Resuscitation 47 (2000) 147–154

The attitudes of cardiac arrest survivors and their family memberstowards CPR courses

Andreas Kliegel, Wolfdieter Scheinecker, Fritz Sterz *, Philip Eisenburger,Michael Holzer, Anton N. Laggner

Uni6ersitatsklinik fur Notfallmedizin, Allgemeines Krankenhaus der Stadt Wien, Wahringer Gurtel 18–20/6/D, 1090 Wien, Austria

Received 17 November 1999; received in revised form 27 March 2000; accepted 7 April 2000

Abstract

Objecti6es: to evaluate self-assessment of first aid knowledge, readiness to make use of it in case of a medical emergency andjudgement of a 1-day CPR course by cardiac arrest survivors, their family members and friends as compared to the general public.Background: the recurrence rate of a cardiac arrest after successful resuscitation is high and most of out-of-hospital cardiac arrestsoccur at the patient’s home. Methods: medical students trained in basic and advanced life support provided 101 members of thetarget group and 94 of a sex and age matched control group with a 1-day course in CPR. Results: after the course, half of theparticipants in both groups considered their knowledge of first aid to be very good or good. The readiness to perform first aidin a medical emergency increased significantly. Of the target group 96% of the participants as compared with the control groupwhere 91% felt confident to recognise a cardiac arrest; 79 versus 68% considered themselves capable to perform CPR if needed.The course was judged as very good in 71 versus 69% and as good in 25 versus 27% with no differences between groups.Conclusion: one-day CPR courses are well accepted by cardiac arrest survivors, their family members and friends and help toreduce fears of reacting in medical emergencies. They seem to be more motivated to gain and use first aid knowledge than others.© 2000 Elsevier Science Ireland Ltd. All rights reserved.

Keywords: Basic life support; Cardiac arrest; Cardiopulmonary resuscitation; Education; Training

www.elsevier.com/locate/resuscitation

1. Introduction

Cardiovascular diseases are the principle causeof death in the developed countries [1]. It is wellknown that the interval between cardiac arrest andCPR being started influences the result of resusci-tation [2]. If a resuscitation attempt is started earlyand carried out effectively by bystanders, the pa-tient’s chance of survival is significantly improvedcompared with those victims, who received ineffec-tive or even no bystander CPR at all [3–10]. Moreadvanced prehospital treatment of cardiac arrestimproves the rates of survival and discharge, andneurological outcome [6,11]. Another important

effect of early CPR is prolongation of the durationof ventricular fibrillation (VF), which maintainsthe chance for successful defibrillation [7].

As most of the out-of-hospital cardiac arrestsoccur at the patient’s home, the persons to bemost likely present are family members or friends[12,13]. It is clear that those who live with, orspend a lot of time with, high-risk cardiac patients,should receive sufficient CPR training.

Of the 677 cardiac arrest patients who weretreated in our department, 36% died within thefirst week and another 25% within 5 years. Thecause of death of those patients was again ofcardiac aetiology in over 50% of the cases (unpub-lished data). These findings and those of the earlierstudies reported that the risk of recurrent cardiacarrest is about 30–50% within 1 year [14,15] led usto identify the target group of our study.

* Corresponding author. Tel.: +43-1-404001964/1952; fax: +43-1-404001965.

E-mail address: fritz.sterz@akh-wien.ac.at (F. Sterz).

0300-9572/00/$ - see front matter © 2000 Elsevier Science Ireland Ltd. All rights reserved.PII: S0300-9572(00)00214-8

A. Kliegel et al. / Resuscitation 47 (2000) 147–154148

The aim of the study was to evaluate self-assess-ment of first aid knowledge, the readiness to makeuse of it in case of a medical emergency and thejudgment of a 1-day CPR course of cardiac arrestsurvivors and their family members and friends,compared to a sex and age matched control group.

