The natural course of neurological recovery following cardiopulmonary resuscitation

Resuscitation 36 (1998) 111 – 122

The natural course of neurological recovery following cardiopulmonary resuscitation E.O. Jørgensen a,*, S. Holm b b

a Medical Department I, Bispebjerg Hospital, Copenhagen Health Ser6ices, Copenhagen DK-2400 NV, Denmark Department of Medical Philosophy and Clinical Theory, Uni6ersity of Copenhagen, Copenhagen DK-2200 N, Denmark

Received 30 May 1996; accepted 11 November 1997

Abstract In 231 patients resuscitated from circulatory arrest of cardiovascular or pulmonary aetiology brain recovery was evaluated by serial neurological and EEG examinations for up to 1 year. One-hundred and sixteen patients never regained consciousness; 115 patients awakened within 30 days, and 40 eventually recovered completely within 90 days. Patients who had electrocortical activity recorded by the immediate post-resuscitation EEG (N= 106), and patients initially without such activity (N = 125) pursued the same course of recovery: during unconsciousness, interrelated EEG and neurological findings featured a phase of intermittent cortical acti6ity with postural or stereotypic motor responses followed by a phase of continuous cortical acti6ity with sequential appearances of delta, theta, and alpha activities on EEG accompanied by stereotypic or defensive motor responses. After awakening, the sequential return of motor, sensory, and mental faculties differentiated an early phase of se6ere disability with orientating eye movements and a bilateral Babinski response from the phase of moderate disability featured by speech, locomotor functions, ability to cope with personal necessities and orientation as to personal data, and a normal plantar response. Finally, orientation as to time, place and role of other persons, and retention and recall, defined the phase of slight/no disability. Abnormal courses were identified by incomplete EEG and neurological recoveries or by the appearance of spikes and sharpwaves in the EEG, or by losses of function. © 1998 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Cardiopulmonary arrest; Resuscitation; Cerebral ischemia; Neurologic examination; Electroencephalography; Patient outcome assessment

1. Introduction Every day, thousands of people all over the world are resuscitated from circulatory arrest. Many are left with neurological deficits. Various means have been recommended to improve the course of brain recovery after resuscitation [1,2]. However, any such recommendation requires a fair knowledge of the natural history of global brain ischaemia. Without such knowledge it is difficult to detect possible ischaemic damage to the brain and to assess whether interventions influence neurological recovery favourably. * Corresponding author.

Many studies have sought predictive variables in victims of circulatory arrest [3–19]. But none of these studies described the hour by hour course of post-ischaemic brain recovery. We have previously described neurological recovery following circulatory arrest in patients who had no cortical activity in their immediate post-resuscitation EEG [20–22]. The sequential features of recovery with time in such a highly selected population of patients need not be applicable to victims of circulatory arrest in general. We therefore compared previous experiences with the findings in a group with circulatory arrest who retained some cortical activity reflected by the initial EEG recording for the purpose of defining common features of post-ischaemic brain recovery.

0300-9572/98/$19.00 © 1998 Elsevier Science Ireland Ltd. All rights reserved. PII S 0 3 0 0 - 9 5 7 2 ( 9 7 ) 0 0 0 9 4 - 4

E.O. Jørgensen, S. Holm / Resuscitation 36 (1998) 111–122

112

Fig. 1. Course of neurological recovery following circulatory arrest in 231 patients; + , present; − , absent.

