- Email: [email protected]
Decision making to cease or to continue cardiopulmonary resuscitation (CPR)
s137
Resuscitation, 17, Suppl. (1989) S137--5147 Elsevier Scientific Publishers Ireland Ltd.
DECISION MAKING CARDIOPULMONARY
TO CEASE OR TO CONTINUE RESUSCITATION (CPR)
R. VAN HOEYWEGHENa, A. MULLIEb, RESUSCITATION STUDY GROUP
L. BOSSAERT”
and THE CEREBRAL
“Department of Intensive Care and Emergency Medicine, University Antwerp - UIA, Universiteitsplein I,2610 Antwerpen and bDepartment of Critical Care Medicine, Algemeen Ziekenhuis Sint Jan, Ruddershove 10,800O Brugge (Belgium)
SUMMARY
CPR should be initiated in any patient who has a cardiac arrest. This might improve overall outcome but implies that CPR is started in patients without any virtual chance for long-term survival (LTS). The aim of this study is, by analysing retrospectively 2713 out-of-hospital cardiac arrests (CA), to identify indices which might be of help in the decision making to continue or to discontinue CPR. In an important number of unsuccessful CPR attempts ALS-time did not exceed 20 min. This occurred more frequently in subgroups where limited chances of LTS are expected on clinical grounds. The decision to cease CPR might have been based on other clinical and/or ethical parameters which were not recorded in the registry. This behavior results in a “self-fulfilling prophecy”. A subset of patients with limited chances for LTS (014051can be identified: patients in electromechanical dissociation (EMD) or asystole on arrival of the mobile intensive care unit (MICU) team, without pupil reaction to light during CPR and with inefficient cardiac massage by the MICU (405/27131.Other patients in EMD or asystole without pupil reaction to light during CPR (1373/2713) but with efficient ECC should be resuscitated for more than 30 min, especially if the patient is gasping during CPR (LTS 27/1373). Patients in EMD or asystole on arrival of the MICU with pupil reaction to light during CPR (236/2713) should have an ALS-time of at least 45 min (LTS 42/2361. Cardiac arrests in ventricular fibrillation (VF) (699/2713) should be resuscitated for at least 45 min, especially when gasping during CPR (LTS 119/699).
Key words: Cardiopulmonary resuscitation - Outcome - Prognosis Address for correspondence: L. Bossaert, Department of Intensive Care and Emergency University Antwerp - UIA. Universiteitsplein 1, B 2610 Antwerpen (Belgium).
0300-9572/89/$03.50 0 1989 Elsevier Printed and Published in Ireland
Scientific
Publishers
Ireland
Ltd.
Medicine,
S138 INTRODUCTION
The initial aim of CPR was to resuscitate the sudden cardiac death victim [l]. At first CPR was limited to the critical care departments of the in-hospital setting where it was performed by skilled physicians. Therefore, the first outcome studies of CPR by Kouwenhoven et al. had a high success rate [2]. From the early sixties on, more people were trained in the techniques and skills of CPR (physicians, paramedics, emergency medical technicians, laymen) and more patients were involved. At present it is recommended that CPR should be initiated in almost any patient suffering a cardiac arrest, except for patients with a known DNR-code (do not reanimate). This widespread use of CPR might improve the chances of initial response to CPR and ensure good neurologic function in patients who did respond [3]. However, the price one has to pay for the rescue of extra lifes without brain damage is that CPR is started in patients who have virtually no chance for long-term survival. This report will not discuss whether to start CPR or not. The aim is to find out if rules can be formulated that are helpful in the decision making to continue or discontinue CPR by studying retrospective 2713 out-of-hospital cardiac arrests. Only limited literature data [4-71 are available on this subject. Most authors focus on save rates and survival, and do not mention experiences with CPR failure. Although, this is the most frequent outcome of all CPR attempts. METHODS
The registration method has been described elsewhere [8]. Several variables related to the pre-arrest, arrest, CPR and post-CPR period have been registered prospectively on a uniform registration form, in all cardiac arrest events attended by the mobile intensive care unit (MICU) teams of the participating emergency medical service (EMS) systems. Outcome was subdivided in 4 classes: Class 1 = immediate CPR failure, i.e. spontaneous circulation can not be restarted; Class 2A = CPCR failure due to cerebral failure, i.e. patients whose circulation can be restored but die within 14 days from cerebral failure or remain in a vegetative state 14 days post-CPR; Class 2B = CPCR failure and cardiac, septic or other failure, i.e. patients whose circulation can be restored but die within 14 days from non-cerebral failure; Class 3 = CPCR successes, i.e. patients awake 14 days post-CPR either normal (good recovery), moderately disabled (disabled but independent) or severely disabled (dependent on others)
PI.
Between January 1st 1983 and December 31st, 1987, the MICU teams of the participating centers registered 3083 out-of-hospital cardiac arrests. Only patients with complete files on predictive parameters were included in the analysis (type of cardiac arrest on arrival of the MICU, type of respiratory arrest during CPR by the MICU, quality of pupil reaction during CPR by the MICU, efficiency of cardiac massage by the MICU and duration of advanced life support = ALS). Type of cardiac arrest on arrival of the MICU was classified in two groups:
5139
ventricular fibrillation (VFl or no VF (electromechanical dissociation = EMD and asystole). Type of respiratory arrest (RAl was either complete respiratory arrest, or gasping occurring at any time during CPR after the arrival of the MICU. Pupil reaction during CPR was described as negative pupil reaction when pupils were dilated, pinpoint or inequal without reaction to light and as positive pupil reaction when normal or unequal pupils did react to light (slow or normal). The quality of CPR by the MICU was scored as efficient (Ql = correct technique with good effect1 or as inefficient (Q2 = correct technique with weak effect; Q3 = bad technique) according to the quality of the palpated carotid or femoral pulse during external chest compression (ECC). The registration form was filled out by a member of the MICU team immediately after the CPR attempt. ALS-time (minutes) is the duration of the reanimation by the MICU only. Although no strict written treatment protocol was used, the methods and techniques of CPR, used by the participating centers was comparable and was based on the generally accepted “Standards and Guidelines for CPR and Emergency Cardiac Care” as published by the American Heart Association (AHA) [lo] and by P. Safar [ll]. No specific recommendations were used for the minimum duration of ALS. The physician of the MICU decided when to cease CPR based on personal knowledge and clinical experience. Based on analysis of the above mentioned parameters of the patients studied, we were able to identify a number of predictors of CPR failure and CPR success, which might be of help in the decision making of the emergency team during their CPR attempt. RESULTS