2. Materials and methods

Questionnaires were sent to 190 survivors out of1153 cardiac arrest patients who had primary suc-cessful resuscitation and were transported to orresuscitated at the acute care unit of the Depart-ment of Emergency Medicine at a University Hos-pital. They were asked about their own and theirfamily members’ interest in a 1-day CPR course.We also asked if they had ever attended CPRcourses before and how long ago those trainingsessions had taken place. To find members for thecontrol group, medical students asked passers-byin the entrance hall of the hospital to read and fillin the same questionnaire that was sent to themembers of the target group. Members of medicaland nursing staff, medical students, patients andpeople fitting our target group were excluded.

Medical students have been trained in first aidand advanced life support in a 5-day course con-taining an introduction to the basics of CPR andpractical training with an interactive video system.Afterwards they took part in advanced life sup-port training sessions in which they were taughtintubation, emergency medication and the use ofdifferent kinds of defibrillators. Subsequently theymet both the patients and their family membersand the members of the control group for a CPRcourse, which took place in private homes or inthe seminar rooms of the hospital.

The course contained both theoretical and prac-tical parts. First, the students explained the meth-ods of diagnosing unconsciousness, respiratoryand cardiac arrest. The basis of 1-person CPR wasdiscussed and followed by a practical demonstra-tion. The participants were then encouraged topractice CPR themselves. As we wished to adaptthe content and the practical training precisely tothe personal requirements of the participants weset no time limit for the duration of the courses.The participants themselves determined the com-pletion of the training when they believed theyneeded no more time for practising and had no

further questions. At the end of the training allhad to answer a written questionnaire regardingthe content of the course and their attitudes to-wards CPR and courses performed in this way.

2.1. Questionnaire

Translation of used questionnaires.1. Preliminary questionnaire.� Name:� Address:� Have you ever attended a Cardiopulmonary

Resuscitation (CPR) course?Possible answers: yes, no

� How long ago was your last CPR course?Possible answers: 1–4 weeks, 1–2 months,2–6 months, 6–12 months

� When and where would you like to attendour CPR course?

� How many persons want to attend thecourse?

2. Questionnaire handed out after the course� Sex:� Age:� Where did your CPR course take place?

Possible answers: at home, in the GeneralHospital of Vienna, at a friend’s house, inrooms of a rescue organisation, somewhereelse

� Were you content with the premises?Possible answers: yes, no

� Have you ever performed first aid before?Possible answers: yes, no

� What is the emergency phone number forrescue service?Possible answers: 123, 122, 141, 1774, other:

� Have you written down that number nearyour telephone at home?Possible answers: yes, no

� Do you have an in-house emergency-callsystem at home?Possible answers: yes, no

� How would you describe your first aidknowledge?Possible answers: very good, good, satisfac-tory, unsatisfactory

� How would you describe your readiness toperform first aid (before the course)?Possible answers: high, rather high, ratherpoor, poor

� Are you afraid of making mistakes?

A. Kliegel et al. / Resuscitation 47 (2000) 147–154 149

Possible answers: no, a little, quite a lot, very� Which reasons do you have not to perform

CPR?Possible answers: own illness, handicap, fear,too exhausting, no reason, other

� Did you understand the principles of CPR?Possible answers: yes, no

� How do you recognise a respiratory arrest?Possible answers: mirror, see-listen-feel,movements of the chest, skin, others

� How long was your CPR course?Possible answers:B2 h,B4 h, B6 h, \6 h

� How would you describe the course?Possible answers: very good, good, satisfac-tory, not satisfactory motivating, useful,confusing, boring, too much theory, too lit-tle practice, enough theory and practice

� Are you motivated to attend further courses?Possible answers: yes, maybe, not yet, no

� Are you now able to recognise, whether aCPR is necessary?Possible answers: yes, no, I don’t know

� Does the fear of AIDS influence your way ofacting?Possible answers: yes, no, I don’t know

� How would you act if you had a semi-auto-matic defibrillator for life rescue?Possible answers: use it; if works as simple asa fire extinguisher I would use it even with-out prior training; I don’t know; never use itwithout prior training

� Why do you think do a lot of people notperform first aid in spite of their knowledge?