2. Materials and methods

2.1. Study population The study was conducted in the coronary and intensive care units of two university hospitals (Glostrup County Hospital and Rigshospitalet) in the Copenhagen area during a period of 8 years. One of us (EOJ) had witnessed the efforts at resuscitation in 613 victims of circulatory arrest but had no influence on treatment decisions. As a principle, all patients received maximal life-sustaining treatment which also included the use of mechanical ventilation. Circulation was restored in 329 patients. Ninety-eight patients were excluded from the study because of central nervous system lesions (N =42), acute poisoning (N=38) or because consciousness was regained immediately (N =8). Ten patients were lost to the study. The remaining 231 patients, initially unconscious, had EEG recorded and neurological signs registered at intervals from the time spontaneous circulation was restored until death or survival for 1 year. One-hundred and six patients had electrocortical activity registered by EEG immediately post-resuscitation, and in the following are referred to as Group I, whereas 125 patients without detectable electrocortical activity initially are referred to as Group II. The population characteristics including age, sex, pre-existing heart disease, aetiology of arrest and location as well as the initial cardiac dysrhythmia have been described in detail previously [21,22].

2.3. Definitions Brain death was defined by brain stem areflexia including irreversible apnoea [23], no electrocortical activity above 2 mV during 30 min of EEG recording, and a block of intracranial circulation by four-vessel cerebral angiography, performed twice at an interval of 20 min [20,24]. Persistent unconsciousness was identified by the irreversible loss of consciousness in patients who retained some brain function until final asystole. Patients were considered conscious when they were able to obey simple orders or answer simple questions, and fully reco6ered when no evident motor, sensory or mental deficit was demonstrated by ordinary clinical examination [22].

2.4. Data analysis The data were ultimately tested by linear regression analyses, and the Chi square or Fisher’s exact tests.

3. Results

3.1. O6erall course of neurological reco6ery Fig. 1 shows that patients in Group I remained unconscious and suffered brain death less often than those in Group II (PB0.001 and 0.01, respectively, Fisher). Persistent deficits after awakening were, however, equally often encountered.

2.2. In6estigati6e techniques

3.2. Findings in patients remaining unconscious

The investigative techniques used and the findings in Group II patients are described in previous reports [21,22].

Twenty-eight patients in Group I and 88 in Group II remained unconscious. The time course of neurological recovery and deterioration and the rates of final asys-

E.O. Jørgensen, S. Holm / Resuscitation 36 (1998) 111–122

113

Fig. 2. Time course of neurological recovery and decay in patients remaining unconscious. The columns indicate the number of patients present at definite time intervals after resuscitation. Different shading patterns signify the presence or absence of brain function. The reduction of patient numbers over time reflects the rate of final asystole. Group I, patients with electrocortical activity in the immediate post-resuscitation EEG; Group II, patients without such initial activity.

tole were similar in the two population subsets (Fig. 2). The proportion of patients (P) presenting during the first 4096 h after resuscitation thus related linearly to the natural logarithm of time (ln t) in both groups (P = − 0.14 ln t+1.12 (r = 0.98, N = 13) in Group I and − 0.15 ln t+1.15 (r = 0.97, N = 13) in Group II, PB 0.0001 in both groups). Prior to the return of cortical activity in the EEG, patients in Group II at first presented with cranial nerve reflexes and miosis reflecting the phase of cranial ner6e reflexes only and thereafter developed motor responsiveness, predominantly decerebrate postural movements featuring the phase of cephalic reacti6ity. The electrocortical activity present or appearing at first could be intermittent, featuring the phase of intermittent cortical acti6ity (ICA).

3.2.1. Phase of intermittent cortical acti6ity Fifteen patients in Group I at first had ICA, with either 0.5–3 Hz (delta) potentials of up to 120 mV (ICA d), or single potentials of 50 – 200 ms duration (spikes and sharpwaves) of up to 150 mV, appearing alone, or superimposed on a delta background (ICA spsh). Activity started as runs of 2 – 20 s, separated by 3 –30 s periods without electrocortical activity. Initially, the highest amplitudes were in the frontal leads (Fig. 3a), but later similar amplitudes were seen in other leads. ICA d or ICA spsh was found in 59 patients in Group II; but, with a lower amplitude than in Group I. ICA d was at first equally often found in the two patient

