Cardiac arrest in Ontario: circumstances, community response, role of prehospital defibrillation and predictors of survival Robert J. Brison, MD, MPH, FRCPC; Judith R. Davidson, MSc; Jonathan F. Dreyer, MD, CM, FRCPC; Gordon Jones, MD, FRCPC; Justin Maloney, MD, CCFP (EM), FRCPC; Douglas P. Munkley, MD, MCFP (EM); H. Michael O'Connor, MD, FRCPC, MMEd; Brian H. Rowe, MD, CCFP (EM), MSc Objectives: To describe the patient characteristics, circumstances and community response in cases of out-of-hospital cardiac arrest; to evaluate the effect on survival of the introduction of prehospital defibrillation; and to identify factors that predict survival. Design: Population-based before-and-after clinical trial. Setting: Five Ontario communities: London, Sudbury, the Greater Niagara region, Kingston and Ottawa. Patients: A consecutive sample of 1510 primary cardiac arrest patients who were transported to hospital by ambulance over 2 years. Intervention: The use of defibrillators by ambulance attendants. Main outcome measures: Patient characteristics (sex and age), circumstances of arrest (place, whether arrest was witnessed and cardiac rhythm), citizen response (whether cardiopulmonary resuscitation [CPR] was started by a bystander, time to access to emergency medical services and time to initiation of CPR), emergency medical services response (ambulance response time, time to initiation of CPR and time to rhythm analysis with defibrillator) and survival rates. Main results: A total of 92.1% of the patients were 50 years of age or older, and 68.3% were men. Overall, 79.6% of the arrests occurred in the home. The average ambulance response time for witnessed cases was 7.8 minutes. The overall survival rate was 2.5%. The survival rates before and after defibrillators were introduced were similar, and the general functional outcome of the survivors did not differ significantly between the two phases. Factors predicting survival included patient's age, ambulance response time and whether CPR was started before the ambulance arrived. Conclusions: The survival rate was lower than expected. The availability of prehospital defibrillation did not affect survival. To improve survival rates after cardiac arrest ambulance response times must be reduced and the frequency of bystander-initiated CPR increased. Once these changes are in place a beneficial effect from advanced manoeuvres such as prehospital defibrillation may be seen.

Objectifs: Decrire les caracteristiques des patients, les circonstances et la reaction de la collectivite dans les cas d'arret cardiaque a l'exterieur de l'h6pital; evaluer les effets, sur la survie, de l'introduction d'interventions de defibrillation avant l'arrivee Ai I'hpital; et identifier les facteurs de prevision de survie. Conception: Epreuve clinique sur population avant-apres. Dr. Brison is an assistant professor ofsurgery and of community health and epidemiology at Queen's University, Kingston, Ont., Ms. Davidson is research coordinator of the Ontario Prehospital Care Study Group, Queen's University, Kingston, Ont., Dr. Dreyer is chief of the Department ofEmergency Medicine and medical director of the Base Hospitalfor Pre-Hospital Care, Victoria Hospital, London, Ont., Dr. Jones is medical director of the Regional Base Hospitalfor Southeastern Ontario, Kingston, Ont., Dr. Maloney is medical director of the Base Hospital/ALS programs, Ottawa General Hospital, Ottawa, Ont., Dr. Munkley is director ofEmergency Medical Services and medical director of the Niagara Region Base Hospital, Greater Niagara General Hospital, Niagara Falls, Ont., Dr. O'Connor is an associate professor of surgery, Division of Emergency Medicine, Queen's University, Kingston, Ont., and Dr. Rowe is an assistant clinical professor at the University of Ottawa, Ottawa, Ont., and research director, Northeastern Ontario Family Medicine and Trauma programs, Sudbury, Ont.

Reprint requests to: Dr. Robert J. Brison, Division of Emergency Medicine, Queen's University, 76 Stuart St., Kingston, ON K7L 3V2 -

For prescribing information see page 260

CAN MED ASSOC J 1992; 147 (2)

191

Cadre: Cinq collectivites de l'Ontario: London, Sudbury, la region metropolitaine de Niagara, Kingston et Ottawa. Patients: Echantillon consecutif de 1 510 victimes d'arret cardiaque primaire transportees a l'h6pital par ambulance pendant 2 annees. Intervention: Utilisation d'un defibrillateur par les ambulanciers. Principales mesures de resultat: Caracteristiques des patients (sexe et age), circonstances de l'arret (lieu, s'il y a eu temoin et rythme cardiaque), reaction des citoyens (y a-t-il eu reanimation cardiorespiratoire [RCR] amorcee par un temoin, temps d'acces aux services medicaux d'urgence et delai d'application de la RCR), reaction des services medicaux d'urgence (temps de reponse de l'ambulance, delai d'application de la RCR et temps de prise de rythme au defibrillateur) et taux de survie. Principaux resultats: Au total, 92,1 % des patients avaient au moins 50 ans et 68,3 % d'entre eux etaient des hommes. Globalement, 79,6 % des arrets se sont produits a la maison. Le temps moyen de reaction des ambulanciers pour les cas a temoin etait de 7,8 minutes. Le taux global de survie etait de 2,5 %. Les taux de survie avant et apres defibrillation etaient analogues et l'issue fonctionnelle generale pour les survivants n'etait pas significativement differente dans l'une ou l'autre phase. Parmi les facteurs permettant de prevoir la survie, mentionnons l'age du patient, le temps de reponse des services ambulanciers et la question de savoir si la RCR a e commencee avant l'arrivee de l'ambulance. Conclusions: Le taux de survie etait inferieur aux previsions. L'existence de defibrillation avant l'hospitalisation n'a eu aucune incidence sur le taux de survie. Pour ameliorer les taux de survie apres un arret cardiaque, il faut reduire les temps de reponse des ambulances et faire en sorte que la RCR soit plus souvent amorcee par une personne presente. Lorsque ces changements seront en place, il est possible qu'on puisse observer un effet benefique decoulant des manoeuvres anticipees, par exemple la defibrillation avant l'hospitalisation.