� Who will you try to motivate to attend CPRcourses?Possible answers: family members, friends,co-workers

� Which persons would you help, family mem-bers, friends, co-workers, well-dressed per-sons, drunken persons, neglected?Possible answers: yes, probably, probablynot, no

� Did the practice with the manikin causedifficulties?Possible answers: no problems, fewproblems, problems, practising was impossi-ble

� I had difficulties with...Possible answers: ventilation, heart massage,no difficulties

� Concerning the whole course I especiallyappreciated that...Possible answers: The students came to myhome. I got clear and understandable in-structions. I feel capable now to performbasic first aid steps. I lost my fear of mouth-to-mouth ventilation. The atmosphere athome was pleasant. The students behavedwell. The training did not put me understress. The training was adapted to myneeds.

� Concerning the whole course I did not likethat...Possible answers: The students seemed to beinsecure. Everything was presented too fast.Too much medical terminology was used.Everything was presented too slowly. I willnot remember anything in an emergency. Iwas made feel insecure. Not all of my ques-tions were answered.

2.2. Statistical analysis

Data are expressed as median and range, per-centages were calculated for frequencies. Groups ofcontinuous data were compared with the Mann–Whitney U-test, for the comparison of frequenciesthe x2-test or Fishers exact, if appropriate, wasused. A P value B0.05 was considered statisticallysignificant.

3. Results

Of the questionnaires initially mailed, 62 werefilled in correctly and returned. In the controlgroup, 25 questionnaires were also completed. Ofthe target group, 55% (n=34) had attended a CPRcourse at least once before. Only two of those (6%)had taken part in CPR training within the pastyear. The courses undertaken by the other 94%(n=32) were on average 22 years (range 3–57years) ago. These findings were similar in thecontrol group. Fifty-two percent (n=13) had at-tended a CPR course, but none of them within thepast year. These courses were on average 19 years(range 2–48 years) ago.

The CPR courses, held by nine medical studentsmostly took place in private homes or in theseminar rooms of the hospital. Interestingly, 31%of the members of the target group compared with

A. Kliegel et al. / Resuscitation 47 (2000) 147–154150

only 15% of the control group chose to be taughtin private homes whereas 51 vs. 76% preferred theseminar rooms of the hospital (P=0.003). One-hundred-and-one members of the target group (30patients, 71 family members/friends) attended thecourses and, the control group consisted of 94persons. They were taught in 29 versus 16 groupswith an average of 3.5 (range: 1–18) versus 5.9(range: 2–12) participants per course. About twothirds (65%, n=66 vs. 67%, n=63) were femalewith an average age of 41 years (13–78 years)versus 38 years (11–72 years). The average age ofthe males was 43 years (16–71 years) versus 44years (12–78 years) with no statistical differencesbetween the groups.

Although the members of the target group wereat high risk of witnessing a medical emergency,either because of their own cardiac disease or theillness of their relatives or friends, 8% (n=8) werenot able to name the correct emergency telephonenumber. In the control group knowledge of thatnumber was even poorer as 15% (n=14) failed togive the correct answer (P=0.09). Thirty percent(n=30) (43%, n=40) declared, that they had notwritten that number near their telephone (P=0.02). Only 3% of the target group (which actuallyrepresented a single household) and none of thecontrol group had installed an in-house emer-gency-call system for elderly persons. Twenty-onepercent (n=21) versus 7% (n=7) stated, that theyhad performed first aid at least once before (P=0.05).

The readiness to help in a medical emergencywas 59% (n=60) (47%, n=44) before our course

(P=0.02). After they had been taught by thestudents, 97 (n=98) versus 90% (n=85) werewilling to help their family members (P=0.04), 91(n=92) versus 87% (n=82) to help friends (P=0.13), 80 (n=81) versus 77% (n=72) co-workers(P=0.11), 74 (n=75) versus 72% (n=68) well-dressed persons (P=0.12), 42 (n=42) versus 41%(n=39) drunken people (P=0.12) and 41 (n=41)versus 37% (n=35) would help the neglected (P=0.10). Six percent (n=6) versus 7% (n=7) statedthat they would not use their knowledge fordrunken persons (P=0.20), 8 (n=8) versus 10%(n=9) or the neglected (P=0.18) (Table 1).