groups, whereas ICA spsh was encountered significantly less frequently in Group I compared with Group II (Table 1). Analysis of possible correlations between EEG configurations and neurological signs in Group I was not feasible due to the small number of paired observations (N= 4). Our previous study of Group II patients documented significant relationships between ICA d or ICA spsh and medium-sized pupils and decorticate or stereotypic reactivity. Half the ICA patients made an incomplete recovery of cranial nerve reflexes mainly due to failure of recovery of the ciliospinal or caloric vestibular reflexes. Spasticity, spasms or multifocal myoclonic or tonic-clonic jerks occurred less often in Group I than in Group II (N= 12 vs 55, PB0.01, Fisher). None of the patients opened their eyes either spontaneously or in response to stimulation. ICA was the best recovery in six patients in Group I, and in 27 in Group II but deterioration occurred with loss of brain functions, either abruptly, or gradually, in reverse order of recovery, and all within 5 months. Cerebral function was lost hours before the occurrence of final asystole (brain death) in one patient in Group I and in 10 in Group II. Nine patients in Group I, and 32 in Group II showed signs of improvement, reflected by the appearance of continuous cortical activity (CCA) in their EEGs within a few min to 127 h (median 9 and 17 h, respectively). They characteristically retained or regained all cranial nerve reflexes.

114

E.O. Jørgensen, S. Holm / Resuscitation 36 (1998) 111–122

Fig. 3. EEG, ECG and recording of spontaneous breathing in a 69-year-old woman resuscitated from circulatory arrest due to myocardial infarction. At 1 h, the EEG showed intermittent delta activity with spikes and sharpwaves (a); she had Cheyne – Stokes’ respiration, and all cranial nerve reflexes but the caloric vestibular ones could be elicited; decorticate posturing occurred spontaneously or following cutaneous stimulation. After 13 h, her EEG showed continuous delta activity (b); the caloric vestibular reflex was still missing, and somatosensory stimulation elicited stereotypic motor responses. The patient deteriorated and lost her electrocortical activity during a second though reversible circulatory arrest. Thereafter, she had spontaneous breathing, coughing-swallowing movements, and decerebrate postural responses. These functions were retained until final asystole, 88 h after the primary resuscitation.

3.2.2. Phase of continuous cortical acti6ity In Group I, 22 patients at some time had continuous cortical activity, at first consisting predominantly of d activity. In six of these, random non-focal spikes and sharpwaves of up to 175 mV emerged on the d background. Ten patients eventually improved from CCA d to CCA dominated by 3 – 7 Hz (theta (t)) activity, and three subsequently had predominantly 8 – 13 Hz (alpha, a)) activity (‘pre-awakening alpha’). Continuous activity appeared at first in the frontal leads, with amplitudes of up to 175 mV (Fig. 3b), but later with similar high amplitudes in other leads. The CCA d and CCA t were not influenced by somatosensory stimulation, whereas the CCA a temporarily showed an increase of 11 – 13 Hz activity of 10 – 35 mV. In Group II, the course of EEG recovery in 44

patients was similar to that of Group I. Eight patients in Group II subsequently had ‘pre-awakening alpha’ activity. A few patients, four from each group, did not respond to somatosensory stimuli, whereas some had stereotypic movements elicited. Spasticity, spasms, and multifocal myoclonic features were found in eight patients in Group I and in 10 in Group II. Eye opening, spontaneously or in response to stimuli (‘open eyes’) was noticed after more than 12 h in half the patients from both population subsets, primarily in those who had recovered all cranial nerve reflexes. After more than 256 h of unconsciousness, one patient in Group I and four in Group II showed periodical changes in behaviour resembling sleep and wakefulness (‘sleepwake cycles’).

E.O. Jørgensen, S. Holm / Resuscitation 36 (1998) 111–122

115

Table 1 EEG findings in patients remaining unconscious Electrocortical configurations At first

Group I (N =28) Group II (N =71)

Subsequently

ICA d

ICA spsh

CCA

CCA spsh

ICA d CCA

CCA spsh

ICA spsh CCA

CCA spsh

11 33

2 26

10 11

3 1

6 16

1 5

2

2 9

ICA d, intermittent delta activity; ICA spsh, intermittent cortical activity with spikes and sharpwaves; CCA, continuous cortical activity with spikes and sharpwaves.