During the registration period data on 3083 cardiac arrest events were recorded. Complete files on prognostic indices were available on 2713 patients (missing: n = 3701.In Table I, patients are classified according to 4 parameters (type of cardiac arrest on arrival of the MICU, type of respiratory arrest during CPR, pupil reaction to light during CPR, quality of external chest compresssion by the MICU). The prevalence of ventricular fibrillation was 26% (699/27131.Of these patients 32% (221/6991were gasping during CPR. Of these 221 gasping patients 116 (55%) had positive pupil reaction to light during CPR. Patients in VF in complete respiratory arrest during CPR (478/699 = 68%) still had pupil reaction to light during CPR in 26%1 of cases (122/4781.Of the 2713 studied patients, 2014 (74%~)were in electrochemical dissociation or asystole on arrival of the MICU. Twelve percent (245/20141of these patients were gasping during CPR of which 36% (87/2451had pupil reaction to light during CPR. The majority of the patients in EMD or asystole was in complete respiratory arrest (1769/ 2014 = 88%). Only 8O/b(149/17691of the patients in this setting had pupil reaction to light during CPR. It was analysed whether these parameters could be used for prediction of long-term survival. Prediction of long-term survival (LTS = class 31using these predictive parameters is summarised in Table II. The best predictor of LTS was the presence of positive pupil reaction to light during CPR. Of the patients with positive
5140 TABLE I CLASSIFICATION OF PATIENTS ACCORDING TO PREDICTIVE PARAMETERS VF = ventricular fibrillation on arrival of the MICU; EMD = electromechanical dissociation on arrival of the MICU; asyst = asystole on arrival of the MICU; RA = complete respiratory arrest during CPR; gasp = gasping during CPR; pupil + = positive pupil reaction to light during CPR; pupil - = negative pupil reaction to light during CPR. Number of patients between brackets. Inefficient ECC by the MICU pupil + (1161 vF(w91/gasping~2211~pupil\RA(4,81
/pupil+ -pupil gasping (2451-----1
EMD/ASYST (20141 LRA
(1769)/pupil 1
(61
(1051
(191
(1221
(41
- (3561 pupi1+ (871 pupil - (1581 + (1221 pupil - (16201
66) (71 (381 (41 (3671
TABLE II PREDICTION OF LONG-TERM SURVIVAL BY TYPE CARDIAC ARREST ON ARRIVAL OF THE MICU, TYPE RESPIRATORY ARREST DURING CPR, PUPIL REACTION TO LIGHT DURING CPR AND EFFICIENCY OF EXTERNAL CHEST COMPRESSION BY THE MICU Number
% long-term survivors
Type of cardiac arresk VF EMDlasystole
699 2014
17.0 3.5
Type of respiratory crrest: Gasping Complete
2247 466
4.5 18.9
Pupil reaction duting CPR: Present Absent
474 2239
23.2 3.5
Quality of ECC: Efficient Inefficient
2202 511
8.4 0.8
Missing data Total
370 3083
7.8 7.1
5141
pupil reaction (n = 474123% ended in class 3. Of the patients gasping during CPR (n = 4661 19% were classified in class 3. Patients in EMD or asystole on arrival of the MICU had a bad prognosis: long-term survial was 3.5Oh (n = 20141. Of the patients presenting VF on arrival of the MICU, 17% were longterm survivors (n = 699). Poor CPR technique or inefficient external chest compression by the MICU is a predictor of unsuccessful outcome. Only 4 out of the subset of 511 patients prere long-term survivors. However, efficient cardiac massage was not a good predictor of survival (LTS: 8.4Ok11. In order to predict long-term survival, the combined influence of the studied parameters on survival was analysed. Patients were classified in a data-matrix (Tables 1111according to these 4 parameters and ALS-time. Most patients (16201 2714 = 60%) belong to the group with bad prognostic parameters (EMDlasystole, complete respiratory arrest and absent pupil reaction during CPR)
TABLE III NUMBER OF LONG-TERM SURVIVORS OF PATIENTS WHOSE CPR EFFORT IS NOT YET CEASED AFTER A CERTAIN DURATION OF ADVANCED LIFE SUPPORT, ACCORDING TO PREDICTIVE PARAMETERS VF = ventricular fibrillation on arrival of the MICU; EMD = electromechanical dissociation on arrival of the MICU; asyst = asystole on arrival of the MICU; RA = complete respiratory arrest during CPR; pupil + = positive pupil reaction to light during CPR; pupil - = negative pupil reaction to light during CPR; 81 = efficient external chest compression by the MICU; &2,3 = inefficient external chest compression by the MICU; ALS-time = duration of advanced life support. No. of patients in parentheses. VF + RA (478)
VF + Gasping@211 ALStime (min) >20 >30 >45 >60
Pupil + (116)
Pupil - (105)
61
62,s
61
331110 8166 4145 2127 l/13
l/6 l/5 l/3 O/l 0
21186 5145 4124 4112 O/l
EMDlAsyst ALS time (min) >20 >30 >45 >60
Pupil+
Pupil + (122)
Pupil-
6293
61
Q&3
Ql
QZ3
2119 l/14 O/9 O/5 013
34/118 9157 5143 l/18 l/7
114 l/3 112 O/l O/l
271300 6/175 31101 0143 o/19
O/56 0143 0123 o/7 O/l
+ Gasping (245)
(87)
EMDlAsyst
Pupil - (158)
(356)
+ RA (1769)
Pupil + (149)
Pupil - (1620)
Ql
Q2,3
Q1
82,s
Ql
6293
Ql
Q2,3
19180 6148 3130 O/16 018
017 016 o/3 0 0
12/120 3172 2145 2130 0111
O/38 on1 o/10 016 O/l
221135 3183 l/56 O/24 0113
o/14 0110 016 O/l 0
17/1253 51712 l/353 o/140 0153
01367 o/177 O/83 0134 o/9
S142
(Table III). In this group of patients overall long-term success rate was 1% (171 16201.Long-term survivors (n = 171had an ALS-time which did not exceed 45 min. In this subset of patients the incidence of inefficient external chest compression by the MICU was 23O/o (367/16201. No long-term survivors were recorded in this subgroup of patients (013671. Of the patients in EMD or asystole on arrival of the MICU and in complete respiratory arrest during CPR, 8% (149/16201had positive pupil reaction to light during CPR. Long-term survival was significantly higher than in patients without pupil reaction to light during CPR: 180/b(22/149) vs. 1% (17/16201.In this group, only one out of the 22 long-term survivors had an ALS-time exceeding 30 min. Patients in EMD or asystole on arrival of the MICU, still gasping and with positive pupil reaction to light during CPR, had long-term survival of 22% (19/871.In case of absent pupil reaction chances for long-term survival were reduced: 8% (12/1581.In this group of patients (EMDlasystole + gasping) 7 of the 31 long-term survivors had an ALS-time of more than 30 min. Of the 2713 studied patients, 699 (26%) were still in VF on arrival of the MICU. In 85 chest compression was considered as inefficient. Of the patients in VF, 51% (356/6991were in complete respiratory arrest and did not have pupil reaction to light during CPR (Table III). In this group, 8% (27/3561were longterm survivors who were all resuscitated within 35 min. Of the 478 patients in VF on arrival of the MICU and in complete respiratory arrest, 122 (26%) had pupil reaction to light. Long-term survival was 290/6 (35/1221.One long-term survivor had an ALS-time of more than 45 min. The highest rate of long-term survivors (29% = 3411161was found in patients in VF on arrival of the MICU, with gasping and pupil reaction to light during CPR. This subset represents 17%1 (11016991of the patients in VF (Table 1111.Two survivors had an ALS-time of more than 45 min. One of these patients was reanimated during more than 1 h. Patients in VF, not gasping and without pupil reaction to light during CPR (105/699 = 150/o),had long-term survival in 22% of cases (23/1051. All long-term survivors were reanimated within 1 h. The same data-matrix of ALS-time and of the studied parameters was also used for prediction of CPR failure. Table IV presents the numbers of CPR failure (class 11in the patients where resuscitation was ceased after a certain ALS time (620 min, 21- 30 min, 31-45 min, 46-60 mini. In 44 cardiac arrest patients in VF on arrival of the MICU, still gasping, with positive pupil reaction to light during CPR and with efficient ECC, the reanimation was discontinued after an ALS-time of less or equal than 20 min. In 8 of these 44 patients the CPR attempt was unsuccessful (class 11.The remaining patients (36/441were either short- or long-term survivors. In another 21 cases, ALS was ceased after 21- 30 min. Fourteen of these were class 1 patients. In an important number of patients in EMD or asystole on arrival of the MICU, in complete respiratory arrest and without pupil reaction to light during CPR, it was noted that the CPR attempts were ceased before 21 min: 43% (541/ 1253) if ECC was efficient and 52% (19013671if ECC was inefficient. Most of these CPR attempts were unsuccessful (in class 1): 780/b (424/5411if ECC was efficient and 97% (185/1901if ECC was inefficient.