Before the training about half the participantsin both the groups considered their knowledge offirst aid to be very good (11 vs. 9%) or good (39vs. 40%). Only 9 versus 6% said that they were notafraid to make mistakes if they had to performfirst aid, with no statistical differences between thegroups (Fig. 1). After CPR training, 96 (n=97)versus 91% (n=86) felt confident to judge whetherthe situation required the performance of CPR ornot (P=0.10) and that they understood the prin-ciples. However, only 79 (n=80) versus 68% (n=64) were confident actually to carry out CPR(P=0.02). The reasons in the six persons of thetarget group, who felt unable to carry out CPRwere fear (n=3), own illness (n=2) and in onecase a medical handicap. The ten members of thecontrol group, who felt unable to perform CPR,put the reason as fear. All were able to namecriteria for the diagnosis of a respiratory arrest.Only 16% (n=16) of the target group, but 26%(n=24) of the control group, said that the fear of

Table 1‘‘Which persons would you help?’’ Target group (cardiac arrest survivors) versus control group (general public)a

No (%)Probably not (%)Probably (%)Yes (%)Group No answer (%)

Family members Target 97 (n=98) 1 (n=1) 2 (n=2)0 00010 (n=9)90 (n=85)Control 0

06 (n=6)91 (n=92)TargetFriends 0 3 (n=3)87 (n=82) 0Control 0 2 (n=2)11 (n=10)

Target 80 (n=81) 7 (n=7)Co-workers 0 0 13 (n=13)17 (n=16)01 (n=1)5 (n=5)77 (n=72)Control

Target 9 (n=9)74 (n=75) 17 (n=17) 0Well dressed persons 0Control 11 (n=10)72 (n=68) 14 (n=13) 3 (n=3) 0

7 (n=7)6 (n=6)22 (n=22)24 (n=24)Drunken persons 42 (n=42)Target7 (n=7) 6 (n=6)41 (n=39) 18 (n=17) 27 (n=25)Control

40 (n=41)Target 9 (n=9)Neglected 8 (n=8)19 (n=19)24 (n=24)12 (n=11)10 (n=9)22 (n=21)19 (n=18)37 (n=35)Control

a For statistics see text.

A. Kliegel et al. / Resuscitation 47 (2000) 147–154 151

Fig. 1. ‘‘Are you afraid of making mistakes in performing CPR?’’ Target group (cardiac arrest survivors) versus control group(general public), P=0.001.

being infected with the human immunoglobinvirus (HIV) would influence their intervention(P=0.03).

After brief information concerning the functionof a semi-automatic external defibrillator, 63 (n=64) versus 67% (n=63) stated that they wouldnever use one without previous training (P=0.10);8% (n=8) versus 13% (n=12) refused to use it atall (P=0.10).

3.1. Judgement of the course

All were content with the premises where thecourses had taken place. Fifty-six percent (56%) ofthe participants were taught in courses that didnot last longer than 2 h. Of those, 11 versus 2%considered the time to be too short (P=0.01).Thirty-four percent versus 41% took part in les-sons that lasted 2–4 h (P=0.06), which wasjudged to be too little by 6 vs. 0% of the concernedpersons (P=0.02). The other courses exceeded 4h, which was considered to be enough byeveryone.

The majority gave a positive judgement on thequality of the CPR courses. Seventy-onepercent (n=72) versus 69% (n=65) thought thecourse was very good (P=0.11), 25 (n=25) ver-sus 27% (n=25) chose the answer ‘good’ (P=0.12). Additionally, 46% (n=46) versus 41%,(n=39) described the course as motivating (P=0.09).