M ore Canadians die of cardiac disease than of any other cause. Each year about 56 000 (21 000 being Ontario residents) do so.' Often, the first sign of cardiac disease is sudden cardiovascular collapse.2 If a pulse is not restored at the scene or during transport to the emergency department, the patient has little chance of recovery.3 Recognizing the importance of prehospital response to cardiac arrest, investigators in many communities are now evaluating the effect of early care given by bystanders and emergency personnel.4-'3 Previous research indicates that the prehospital circumstances most conducive to survival after cardiac arrest include the presence of a witness,4,'4 an electrocardiographic rhythm of ventricular fibrillation as opposed to another type of dysrhythmia,4"3"4 a short time to cardiopulmonary resuscitation (CPR),7 a short time to definitive care7 and the initiation of CPR by a bystander.'5'16 In about 50% to 75% of cases sudden cardiac arrest is initially associated with ventricular fibrillation or other ventricular tachydysrhythmia. 17-22 Early defibrillation is believed to be crucial to the survival of patients with ventricular fibrillation23 and may be more important than early drug therapy or intubation.24 However, results of prehospital defibrillation trials have not been consistent. Although some controlled studies have shown significant increases in survival with prehospital systems providing early defibrillation,25'26 others have failed to do So.27,28 192

CAN MED ASSOC J 1992; 147 (2)

Except in a few large urban centres in Ontario, defibrillation has not been available until the patient arrives in the emergency department. In 1987 the Ontario Ministry of Health introduced defibrillation to the prehospital care system in certain communities. This provided an opportunity to assess the effect on survival of the addition of defibrillation to a system that had previously provided only basic cardiac life support (BCLS). Although some authors have examined the prehospital treatment of cardiac arrest in Canada,9"'922'29 there has been no report on the effectiveness of defibrillation in Canadian systems. In addition, there is a need for Canadian data, based on a large patient sample, describing the actual circumstances of out-of-hospital arrest and the nature of the emergency medical services response. The objectives of this study were to describe the patient characteristics, circumstances and community response in cases of out-of-hospital cardiac arrest in five Ontario communities, to evaluate the effect of the introduction of prehospital defibrillation on survival, and to identify prehospital factors that predict survival following out-of-hospital cardiac arrest in these regions.

Methods The study population consisted of all people who had experienced cardiac arrest and had been transported to hospital by ambulance over a 2-year period (between March 1986 and July 1989) in five LE 15 JUILLET 1992

Ontario communities: London, Sudbury, the Greater Niagara region, Kingston and Ottawa. The records of all pulseless, apneic adult patients were examined. Patients with the following conditions were excluded: poisoning, trauma, suicide, smoke inhalation, drowning, electrocution, hypovolemia, advanced cancer, kidney failure, abdominal or thoracic aortic aneurysm, subarachnoid hemorrhage, cerebral vascular accident, serious infectious disease, drug-related or alcohol-related arrest, obstructed airway, severe asthma attack and congenital heart defect. Information on the patient's condition was obtained from death certificates, emergency department charts and hospital medical records. Each participating centre had a population that varied from 120 000 to 600 000. Ambulance services in the five centres serve a combined population of about 1.5 million. All emergency medical attendants had been trained in BCLS, which includes airway maintenance, ventilation (with bag-valve-mask systems, positive pressure devices or pocket masks) and external cardiac compression. By the second year of the study the attendants had also received 7 to 8 hours of training in the use of manual, semiautomatic or automatic defibrillators. Training included review of cardiac anatomy and physiology, cardiac rhythm interpretation, recognition of cardiac arrest, defibrillator operation, protocol review and testing. Before defibrillators were introduced people with cardiac arrest received only BCLS from emergency medical attendants. The following year, in addition to providing BCLS, attendants used defibrillators to monitor the cardiac rhythm and, if appropriate, to deliver electrical shocks. The defibrillation protocol involved attachment of the defibrillator at the scene followed immediately by an initial cycle of assessment. At all study sites a cycle consisted of a maximum of three rhythm analyses, each followed by a shock, if appropriate, at 200, 200 and 360 J respectively. Shocks were indicated for ventricular fibrillation and, at one centre (with manual defibrillators), asystole. Additional cycles were performed (depending on the local protocol) under specified conditions, such as refibrillation or advice from the base hospital physician. Data were collected for a 1-year period before (phase 1) and after (phase 2) the introduction of defibrillators. Because defibrillators were introduced at different times in the five centres, the dates of the 2-year period differed among centres; however, all 2-year spans fell between March 1986 and July 1989. For both phases information was obtained by reviewing ambulance call reports, dispatch records, emergency department records, death certificates and inpatient medical records. Data recorded for the study included the patient's age and sex, location of the arrest, whether the arrest was witnessed, whether JULY 15, 1992

