Ethics in treatment decisions during out-of-hospital resuscitation

RESUSCITATlON

Resuscitation33 (1997)2455256

ELSEVIER

Ethics in treatment decisions during out-of-hospital

resuscitation

Anne-Cathrine Naess”,*, Eldri Steenb, Petter Andreas SteenC “Center for Medical Ethics, Gaustadalleen 21, N-0371 Oslo Norway ‘Work Research Institute, Stensberggaten 29, P.O. Box 8171 Dep., N-0034 Oslo Norwuy ‘Department of Anesthesiology, Ulievaal University Hospital and Universit,v of’ Oslo, N-0407 Oslo Norwq

Received22 November1995;revised20 May 1996;accepted28 May 1996

Abstract Paramedics in Oslo are allowed to make decisions about withholding or terminating cardiopulmonary resuscitation (CPR). In order to elicit the criteria used, 35 paramedics and nine doctors were interviewed after 70 episodes of cardiac arrest outside-of-hospital. CPR was not attempted in 21 patients, and discontinued in the field in 28 patients. Spontaneous circulation was restored in 15 patients, and six patients were transported to hospital with ongoing CPR. Both prognostic and ethical criteria were used without a clear borderline. Signs considered to indicate good prognosis such as VF, gasps, contracted pupils, or normal skin color always led to start of CPR. Bystander CPR was continued even when the professional thought the effort was futile, partly to encourage the bystanders. The social status of the patient did not affect the decisions, and advanced age only when combined with important criteria such as arrest times or the relatives’ wishes. The only apparent difference between paramedics and doctors was that the reputation of the EMS system influenced only the paramedics. All paramedics had long experience which influenced their decisions, which were based on a rapidly composed broad picture of the patient’s situation. All presentedserious ethical considerations about life and death indicating that they did not make these decisions lightly. Copyright Q 1997 Elsevier Science Ireland Ltd. Keymrds:

Heart arrest; Cardiopulmonary resuscitation; Emergency medical services; Paramedics; Ethics

1. Introduction

Society places increasing emphasis on patient autonomy in health care decisions, and recognizes the patient’s right to accept or refuse medical therapy [l]. There appear to be significant discrepancies between the decisions of patients, families, and physicians regarding the use of life sustaining treatment in hypothetical situations [2-51. Nevertheless, only 14-36 percent of in-hospital patients in US surveys participate in decisions about their own do-not-resuscitate (DNR) orders [1,6-91. Sudden cardiac arrest out-of-hospital is usually unexpected. Therefore patient wishes cannot normally be elicited before treatment decisions are made. The deci* Correspondingauthor.

sions whether to start, withhold or discontinue cardiopulmonary resuscitation (CPR) therefore usually have to be made by the health personnel without certain knowledge of patients’ wishes. fn many emergency medical services (EMS) the decisions are made only by doctors. The EMS personnel in such services always start CPR and usually bring the patient to the hospital. Sometimes CPR is discontinued in the field after radiocommunication with a doctor. The service examined by us in Oslo, Norway permits paramedics to make these decisions. We were interested in eliciting the criteria actually used by the paramedics in the Oslo EMS-system when making decisions about CPR. We were also interested in finding out whether these criteria tended to differ from the criteria used by the doctors on the physician manned ambulance and if they were affected by the length of experience.

0300-9572/97/$17.00 Copyright 0 1997ElsevierScienceIreland Ltd. All rights reserved PII SOiOn-95?‘(96)01023-4

246

A.-C. Naess et al. 1 Resuscitation 33 (1997) 245-256

In an attempt to avoid influence by the interviewees’ notion of socially palatable criteria, the data were obtained prospectively by in-depth interviews subsequent to actual emergency situations.

2. Materials and methods 2. I. The Oslo EMS system The EMS system in Oslo is a one-tiered centralized community run system for a population of 470 000. The median interval from dispatch to arrival at patient is 6 min. The dispatch center is staffed by nurses and paramedics. A maximum of two operators were involved with a median interval for dispatch call processing of one min. A formal protocol for dispatching was not in use at the time of this study, and dispatcher CPR instructions were not offered to callers. Each response team consisted of two paramedics, one team also included an anesthesiologist on weekdays between 07:30 h and 22:00 h. In 1990 there were 461 unexpected, sudden deaths outside-of-hospital in Oslo (personal communication, A. Skulberg, MD PhD, Oslo EMS-system). CPR was attempted in approximately two thirds of the cases [lo]. Of the arrests of primary cardiac origin [II], 36% were admitted to the hospital intensive care unit with spontaneous circulation, and 12% were discharged from hospital alive [12]. Since 1975, the paramedics have had two years of training with a final certification examination. All paramedics work under the supervision of an anesthesiologist 2 weeks per year. From 1978 to 1992 they performed advanced cardiac life support (ACLS) according to the standards and guidelines of the American Heart Association [13- 151 and from 1992 the guidelines of the European Resuscitation Council [16] including endotracheal intubation, defibrillatory shocks and intravenous drug therapy. The personnel follow standing orders and do not need to contact a base station to obtain permission before initiating or discontinuing therapy. 2.2. Entry criteria All paramedics, residents and staff anesthesiologists in the Oslo EMS-system who had been called to patients in cardiac arrest were eligible for inclusion. The study subjects were interviewed when not otherwise busy, e.g. on an ambulance mission or in the operating room (doctors) at the time when the interviewers arrived.