s143 DISCUSSION
The selection of the studied parameters (type cardiac arrest on arrival of the MICU, type respiratory arrest during CPR, quality of the pupil reaction to light during CPR, efficiency of external chest compression by the MICU and ALStime) is based on current clinical practice and experience. Prognostic indices should be available in any patient, should be easy to determine and should preferably be dichotomous with clear distinction in outcome. Literature presents more parameters for decision making in CPR: e.g. underlying disease, age, response time, witnessed or not, bystander CPR, etc [4 - 71.These parameters might not be correctly assessable and are not always available at the time of resuscitation. Therefore we preferred not to select a subgroup of cardiac arrest patients of cardiac origin for this analysis. The type of cardiac arrest on arrival of the MICU is widely accepted as an important predictor for outcome [5-7,12- 141.Electromechanical dissociation and asystole have a similar outcome which is significantly worse than the outcome of ventricular fibrillation. The type of respiratory arrest is not as widely used as a clinical predictor but a strong correlation with survival has been demonstrated in our study population. Furthermore, gasping is easy to determine. Quality of pupils and pupil reaction to light is also generally accepted as a good predictor for neurologic status and outcome [15- 171.In our study population it seemed to be the best predictor for long-term survival. As mentioned by some authors [18], the only time interval which can be estimated accurately is the duration of the reanimation period by the MICU. ALStime has already been used as a predictive parameter [4,5,13,14]. In our registry, inefficient external chest compression revealed to be a predictor of bad outcome. During the 5-year registration period only 4 out of 511 patients (1%) with inefficient ECC were long-term survivors. Registration of efficiency of external chest compression by the MICU might be biased by the fact that a member of the MICU team is filling out the registration form, at a time when the resuscitation efforts have finished. As such, the evaluation is not objective and immediate outcome of the CPR attempt is known. Because of these reasons and because of the small number of patients in VF on arrival of the MICU with inefficient external chest compression by the MICU, this parameter is only used for analysis of prognosis of patients in EMD or asystole. Table IV illustrates that in many cases the CPR effort is ceased rather early. Even in the subgroup with the highest potential success rate (VF, gasping, pupil reaction, ECC Qll ALS was ceased after 20 min or less in 7% (8/110) of cases. In 52% (190/3671of the subgroup without long-term survivors (EMD/ asyst, RA, no pupil reaction, ECC Q2 or Q31the ALS-time did not exceed 20 mon. Only 5 out of these 190 patients were successful resuscitations (class 2A and 2Bl. These unsuccesful CPR attempts might not have received an optimal treatment because CPR was maybe ceased prematurely. In this way a self-fulfilling prophency is created. Due to the early cessation cardiac arrests with bad prognostic parameters do not have the opportunity to demonstrate that longterm survival is possible. These observations might suggest that inefficient
s144 TABLE IV NUMBER OF PATIENTS WITH UNSUCCESSFUL CPR OUTCOME WHOSE RESUSCITATION WAS CEASED AFTER A CERTAIN DURATION OF ADVANCED LIFE SUPPORT, ACCORDING TO PREDICTIVE PARAMETERS VF = ventricular fibrillation; RA = complete respiratory arrest during CPR; pupil + = positive pupil reaction to light during CPR; pupil - = negative pupil reaction to light during CPR; Ql = efficient external chest compression by the MICU; Q2,3 = inefficient external chest compression by the MICU; ALS-time = duration of advanced life support. VF + Gasping(221) ALStime (min) *20 20-30 31-45 46-60
Pupil + (116)
Pupil + (122)
Pupil - (356)
62,s
62,s
(110)
&2,3
Q1
Q2,a
(6)
(19)
(4)
(300)
(56)
a/44 14121 12/18 13/14
l/l l/2 l/2 l/l
8141 17/21 11/12 8/12
l/l 313 313
16132 16127 14115 17/19
7161 4114 16125 6/11
l/l l/l l/l
EMD/Asyst
Pupil - &?I??)