CPR was started before the ambulance arrived (and by whom), vital signs at the scene and in the emergency department, and outcome. The time of the following events was recorded: collapse (estimate), initiation of CPR, receipt of call by emergency medical services, crew notification, arrival at and departure from scene, and arrival in emergency department. For witnessed cases the time of collapse was that obtained from witnesses and recorded by ambulance officers. For unwitnessed arrests this time was estimated, if possible, from ambulance and emergency department records. The following time intervals were calculated: time to access to emergency medical services (time of collapse to time call received by emergency medical services), time to CPR (time of collapse to time CPR initiated), response time (time call received to time of arrival of ambulance at scene), time at scene (time of arrival of ambulance at scene to time of departure) and time to emergency department (time of collapse to time of arrival of ambulance at emergency department). After defibrillators were introduced tape recordings (from recorders located in semiautomatic and automatic defibrillators) were used to identify cardiac rhythms occurring at the scene and to confirm rhythm interpretation and delivery of shocks. The cardiac rhythm on arrival in the emergency department was obtained from emergency department charts. Information on survival was obtained from emergency department charts and (for patients who were admitted) hospital medical records. The data from all the centres were entered into a computerized database. We computed descriptive measures of the data using the database software.30 In calculating all proportions, we used as the denominator the number of known cases for the variable under consideration. Factors potentially associated with survival were compared between the two phases by means of x2 tests and t-tests. Survival and outcome measures were compared between the phases by means of x2 tests and logistic regression. To identify factors associated with survival we performed a two-step analysis. First, univariate analyses were performed for each variable, patients being grouped according to whether they had survived. Second, variables predictive of survival with p values of 0. 10 or less in univanate analyses were considered in multivariate logistic models.3'

Results Descriptive data Over the 2 years 1510 cases met the study criteria, for a crude rate of 5.0 cardiac arrests per CAN MED ASSOC J 1992; 147 (2)

193

10 000 population per year. The estimated annual incidence rate varied from 4.2 to 8.2 per 10 000 population among the centres. Male patients outnumbered female patients by 2.2:1. The mean age was 68.1 (standard deviation 12.4) years (extremes 22 and 105 years); 92.1% of the patients were aged 50 years or more. A total of 79.6% of the arrests occurred in the home. Overall, 47.7% of the arrests were not witnessed, 44.1% were witnessed by a bystander, and 8.2% were witnessed by ambulance personnel. Initial electrocardiographic recordings at the scene (in 709 cases in phase 2) showed the following distribution of rhythms: ventricular fibrillation in 41.3% of cases, asystole in 41.3%, agonal bradydysrhythmia in 7.8%, supraventricular rhythms in 1.8% and other (pacemaker spikes, idioventricular rhythms, third-degree heart block, other type of heart block or junctional rhythms) in 7.5%. In the 648 cases witnessed by bystanders 26.3% of the patients received CPR before the ambulance arrived. The proportion varied from 18.4% to 29.5% among the centres. The frequency of provision of CPR by bystanders varied depending on where the arrest occurred. The rate was 58% for arrests that occurred in the workplace, 51.5% for arrests that occurred in public places and 18.0% for arrests that occurred at home. The mean length of time for the various events following access to emergency medical services is shown in Table 1. To eliminate the confounding influence of various collapse-to-discovery times in unwitnessed cases, times are shown for bystanderwitnessed cases only. A total of 9.5% of the patients survived to be admitted to hospital, 2.5% survived to be discharged from hospital and 1.6% were discharged at their previous level of functioning. The rates varied from 5.4% to 19.2%, 0.5% to 4.4% and 0.5% to 3.2% respectively among the centres.

Effect ofdefibrillation The characteristics of the patients and the circumstances of arrest in phases 1 and 2 are shown in Table 2. Four factors were found to differ between the phases: there was a tendency for more arrests to be witnessed by a bystander in phase 1 than in phase 2 (p < 0.06), 911 service was used more often in phase 2 than in phase 1 (p < 0.001) (two centres had 911 service in phase 2 v. one centre in phase 1), more patients in phase 2 than in phase 1 received CPR from a firefighter or police officer before the ambulance arrived (owing primarily to enhanced fire department response in phase 2 at two centres) (p < 0.001), and the mean time at the scene was longer in phase 2 than in phase 1 (p < 0.001). The outcome in the two phases is shown in Table 3. We used multivariate logistic regression to compare survival between the phases, adjusting for factors that differed between the phases. (Time at the scene was not adjusted for because it apparently had a direct relation to phase; i.e., it increased with the use of defibrillators in phase 2. Each of the other factors was added separately to the basic model, which examined the influence of phase on survival.) Neither the basic model nor any of the adjusted models showed a statistically significant difference in survival between the phases. There was no significant difference between the phases in the proportion of survivors discharged at

their previous level of functioning. There were fewer patients in ventricular fibrillation (p < 0.0001) and more patients in asystole (p < 0.001) on arrival in the emergency department in phase 2 than in phase 1. There was no significant difference between the phases in the frequency of supraventricular rhythms or the frequency of agonal rhythms on arrival. The number of patients with a detectable pulse at the scene on ambulance arrival or departure did

iu ,u s ii. e cs Mear u r n, aese ntn.ssed ov OvSt?rse9tn tar t.-lr(lit'-

rable 'Arrest

- .. 1.