2.3. Data collection instrument The data were collected at in-depth interviews. After defining the major objectives of the study, an extensive interview guide with 10 main questions and 11 sub-questions was set up (Appendix A). The guide was partially structured with no set alternative answers. Specific opening questions concerning each topic were included and all respondents were encouraged to elaborate freely on each topic. The aims of the study and a description of the proposed interviewing process were presented and discussed with the paramedics and doctors and all received written information. It was clarified that participation was voluntary. The subjects could withdraw from the study at will. All information would be presented exclusively as anonymous group information. The study was approved by the regional committee for medical research ethics, the local paramedics organization and by the EMS administration. It was agreed that informed consent would be verbal, as a signed consent was thought to put unneccessary pressure on the participants. Two interviewers were involved. One was a resident in anesthesiology with 3.5 years of training and 2 years of experience in the EMS system. The other was an educational psychologist with a previous degree and practice as a registered nurse (RN). The latter had taken a special interest in the EMS-system for many years and followed the ambulance crew on somemissions. To coordinate the technique, three pilot interviews were performed by the two interviewers together. The interviewees could choose whether or not an interview should be recorded on tape, but all consented to tape recording of the interview. 2.4. Data collection The interviews took place between December 1992 and May 1993 during regular working hours in the ambulance quarters or in the department of anesthesiology (for the doctors). Approximately 50% of the cardiac arrest incidents during that period were followed by an interview, each lasting 20-60 min, as soon as possible and usually within 2 weeks of the actual event. Most of the paramedics worked in fixed pairs and they could choose to be interviewed together or separately. Four of the paramedics were interviewed in pairs. Nine doctors and 35 paramedics were interviewed (Table 1). The signs and symptoms used for the initial diagnosis of cardiac arrest are listed in Table 2. In 21 cases (30%) CPR was not attempted. In 28 cases (40%) CPR was started and discontinued in the field. In 15 cases (21%) spontaneous circulation was restored in the field and in six cases (9%) the patient was transported to the hospital with ongoing CPR.

Table 1 Respondents’

duration

of work experience in the Oslo EMS-system

Respondent MD 6 3 Paramedics 7 i 14 I1

Number

I 7 >I0

2 (F). 4 (M) I (F), 2 (M)

3 4 5 9 10 19 >20

2 (Ml 8 CM) 14 CM) 11 W

Table 3 Criteria used for treatment decisions The patient perspective

The tapes were transcribed to verbatim written accounts. The interviewer then listened to all tapes and added notes on non-verbal communication such as laughter, crying and silence. Each interview text was broken up into quotations and the information stratified pertaining to the key questions in the interview-guide (see Appendix). These quotations were sorted and stored in FileMakerTM-Pro data-base program. In the following, quotations are presented in small letters, paramedics identified by (P), doctors by (MD).

3. Results

I. Prognostic criteria Ventricular fibrillation registered on ECC; Gasps or small movements Contracted pupils or normal skin co101 Age Time intervals Cardiac arrest witnessed by the ambulance personeli Bystander CPR 3. Ethical criteria ‘The patients right to live or right to dtc Age

The bystander perspective ____ __._. - _.._ .__----.

The MD,paramedic

Some signs that indicated a good prognosis always led to start of CPR by paramedics and doctors.

Table 2 Criteria for initially

diagnosing cardiac arrest

Criterton

Number

Immediate picture of ‘lifeless’ patient The patient was extremely pale or blue and:or had no pulse andior had dilated pupils Others had already stated cardiac arrest and started CPR Patient had agonal gasps ECG-registration NC) answer

34 16

~-

-.

~.

CPR)

perspective

The reputation of the EMS-system Burden of responstbility Experiences from previous cascs Need for practice ~~~ _-~~~.. ~~~-..__

A similar and limited set of criteria was used by both the paramedics and doctors when deciding whether or not to start CPR, and deciding whether to discontinue CPR in the field or to continue CPR during transport to hospital. These criteria are therefore presented as one group. The criteria relate to four different perspectives, independent of the treatment decision made (Table 3).

~-

Expectations from bystanders (Bystander Forced to start CPR by bystanders

..-~_.-~----

The community

-..--.--~.-

-~

perspective

Social status of the paticnr Attempted \uicidc Ongoing CPR durmg transport

3.1.1.1. Initid rhythm. All stated that they always started recuscitation in cases of confirmed ventricular fibrillation. Two doctors explained that if the patient had continuous ventricular fibrillation it was hard to cease CPR efforts. Continuous or recurring ventricular fibrillation was felt to be more of a practical problem by the paramedics, becausethey felt a need for a doctor at the scene to decide whether to stop CPR. Other factors that always led to initiation of CPR were patient gasps or small apparent movements of the jaw (four cases), constricted pupils or normal looking skin color (16 cases). If the patient arrested while the paramedics were present, CPR was started in two out of three cases. The third of these was ;.Lpatient with terminal cancer.

I2

If you have had a emotionally involved without hesitation.

4 3 1 (P)

conversationwith

the in a way that makes

patient. you get you start C’PR

248

A.-C. Naess et al. / Resuscitation 33 (1997) 245-256

Table 4 Patient age and assosiation to treatment decision” Decision

N

Mean age (S.D.)

Range

Age unknown

CPR started, primarily successful CPR started-ceased in the field CPR not started

15 21 15

68 (19) 61 (20) 69 (15)

14-89 25-96 30-88

6 1

6

“One way analysis of variance, F = 1.12, P = 0.333.

3.1.1.2. Time intervals, Time intervals, either estimated collapse-to-first CPR attempt interval [l l] or known dispatch-response interval [ 1l] were mentioned during 50 interviews as influencing the decision. In 23 of the incidents, short arrest time was mentioned, although this was not neccessarily the only criterion used. In only one of these cases resuscitation was not started, because the patients wife said that she was relieved that he was allowed to die as he was very ill. Long arrest time ( > 10 min to several hours) was mentioned in 29 interviews. In 16 cases,long arrest time was a reason for not starting CPR, seven without specifying in min. Seven interviewees said they knew time passed quickly, but that it sometimes was hard to know exactly how long the patient had been without circulation. Some said that they most often started CPR immediately to save time and then asked relatives or bystanders for more information about the patient before deciding whether or not to continue CPR. If I don’t start CPR, time will run out and I can never make up for the time I have lost. m