6293 (7) 10132 12/18 9/14 818
415 415 314 2/2
+ Gasping (245)
Pupil + (87)
(min) 620 21-30 31-45 46-60
Pupil - (105)
Ql
EMD/Asyst ALS time
VF + RA (4’78)
70/115 62174 44158 24124
11113 20120 15/16 6/6
+ RA (1769)
Pupil + (149)
Pupil - (1620)
Q&s
6293
&2,3
(38)
(14)
(367)
17117 ll/ll 414 515
11/52 22129 26132 8/11
314 314 515 l/l
4241541 3281358 197/213 80187
185/196 94194 49149 24125
external chest compression could be considered as motive to stope CPR instead of an incentive to intensify the CPR efforts. It is most likely that in these circumstances the decision to stop CPR was based on other relevant conditions which were not recorded in this registry. For prediction of bad outcome, only patients in electromechanical dissociation and asystole on arrival of the MICU (Table Vl are taken into consideration. Positive pupil reaction during CPR is found in 236/2014 patients and negative pupil reaction in 177812014patients. Long-term survival is 17% (411 236) in patients with positive pupil reaction vs. 2Orb(29117781in patients with negative pupils. Considering patients without pupil reaction, the chances for long-term survival are extremely limited (O/4051if external chest compression by the MICU is inefficient. When external chest compression is efficient CPR success can be obtained even after 30 min of ALS-time: 3 out of 29 long-term survivors in this group required more than 30 min of ALS to be successfully resuscitated. Two of these patients were gasping and both had an ALS-time of 46 min. A third patient suffered a cardiac arrest due to intoxication, was in complete respiratory arrest and had an ALS-time of 45 min. Considering
s145 TABLE V PROGNOSIS OF BAD OUTCOME EMD = electromechanical dissociation on arrival of the MICU; asyst = asystole on arrival of the MICU; pupil + = positive pupil reaction to light during CPR; RA = complete respiratory arrest during CPR; gasp = gasping during CPR; Ql = efficient external chest compression by the MICU; Q2,3 = inefficient external chest compression by the MICU; ALS-time = duration of advanced life support. No. of patients in parentheses. pupi/ST
p”pil_
(4112361
(29/17781 Ql A\Q23 (29113731
I
(3/4911
ALS-time: > 30
1
GASP’ (21551
RA (114361
I ALS-time: > 45
(014051
I (O/1741
(21361
(O/411
patients in EMD or asystole with positive pupil reaction to light during CPR, ALS-time should be continued to at least 30 min: 4 of the 41 long-term survivors required an ALS-time of more than 30 min. All long-term survivors were resuscitated within 45 min. For prediction of good outcome, the subset of patients in ventricular fibrillation on arrival of the MICU is studied (Table VI). Patients in VF have good chances for long-term survival if pupil reaction to light is present during TABLE VI PROGNOSIS OF GOOD OUTCOME VF = ventricular fibrillation on arrival of the MICU; pupil + = positive pupil reaction to light during CPR; pupil - = negative pupil reaction to light during CPR: RA = complete respiratory arrest during CPR; gasp = gasping during CPR; ALS-time = duration of advanced life support. VF (119/6991 Pupil+ ’ (69/238)
\Pupil(50/461)
I
I
(4/671
(31471
ALS-time: > 45 GAS< (2/281
\RA (l/191
GASP f4/171
/l
RA (01501
5146
CPR (69/238 = 29%). Most of these successful resuscitations occur within 45 min (66/691ALS-time. In only 3 patients ALS-time exceeded 45 min. Two of these patients were gasping and had an ALS-time of 60 and 65 min. The third patient was not gasping. It concerned a 14-year-old girl that drowned and was found in profound hypothermia (24OCl.Complete ALS-time was more than 3 h (1 h artificial circulation). Patients in VF on arrival of the MICU, without pupil reaction to light during CPR still have a rather good prognosis (LTS: 50/461 = 11%). In 4 patients out of the 50 long-term survivors, an ALS-time of 60 min was required. These 4 patients were gasping during CPR. CONCLUSION
In our patient material (n = 27131a subset of 405 patients (15%) can be identified with no long-term survival. These patients were characterised by: EMD or asystole on arrival of the MICU, no pupil reaction to light during CPR, inefficient external chest compression. The question is whether this poor outcome is the result of a self-fulfilling prophecy or the result of a decision making process to stop CPR based on other clinical and ethical parameters that were not studied in this registry. Patients in EMD or asystole without pupil reaction to light during CPR but with efficient ECC (510/6of all patients, 1373/2713) still have a long-term survival of 2%. Therefore, the resuscitative effort should be continued for more than 0.5 h, especially if the patient is gasping during CPR. Patients in EMD or asystole with pupil reaction to light during CPR (90/oof all patients, 236/27131 have a long-term survival of 18%. In this subset of patients ALS should not be stoppped before at least 45 min. In cardiac arrest patients in VF on arrival of the MICU (26% of all patients, 699/27131,pupil reactions to light during CPR are indicative for long-term survival: if pupil reaction is positive, LTS is 28%, if negative only 11%. In this subset of patients ALS should be continued for at least 45 min. This analysis indicates that during the initial phase of resuscitations it is not possible to identify reliable indicators for CPR failure. The studied parameters are only helpful elements in the decision making to continue the CPR efforts. REFERENCES 1 2 3 4
5
J.H. Talbott, Introduction. In: Fundamentals of Cardiopulmonary Resuscitation. J.R. Jude and J.O. Elam. F.A. Davis, Philadelphia, 1965. W.B. Kouwenhoven, J.R. Jude and G.G. Knickerbocker, Closed-chest cardiac massage, J. Am. Med. Assoc., 173 (1960) 1064- 1067. L.J. Blackhall. Must we always use CPR? N. Engl. J. Med., 317 (1987) 1281- 1285. R.S. Pionkowski, B.M. Thompson, H.W. Gruchow, C. Aprahamian and J.C. Darin, Resuscitation time in ventricular fibrillation - a prognostic indicator. Ann. Emerg. Med., 12 (1983) 733738. J.P. Smith and B.I. Bodai, Guidelines for discontinuing prehospital CPR in the emergency department - a review. Ann. Emerg. Med., 14 (1985) 1093- 1098.
s147 6
7 8 9 10 11
12 13
14
15 16 17 18
C. Aprahamian, B.M. Thompson, H.W. Gruchow, J.R. Mateer, J.F. Tucker, H.A. Stueven and J.C. Darin, Decision making in prehospital sudden cardiac arrest, Ann. Emerg. Med., 15 (19861 445-449. M. Eisenberg, A. Hallstrom and L. Bergner, The ACLS score. Predicting survival from out-ofhospital cardiac arrest, J. Am. Med. Assoc., 246 (1981150- 52. The Cerebral Resuscitation Study Group. The Belgian Cardiopulmonary Cerebral Resuscitation Registry: Form Protocol, Resuscitation, 17 (Suppl.l(l9891 S5- SlO. B. Jennett and M. Bond, Assessment of outcome after severe brain damage: a practical scale, The Lancet, March l(1975) 480 - 484. Standards and Guidelines for CPR and ECC, J. Am. Med. Assoc., 244 (1980)453-509. P. Safar, Cardiopulmonary Cerebral Resuscitation. Prepared for the World Federation of Societies of Anaesthesiologits, 2nd edn.. A. Stavanger and Laerdal, (Eds.1,WB Saunders. London Philadelphia Toronto, 1988. R. Roth, R.D. Stewart, K. Rogers and G.M. Cannon, Out-of-hospitalcardiac arrest: factors associated with survival, Ann. Emerg. Med., 13 (19841237-243. M. Eisenberg, L. Bergner and A. Hallstrom, Paramedic programs and out-of-hospital cardiac arrest: I. Factors associated with successful resuscitation, Am. J. Public Health, 69 (197911049 - 1051. R.J. Myerburg, C.A. Conde, R.J. Sung, A. Mayorga-Cortes, S.M. Mallon, D.S. Sheps, R.A. Appel and A. Castellanos, Clinical, electrophysiologic and hemodynamic profile of patients resuscitated from prehospital cardiac arrest, Am. J. Med., 68 (1980)568-576. P. Sobotka, and E. Gebert, Effect of complete brain ischemia on pupillary changes, Acta Anaesth. Stand., 16 (19721112-116. B.D. Snyder, M. Ramirez-Lassepas and D.M. Lippert, Neurologie status and prognosis after cardiopulmonary arrest: I. A restrospeetive study, Neurology, 27 (1977)807 - 811. J.E. Steen-Hansen, N.N. Hansen, P. Vaagenes and B. Schreiner, Pupil size and light reactivity during cardiopulmonary resuscitation: a clinical study, Crit. Care Med., 16 (1988)69-70. B.M. Thompson, H.A. Stueven, J.R. Mateer, C.C. Aprahamian, J.F. Tucker and J.C. Darin, Comparison of clinical CPR studies in Milwaukee and elsewhere in the United States, Ann. Emerg. Med., 14 (19851750- 754.