HeSp(1 Stle r'im

to

carioi[)uImoniar

r ;

esuscitation 'CPP 2PR started b.Ly

Bystander icl 8 o3loce officer i: retilohtel tm b u;anLe atienidar. Hes p on se 't e 3 1n e a t scee rie 638: -le t( fi.s anaiys s Nit :Jefbr'IlatL :-.; me t arrlV;v ir: ernerge.

.4. %. I320t:

0.

'1eparnnt

194

CAN MED ASSOC J 1992; 147 (2)

LE 1 5 JUILLET 1992

=

not differ significantly between phase 1 and phase 2 (7.0% v. 7.5% and 3.7% v. 5.4% respectively). However, on arrival in the emergency department significantly fewer patients in phase 1 than in phase 2 had a detectable pulse (1.7% v. 4.1%) (p= 0.01).

Factors predicting survival We compared the patient characteristics and circumstances of arrest for the survivors and nonsurvivors to select variables appropriate for predicting survival in a multifactor model (Table 4). Sex and

Table 2: Patient characteristics and circumstances of arrest before (phase 1) and after (phase 2) introduction of prehospital defibrillation Variable

Mean age (and SD), yr Sex, no. of patients Female Male Location of arrest, no. of patients Home Public place or vehicle Ambulance Unknown Who witnessed arrest, no. of patients Bystander Ambulance attendants No one Unknown 911 service used, no. of patients Yes No Who began CPR, no. of patients Bystander Firefighter or police officer Ambulance attendant Unknown Mean time to access to emergency medical services (and SD), min Mean time to CPR (and SD), min Mean response time (and SD), min Mean time at scene (and SD), min Mean time to emergency department (and SD), min 'NS

=

Phase 1 (n = 754)

Phase 2 (n = 756)

p value*

68.0 (12.2)

68.2 (12.6)

NS NS

251 503

228 528

602 124 26 2

598 137 21

340 58 322 34

308 62 379 7

137 617

359 397

122

145

92 456 84

160 380 71

18.6 (59.6)

19.1 (64.0)

NS

25.5 (58.9)

25.4 (58.5)

NS

7.8 (4.4)

7.6 (4.3)

NS

8.0 (3.8)

10.9 (4.4)

< 0.001

40.6 (60.5)

43.9 (64.0)

NS

NS

0.06

< 0.001 < 0.001

not significant.

Table 3: Outcome in the two phases =

Outcome Died in emergency department Died in hospital after admission Discharged To institution Home, altered level of functioning Home, previous level of functioning JULY 15, 1992

No. (and O/o) of patients Phase 1 Phase 2

689 (91)

677 (90)

49 (6) 16 (2) 1

57 (8) 22 (3) 3

6

4

9

15 CAN MED ASSOC J 1992; 147 (2)

195

mean time at the scene were not found to differ significantly between the two groups in univariate analysis. All the other variables were considered in fitting multivariate logistic regression equations to the data, survival to hospital discharge being used as the dependent variable. Significant factors from the univariate analysis fell into three categories: patient characteristics (age), circumstances at the scene and time factors. Owing to the small number of survivors, for statistical reasons only one variable of each type could be included in a multivariate model at the same time. The final model was derived by choosing one factor from each group on the basis of statistical significance and usefulness for planning emergency medical services. The following logistic regression model best described the data. ln[Y/(1

-

Y)] = 1.327

0.038a + 1.229b, + 2.376b2

-

Ii' d

r

Vi

-si

u.

.:

0.2469c

,

,--

.-:

(,,IdI *'

HCI -C

-

where ln = natural logarithm; Y = probability of survival; a = age, in years; b, = 1 when CPR was started by fire department or police personnel, otherwise 0; b2 = 1 when CPR was started by a bystander, otherwise 0; and c = response time in minutes. The factors in the model are age, who started CPR and response time. Each term was significant below the 0.05 level. A total of 1335 cases for which values for all three variables were available were used to create the model. It can be used to estimate the probability of survival in people who experience cardiac arrest for whom age, information on the provision of CPR and the response time are available. In general the model illustrates that survival decreases with age and with increasing response time. In addition, who starts CPR is important: relative risk estimates calculated from this variable indicated that, compared with no CPR until the ambulance arrived, the chances of survival increased by a factor of 10.8 when CPR was started by a bystander and by 3.4

r

l

-Jf *-aE

-.

eiq *"v ea]b

F'lC-l Fi

i rn er-

,

c;| i| .

S

196

CAN MED ASSOC J 1992; 147 (2)

LE 15 JUILLET 1992

when CPR was started by a firefighter or police officer.