Onk of the doctors thought that time intervals were the legitimate criteria, which everybody would have to accept. One paramedic claimed that he did not consider time intervals at all when making decisions about CPR. 3.1.1.3. Age. Age per se was not statistically related to whether or not CPR was started or discontinued, as tested by one way analysis of variance (Table 4). In 19 casespatient age was unknown at the time of treatment decision. Three interviewees found it difficult to estimate age in ctis.es‘of cardiac arrest. One thought it improper to ask relatives how old the patient was. Age wasi l&ever, the second most frequent criterion spontaneously nientioned, in 45 interviews graded as ‘very old’, ‘rather old’, ‘not so old’ or ‘rather young’. There was no .c,onsi,stencyin the definition of ‘old age’ between the intervieyees or even for the same interviewee in different situations. A 69-year-old patient was defined as old, while a 70-year-old was called ‘rather young’. In 17 of the interviews the chronologic age was

mentioned, but not used as a criterion. Twelve of these 17 patients were 69-86 years old, but not classified as old. While old age was never reported as a sufficient reason not to start or cease CPR, young age was mentioned as a reason to start CPR. Six respondents thought old age combined with knowledge about malignant disease, mental disability, severe illness or knowledge of duration of arrest a reason to ceaseCPR. Two stated that the time spent on CPR efforts without successful restoration of spontaneous circulation, tended to be inversely related to the age of the patient. To discontinue CPR in younger patients was difficult even if all attempts failed. Three respondents considered it important that all patients be given equal opportunities. One of the doctors described the gratitude of a 96-year-old patient who could return to live with his wife in their home. The paramedics sometimes received flowers and letters from patients after CPR which encouraged them to start CPR in spite of old age. If you are 96 years old and fit, old age is not a reason to abstain from CPR. (MD)

3.1.1.4. Bystander CPR. In the 24’ incidents where bystander CPR was started before the arrival of the ambulance, CPR efforts were always continued. Five respondents stressed the positive feedback to the bystanders achieved by continuing the CPR efforts and assuring them that their effort was important for a good outcome. CPR was continued regardless of whether the preceding bystander CPR seemedeffective. The most important reason for me to start CPR was that a bystander had started basic CPR. m It is a way of saying “thank you for starting”.

m

3.1.2. Ethical criteria 3.1.2.1. The patients right to live or right to die. Three respondents claimed that one criterion for starting CPR is that all people have a right to live because life is sacred.

In six cases where CPR was not started, the ambulance personnel felt that the patient had a right to die without interference. Five respondents used the right to die as a criterion to end CPR at the site, six spontaneously discussed the patient’s right to live and the right to die with dignity in general, without tying the arguments to the decision they made in the actual incident. If the patient was old and seriously ill, the respondents emphasized the right to ‘die a natural death’, ‘die peacefully’ or ‘die with dignity’. They also found it important not to treat death as an illness and explained how they imagined it to be unpleasant for the relatives if death was ‘technical’. The decision should be based on medical findings and prognosis. If the prognosis is bad. it might be better for the patient to die at home with the relatives bidding farewell. than bringing everything into the technical world of the hospital. iP)

3.2.2. Ford

to Start CPR b.v bystund~vx

Three paramedics mentioned threats from bystanding friends or relatives as a reason to start CPR. These threats came from people under the influence of alcohol or drugs, and two were underlined by a knife or a pistol. The paramedics always carried out CPR on these missions. They said that these cvperiences were frightening and a burden as they tried to do their job.

3.3.1. Tlw repzrtution of’ tlzc EM.~-.q:~ter~

Nine paramedics mentioned consideration of the reputation of the EMS-system in Oslo as an important factor, seven discussed the reputation of the EMS-system in general. This might inffuence their decision

either towards starting or not starting C’PR, depending on what they perceived to be the relatives’ expectations. It was important to retain the relatives confidence in the EMS system. It is important for the future what the relatives think of you and the EMS service. It is important chat we take care of the relatives. It is important

that 1 feel 1 have clone a

good ioh when I leave them. (P)

In 29 cases, the respondents said spontaneously that expectations expressed by the relatives or bystanders affected their decision to start CPR, and in 31 of the casesthe respondents said that the presence of relatives or bystanders did not affect their decision. They did not distinguish between verbally expressed expectations and the cry for help implicit in calling for an ambulance. If the relatives expressed hope for survival or asked the personnel to start CPR, often by crying or begging, all paramedics complied with this. The respondents sometimes found relatives as an emotional

burden in the

emergency situation. Sometimes I start CPR to give the relatives time to understand the gravity of the situation even if I’m confident that the attempt is futile. (PI

The relatives’ expectations could also influence the decision to abstain from or to cease CPR. If the relatives feared a permanent vegetative state or more suffering or if they provided information about severe illness, the paramedics would stop CPR if they inferred a bad prognosis for the patient. Sometimes the patient was transported to hospital during ongoing CPR to avoid facing relatives with the information that CPR was futile.

Paramedics and doctors said that if relatives already had accepted that the patient was ‘dead’ on their arrival, they would most likely not start CPR if the patient was elderly or severely ill.

I feel that people expect us to do something, and if you don’t do anything it will hurt the reputation of the EMSsystem. (P)

3.3.2. Burden of rcsponsibilit~~

Four paramedics said that deciding whether or not to start CPR was a great responsibility. sometimes felt as a burden. Awareness of this responsibility sometimes made them start CPR so that no one would criticize them afterwards even if they thought CPR would be futile. If the relatives discover that you are not a doctor. you can all the get into trouble. We’re working in a ‘grcy 70~ time. The relatives say: how can you decide not to start CPR when you are only paramedics’? (P) Only one of the doctors mentioned responsibility criterion. 3.3.3. Experience ,fiom pretkus

as ;I

ct~ws

In 32 interviews, the respondent referred to experience from previous incidents when they decided whether or not to start CPR. Once I was in doubt whether to star1 CPR or not. I started and the patient walked out of the hospital three days later. So if I am in doubt I a1~iay.sstart. (PI

A.-C. Naess et al. / Resuscitation 33 (1997) 245-256

250

The effect of experience was described in two different ways. As producing the ability within seconds to ‘size up’ the whole situation: I have worked as a paramedic for 23 years, and I feel that I can learn a lot just from the expression of the patient’s face.

Secondly, as being able rapidly to integrate time intervals and medical findings to come up with a probable prognosis for the patient, combined with considerations for the relatives. I have to use both reason and compassion when I decide to start CPR or not. (P> 3.3.4. Need for CPR practice

2

Both paramedics and doctors said that they sometimes needed to practice CPR or had students with them who needed practice.