Practice with the manikin caused difficulties in33 (n=33) versus 34% (n=32). These occurredmainly with the ventilation aspect and rarely withchest compressions (P=0.12).

The interest in further courses was significantlyhigher in the target group than in the control

group and also varied between the age groups(Fig. 2). It was highest among the older persons.Of all the participants, 87 vs. 81% said (P=0.08),that they would try to motivate their family mem-bers to attend CPR courses, 85 vs. 78% wouldrecommend them to their friends (P=0.06) and 63vs. 61% to their co-workers (P=0.11).

4. Discussion

Our results suggest that cardiac arrest survivorsand their family members and friends show con-siderable interest in learning CPR in our coursesand that afterwards the majority was ready to usetheir knowledge and skills in medical emergencies.The training seemed to reduce prior fears to react.

As in many other countries, CPR courses arerequired in Austria also for the acquisition of adriving licence. Some companies require a certifi-

Fig. 2. ‘‘I am motivated to attend further CPR courses.’’Target group (cardiac arrest survivors) versus control group(general public), P=0.012.

A. Kliegel et al. / Resuscitation 47 (2000) 147–154152

cate of a CPR course as a condition for employ-ment. Some companies organize their own first aidcourses. Therefore, it is reasonable to believe thata high percentage of the population should be ableto know at least the basic elements of first aidincluding CPR. Nevertheless, the proportion oflay persons who attempt CPR, if they witness acardiac arrest, is not as high as it would be desir-able [16]. In comparison, the proportion of healthcare workers, such as ambulance or hospital per-sonnel, who try to help in an emergency while theyare off duty is well above average [17]. Our studyshows that there are persons who do not evenknow the emergency phone number and it is con-ceivable that this group could increase with theexcitement in an emergency case.

Previous studies have discussed the efficacy andvalue of the present models of CPR training[18,19]. As family members of cardiac patients areat greater risk of being at hand when a suddencardiac arrest occurs, the logical consequenceshould be that they should take CPR courses morefrequently and in a higher proportion than others.In fact, family members of high risk patients donot even attend CPR courses as much as thosewho are not related to members of a risk group[19].

When family members were asked for their rea-sons for not attending CPR training, the mainanswers given were lack of time and that it did notoccur to them to do so [20]. As we achieved goodresults in our courses of 2–6 h duration, the firstreason seems to be incomprehensible. The secondreason that was mentioned, may be solved bybetter information, given, for example, by thetreating physician, through information campaignsor by public health officials.

Though physicians who treat cardiac patientsare aware of the importance of CPR training forfamily members, only a few doctors explicitly rec-ommend it [19,21]. Dracup et al. [22] showed thatthe doubts of many physicians concerning theability of family members of cardiac patients tolearn CPR effectively, were unfounded. Of the 83family members of high risk patients who theytrained in CPR, 81% successfully learned anddemonstrated appropriate CPR skills. This is com-parable to our results, where 79% of the targetgroup felt confident of being able to perform CPR.However, our results might be not comparable,because of the wording of the question. ‘‘Which

reasons do you have for not performing CPR?’’The possible answers were: ‘‘own illness’’, ‘‘handi-cap’’, ‘‘fear’’, ‘‘too exhausting’’, others and noreason. As 15 members of this group did notanswer the question at all, we suspect, that theyonly read the question without looking at all ofthe answers including ‘‘no reason’’, and so did notchoose any of them.

Although family members of cardiac patients donot seek CPR courses [20], our results indicatethat most of them are ready to learn or improvetheir skills. The majority is also willing to usethem if required. Compared with the controlgroup, those who had already been confrontedwith a cardiac arrest of a relative or a friend, orhad been resuscitated themselves, seemed to havea better approach to the whole subject. Obviously,they accepted the danger of a cardiac emergencyas something that could always happen in theirclose surroundings, as they chose to invite thestudents to their homes more often than the mem-bers of the control group.