Discussion The resuscitation rate for these Ontario communities is substantially lower than the rates observed in systems that offer prehospital advanced cardiac life support (ACLS) or paramedic services, and it is among the lowest rates reported for systems that provide BCLS.32 The survival rate, 2.5%, is the lowest found in Canadian reports. The rate of survival following out-of-hospital cardiac arrest with prehospital BCLS was found to be 4.0% in Winnipeg29 and 8.8% in Halifax-Dartmouth.9 With ACLS the rate was 9.3% in New Westminster, BC.22 The low survival rate in our study probably cannot be blamed on chance, given the large sample. Fewer arrests were witnessed, but otherwise the patient profile and circumstances of arrest are similar to those of previous studies.4,6,0,12,20,21,24,25,27,28,33-39 However, two prehospital response factors are noticeably different from those in most previous reports. The rate of bystander-initiated CPR is among the lowest of reported values.32,38-4' The mean ambulance response time (8 minutes) is longer than that reported for other systems (2 to 7.3 minutes).9 lo 16,20,25,33,35,42 The introduction of defibrillation to the prehospital BCLS system did not markedly improve the survival rate or outcome status of patients with cardiac arrest in the five communities studied. Although the number of prehospital resuscitations increased in phase 2 (a significantly greater proportion of patients had a pulse on arrival in the emergency department), the number of patients surviving to hospital discharge was similar in the two phases. Our findings are consistent with those in studies from rural Minnesota, which showed no effect on survival of the availability of prehospital defibrillation.27,28 These reports stated that in rural and small communities unavoidably long response times (and therefore times to defibrillation) may cancel the potential benefits of prehospital defibrillation. Hargarten and colleagues23 demonstrated that with every minute of delay between arrest and administration of the first shock the chance of obtaining a perfusing rhythm with defibrillation declines steeply. After 3 minutes the success rate of defibrillation plateaus at a low level. It is evident that rapid response is necessary for successful defibrillation. Studies showing a positive effect of defibrillation have involved emergency medical services systems capable of providing shocks in less than 10 minutes (on average) from the time of collapse.25'26'33 Eisenberg and associates33 found that JULY 15, 1992

19% of patients survived after prehospital defibrillation was introduced; Stults and coworkers25 observed a survival rate of 11% in communities providing BCLS plus defibrillation. In our study the mean time from collapse to rhythm analysis with the defibrillator for witnessed cases was 13.1 minutes, clearly longer than times in previous studies reporting positive results. A brief response time is not the sole prerequisite for improved survival with prehospital defibrillation. In Stockport, England,43 a low survival rate (0.9%) was observed with automatic external defibrillator pacing, even with an average response time of 4.5 minutes. The authors suggested that survival may also depend on community response, including early call for help and administration of CPR by a bystander. Maximal survival rates are observed in communities with well-developed emergency medical services systems, involving early emergency response (BCLS or BCLS plus defibrillation) followed by arrival of paramedics.?0 Eisenberg and associates32 suggested that it is the early provision of CPR in these systems that enhances the effectiveness of other procedures, including defibrillation. The factors most important in predicting survival in our study were age, who initiates CPR and response time. Improving the odds of survival therefore depends on optimizing the last two, modifiable factors. Promoting initiation of CPR before the ambulance arrives and reducing response times could result in substantial improvements in survival with defibrillation. On a practical level, planners of emergency medical services in the study regions would be wise to promote CPR training programs for citizens. In addition, ways of reducing ambulance response times should be investigated. Response time can be reduced by paying attention to organizational factors such as geographic location and number of ready ambulances, staffing and systems of routing. Although the type of defibrillator (automatic, semiautomatic or manual) and specific resuscitation procedures varied among the centres, we do not believe this weakens the validity of our results. Previous reports indicate that defibrillator type does not have an important effect on either time to defibrillation" or survival rate.45 The poor survival rate at all five centres suggests that any variation in the effect of defibrillator type on survival was slight. The study design did not permit random allocation of defibrillation to patients because the Ontario Ministry of Health had decided to introduce prehospital defibrillation in these communities on the basis of published reports of defibrillator efficacy. We took this opportunity to study the effect of prehospital defibrillation in this setting by collecting data for the 1-year periods before and after the introduction CAN MED ASSOC J 1992; 147 (2)

197

of defibrillators. Data quality was monitored through site visits and regular meetings of the investigators. Care was taken in identifying eligible patients to accumulate data only for cases of primary cardiac arrest - those most likely to benefit from such an intervention. In addition, factors that differed between phases and that could have affected survival were adjusted for in our analysis. We therefore believe that the nonrandom design of this population-based study is unlikely to have added biases that would alter the interpretation of our results. Finally, at the inception of our study it was thought that an improvement in the rate of survival to hospital discharge from an expected 2.5% to at least 5% was needed to warrant the effort of introducing automatic defibrillators into the prehospital care setting. We felt, from the results of previous studies assessing these devices, that a survival rate of 5% was a conservative expectation. Our study had a power of 87% to detect such a difference.