All patients deteriorated within 6 months and functional losses were either abrupt, or gradual in reverse order of recovery. Brain death was eventually diagnosed in two patients in Group I and in 10 in Group II while the remaining patients all retained some brain function until final asystole.

3.3. Findings in patients regaining consciousness Seventy-eight patients in Group I and 37 in Group II eventually recovered consciousness. The demonstration of the conscious state sometimes required vigorous, repeated stimulation. Fig. 4 shows that the rate of recovery during the first 256 h in the two groups was the same. The proportion of patients who woke up (P) was related to the natural logarithm of time (ln t) as follows: P=0.11 ln t + 0.23 (r =0.98, N = 59) in Group I, and 0.16 ln t− 0.23 (r =0.95, N =5) in Group II (PB 0.0001 in both groups).

3.3.1. During unconsciousness The steps of recovery during postischaemic unconsciousness resembled those described in patients remaining unconscious. Thus, cortical activity started either intermittently or continuously. The electrocortical pattern seen initially and at the time of awakening is shown in Table 2. The intermittent cortical activity consisted mainly of d potentials in both groups as illustrated by Fig. 5a (phase of ICA). The ICA improved to continuous cortical activity with similar delays in the two groups (median 20 and 16 h, respectively), and the patterns could evolve from CCA d via CCA t to CCA a in that order (phase of CCA). In both groups CCA d, as well as CCA t, and CCA a started with the highest amplitudes in the frontal leads, or with equally high amplitudes in additional leads (Fig. 5b–d). ‘Pre-awakening alpha’ activity was found in 15 patients in Group I and in six in Group II; eventually, this activity was replaced by a activity of highest amplitude in the occipital leads corresponding to the pattern found in normal, alert adults (‘postawakening alpha’), either before (three patients in Group I) or after the return of consciousness (12 pa-

tients in Group I and three in Group II). Three Group II patients actually retained their ‘pre-awakening alpha’ activity until final asystole hours to days after consciousness had returned. Random, non-focal spikes and sharpwaves were seen (Table 2). The neurological signs associated with ICA and CCA were similar in the population subsets (Table 3). In contrast to the findings in patients who remained unconscious, all patients made a complete recovery of cranial nerve reflexes. Table 3 also demonstrates that decorticate posturing was commonly encountered in patients with ICA with or without spsh (ICA versus CCA d in the two groups, P B 0.001 and 0.05, respectively, Fisher) and that CCA d was usually accompanied by stereotypic reactivity (ICA versus CCA d, ns, CCA d versus CCA t in Group I and II, PB 0.01, Fisher). With the advent of dominant t or a activities, an increasing number of patients moved trunk or limbs following stimulation, as if they tried to locate and remove or avoid the stimulus (defensive reactivity). The EEG activity related significantly to this motor response (CCA d vs CCA t in the two groups, PB0.01 and 0.001, respectively, Fisher). ‘Open eyes’ were not found in patients with ICA; but this was noted in some patients with CCA (four in Group I, and three in Group II) after more than 96 and 200 h of unconsciousness, respectively. Spasticity, spasms or myoclonic features were noticed in 14 patients in Group I and six in Group II. One patient from each group had generalized convulsions which appeared without consistent correlation to any particular EEG configuration.