Three doctors claimed that it was important to drill the ambulance team regularly, but they had not used this as the only criterion for starting CPR. 3.4. The community perspective 3.4.1. Social status of the patient

This was never mentioned as a reason for deciding whether or not to start CPR. 12 respondents said that they were aware that the social status of the patient could influence their decision and claimed that they had given much thought to the problem of a negative influence. They asserted that it did not affect their decision in actual incidents. Two said that they had made up their minds that social status should never matter. One of the paramedics said that this was a way of coping when working with patients of widely differing social status. Considerations for an equal right to life were emphasised when the patient was from a lower social class, a drug addict or an HIV-carrier. Some paramedics said that in general they were afraid of judging people by the way they lived. 3.4.2. Attempted suicide

Attempted suicide was discussed spontaneously only in a few of the interviews. Four paramedics claimed that they always started CPR even if a patient clearly had attempted suicide. They were of the opinion that attempted suicide usually was an impulsive act and a cry for help. Help should be given accordingly and CPR started. One paramedic claimed that for him it was a question of principle, he would always start resuscitation without considering circumstances or motives.

3.4.3. Ongoing CPR during transport

Five paramedics and one doctor described emergencies where the personnel decided to transport the patient with ongoing CPR. 4. Discussion

In Oslo, paramedics are allowed to start advanced life support, and to withhold or cease CPR-efforts when resuscitation is considered impossible. There are no official criteria guidelines for when to withhold or cease CPR-efforts in the field. Permitting emergency personnel to terminate resuscitation efforts in the field without hospital transport is also becoming more frequent in the United States [ll]. In published reports from 29 cities worldwide [17], 13 were reported to allow withholding or discontinuing CPR where resuscitation was considered impossible. The criteria for ‘impossible resuscitation’ were not defined. 4.1. The in-depth interview

We used in-depth interviews, as standardized questionnaires and constructed cases as a basis for the interview process can bring forward responses more easily which are thought to be socially desirable [l]. Three aspects were considered to enhance the reliablility and validity of the interviews. (1) The conceptual framework during the interview Both interviewers had good knowledge of prehospital emergency medicine in general and of the Oslo EMS-system in particular. Only two interviewers participated, as a small number of interviewers increases the reliability of the method [18]. (2) The communication process Emotional reactions expressed very frequently by the respondents such as aggression, sadness or emotion, indicated reflections at a highly personal level. (3) The process of interpretation during the interviews and during the analysis of the text The semi-structured interview made it possible to quantify the data while elaboration of the answers gave qualitative results. The processing of the interviews was made by a single individual to achieve consistent classification of the answers v91. 4.2. Criteria used for treatment decisions

In the present study the decision whether or not to start or to discontinue CPR was based on a variety of factors which theoretically could be divided into prog-

nostic and ethical criteria, though the borderline often was difficult to define. Thus bystander CPR affects outcome [20,21], but can also imply bystander expectations for a continued effort by the ambulance personnel. The patient prognosis was not evaluated systematically. The time frame is extremely short, and the information available on arrival of the ambulance team is incomplete at best. The paramedics appeared to build total pictures of each patient’s situation within seconds from the bodily status and information from bystanders about previous medical history before the final decision was made. Other criteria such as age, relatives as proxies, legal and social aspects or the reputation of the EMS system also influenced the picture. The final picture reflected the respondent’s view of the value and quality of life, the patients’ right to live and a dignified death. This decision making process is typical of experts [22,23], who perceive the situation as a whole, use past concrete situations as paradigms and move to the accurate resolution of the problem without wasteful consideration of a large number of irrelevant options. We were not able to detect any effect on the decision-making process depending on the seniority of the paramedics. This could be due to the fact that most in the group were very experienced; only two had served less than five years as a paramedic (Table I). We presume that there is a continuous learning process in the EMS-system, but the paramedic learning curve is probably at its steepest during the first few years in the service. The interview guide was not suited for detecting minor differences in the decision making process. Additionally, novice paramedics always teamed up with experienced paramedics and person to person tuition is an important aspect of developing expertise, although not all novices will become experts [22]. Only three doctors had long experience. but their description of the decision making process did not differ from their less experienced colleagues. The doctors put much more emphasis on time intervals as prognostic factors than the paramedics. They did not refer much to previous personal experience and did not describe a process of rapidly grasping the whole situation. Only the paramedics were concerned about the reputation of the EMS-system and the legal implications of their responsibility in their medical decision making, including choices about CPR. The paramedics emphasized the importance of having time for, and being able to show compassion when taking care of mourning relatives. This was not emphasized in the same way by the doctors.

Was any single criterion alone sufficient or required for a decision’? The presence of VF, contracted pupils or ‘the right to live’ alone was always sufficient foi initiating and continuing CPR. Other criteria such as old age or the right to die with dignity were never used in isolation, No single criterion was required to withhold or discontinue CPR effort in the field. Although VF always resulted in initiation of CPR. a confirmed absence of VF was not neccessary for withholding CPR. In some cases where relatives expressed a strong wish that the patient should be allowed to die and be relieved from a great suffering, the personnel refrained from connecting the ECG. In a study attempting to grade various states of ilhiess defined by degrees of disability and subjective distress, Rosser and Kind [24] found that this did not correlate with personal characteristics such as age, social class, past experience with severe illness or pain. It did correlate with the current experience of illness. There was also a good correlation between the values obtained from nurses and patients, while the correlation between doctors’ and patients’ values wds poor. When deciding whether to start or withhold CPR based upon an impression that the patient should be allowed to die and be relieved from great suffering. this should probably be based upon a strong indication that the pre-arrest state was felt as a situation worse than death by the patient, It should not be based upon a hypothesis that the post-arrest situation in survivors will be a deterioration to a new state worst than death. There is good information both from Oslo [25] and other EMS systems [26] that the great majority of patients Mho survive CPR return to their me-arrest condition. Chronic pain. coma and severe physrcal or mental dysfunction is considered a state worse than death by many [19]. When possible. this should be discussed with patients in depth to elicit their values concerning life and death, with a view to create an advance directive. Although such directives are getting more common in US hospitals, they are less common out-k)f-hospital and outside the US. None were presented in our study, and in no interview were we presented with information that the patient had discussed DNR orders or life sustaining treatment with their physician or close family members. This is not surprising. Out-of-hospital cardiac arrest is usually a sudden, unexpected event. and advance directives can therefore usually not be expected. Even in the IJS where the movement towards patient. autonomy is stronger than in Norway. only 20. 38 percent of in-hospital patients with AIDS. cancer OI other diseaseswith poor prognosis had had such discussions. according to a recent literature re\icw [l]. Only half of the patients who had decided lo forego life