Resuscitation, 17, Suppl. (1989) S137--5147 Elsevier Scientific Publishers Ireland Ltd.
DECISION MAKING CARDIOPULMONARY
TO CEASE OR TO CONTINUE RESUSCITATION (CPR)
R. VAN HOEYWEGHENa, A. MULLIEb, RESUSCITATION STUDY GROUP
L. BOSSAERT”
and THE CEREBRAL
“Department of Intensive Care and Emergency Medicine, University Antwerp - UIA, Universiteitsplein I,2610 Antwerpen and bDepartment of Critical Care Medicine, Algemeen Ziekenhuis Sint Jan, Ruddershove 10,800O Brugge (Belgium)
SUMMARY
CPR should be initiated in any patient who has a cardiac arrest. This might improve overall outcome but implies that CPR is started in patients without any virtual chance for long-term survival (LTS). The aim of this study is, by analysing retrospectively 2713 out-of-hospital cardiac arrests (CA), to identify indices which might be of help in the decision making to continue or to discontinue CPR. In an important number of unsuccessful CPR attempts ALS-time did not exceed 20 min. This occurred more frequently in subgroups where limited chances of LTS are expected on clinical grounds. The decision to cease CPR might have been based on other clinical and/or ethical parameters which were not recorded in the registry. This behavior results in a “self-fulfilling prophecy”. A subset of patients with limited chances for LTS (014051can be identified: patients in electromechanical dissociation (EMD) or asystole on arrival of the mobile intensive care unit (MICU) team, without pupil reaction to light during CPR and with inefficient cardiac massage by the MICU (405/27131.Other patients in EMD or asystole without pupil reaction to light during CPR (1373/2713) but with efficient ECC should be resuscitated for more than 30 min, especially if the patient is gasping during CPR (LTS 27/1373). Patients in EMD or asystole on arrival of the MICU with pupil reaction to light during CPR (236/2713) should have an ALS-time of at least 45 min (LTS 42/2361. Cardiac arrests in ventricular fibrillation (VF) (699/2713) should be resuscitated for at least 45 min, especially when gasping during CPR (LTS 119/699).
Key words: Cardiopulmonary resuscitation - Outcome - Prognosis Address for correspondence: L. Bossaert, Department of Intensive Care and Emergency University Antwerp - UIA. Universiteitsplein 1, B 2610 Antwerpen (Belgium).
0300-9572/89/$03.50 0 1989 Elsevier Printed and Published in Ireland
Scientific
Publishers
Ireland
Ltd.
Medicine,
S138 INTRODUCTION
The initial aim of CPR was to resuscitate the sudden cardiac death victim [l]. At first CPR was limited to the critical care departments of the in-hospital setting where it was performed by skilled physicians. Therefore, the first outcome studies of CPR by Kouwenhoven et al. had a high success rate [2]. From the early sixties on, more people were trained in the techniques and skills of CPR (physicians, paramedics, emergency medical technicians, laymen) and more patients were involved. At present it is recommended that CPR should be initiated in almost any patient suffering a cardiac arrest, except for patients with a known DNR-code (do not reanimate). This widespread use of CPR might improve the chances of initial response to CPR and ensure good neurologic function in patients who did respond [3]. However, the price one has to pay for the rescue of extra lifes without brain damage is that CPR is started in patients who have virtually no chance for long-term survival. This report will not discuss whether to start CPR or not. The aim is to find out if rules can be formulated that are helpful in the decision making to continue or discontinue CPR by studying retrospective 2713 out-of-hospital cardiac arrests. Only limited literature data [4-71 are available on this subject. Most authors focus on save rates and survival, and do not mention experiences with CPR failure. Although, this is the most frequent outcome of all CPR attempts. METHODS
The registration method has been described elsewhere [8]. Several variables related to the pre-arrest, arrest, CPR and post-CPR period have been registered prospectively on a uniform registration form, in all cardiac arrest events attended by the mobile intensive care unit (MICU) teams of the participating emergency medical service (EMS) systems. Outcome was subdivided in 4 classes: Class 1 = immediate CPR failure, i.e. spontaneous circulation can not be restarted; Class 2A = CPCR failure due to cerebral failure, i.e. patients whose circulation can be restored but die within 14 days from cerebral failure or remain in a vegetative state 14 days post-CPR; Class 2B = CPCR failure and cardiac, septic or other failure, i.e. patients whose circulation can be restored but die within 14 days from non-cerebral failure; Class 3 = CPCR successes, i.e. patients awake 14 days post-CPR either normal (good recovery), moderately disabled (disabled but independent) or severely disabled (dependent on others)
PI.
Between January 1st 1983 and December 31st, 1987, the MICU teams of the participating centers registered 3083 out-of-hospital cardiac arrests. Only patients with complete files on predictive parameters were included in the analysis (type of cardiac arrest on arrival of the MICU, type of respiratory arrest during CPR by the MICU, quality of pupil reaction during CPR by the MICU, efficiency of cardiac massage by the MICU and duration of advanced life support = ALS). Type of cardiac arrest on arrival of the MICU was classified in two groups:
5139
ventricular fibrillation (VFl or no VF (electromechanical dissociation = EMD and asystole). Type of respiratory arrest (RAl was either complete respiratory arrest, or gasping occurring at any time during CPR after the arrival of the MICU. Pupil reaction during CPR was described as negative pupil reaction when pupils were dilated, pinpoint or inequal without reaction to light and as positive pupil reaction when normal or unequal pupils did react to light (slow or normal). The quality of CPR by the MICU was scored as efficient (Ql = correct technique with good effect1 or as inefficient (Q2 = correct technique with weak effect; Q3 = bad technique) according to the quality of the palpated carotid or femoral pulse during external chest compression (ECC). The registration form was filled out by a member of the MICU team immediately after the CPR attempt. ALS-time (minutes) is the duration of the reanimation by the MICU only. Although no strict written treatment protocol was used, the methods and techniques of CPR, used by the participating centers was comparable and was based on the generally accepted “Standards and Guidelines for CPR and Emergency Cardiac Care” as published by the American Heart Association (AHA) [lo] and by P. Safar [ll]. No specific recommendations were used for the minimum duration of ALS. The physician of the MICU decided when to cease CPR based on personal knowledge and clinical experience. Based on analysis of the above mentioned parameters of the patients studied, we were able to identify a number of predictors of CPR failure and CPR success, which might be of help in the decision making of the emergency team during their CPR attempt. RESULTS