3.3.2. After awakening Consciousness returned significantly earlier in Group I than in Group II (a few min to 335 h, median 6 h versus 11–720 h, median 50 h, PB0.0001, Mann– Whitney). At 24 h, 57 patients in Group I had regained consciousness as compared to only 14 in Group II (73 vs 38%, PB 0.001, Fisher). The features of post-awakening recovery were similar in the two patient groups. But after awakening, the EEG and neurological recoveries occurred indepen-

116

E.O. Jørgensen, S. Holm / Resuscitation 36 (1998) 111–122

Fig. 4. Course of neurological recovery in patients regaining consciousness. The parts of columns below the midline indicate the number of unconscious patients, and the parts of columns above the midline the number of conscious patients. The reduction in the number of unconscious patients over time reflects improvement, whereas reduction in the number of conscious patients indicates the rate of final asystole. Areas with different shading patterns signify whether neurological recovery was complete or not.

dently, in contrast to the findings during the period of unconsciousness. Three steps of post-awakening recovery could be distinguished, however, because patients regained elementary motor, sensory and mental faculties in a definite order: At first, there was a phase of severe disability; then followed the phase of moderate disability, and finally, full faculties were restored, with only slight or no impairment. The number of patients with persistent or transient disability appears in Table 4.

3.3.2.1. Phase of se6ere disability. The severely disabled patients had defensive motor responses as they were able to react to single, simple commands by moving their head or limbs. In addition, eye orientation (fixation and eye following of objects and persons) was invariably present, whereas other faculties were absent or impaired. Motor functions such as sitting, standing and walking were absent and patients were unable to eat, wash, or to control their bladder and bowel functions (no locomotor function and no ability to cope with personal necessities). Some patients were eventually able to say a few words (speech), some were aware of some personal data (auto-orientation), but none seemed to have any idea of time, place, or the role of other persons (no allo-orientation). They had no capacity to retain or recall experiences (no retention and recall). The Babinski response was elicited in 22 pa-

tients, unilaterally in seven, and five of these were hemiplegic. Tonic-clonic jerks or extensor fits were observed in three patients. In Group I, 31 patients remained severely disabled throughout, six had CCA d early after awakening, one improved to present CCA t, and four eventually presented CCA a. Another 16 patients initially had CCA t, and two of these subsequently showed CCA a. The remaining nine patients had CCA a throughout. Similar features were found in Group II, and the findings in both groups were independent of whether severe disability was transient or persistent. All patients with severe disability died within 6 months except one patient from Group I who was alive at 1 year. The phase of severe disability among those patients who eventually improved to moderate disability was significantly longer in 20 patients in Group II than in 47 patients in Group I (23–2952 h (median 207 h) and 1–2665 h (median 47 h), respectively; P B 0.01, Mann– Whitney).

3.3.2.2. Phase of moderate disability. During this phase patients regained auto-orientation and speech, though aphasic disorders were common. The locomotor functions and ability to cope with personal necessities also returned. Allo-orientation or retention and recall were rarely restored and if so, obviously impaired. The pa-

E.O. Jørgensen, S. Holm / Resuscitation 36 (1998) 111–122

117

Table 2 EEG findings in patients regaining consciousness Electrocortical configurations At first

Group I (N =78) Group II (N = 37)

At the time of awakening

ICA

CCA d

CCA t

CCA a

CCA d

CCA t

CCA a

12 (1) 30 (3)

13 (1) 7 (2)

33 (7)

20 (5)

4 (1) 2 (1)

25 (3) 26 (4)

49 (6) 9 (4)

Abbreviations as for Table 1; d, delta; t, theta; a, alpha. The figures in parentheses indicate the number of EEGs with spikes and sharpwaves included.

tients could be apathetic, confused or emotionally labile, with outbursts of euphoria or aggressiveness. The Babinski response was elicited in 10 patients, in four it was unilateral and combined with hemiplegia. The persistently disabled patients either remained in the medical ward or were transferred to psychiatric institutions. A few (four from Group I and one from Group II) were discharged home. Nine patients in Group I, and three in Group II died during the year of observation. Further improvement to no or slight disability was observed in 26 patients in Group I after 8 – 1410 h (median 286 h), and in 14 in Group II after 49 – 1224 h (median 314 h).