Conclusions The introduction of automatic external defibrillation into a prehospital care system may not improve outcome in cases of out-of-hospital cardiac arrest unless other factors are optimized. Early defibrillation is important in patients who have been sustained by CPR. To enhance Ontario's prehospital services for cardiac arrest patients, ambulance response time and time to defibrillation must be reduced, and the delivery of CPR by bystanders and by police officers and firefighters must be promoted. We appreciate the valuable contributions and support of the following base hospital program directors: Andrew Duncan (Niagara), Brian Field (Kingston), Dallas La Barre (Sudbury), Ronald Sharpe (Ottawa) and J. Edward Slomer (London). For assistance with data collection we thank Joanne Adams (Sudbury), Mary Ellen Logan (Niagara) and Pamela Sheehan (Ottawa). For assistance with data entry we thank Andrea Crowdis, Brenda Fowler and Paul McElligott. All of these people and ourselves constituted the Ontario Prehospital Care Study Group. This research was supported by Emergency Health Services research grant 00888N from the Ontario Ministry of Health.

References 1. Statistics on Number of Deaths - Canada and Ontario, Heart and Stroke Foundation of Ontario, Toronto, 1991 2. American Heart Association: Standards and guidelines for cardiopulmonary resuscitation and emergency cardiac care. JAMA 1986; 21: 2841-3044 3. Lewis LM, Ruoff B, Rush C et al: Is emergency department resuscitation of out-of-hospital cardiac arrest victims who arrive pulseless worthwhile? Am J Emerg Med 1990; 8: 118198

CAN MED ASSOC J 1992; 147 (2)

120 4. Mullie A, Lewi P, Van Hoeyweghen R et al: Pre-CPR conditions and the final outcome of CPR. Resuscitation 1989; 17 (suppl): SI l-S17 5. Bjorklund P, O'Rourke MF: Pre-hospital emergency care: evaluation of an Australian system. Aust NZ J Med 1984; 14: 419-423 6. Einarsson 0, Jakobsson F, Sigurdsson G: Advanced cardiac life support in the prehospital setting: the Reykjavik experience. J Int Med Res 1989; 225: 129-135 7. Eisenberg MS, Bergner L, Hallstrom A: Cardiac resuscitation in the community: importance of rapid provision and implications for program planning. JAMA 1979; 241: 1905-1907 8. Eitel DR, Walton SL, Guerci AD et al: Out-of-hospital cardiac arrest: a six-year experience in a suburban-rural system. Ann Emerg Med 1988; 17: 808-812 9. Goldstein DH, MacKenzie BR, Merchant N et al: Effectiveness of a non-amalgamated ambulance service on out-of-hospital sudden cardiac death. Can J Cardiol 1987; 3: 66-69 10. Gudjonsson H, Baldvinsson E, Oddsson G et al: Results of attempted cardiopulmonary resuscitation of patients dying suddenly outside the hospital in Reykjavik and the surrounding area, 1976-1979. Acta Med Scand 1982; 212: 247-251 11. Hunt RC, McCabe JB, Hamilton GC et al: Influence of emergency medical services systems and prehospital defibrillation on survival of sudden cardiac death victims. Am J Emerg Med 1989; 7: 68-82 12. Jakobsson J, Nyquist 0, Rehnqvist N: Cardiac arrest in Stockholm with special reference to the ambulance organization. Acta Med Scand 1987; 222: 117-122 13. Roth R, Stewart RD, Rogers K et al: Out-of-hospital cardiac arrest: factors associated with survival. Ann Emerg Med 1984; 13: 237-243 14. Hearne TR, Cummins RO: Improved survival from cardiac arrest in the community. PACE 1988; 11: 1968-1973 15. Bossaert L, Van Hoeyweghen R, Cerebral Resuscitation Group: Bystander cardiopulmonary resuscitation (CPR) in out-of-hospital cardiac arrest. Resuscitation 1989; 17 (suppl):

S55-S69 16. Cummins RO, Eisenberg MS: Prehospital cardiopulmonary resuscitation: Is it effective? JAMA 1985; 253: 2408-2412 17. Enns J, Tweed WA, Donen N: Prehospital cardiac rhythm deterioration in a system providing only basic life support. Ann Emerg Med 1983; 12: 478-481 18. Myerburg RJ, Conde CA, Sung RJ et al: Clinical, electrophysiologic and hemodynamic profile of patients resuscitated from prehospital cardiac arrest. Am JMed 1980; 68: 568-576 19. Adamson F, Munkley D, Reed WC: Out-of-hospital cardiac arrest in a Canadian mixed density population. Emerg Prehosp Med 1988; 2 (5): 15-18 20. Wilson BH, Severance HW, Raney MP et al: Out-of-hospital management of cardiac arrest by basic emergency medical technicians. Am J Cardiol 1984; 53: 68-70 21. Olson DW, LaRochelle J, Fark D et al: EMT-defibrillation: the Wisconsin experience. Ann Emerg Med 1989; 18: 806811 22. Vertesi L: The paramedic ambulance: a Canadian experience. Can MedAssoc J 1978; 119: 25-29 23. Hargarten KM, Stueven HA, Waite EM et al: Prehospital experience with defibrillation of coarse ventricular fibrillation: a ten-year review. Ann Emerg Med 1990; 19: 157-162 24. Martin TG, Hawkins NS, Weigel JA et al: Initial treatment of ventricular fibrillation: Defibrillation or drug therapy? Am J Emerg Med 1986; 6: 113-119 25. Stults KR, Brown DD, Schug VL et al: Prehospital defibrillation performed by emergency medical technicians in rural communities. N Engl J Med 1984; 310: 219-223 26. Weaver WD, Copass MK, Bufi D et al: Improved neurological recovery and survival after early defibrillation. Circulation 1984; 69: 943-948 27. Bachman JW, McDonald GS, O'Brien PC: A study of LE 15 JUILLET 1992

28. 29.