252

A.-C. Naess et al. 1 Resuscitation 33 (1997) 245-2.56

sustaining treatment had informed their physician of this, and only 13 percent of patients with a high risk of cardiac arrest had discussed CPR with their physician. Life expectancy, hope for improvement, dependence upon others and enjoyment of living were important factors for patients regardless of their personal prognosis in the study of Pearlman et al. [19]. This paper suggests that there is a need for in-depth communication to establish what patients’ preferences are. The concern of patients with prognosis, death with dignity, manner and location of death or becoming permanently comatose [5] agree with some of the criteria expressed by the respondents in our study. We have no information on whether the threshold for an unacceptable state of living and ways of dying agree between patients, bystanders and EMS personnel. 4.3.1. Time intervals

The prognosis for survival after cardiac arrest is inversely proportional to the interval between the arrest to the start of CPR, to defibrillation and to the start of advanced life support [27-291. That dispatch-response intervals were usually mentioned in the interviews was not surprising as this is prominent information required in the paramedic reports and this or the call-response interval is most frequently reported [11,12,21,29]. As stated by one respondent, a long arrest time was often accepted as an objective criterion for not starting CPR. Several respondents emphasized that objective calculation of arrest time is not possible and time intervals should be interpreted with caution. They frequently found inaccuracies in bystander reported time intervals and some of them were therefore inclined to disregard bystander information. No respondent stated that he/she had an exact time limit for not starting CPR, but > 10 min dispatch-response interval was referred to as a ‘too long arrest time’ by all who made a statement on what they theoretically thought to be futile. For cases without bystander CPR this agree with information based on 1667 cases of ventricular fibrillation in King County [29]. From that study a zero survival rate can be calculated for a dispatch-response interval of > 10 min with two min added from the arrest until the ambulance is dispatched and 0.5 min from the paramedics arrival at the scene until the first defibrillation attempt. If bystander CPR was started within three min, on the other hand, survival was possible with an arrest-defibrillation interval of 18-20 min. It is important to note that time intervals were used in different ways by the respondents. Sometimes ‘only ten min’ without circulation combined with strongly expressed expectations from relatives was a good reason to start CPR. In other interviews ‘ten min or more’ were given as a reason not to start CPR when com-

bined with old age or no expressed expectations of treatment from relatives. This could be interpreted in two ways, either a true rapid composition of a total picture or a tailoring of the criteria to fit a decision already made. It was our impression that the time intervals were used in both ways by the respondents. 4.3.2. Age

Patient age was not a single criterion for decisions or not systematically related to the three treatment choices (Table 4). The lack of consistency in the definition of ‘old age’ could also reflect that the personnel placed little emphasis on this factor. They may define age in a way that fits a decision already made based on other grounds. Age is not necessarily related to the patient’s mental or physical condition. The personnel were more concerned about the patient’s everyday function than the chronological age. This is consistent with a study [30] in a county hospital, which showed that the use of DNR-orders was not correlated with age, while patients with impaired functions were more likely to have a DNR-order. Our respondents also based withholding of treatment on the considered outcome of survivors. This contrasts with findings by Charlson et al. [31], where patient age was one factor which tended to restrict intervention by physicians in patients admitted to hospital, despite a follow up study showing that age was not an independent predictor of mortality. It has been disputed whether patient age is [32-351 or is not [3537] a factor in post-CPR mortality. Reports show that the patient’s own wishes for CPR depend on information of survival chances and on patient age [38] and older people consider old age a more important factor in withholding treatment than younger people [19]. Precise patient age is often unknown in emergency situations and difficult to estimate accurately even for professionals. Several respondents said that it was unethical to label the patient as too old to receive CPR. The sudden deaths of children and young adults appear meaningless and is always hard to accept. Deciding not to start or to cease CPR of a child in the field will intensify the feeling of responsibility. This might be why youth was mentioned as a reason to start CPR regardless of factors that might indicate poor outcome. 4.4. The bystander/relative

perspective

4.4.1. Expectations from bystanders

Expectations from bystanders was considered as important. CPR started by bystanders was always continued regardless of its quality, although recent data from the Oslo EMS system suggestspoor quality CPR not to improve outcome compared with no CPR [lo]. While the time of onset or duration of bystander CPR was not mentioned as a prognostic factor in the

interviews, the ethical and social aspects were emphasized. The respondents interpreted bystander CPR as a moral imperative for them to take over and continue CPR. Both paramedics and doctors claimed that they were influenced by the presence of bystanders independent of the relationship between the bystander and the patient. The bystanders could indicate a wish for treatment or for the EMS personnel to stand back because of known illness or the patient’s right to die in peace. The call for an ambulance was in itself interpreted as a bystander cry for help. and included nurses in a nursing home treating a terminal cancer patient. In other situations it was experienced as a disclaimer of responsibility. No respondents reported that bystanders who were strangers to the patient tried to influenced them to cease or not start CPR. A wish by relatives for the EMS personnel to hold back was always based on reported knowledge of illness, or known or presumed patient wish. The EMS personnel had no time to get this information confirmed from a third party before they decided whether or not to start CPR. No respondent reported that they doubted the sincerity of relatives’ expressed wish or even their right to express a wish for the patient. The patient had apparently told the relatives that he was afr:ud ol‘ waking up \vith 21brain injury (I’) Concern for the patient and the relatives are closely interwoven when the patient is unable to communicate. ( I) The relatives become informal proxies and it is difficult to know what emphasis to place on this role in the acute setting The relatives’ knowledge of patient values and wishes might be unclear to the ambulance personnel. Seckler et al [3] found that only 16% of patients had discussed resuscitation preferences with their family, and Zweibel et al. [5] that only 10% of older out-patients had given family members specific instructions about future medical care. An advance directive should be recent since patients change their opinion of the quality of life as their own illness changes [24]. It could be argued that preferences expressed by relatives are better than nothing. Seckler et al [3] found that family members achieved a concordance with the resuscitation choices of their elder relatives, while physicians were not significantly better than chance. On the other hand, Suhl et al. 1391found that surrogate decision making regarding life support is no better than random chance. (2) Relatives might present information about the past medical history of the patient. which may be sketchy and more or less correct.