During the registration period data on 3083 cardiac arrest events were recorded. Complete files on prognostic indices were available on 2713 patients (missing: n = 3701.In Table I, patients are classified according to 4 parameters (type of cardiac arrest on arrival of the MICU, type of respiratory arrest during CPR, pupil reaction to light during CPR, quality of external chest compresssion by the MICU). The prevalence of ventricular fibrillation was 26% (699/27131.Of these patients 32% (221/6991were gasping during CPR. Of these 221 gasping patients 116 (55%) had positive pupil reaction to light during CPR. Patients in VF in complete respiratory arrest during CPR (478/699 = 68%) still had pupil reaction to light during CPR in 26%1 of cases (122/4781.Of the 2713 studied patients, 2014 (74%~)were in electrochemical dissociation or asystole on arrival of the MICU. Twelve percent (245/20141of these patients were gasping during CPR of which 36% (87/2451had pupil reaction to light during CPR. The majority of the patients in EMD or asystole was in complete respiratory arrest (1769/ 2014 = 88%). Only 8O/b(149/17691of the patients in this setting had pupil reaction to light during CPR. It was analysed whether these parameters could be used for prediction of long-term survival. Prediction of long-term survival (LTS = class 31using these predictive parameters is summarised in Table II. The best predictor of LTS was the presence of positive pupil reaction to light during CPR. Of the patients with positive
5140 TABLE I CLASSIFICATION OF PATIENTS ACCORDING TO PREDICTIVE PARAMETERS VF = ventricular fibrillation on arrival of the MICU; EMD = electromechanical dissociation on arrival of the MICU; asyst = asystole on arrival of the MICU; RA = complete respiratory arrest during CPR; gasp = gasping during CPR; pupil + = positive pupil reaction to light during CPR; pupil - = negative pupil reaction to light during CPR. Number of patients between brackets. Inefficient ECC by the MICU pupil + (1161 vF(w91/gasping~2211~pupil\RA(4,81
/pupil+ -pupil gasping (2451-----1
EMD/ASYST (20141 LRA
(1769)/pupil 1
(61
(1051
(191
(1221
(41
- (3561 pupi1+ (871 pupil - (1581 + (1221 pupil - (16201
66) (71 (381 (41 (3671
TABLE II PREDICTION OF LONG-TERM SURVIVAL BY TYPE CARDIAC ARREST ON ARRIVAL OF THE MICU, TYPE RESPIRATORY ARREST DURING CPR, PUPIL REACTION TO LIGHT DURING CPR AND EFFICIENCY OF EXTERNAL CHEST COMPRESSION BY THE MICU Number
% long-term survivors
Type of cardiac arresk VF EMDlasystole
699 2014
17.0 3.5
Type of respiratory crrest: Gasping Complete
2247 466
4.5 18.9
Pupil reaction duting CPR: Present Absent
474 2239
23.2 3.5
Quality of ECC: Efficient Inefficient
2202 511
8.4 0.8
Missing data Total
370 3083
7.8 7.1
5141
pupil reaction (n = 474123% ended in class 3. Of the patients gasping during CPR (n = 4661 19% were classified in class 3. Patients in EMD or asystole on arrival of the MICU had a bad prognosis: long-term survial was 3.5Oh (n = 20141. Of the patients presenting VF on arrival of the MICU, 17% were longterm survivors (n = 699). Poor CPR technique or inefficient external chest compression by the MICU is a predictor of unsuccessful outcome. Only 4 out of the subset of 511 patients prere long-term survivors. However, efficient cardiac massage was not a good predictor of survival (LTS: 8.4Ok11. In order to predict long-term survival, the combined influence of the studied parameters on survival was analysed. Patients were classified in a data-matrix (Tables 1111according to these 4 parameters and ALS-time. Most patients (16201 2714 = 60%) belong to the group with bad prognostic parameters (EMDlasystole, complete respiratory arrest and absent pupil reaction during CPR)
TABLE III NUMBER OF LONG-TERM SURVIVORS OF PATIENTS WHOSE CPR EFFORT IS NOT YET CEASED AFTER A CERTAIN DURATION OF ADVANCED LIFE SUPPORT, ACCORDING TO PREDICTIVE PARAMETERS VF = ventricular fibrillation on arrival of the MICU; EMD = electromechanical dissociation on arrival of the MICU; asyst = asystole on arrival of the MICU; RA = complete respiratory arrest during CPR; pupil + = positive pupil reaction to light during CPR; pupil - = negative pupil reaction to light during CPR; 81 = efficient external chest compression by the MICU; &2,3 = inefficient external chest compression by the MICU; ALS-time = duration of advanced life support. No. of patients in parentheses. VF + RA (478)
VF + Gasping@211 ALStime (min) >20 >30 >45 >60
Pupil + (116)
Pupil - (105)
61
62,s
61
331110 8166 4145 2127 l/13
l/6 l/5 l/3 O/l 0
21186 5145 4124 4112 O/l
EMDlAsyst ALS time (min) >20 >30 >45 >60
Pupil+
Pupil + (122)
Pupil-
6293
61
Q&3
Ql
QZ3
2119 l/14 O/9 O/5 013
34/118 9157 5143 l/18 l/7
114 l/3 112 O/l O/l
271300 6/175 31101 0143 o/19
O/56 0143 0123 o/7 O/l
+ Gasping (245)
(87)
EMDlAsyst
Pupil - (158)
(356)
+ RA (1769)
Pupil + (149)
Pupil - (1620)
Ql
Q2,3
Q1
82,s
Ql
6293
Ql
Q2,3
19180 6148 3130 O/16 018
017 016 o/3 0 0
12/120 3172 2145 2130 0111
O/38 on1 o/10 016 O/l
221135 3183 l/56 O/24 0113
o/14 0110 016 O/l 0
17/1253 51712 l/353 o/140 0153
01367 o/177 O/83 0134 o/9
S142
(Table III). In this group of patients overall long-term success rate was 1% (171 16201.Long-term survivors (n = 171had an ALS-time which did not exceed 45 min. In this subset of patients the incidence of inefficient external chest compression by the MICU was 23O/o (367/16201. No long-term survivors were recorded in this subgroup of patients (013671. Of the patients in EMD or asystole on arrival of the MICU and in complete respiratory arrest during CPR, 8% (149/16201had positive pupil reaction to light during CPR. Long-term survival was significantly higher than in patients without pupil reaction to light during CPR: 180/b(22/149) vs. 1% (17/16201.In this group, only one out of the 22 long-term survivors had an ALS-time exceeding 30 min. Patients in EMD or asystole on arrival of the MICU, still gasping and with positive pupil reaction to light during CPR, had long-term survival of 22% (19/871.In case of absent pupil reaction chances for long-term survival were reduced: 8% (12/1581.In this group of patients (EMDlasystole + gasping) 7 of the 31 long-term survivors had an ALS-time of more than 30 min. Of the 2713 studied patients, 699 (26%) were still in VF on arrival of the MICU. In 85 chest compression was considered as inefficient. Of the patients in VF, 51% (356/6991were in complete respiratory arrest and did not have pupil reaction to light during CPR (Table III). In this group, 8% (27/3561were longterm survivors who were all resuscitated within 35 min. Of the 478 patients in VF on arrival of the MICU and in complete respiratory arrest, 122 (26%) had pupil reaction to light. Long-term survival was 290/6 (35/1221.One long-term survivor had an ALS-time of more than 45 min. The highest rate of long-term survivors (29% = 3411161was found in patients in VF on arrival of the MICU, with gasping and pupil reaction to light during CPR. This subset represents 17%1 (11016991of the patients in VF (Table 1111.Two survivors had an ALS-time of more than 45 min. One of these patients was reanimated during more than 1 h. Patients in VF, not gasping and without pupil reaction to light during CPR (105/699 = 150/o),had long-term survival in 22% of cases (23/1051. All long-term survivors were reanimated within 1 h. The same data-matrix of ALS-time and of the studied parameters was also used for prediction of CPR failure. Table IV presents the numbers of CPR failure (class 11in the patients where resuscitation was ceased after a certain ALS time (620 min, 21- 30 min, 31-45 min, 46-60 mini. In 44 cardiac arrest patients in VF on arrival of the MICU, still gasping, with positive pupil reaction to light during CPR and with efficient ECC, the reanimation was discontinued after an ALS-time of less or equal than 20 min. In 8 of these 44 patients the CPR attempt was unsuccessful (class 11.The remaining patients (36/441were either short- or long-term survivors. In another 21 cases, ALS was ceased after 21- 30 min. Fourteen of these were class 1 patients. In an important number of patients in EMD or asystole on arrival of the MICU, in complete respiratory arrest and without pupil reaction to light during CPR, it was noted that the CPR attempts were ceased before 21 min: 43% (541/ 1253) if ECC was efficient and 52% (19013671if ECC was inefficient. Most of these CPR attempts were unsuccessful (in class 1): 780/b (424/5411if ECC was efficient and 97% (185/1901if ECC was inefficient.