30. 31. 32.

33.

34.

35. 36.

out-of-hospital cardiac arrests in northeastern Minnesota. JAMA 1986; 256: 477-483 Vukov LF, White RD, Bachman JW et al: New perspectives on rural EMT defibrillation. Ann Emerg Med 1988; 17: 318321 Tweed WA, Bristow G, Donen N: Resuscitation from cardiac arrest: assessment of a system providing only basic life support outside of hospital. Can Med Assoc J 1980; 122: 297300 R:BASE, Microrim Inc, Redmond, Wash, 1988 EGRET, version 2, Statistical and Epidemiological Research Corp, Seattle, 1986 Eisenberg MS, Horwood BT, Cummins RO et al: Cardiac arrest and resuscitation: a tale of 29 cities. Ann Emerg Med 1990; 19: 179-186 Eisenberg MS, Copass MK, Hallstrom AP et al: Treatment of out-of-hospital cardiac arrests with rapid defibrillation by emergency medical technicians. N Engl J Med 1980; 302: 1379-1383 Rosman HS, Goldstein S, Landis JR et al: Clinical characteristics and survival experience of out-of-hospital cardiac arrest victims without coronary heart disease. Eur Heart J 1988; 9: 17-23 Weaver WD, Hill D, Fahrenbruch CE et al: Use of the automatic external defibrillator in the management of out-ofhospital cardiac arrest. N Engl J Med 1988; 319: 661-666 Jakobsson J, Rehnqvist N, Nyquist 0: One year's experience of early defibrillation in Stockholm. J Int Med Res 1989; 225: 297-301

selegiline hydohrd

Rx Summary Antiparkinson Agent Indications and clinical use: As an adjunct to levodopa (with or without a decarboxylase inhibitor) in the management of the signs and symptoms of Parkinson's disease. In newly diagnosed patients before symptoms begin to affect the patient's social or professional life, at which time more efficacious treatment becomes necessary.

Contraindications: In patients with known hypersensitivity to Eldepryl, Eldepryl should not be used in patients with active peptic ulcer, extrapyramidal disorders such as excessive tremor or tardive dyskinesia, or patients with severe psychosis or profound dementia. Eldepryl should not be used with meperidine (Demerol or other trade names). This contraindication is often extended to other opioids. Warnings (Selective vs non-selective inhibition of MAO-B): Eldepryl should not be used at daily doses exceeding those recommended (10 mg/day) because of the risks associated with non-selective inhibition of MAO. It is prudent, in general, to avoid the concommitant use of Eldepryl and fluoxetine (Prozac). Warnings to patients: Patients should be advised of the possible need to reduce levodopa dosage after the initiation of Eldepryl therapy. The patients should be advised not to exceed the daily dose of 10 mg. The risk of using higher doses of Eldepryl should be explained, and a brief description of the "hypertensive crisis" ("cheese reaction") provided. Precautions: Some patients given Eldepryl may experience an exacerbation of levodopa associated side effects, presumably due to the increased amounts of dopamine reacting with supersensitive post-synaptic receptors. These effects may often be mitigated by reducing the dose of levodopa by 10-30%. NURSING MOTHERS: It is not known whether Eldepryl is excreted in human milk. Because many drugs are excreted in human milk, consideration should be given to discontinuing the use of all but absolutely essential drug treatments in nursing women. PEDIATRIC USE: The effects of Eldepryl in children under 18 have not been evaluated.

JULY 15,1992

37. Litwin PE, Eisenberg MS, Hallstrom AP et al: Location of cardiac arrest and its effects on survival from cardiac arrest. Ann Emerg Med 1987; 16: 787-791 38. Ritter G, Wolfe RA, Goldstein S et al: The effect of bystander CPR on survival of out-of-hospital cardiac arrest victims. Am Heart J 1985; 110: 932-937 39. Guzy PM, Pearce ML, Greenfield S: The survival benefit of bystander cardiopulmonary resuscitation in a paramedic served metropolitan area. Am J Public Health 1983; 73: 766769 40. Lund I, Skulberg A: Cardiopulmonary resuscitation by lay people. Lancet 1976; 2: 702-704 41. Thompson RG, Hallstrom AP, Cobb LA: Bystander-initiated cardiopulmonary resuscitation in the management of ventricular fibrillation. Ann Intern Med 1979; 90: 737-740 42. Pressley JC, Raney MP, Wilson BH et al: Assessment of out-of-hospital resuscitation. Am J Emerg Med 1984; 2: 215216 43. Gray AJ, Redmond AD, Martin MA: Use of the automatic external defibrillator-pacemaker by ambulance personnel: the Stockport experience. BMJ 1987; 294: 1133-1135 44. Papa FJ: Time to defibrillation: a controlled laboratory study comparing three automated and semi-automated defibrillators. J Emerg Med 1987; 7: 163-167 45. Stults KR, Brown DD, Kerber RE: Efficacy of an automated external defibrillator in the management of out-of-hospital cardiac arrest: validation of the diagnostic algorithm and initial clinical experience in a rural environment. Circulation 1986; 74: 701-709