(3) The relatives may express their own wishes and concerns either explicitly or projected onto the patient. Suhl [39] and Uhlmann et al. [40] found an overestimation by surrogates and spouses of patients wishes for life sustaining treatment. This tendency can be viewed as a safeguard. Other investigators have found that families and physicians are correct in their assumptions of the patient’s preferences 52 to 90 percent of the time [2 -41 (4) The relatives might become an emotional challenge for the personnel. dependinjl on the decision made. If CPR is not started or ceased on the site. the personnel have the task of euplaining, listening and comforting It could appear that the ambulance personnel tend to follow the bystander wish for CPR, but if they felt certain that the effort would be futile. it would be only half-hearted and terminated rather rapidly. The personnel did not question the role or the motivation of the relative acting as a proxy. but automatically gave the relatives this role in the decision-making process. This was more than ‘passing the buck’. It appeared to be based upon long experience working with frightened and mourning relatives. The wishes of the relatives were considered in relation to value judgements of the personnel like *death with dignity’, ‘time to realize the loss of ;i death’ and ‘respect lilr the relative’, ‘tcchnifying mourning relative’. We be1icL.e that there still are strong reasons for family members to have a central role in the decision making process. Family members are most affected by such decisions. and as ;I cuiture wc value the integrity of the family unit.

Concern for the community and the use of scarce health care resources was never used as criteria fol deciding whether or not to start CPR. Respondents who commented on prioritizing said that it was important to them that their patient should always get a fair and respectful treatment. No respondent took other patients in need of help into their consideration. ‘The dispatch system is organized with a central distribution of resources and the single paramedic team does not have to prioritize between patients. One paramedic said that patients in general ought to be held more responsible for self-inflicted illness. He claimed that if the patients suffering was self-inflicted, he would make only a limited effort to pull the patient through.

254

A.-C. Naess et al. / Resuscitation 33 (1997) 245-2.56

4.5.2. Social status

That social status of the patient did not affect the decisions is consistent with the finding that the writing of do-not-resuscitate orders in hospital patients was not influenced by socio-economic status 130,411. The Oslo EMS treats approximately 1000 patients per year for respiratory arrest after a drug overdose, i.e. 10% of all life threatening emergencies.Thus the EMSpersonnel spend much time giving help to patients with low social status and apparently the respondents were afraid of having a bias against patients with low social status. 4.53. Ongoing CPR during transport

If CPR is started and later ceased, this will give the relatives more time, however brief, to understand the gravity of the situation while making it apparent that a real effort is being made to save the patient. This impression is probably strengthened by transporting the patient to the hospital with continued CPR if the initial effect is unsuccessful. At present this appears only to have a ‘good appearance’ effect, as the present technique for manual CPR during transport is inadequate [42] and the prognosis is dismal with 0.4%-0.6% suvival in two large studies [43,44]. With the risks associated with rapid emergency transport by ambulance [45,46], particular with ongoing CPR, it must be reasonable to recommend on-scene termination of resuscitative efforts, unless the distance is short and the EMS service is not equipped for advanced cardiac life support. In a case with children bystanders the paramedics were encouraged by the children to continue the CPR efforts. Once they had the patient in the ambulance they were so absorbed in the CPR effort that they continued during the transport. Another effect of transporting the patient to the hospital with continued CPR might be to hand over the responsibility for handling the relatives to the hospital staff. 4.5.4. Attempted suicide

Because attempted suicide was discussed at only five interviews, it was not possible to draw any conclusions concerning this factor in decision making. In the emergency situation there may be a doubt as to whether the patient had been exposed to an accident or had attempted suicide, and so CPR is started. 4.6. The paramedic/MD

perspective

4.6.1. The reputation of the EMS system

In Norway the paramedics are not yet recognized by law as a profession. Their professional identity is therefore related to the informal approval of their work, and it appeared important to them that the relatives should

be able to see that they make considered decisions based on observations at the site as well as ethical value judgements. In cases where it is clear from the onset that the effort will be futile, starting CPR can be thought of as an unnecessary and undignified procedure. Several respondents mentioned these aspects of CPR and considered alleviating stress and pain expressed by the relatives as a more important job. Holmberg et al. [47] underline the importance of protecting the in-hospital medical personell from the stress of taking part in futile resuscitation procedures and emphasize that taking part in futile procedures repeatedly is a highly demoralizing experience. 4.6.2. Burden of responsibility

Deciding whether or not to start CPR is deciding about life and death within seconds and was experienced by some of the paramedics as a great responsibility. To avoid criticism they were very aware of the signals from the relatives and this was another reason to start CPR even if they felt that it was futile. The doctors did not experience pressure from relatives as a request or an order. They probably felt that they had more authority with which to meet pressure and criticism from relatives, and therefore could be more independent in their decision making. In conclusion the paramedics in Oslo used both prognostic and ethical criteria in their CPR decisions. They had gived the ethical aspects of resuscitation much thought and used reason and compassion in each single incident. The criteria differed very little from those used by the doctors, exept that only paramedics mentioned the reputation of the EMS-system. EMS personnel who make decisions at their own discretion must have a broad base of experience. In Oslo most of the paramedics had extensive experience and a large number of episodes had made lasting impressions. These special episodes were part of the knowledge on which they based their decisions in the stressed situation of facing sudden cardiac arrest.