Although only 59% of the participants of thehigh risk group were ready to intervene in amedical emergency before the training, the propor-tion of the control group were even less. Thistendency remained after the course, though boththe groups showed an impressive increase in theirreadiness to help (Table 1). The differences be-tween the two groups were also noticeable in theirattitudes towards refresher courses, and their will-ingness to motivate other people was probably dueto a higher personal interest and motivation in theparticipants of our target group. The fact that themajority of the high risk group and even a higherpercentage of the control group were at leastslightly afraid of making mistakes (Fig. 1) indi-cates, that CPR courses should contain only essen-tial information which can be easily understoodand remembered by the participants. We reducedthe theoretical part of our courses to basic infor-mation with emphasis on the diagnosis and treat-ment of respiratory and cardiac arrest. We omitteddetailed pathophysiological aspects of differentcauses of death so that the participants were notoverloaded with information and seemed to bemore confident of performing CPR if necessary.The importance of simplifying the contents ofCPR courses has already been stressed by otherauthors [23]. We think that the small number ofparticipants per course, with an average of 3.5 vs.

A. Kliegel et al. / Resuscitation 47 (2000) 147–154 153

5.9, and the fact that the participants knew eachother well also had a decisive influence on thegeneral positive evaluation of the training. Thethreshold of inhibition to ask questions and toperform new learned skills may be quite a bitlower than in public courses. The informal atmo-sphere made the whole course more pleasant forthe participants and for the students and, there-fore, possibly more effective. These considerationscould be valid for first aid courses in general.

The interest in further courses was significantlyhigher in our target group especially in olderparticipants (Fig. 2). Though over 90% of theparticipants were trained in courses that did notlast longer than 4 h, the majority was content andthought that the time was enough.

Our results suggest that cardiac arrest survivors,their family members and friends can learn CPRand that their attitudes towards such training isvery positive. Therefore, effort should be increasedtowards informing members of the target groupabout the possibility and importance of suchcourses. As CPR skill retention is very poor[20,24], it must be emphasised that they should bemotivated to refresh their skills and knowledgefrom time to time.

Acknowledgements

Thanks to all the students, who made this pro-ject possible (R. Drescher, V. Gies, M. Grabner,Ch. Gruber, St. Krassnitzer, St. Majtenyi, E.Pfeifer, B. Pohl, U. Pollheimer, A. Reichel, G.Reichel, B. Schmidt, N. Schrammel, St. Trauner,P. Vychytil, M. Wlk). This work was made possi-ble through a generous contribution of MedtronicAustria GesmbH (Dr Robert Nitsche). MichaelHolzer was supported by BIOMED2 EuropeanCommission, DG XII for Science Research andDevelopment, Directorate Life Science and Tech-nologies, Biomedical and Health Research Divi-sion (c : BMH4-CT96-0667).

References

[1] Sans S, Kesteloot H, Kromhout D. The burden ofcardiovascular diseases mortality in Europe. Task forceof the European Society of Cardiology on cardiovascu-lar mortality and morbidity statistics in Europe. EurHeart J 1997;18:1231–48.

[2] Larsen MP, Eisenberg MS, Cummins RO, HallstromAP. Predicting survival from out-of-hospital cardiac ar-rest: a graphic model. Ann Emerg Med 1993;22:1652–8.

[3] Cobb LA, Hallstrom AP. Community-based cardiopul-monary resuscitation: what have we learned? Ann NewYork Acad Sci 1982;382:330–42.

[4] Lund I, Skulberg A. CPR by lay people. Lancet1976;2:702–4.

[5] Thompson DG, Hallstrom AP, Cobb LA. Bystanderinitiated cardiopulmonary resuscitation in the manage-ment of ventricular fibrillation. Ann Int Med1979;90:737–40.

[6] Copley DP, Mantel JA, Rogers WJ, Russel RO Jr,Rackley CE. Improved outcome for prehospital car-diopulmonary collapse with resuscitation by bystanders.Circulation 1977;56:901–5.