s143 DISCUSSION
The selection of the studied parameters (type cardiac arrest on arrival of the MICU, type respiratory arrest during CPR, quality of the pupil reaction to light during CPR, efficiency of external chest compression by the MICU and ALStime) is based on current clinical practice and experience. Prognostic indices should be available in any patient, should be easy to determine and should preferably be dichotomous with clear distinction in outcome. Literature presents more parameters for decision making in CPR: e.g. underlying disease, age, response time, witnessed or not, bystander CPR, etc [4 - 71.These parameters might not be correctly assessable and are not always available at the time of resuscitation. Therefore we preferred not to select a subgroup of cardiac arrest patients of cardiac origin for this analysis. The type of cardiac arrest on arrival of the MICU is widely accepted as an important predictor for outcome [5-7,12- 141.Electromechanical dissociation and asystole have a similar outcome which is significantly worse than the outcome of ventricular fibrillation. The type of respiratory arrest is not as widely used as a clinical predictor but a strong correlation with survival has been demonstrated in our study population. Furthermore, gasping is easy to determine. Quality of pupils and pupil reaction to light is also generally accepted as a good predictor for neurologic status and outcome [15- 171.In our study population it seemed to be the best predictor for long-term survival. As mentioned by some authors [18], the only time interval which can be estimated accurately is the duration of the reanimation period by the MICU. ALStime has already been used as a predictive parameter [4,5,13,14]. In our registry, inefficient external chest compression revealed to be a predictor of bad outcome. During the 5-year registration period only 4 out of 511 patients (1%) with inefficient ECC were long-term survivors. Registration of efficiency of external chest compression by the MICU might be biased by the fact that a member of the MICU team is filling out the registration form, at a time when the resuscitation efforts have finished. As such, the evaluation is not objective and immediate outcome of the CPR attempt is known. Because of these reasons and because of the small number of patients in VF on arrival of the MICU with inefficient external chest compression by the MICU, this parameter is only used for analysis of prognosis of patients in EMD or asystole. Table IV illustrates that in many cases the CPR effort is ceased rather early. Even in the subgroup with the highest potential success rate (VF, gasping, pupil reaction, ECC Qll ALS was ceased after 20 min or less in 7% (8/110) of cases. In 52% (190/3671of the subgroup without long-term survivors (EMD/ asyst, RA, no pupil reaction, ECC Q2 or Q31the ALS-time did not exceed 20 mon. Only 5 out of these 190 patients were successful resuscitations (class 2A and 2Bl. These unsuccesful CPR attempts might not have received an optimal treatment because CPR was maybe ceased prematurely. In this way a self-fulfilling prophency is created. Due to the early cessation cardiac arrests with bad prognostic parameters do not have the opportunity to demonstrate that longterm survival is possible. These observations might suggest that inefficient
s144 TABLE IV NUMBER OF PATIENTS WITH UNSUCCESSFUL CPR OUTCOME WHOSE RESUSCITATION WAS CEASED AFTER A CERTAIN DURATION OF ADVANCED LIFE SUPPORT, ACCORDING TO PREDICTIVE PARAMETERS VF = ventricular fibrillation; RA = complete respiratory arrest during CPR; pupil + = positive pupil reaction to light during CPR; pupil - = negative pupil reaction to light during CPR; Ql = efficient external chest compression by the MICU; Q2,3 = inefficient external chest compression by the MICU; ALS-time = duration of advanced life support. VF + Gasping(221) ALStime (min) *20 20-30 31-45 46-60
Pupil + (116)
Pupil + (122)
Pupil - (356)
62,s
62,s
(110)
&2,3
Q1
Q2,a
(6)
(19)
(4)
(300)
(56)
a/44 14121 12/18 13/14
l/l l/2 l/2 l/l
8141 17/21 11/12 8/12
l/l 313 313
16132 16127 14115 17/19
7161 4114 16125 6/11
l/l l/l l/l
EMD/Asyst
Pupil - &?I??)