Acknowledgements

This study was supported by the Laerdal Foundation for Acute Medicine, the Ullevaal Hospital Research Forum and the Research Council of Norway. All the paramedics, residents and staff anesthesiologists at the Oslo EMS made this study possible and Mona Monsen provided excellent journal administration support. Special thanks to Knut Erik Tranoy, Professor Emeritus in Medical Ethics, for critical comments to earlier drafts of this article, and to paramedic Dag Moen for being a valuable liason with the paramedic group.

Interviewee data: Age: 1 Paramedics: (a) Formal education/courses. (b) Years as paramedic in the EMS system in Oslo/elsewhere. 1 Doctors: (a) Years as resident/staff in anaesthesiology. (b) Years working with the EMS system in Oslo. (c) Other relevant service. Remember: ‘Peel the onion ‘/he

General question: If there is anything like an “ordinary CPR situation’*, would you call this one ordinary? If not, what made this situation special? General question: Is there anything you would have done differently if you could do ii a11 again’? What? Why? 8 Do you usually talk about the CPR afterwards? With whom? Is it important for you to talk about the CPR afterwards?HrlJ? the interrir~l~w to bc us .vpec& us

.specjfic ;u.vk ‘what reu&.. . . ’

possible!

2.1 Can you tell me about the mission (resuscitation If they started) as you remember it right now? 2.2 Do you know if anybody had observed the patient getting ill or collapsing? 2.3 Had anybody started CPR before you got to the site, and if someone had started, who were they? 3 Did the patient have both a cardiac and a respiratory arrest? If ‘yes’, how did you decide?Checking pulse;

General question: Do you wish that anything should be altered on the cardiac arrest missions‘? What for instance? Why‘? 9 How do you find CPR compared with other work at the ambulance? Do you think it is . .. ...than . .. ... &sr inter-

Appendix A. Interview guide

ECG ~pupils,‘respirator},

@rtslother.

3.1 Can you remember why you started/did not start CPR?Check the importance qf! Age/information about illnr.rs /short intcrual since patient became ill/probable clurrrtion of wrest / rc3pon.w interwllhystander-no bJ>.ctunder C’PR cardiuc urrrst in private home - in public: c,qwtutions of’ people present (ut home, in public, man). .ywctutors, ,fim~il~~, heulth pcwonell)/sociul stutus (Solid’ c,itiren, ?soung mother.
4.2 What was in reality the most important for you there and then? Can you say something more about it? 4.3 Was there a certain finding that made you decide or was there anything else in the situation that influenced vou? 5 Di you remember how you felt? 6 If you considered the CPR as failing, did you end CPR at the site or did you transport the patient under ongoing CPR?The primary result of the CPR is categorized in the EMS journul.

7.1 Who made the decisions? Do you have a system for decision- making on the team? 7.3 Do vou _I remember what made you go on/terminate CPR?PuIyr ‘ECG “pupils infiwmationJiom in othersrclutiws

:ul[ ,firctor.v under p. 3.

7.3 What do you think was the deciding factor when you made up your mind what to do?Considerution ,for r~c~lrrtit~r~.s-h~~.stuii~l~~rs )_I teaching

uspects.

General question: How do you feel in general about decisions to terminate CPR in the field compared with transportation to the hospital under ongoing CPR? 7.4 How do you feel about the decision to cease CPR in general? Can you explain why? 7.5 How was it to make a decision in this case? 7.6 If you should compare deciding to cease CPR with deciding not to start at all. how do you find that?

rirrlws

oliw uords)

10.1 How did you experience being interviewed...... 10.2 ...and talking about the mission? 10.3 Was it anything you found especially important to mention? What? General question: Is there anything else you want to talk about that concerns your job or this interview? All text in italics was notes/checkpoints for the interviewer.

References [iI Layson R. Adelman HM. Wallach PM. Pfeifcr MP. Johnston S. McNutt RA. Discussions about the use of life-sustaining treatments: A literature review of physicians‘ and patients’ attitudes and practices. J Clin Ethics 1904: 5 (3): 195 33. VI Hare J. Pratt C. Nelson C. Agreement between patients and their self-selected surrogates on difficult medical decisions. 4rch Intern Med 1992: 1.52: 1049--1054. VI Seckler AB. Mcier DE, Mulvihill M. Pari.z BEC. Substituted judgemcnt: How accurate are proxy predictjones” Ann Interrl Med 19Y1: 115 (2): 91 ~95. [41 Starr JT. Pearlman RA. Uhlmann RF. Quahty of life and resuscitation decisionb in elderly patients. .I (irn Intern Med 1986: I: 373 -379. [51 Zweibcl NR. Cassel C‘K. Treatment choices at the end of life: a comparison of decisions by older patient> and their physiciunselected proxies. Gerontologist 1989: 29 (5): 015 621, 1’4 Bedell SE. Pelle D. Maher Pt.. Clear) PD. i)o-not-resuscitate orders for critIcally ill patients in the hospital J 1111Mc‘d Assoc 1986. 256 (7): 233 2.37. [71 Evans AL, Brody BA The do-not-rehuscitate orders m teaching hospitals. J Am Mcd Assoc 19X5; 153 ( IS): 273(> 2230 VI Lo B. Saika G. Strut1 W, Thomas I:. Showst;\ck .I Ilo not resuscitate decisions. / \ prospective srudq a: Ihrc: te;lching ho
256