[7] Cummins RO, Eisenberg MS, Hallstrom A, Litwin PE.Survival of out-of-hospital cardiac arrest with early ini-tiation of cardiopulmonary resuscitation. Am J EmergMed 1985;3:114–9.

[8] Ritter G, Wolfe RA, Goldstein S, Landis JR, Vasu CM,Acheson A, Leighton R, Medendrop SV. The effect ofbystander CPR on survival of out-of-hospital cardiacarrest victims. Am Heart J 1985;110:932–7.

[9] Wik L, Steen BA, Bircher NG. Quality of bystandercardiopulmonary resuscitation influences outcome afterprehospital cardiac arrest. Resuscitation 1994;28:195–203.

[10] Crone PD. Auckland Ambulance Service cardiac arrestdata 1991–1993. New Zealand Med J 1995;108:297–9.

[11] Nielsen JR, Gram L, Larsen CF, Lybecker H, AndersenC, Frandsen F, Jorgensen HR, Haghfelt T. Intensifiedprehospital treatment of heart arrest increases the num-ber of survivors with good cerebral function. UgeskrLaeger 1990;152:1901–4.

[12] Goldstein S, Friedman L, Hutchinson R, Canner P,Romhilt D, Schlant R, Sobrino R, Verter J, WassermanA. Timing, mechanism and clinical setting of witnesseddeaths in post myocardial infarction patients. J Am CollCardiol 1984;3:1111–7.

[13] Lombardi G, Gallagher J, Gennis P. Outcome of out-of-hospital cardiac arrest in New York City. The pre-hospi-tal arrest survival (PHASE) study. J Am Med Assoc1994;271:678–83.

[14] Goldstein S, Landis JR, Leighton R, Ritter G, VasuCM, Lantis A, Serokman R. Characteristics of the resus-citated out-of-hospital cardiac arrest victims with coro-nary heart disease. Circulation 1981;64:977–84.

[15] Kass LE, Eitel DR, Sabulsky NK, Ogden CS, Hess DR,Peters KL. One-year survival after prehospital cardiacarrest: the Utstein Style applied to a rural-suburbansystem. Am J Emerg Med 1994;12:17–20.

[16] Ekstrom L, Herlitz J, Wennerblom B, Axelsson A, BangA, Holmberg S. Survival after cardiac arrest outsidehospital over a12-year period in Gothenburg. Resuscita-tion 1994;27:181–7.

[17] Axelsson A, Herlitz J, Ekstrom L, Holmberg S. By-stander-initiated cardiopulmonary resuscitation out-of-hospital. A first description of the bystanders and theirexperiences. Resuscitation 1996;33:3–11.

A. Kliegel et al. / Resuscitation 47 (2000) 147–154154

[18] Brennan RT, Braslow A. Are we training the rightpeople yet? A survey of participants in public cardiopul-monary resuscitation classes. Resuscitation 1998;37:21–5.

[19] Goldberg RJ, Gore JM, Love DG, Ockene JK, DalenJE. Layperson CPR are we training the right people?Ann Emerg Med 1984;13:701–4.

[20] Dracup K, Moser DK, Guzy PM, Taylor SE, MarsdenC. Is cardiopulmonary resuscitation training deleteriousfor family members of cardiac patients? Am J PublicHealth 1994;84:116–8.

[21] Mandel LP, Cobb LA, Weaver WD. Cardiopulmonary

resuscitation training for patients’ families: do physiciansrecommend it? Am J Public Health 1987;77:727–8.

[22] Dracup K, Moser Heaney D, Taylor SE, Guzy PM,Breu C. Can family members of high-risk cardiac pa-tients learn cardiopulmonary resuscitation? Arch IntMed 1989;149:61–4.

[23] Kaye W, Mancici ME. Teaching adult resuscitation inthe US — time for a rethink. Resuscitation1998;37:177–87.

[24] Weaver FJ, Ramirez AG, Dorfman SB, Raizner AE.Trainees retention of CPR: how quickly they forget. JAm Med Assoc 1979;241:901–3.

.