6293 (7) 10132 12/18 9/14 818
415 415 314 2/2
+ Gasping (245)
Pupil + (87)
(min) 620 21-30 31-45 46-60
Pupil - (105)
Ql
EMD/Asyst ALS time
VF + RA (4’78)
70/115 62174 44158 24124
11113 20120 15/16 6/6
+ RA (1769)
Pupil + (149)
Pupil - (1620)
Q&s
6293
&2,3
(38)
(14)
(367)
17117 ll/ll 414 515
11/52 22129 26132 8/11
314 314 515 l/l
4241541 3281358 197/213 80187
185/196 94194 49149 24125
external chest compression could be considered as motive to stope CPR instead of an incentive to intensify the CPR efforts. It is most likely that in these circumstances the decision to stop CPR was based on other relevant conditions which were not recorded in this registry. For prediction of bad outcome, only patients in electromechanical dissociation and asystole on arrival of the MICU (Table Vl are taken into consideration. Positive pupil reaction during CPR is found in 236/2014 patients and negative pupil reaction in 177812014patients. Long-term survival is 17% (411 236) in patients with positive pupil reaction vs. 2Orb(29117781in patients with negative pupils. Considering patients without pupil reaction, the chances for long-term survival are extremely limited (O/4051if external chest compression by the MICU is inefficient. When external chest compression is efficient CPR success can be obtained even after 30 min of ALS-time: 3 out of 29 long-term survivors in this group required more than 30 min of ALS to be successfully resuscitated. Two of these patients were gasping and both had an ALS-time of 46 min. A third patient suffered a cardiac arrest due to intoxication, was in complete respiratory arrest and had an ALS-time of 45 min. Considering
s145 TABLE V PROGNOSIS OF BAD OUTCOME EMD = electromechanical dissociation on arrival of the MICU; asyst = asystole on arrival of the MICU; pupil + = positive pupil reaction to light during CPR; RA = complete respiratory arrest during CPR; gasp = gasping during CPR; Ql = efficient external chest compression by the MICU; Q2,3 = inefficient external chest compression by the MICU; ALS-time = duration of advanced life support. No. of patients in parentheses. pupi/ST
p”pil_
(4112361
(29/17781 Ql A\Q23 (29113731
I
(3/4911
ALS-time: > 30
1
GASP’ (21551
RA (114361
I ALS-time: > 45
(014051
I (O/1741
(21361
(O/411
patients in EMD or asystole with positive pupil reaction to light during CPR, ALS-time should be continued to at least 30 min: 4 of the 41 long-term survivors required an ALS-time of more than 30 min. All long-term survivors were resuscitated within 45 min. For prediction of good outcome, the subset of patients in ventricular fibrillation on arrival of the MICU is studied (Table VI). Patients in VF have good chances for long-term survival if pupil reaction to light is present during TABLE VI PROGNOSIS OF GOOD OUTCOME VF = ventricular fibrillation on arrival of the MICU; pupil + = positive pupil reaction to light during CPR; pupil - = negative pupil reaction to light during CPR: RA = complete respiratory arrest during CPR; gasp = gasping during CPR; ALS-time = duration of advanced life support. VF (119/6991 Pupil+ ’ (69/238)
\Pupil(50/461)
I
I
(4/671
(31471
ALS-time: > 45 GAS< (2/281
\RA (l/191
GASP f4/171
/l
RA (01501
5146
CPR (69/238 = 29%). Most of these successful resuscitations occur within 45 min (66/691ALS-time. In only 3 patients ALS-time exceeded 45 min. Two of these patients were gasping and had an ALS-time of 60 and 65 min. The third patient was not gasping. It concerned a 14-year-old girl that drowned and was found in profound hypothermia (24OCl.Complete ALS-time was more than 3 h (1 h artificial circulation). Patients in VF on arrival of the MICU, without pupil reaction to light during CPR still have a rather good prognosis (LTS: 50/461 = 11%). In 4 patients out of the 50 long-term survivors, an ALS-time of 60 min was required. These 4 patients were gasping during CPR. CONCLUSION
In our patient material (n = 27131a subset of 405 patients (15%) can be identified with no long-term survival. These patients were characterised by: EMD or asystole on arrival of the MICU, no pupil reaction to light during CPR, inefficient external chest compression. The question is whether this poor outcome is the result of a self-fulfilling prophecy or the result of a decision making process to stop CPR based on other clinical and ethical parameters that were not studied in this registry. Patients in EMD or asystole without pupil reaction to light during CPR but with efficient ECC (510/6of all patients, 1373/2713) still have a long-term survival of 2%. Therefore, the resuscitative effort should be continued for more than 0.5 h, especially if the patient is gasping during CPR. Patients in EMD or asystole with pupil reaction to light during CPR (90/oof all patients, 236/27131 have a long-term survival of 18%. In this subset of patients ALS should not be stoppped before at least 45 min. In cardiac arrest patients in VF on arrival of the MICU (26% of all patients, 699/27131,pupil reactions to light during CPR are indicative for long-term survival: if pupil reaction is positive, LTS is 28%, if negative only 11%. In this subset of patients ALS should be continued for at least 45 min. This analysis indicates that during the initial phase of resuscitations it is not possible to identify reliable indicators for CPR failure. The studied parameters are only helpful elements in the decision making to continue the CPR efforts. REFERENCES 1 2 3 4
5
J.H. Talbott, Introduction. In: Fundamentals of Cardiopulmonary Resuscitation. J.R. Jude and J.O. Elam. F.A. Davis, Philadelphia, 1965. W.B. Kouwenhoven, J.R. Jude and G.G. Knickerbocker, Closed-chest cardiac massage, J. Am. Med. Assoc., 173 (1960) 1064- 1067. L.J. Blackhall. Must we always use CPR? N. Engl. J. Med., 317 (1987) 1281- 1285. R.S. Pionkowski, B.M. Thompson, H.W. Gruchow, C. Aprahamian and J.C. Darin, Resuscitation time in ventricular fibrillation - a prognostic indicator. Ann. Emerg. Med., 12 (1983) 733738. J.P. Smith and B.I. Bodai, Guidelines for discontinuing prehospital CPR in the emergency department - a review. Ann. Emerg. Med., 14 (1985) 1093- 1098.
s147 6
7 8 9 10 11
12 13
14
15 16 17 18
C. Aprahamian, B.M. Thompson, H.W. Gruchow, J.R. Mateer, J.F. Tucker, H.A. Stueven and J.C. Darin, Decision making in prehospital sudden cardiac arrest, Ann. Emerg. Med., 15 (19861 445-449. M. Eisenberg, A. Hallstrom and L. Bergner, The ACLS score. Predicting survival from out-ofhospital cardiac arrest, J. Am. Med. Assoc., 246 (1981150- 52. The Cerebral Resuscitation Study Group. The Belgian Cardiopulmonary Cerebral Resuscitation Registry: Form Protocol, Resuscitation, 17 (Suppl.l(l9891 S5- SlO. B. Jennett and M. Bond, Assessment of outcome after severe brain damage: a practical scale, The Lancet, March l(1975) 480 - 484. Standards and Guidelines for CPR and ECC, J. Am. Med. Assoc., 244 (1980)453-509. P. Safar, Cardiopulmonary Cerebral Resuscitation. Prepared for the World Federation of Societies of Anaesthesiologits, 2nd edn.. A. Stavanger and Laerdal, (Eds.1,WB Saunders. London Philadelphia Toronto, 1988. R. Roth, R.D. Stewart, K. Rogers and G.M. Cannon, Out-of-hospitalcardiac arrest: factors associated with survival, Ann. Emerg. Med., 13 (19841237-243. M. Eisenberg, L. Bergner and A. Hallstrom, Paramedic programs and out-of-hospital cardiac arrest: I. Factors associated with successful resuscitation, Am. J. Public Health, 69 (197911049 - 1051. R.J. Myerburg, C.A. Conde, R.J. Sung, A. Mayorga-Cortes, S.M. Mallon, D.S. Sheps, R.A. Appel and A. Castellanos, Clinical, electrophysiologic and hemodynamic profile of patients resuscitated from prehospital cardiac arrest, Am. J. Med., 68 (1980)568-576. P. Sobotka, and E. Gebert, Effect of complete brain ischemia on pupillary changes, Acta Anaesth. Stand., 16 (19721112-116. B.D. Snyder, M. Ramirez-Lassepas and D.M. Lippert, Neurologie status and prognosis after cardiopulmonary arrest: I. A restrospeetive study, Neurology, 27 (1977)807 - 811. J.E. Steen-Hansen, N.N. Hansen, P. Vaagenes and B. Schreiner, Pupil size and light reactivity during cardiopulmonary resuscitation: a clinical study, Crit. Care Med., 16 (1988)69-70. B.M. Thompson, H.A. Stueven, J.R. Mateer, C.C. Aprahamian, J.F. Tucker and J.C. Darin, Comparison of clinical CPR studies in Milwaukee and elsewhere in the United States, Ann. Emerg. Med., 14 (19851750- 754.