Laboratory Tests: No specific laboratory tests are esential for management of patients on Eldepryl. Transient or continuing abnormalities with tendency for elevated values in liver function tests have been described in long term therapy. Although serious hepatic toxicity has not been observed, caution is recommended in patients with a history of hepatic dysfunction. Periodic routine evaluation of all patients is however appropriate. Drug Interactions: The occurence of stupor, muscular rigidity, severe agitation and elevated temperature has been reported in a man receiving selegiline and meperidine, as well as other medications. These symptoms were resolved over days when the combination was discontinued. This case is typical of the interaction of meperidine and MA01s. Other than the possible exacerbation of side effects in patients receiving levodopa therapy, no interactions attributed to the cornbined use of ELDEPRYL and other drugs have been reported. It is also prudent to avoid the combination of ELDEPRYL and fluoxetine (Prozac). Use during Pregnancy: The use of Eldepryl during pregnancy has not been established. Therefore, Eldepryl should be given to a pregnant woman only if the potential benefits outweigh the potential risks. Adverse reactions: A) IN COMBINATION WITH LEVODOPA THE SIDE EFFECTS OF ELDEPRYL ARE USUALLY THOSE ASSOCIATED WITH DOPAMINERGIC EXCESS. ELDEPRYL MAY POTENTIATE THE SIDE EFFECTS OF LEVODOPA, THEREFORE ADJUSTMENT OF THE DOSAGE OF LEVODOPA MAY BE REQUIRED. ONE OF THE MOST SERIOUS ADVERSE REACTIONS REPORTED WITH ELDEPRYL USED AS AN ADJUNCT TO LEVODOPA THERAPY ARE HALLUCINATIONS/CONFUSION, PARTICULARLY VISUAL HALLUCINATIONS. Other reactions include nausea, dizziness, faintness, abdominal pain, dry mouth, vivid dreams, dyskinesias and headache. B) IN MONOTHERAPY The incidence of adverse reactions occurring in trials using Eldepryl as monotherapy has not been fully reported to date. Serious adverse reactions include depression, chest pain, myopathy and diarrhea. Other reported adverse reactions include insomnia, headache, nausea, dizziness and vertigo. In prospective clinical trials, the following adverse effects (listed in decreasing order of frequency), led to the discontinuation of Eldepryl: Nausea, hallucinations, confusion, depression, loss of balance, insomnia, orthostatic hypotension, increased akinetyic involuntary movements, agitation, arrhythmia, bradykinesia chorea, delusions, hypertension, new or increased angina pectoris and syncope. Events reported only rarely as a cause of discontinuation

of treatment include anxiety, drowsiness/lethargy, nervousness, dystonia, increased episodes of freezing, increased tremor, weakness, excessive persperation, constipation, weight loss, buming lips/mouth, ankle edema, gastroitestinal bleeding and hair loss. Dosage: The recommended dosage of Eldepryl as monotherapy in newly diagnosed patients, or as adjunct to levodopa (usually with a decarboxylase inhibitor) is 10 mg per day administered as divided doses of 5 mg each taken at breakfast and lunch. When ELDEPRYL adjunctive therapy is added to the existing levodopa therapeutic regime, a reduction, usually of 10 to 30% in the dose of levodopa (in some instances a reduction in the dose of Eldepryl to 5 mg/day) may be required during the period of adjustment of therapy or in case of exacerbation of adverse effects. Doses higher than 10 mg per day should not be used. There is no evidence that additional benefit will be obtained from the administation of higher doses. Furthermore, higher doses will result in a loss of selectivity of Eldepryl towards MAO-B with an increase in the inhibition of type MAO-A. There is an increased risk of adverse reactions with higher doses as well as an increased risk of hypertensive episode ("cheese reaction")

Supplied:

Eldepryl 5 mg tablets, available in bottles of 60 tablets. References: 1. The Parkinson Study Group. Effect of Deprenyl on the Progression of Disability in Early Parkinson's Disease. New Eng Journ 321, 1364-1371, November 1989. 2. Eldepryl (selegiline hydrochloride) Product Monograph, December 1990.3. Myllyla VV, Sotaniemi KA, Vuorinen JA, Heinonen EH. Selegiline as initial treatment in de novo parkinsonian patients. Neurology 1992; 42, 339-343. 4. Tetrud JW, Langston JW. The Effect of Deprenyl (Selegiline) on the Natural History of Parkinson's Disease. Science, August 1989, vol. 245, 519-522. 5. Myllyla VV, Sotaniemi KA, Vuorinen J. Heinonen EH. Selegiline (deprenyl) as primary treatment in Parkinson's disease. Selegiline therapy in early Parkinson's disease. July 1990, 19-24. 6. Langston JW in Lees A. Deprenyl in Parkinson's Disease: Guidelines for Clinicians. North American Round Table Series, No. 1, 1988, 1-26. 7. DuVoisin RC in Lees A. Deprenyl in Parkinson's Disease: Guidelines for Clinicians. North American Round Table Series, No. 1,1988, 1-26. Product Monograph available upon request.

in PURDUE FREDERICK INC. 378 Roncesvalles Ave., Toronto, Ontario TEL: (416) 537-4372 CAN MED ASSOC J

1992; 147 (2)

199