unwitnessed arrests are essential in template presentation. Not all recommended core times could be obtained due to the retrospective nature of the study. There are 16 core and supplemental times in the Utstein recommendations [7]. Even in diligent record keeping it may be difficult to obtain all times reliably, and the number of recommended times to record should probably be reduced in order to allow uniform data reporting. The development of EMS systems and CPR have markedly increased the quantity and quality of survival from out-of-hospital cardiac arrest in
30 (1995) 141-150
149
adult population. The sametrend cannot be clearly seenin paediatric patients major reasons probably being long delay from the collapse to the initiation of CPR due to the high proportion of unwitnessed arrests and the different aetiology of cardiac arrest. Because these underlying prognostic factors cannot be modified no major improvement in survival is likely to occur in developed EMS systemsin the near future. However, based on the fact that injury is the leading cause of paediatric death above the age of 1 year, programmes aimed to prevent childhood injuries and drownings should receive vigorous attention [24,25]. As a medical system resource ALS has been shown to be associatedwith lower paediatric trauma death rates [26]. In addition to trauma prevention programmes, the teaching of bystander initiated CPR and the art of swimming should be further encouraged. conclusions Survival from paediatric cardiac arrest has remained low probably owing to the aetiology and high proportion of unwitnessed arrests. The overall survival rate in the present study was 9.6%, survival after attempted resuscitation 14.7% and 0% when resuscitation was attempted in witnessed arrest of cardiac origin. Asystole was the most common initial rhythm and the four leading causes for cardiac arrest were SIDS, trauma, airway related arrest and (near)drowning. Those five who survived to discharge were also alive at 1 year and four of them had favourable outcome. The Utstein style for recording adult out-of-hospital cardiac arrests was found applicable also in paediatric cases. Acknowledgements
The authors thank Per Rosenberg, MD, PhD, for valuable comments, Anneli OjajCrvi, MSc, for statistical review, Leila Saari, MICU nurse and Ville Voipio, assistant, for help in data collection. References [II
Eisenbexg M, Bergner L, Hallstrom A. Epidemiology of cardiac arrest and resuscitation in children. Ann Emerg Med 1983; 12: 612-674.
150
hf. Ku&ma et al. /Resuscitation
[2] Lewis JK, Minter MG, Eshelman SJ, Witte MK. Outcome of pediatric resuscitation. Ann Emerg Med 1983; 12: 291-299. [3] Torphy DE, Minter MG, Thompson BM. Cardiorespiratory arrest and resuscitation of children. AJDC 1984; 138: 1099-1102. [4] Applebaum D. Advanced prehospital care for pediatric emergencies.Ann Emerg Med 1985; 14: 656-659. [5] Zaritsky A, Nadkami V, Getson P, Kuehl K. CPR in children. Ann Emerg Med 1987; 16: 1107-l 11I. [6] Losek JD, HermesH, Glaeser P, Hendley G, Nelson DB. Prehospital care of the pulseless,nonbreathing pediatric patient. Am J Emerg Med 1987; 5: 370-374. [7] Cummins RO, Chamberlain DA, Abramson NS et al. Recommendedguidelines for uniform reporting of data from out-of-hospital cardiac arrest: the Utstein style. Ann Emerg Med 1991;20: 861-874. [S] Eisenberg MS, Cummins RO, Larsen MP. Numerators, denominators, and survival rates: reporting data from out-of-hospital cardiac arrest. Am J Emerg Med 1991;9: 544446. [9] Valenzuela TD, Spaite DW, Meislin HV, Clark LL, Ewy GA. Case and survival definitions in out-of-hospital cardiac arrest: effect on survival rate calculation. J Am Med Assoc 1992;261: 272-274. [IO] American Heart Association. Guidelines for cardiopulmonary resuscitation and emergency cardiac care. Recommendationsof the 1992national conference J Am Med Assoc 1992;268: 2251-2275. [ll] Bloom HJG, Wallace ENK, Henk JM. The treatment and prognosis of medulloblastoma in children: a study of 82 verified cases.Am J Roentgen01 1969; 105: 43-62. [12] Glauser TA, Packer RJ. Cognitive deficits in long-term survivors of childhood brain tumors. Child’s Nerv Syst 1991;7: 2-12. [13] Myerburg RJ. Sudden cardiac death: epidemiology, causesand mechanisms.Cardiology 1987;74: suppl. 2-9. [I41 Neuspiel DR, Kuller LH. Sudden and unexpected natural death in childhood and adolescence.J Am Med Assoc 1985;254: 1321-1325.
30 (1995) 141-150
I151 Becker LB, Smith DV, Rhodes KV. Incidence of cardiac arrest: a neglected factor in evaluating survival rates. Ann Emerg Med 1993;22: 86-91. [I61 Safranek DJ, Eisenberg MS, Larsen MP. The epidemiology of cardiac arrest in young adults. Ann Emerg Med 1992;21; 1102-l 106. [171 Eisenberg MS, Cummins RO, Litwin PE, Hallstrom AP. Out-of-hospital cardiac arrest: Significance of symptoms in patients collapsing before and after arrival of paramedics. Am J Emerg Med 1986;4: 116-120. Walsh CK, Krongrad E. Terminal cardiac eletrical activMl ity in pediatric patients. Am J Cardiol 1983;5I: 557-561. I191 Becker LB, Gstrander MP, Barrett J, Gondos GT. Outcome of CPR in a large metropolitan area - where are the survivors. Ann Emerg Med 1991;20: 355-361. PO1 Litwin PE, Eisenberg MS, Hallstrom AP, Cummins RO. The location of collapse and its effect on survival from cardiac arrest. Ann Emerg Med 1987; 16: 787-791. Lombardi G, Gallagher J, Gennis P. Outcome of out-ofWI hospital cardiac arrest in New York City. The prehospital arrest evaluation (PHASE) study. J Am Med Assoc 1994;271: 678-683. Kimman GP, Ivens EM, Hartman JA, Hart HN, SiWI moons ML. Long-term survival after successfulout-ofhospital resuscitation. Resuscitation 1994,28: 227-232. 1231 Orlowski JP. Effectivenessof pediatric cardiopulmonary resuscitation (editorial). AJDC 1984; 138: 1097. 1241 Hoekelman RA, Pless IB. Decline in mortality among young Americans during the 20th century: Prospects for reaching national mortality reduction goals for 1990. Pediatrics 1988;82: 582-595. 1251 Chameides L. CPR challenges in pediatrics. Ann Emerg Med 1993;22 (pt2): 388-392. WI Rutledge R, Smith CY, Axizkhank RG. A populationbased multivariate analysis of the association of country demographic and medical system factors with per capita pediatric trauma death rates in North Carolina. Ann Surg 1994,219: 205-210.