A.-C. Naess et al. I Resuscitation 33 (1997) 245-256

[12] Dybvik T, Strand T, Steen PA. Buffer therapy during out-ofhospital cardiopulmonary resuscitation. Resuscitation 1995; 29: 89-95. [I31 Standards for cardiopulmonary resuscitation (CPR) and emergency cardiac care (ECC). J Am Med Assoc 1986; 255: 29055 2984. [14] Standards for cardiopulmonary resuscitation (CPR) and emergency cardiac care (ECC). J Am Med Assoc 1980; 244: 453-509. [151 Standards for cardiopulmonary resuscitation (CPR) and emergency cardiac care (ECC). J Am Med Assoc 1974; 227: 8333 868(suppl). 1161European Resuscitation Council. Guidelines for basic and advanced life support. Resuscitation 1992; 24: 103-122. u71 Eisenberg MS, Horwood BT, Cummins RO, Reynolds-Haertle R, Hearne TR. Cardiac arrest and resuscitation: a tale of 29 cities. Ann Emerg Med 1990; 19: 1799186. [I81 Steen E. Helsevesenet og forandringsprosesser. Oslo: Work Research Institute, 1993. Arbeidsforskningsinstituttets rapportserie. iI91 Pearlman RA, Cain KC, Patric DL et al. Insight pertaining to patient assessmentsof states worse than death. J Clin Ethics 1993; 4 (1): 33-41. PO1 Cobb LA, Hallstrom AP. Community-based cardiopulmonary resuscitation: What have we learned? In: Cardiopulmonary Resuscitation. New York: New York Academy of Sciences, 1982; 330-342. Pll Cummins RO, Eisenberg MS, Hallstrom AP. Survival of Out-ofhospital cardiac arrest with early initiation of cardiopulmonary resuscitation. Am J Emerg Med 1985; 3 (2): 1144119. P21 Benner P. From novice to expert. Exellence and power in clinical nursing practice, 1st ed. New York: Addison-Wesley, 1984. u31 Dreyfus SE, Dreyfus HL. The scope, limits, and training implications of the three models of aircraft pilots emergency response behavior. University of California at Berkeley, 1979. ~241 Rosser R, Kind P. A scale of valuations of states of illness: is there a social consensus?Int J Epidemiol 1978; 7 (4): 347-358. v51 Forsman M, Aarseth HP, Nordby HK, Skulberg A, Steen PA. Effects of Nimodipine on cerebral blood flow and cerebrospinal fluid pressure after cardiac arrest: Correlation with neurologic outcome. Anesth Analg 1989; 68: 436-43. WI Edgren E, Hedstrand U, Kelsey S, Sutton-Tyrrell K, Safar P. Assessment of neurological prognosis in comatose survivors of cardiac arrest. BRCT I study group. Lancet 1994; 343: 1055-9. v71 Mullie A, van Hoeyweghen R, Quets A, group Tcrs. Influence of time intervals on outcome of CPR. Resuscitation 1989; 17 Suppl.: S23-S33. WI Roth R, Stewart RD, Rogers K, Cannon GM. Out-of-hospital cardiac arrest: factors assiciated with survival. Ann Emerg Med 1984; 13: 237-243. t291 Larsen M, Eisenberg M, Cummins R, Hallstrom A. Predicting survival from cardiac arrest: a graphic model. Ann Emerg Med 1993; 22: 1652-8.

[30] Levy MR, Lambe ME, Shear CL. Do-not-resuscitate orders in a county hospital. West J Med 1984; 140 (1): 111-t 13. [31] Charlson ME, Sax FL, MacKenzie R, Fields SD, Braham RL, Douglas G. Resuscitation: how do we decide? J Am Med Assoc 1986; 255 (IO): 131661322. [32] Herlitz J, Ekstram L, Wennerblom B, Axelsson A, Bang A, Nolmberg S. Hospital mortality after out-of-hospital cardiac arrest among patients found in ventricular fibrillation. Resuscitation 1995; 29: 11-21. [33] Hoe1 TM, Steen PA. Legeambulansen i Oslo. Tidsskr Nor Laegeforen 1985; 105 (27): 1860-2. [34] Rogove HJ, Safar P, Sutton-Tyrrell K, Abramson NS, group TbrctIaIs. Old age does not negate good cerebral outcome after cardiopulmonary resuscitation: analysis from the brain resuscitation clinical trials. Crit Care Med 1995; 23 (1): 18-25. [35] Tresch DD, Thakur RK, Hoffmann RG, Olson D, L BH. Should the elderly be resuscitated following out-of-hospital cardiac arrest? Am J Med 1989; 86: 145-150. [36] Mullie A, Lewi P, van Hoeyweghen R, group Tcrs. PreCPR-condidtions and the final outcome of CPR. Resuscitation 1989; 17 Suppl.: sll-s21. [37] Longstreth WT, Jr., Cobb LA, Fahrenbruch CE, Copass MK. Does age affect outcomes of out-of-hospital cardiopulmonary resuscitation? J Am Med Assoc 1990; 264 (16): 2109-2110. [38] Murphy DJ, Burrows D, Santilli S et al. The influence of the probability of survival on patients’ preferences regarding Cardiopulmonary resuscitation. N Engl J Med 1994; 330 (8): 545549. [39] Suhl J, Simons P, Reedy T, Garrick T. Myth of susbstituted judgement. Arch Intern Med 1994; 154: 90-96. [40] Uhlmann RF, Pearlman RA, Cain KC. Physicians’ and spouses’ predictions of elderly patients resuscitation preferences. J Geronto1 1988; 43 (5): Ml155121. [41] Youngner SJ, Lewandowski W, McClish DK, Juknialis BW, Coulton C, Barlett ET. The do-not-resuscitate orders. Incidence and implikations in a medical intensive care unit. J Am Med Assoc 1985; 253 (I): 54-57. [42] Stapelton ER. Comparing CPR during ambulance transport. J Emerg Med Services 1991; Sep.: 63-72. [43] Kellermann AL, Hackman BB, Somes G. Predicting the outcome of unsuccessfulprehospital advanced cardiac life support. J Am Med Assoc 1993; 270 (12): 143331436. [44] Bonnin MJ, Pepe PE, Kimball KT, Clark PS, Jr. Distinct criteria for termination of resuscitation in the out-of-hospital setting. J Am Med Assoc 1993; 270 (12): 145771462. [45] Auerbach PS, Morris JA, Phillips JB, Jr., Redinger SR, Vaughn WK. An analysis of ambulance accidents in Tennessee. J Am Med Assoc 1987; 258 (11): 148771490. [46] Gray WA. Prehospital resuscitation. The good, the bad, and the futile. J Am Med Assoc 1993; 270 (12): 1471-1472. [47] Holmberg S, Ekstrsm L. Ethics and practicalities of resuscitation. Resuscitation 1992; 24: 239-244.