3.3.2.3. Phase of slight/no disability. Patients recovering fully had no evidence of motor, sensory or mental deficits on ordinary clinical examination. However, some had memory disorders, and suffered from depression or anxiety. All experienced improvement during the year of observation. In Group II one patient complained of a minor weakness of his right arm; a remnant of a total hemiplegia and atypical plantar responses were found in three other patients. Forty patients made a complete recovery, but two in Group I, and one in Group II died by the end of the 1-year study period. 3.3.2.4. Time course of post-awakening reco6ery. Recovery of post-awakening faculties during the first 8192 h after resuscitation was related linearly to the logarithm of time (ln t). For example, the rate of recovery of the ability to sit up was in Group I = 3.58 ln t −6.37 (r= 0.98, N=11) and in Group II = 5.69 ln t −19.2 (r= 0.99, N= 6); in both groups PB0.0001. Locomotor functions and ability to cope with personal necessities, and auto-orientation were restored later but more rapidly in Group II than in Group I. Thus, no patient in Group II could eat without being assisted by 32 h, whereas 13 (17%) of the patients in Group I could. However, late recovered functions such as allo-orientation and retention and recall returned at the same rate in the two groups.

In the study population as a whole, consciousness returned not later than 30 days, and improvement from severe to moderate disability invariably occurred within 128 days. Restitution of full faculties was never delayed more than 90 days after resuscitation.

3.4. Salient features of post-ischaemic brain reco6ery The course of recovery during the first post-resuscitation year can be epitomized as shown in Fig. 6. Early recovery is characterised by spontaneous respiratory movements, miosis and cranial nerve reflexes (phase of exclusi6e presence of cranial ner6e reflexes). This phase as well as the following phase of cephalic reacti6ity, featured by the presence of motor responsiveness is defined by the lack of the caloric vestibular reflex and of detectable cortical activity in the EEG (strictly defined according to Jørgensen [20]). Recovery progress is recognized by related appearances of neurological signs and EEG configurations which initially determined the phase of intermittent cortical acti6ity, also featuring medium-sized pupils and decorticate posturing, progressing to the phase of continuous cortical acti6ity with stereotypic reactivity. After awakening, neurological and EEG recovery occurs independently. But, a fixed sequence of return of elementary motor, sensory and mental faculties differentiates the phase of se6ere and moderate disability from the phase of slight/no disability. The severely disabled patient has communicating motor responses, orientating eye movements, and a bilateral Babinski response. The moderately disabled patient is identified by the presence of speech, auto-orientation, and locomotor abilities. The slightly or not disabled patient presents with allo-orientation, retention and recall. Some patients with abnormal courses can make some of the steps of recovery mentioned above. Among these, the unfavourable course is identified during unconsciousness by an incomplete recovery of the cranial nerve reflexes or by spikes and sharpwaves in the EEG; and by losses of function. After awakening, abnormal courses are characterized either by inadequate recovery progress or by function losses.

118

E.O. Jørgensen, S. Holm / Resuscitation 36 (1998) 111–122

Fig. 5. Serial recordings of EEG, EKG, and artificial ventilation in a 59-year-old man who was resuscitated from circulatory arrest complicating myocardial infarction. Immediately upon re-establishment of circulation there was no detectable cortical activity in the EEG; but all cranial nerve reflexes and decerebrate posturing could be elicited. Three h after resuscitation, intermittent delta activity appeared with high amplitudes in the frontal leads (a), and cutaneous stimulation resulted in decorticate posturing. After 7 h, the EEG showed continuous cortical delta activity (b), and neurological examination revealed stereotypic re-activity. At 17 h, the EEG displayed continuous theta activity with highest amplitudes in the frontal leads (c), and at 33 h, continuous alpha activity dominated more of the leads (d). The patient regained consciousness after 29 days. Gradually, he was able to utter syllables; but, he remained unable to sit, stand or walk, and required assistance to cope with all personal necessities, until final asystole, 79 days after the initial resuscitation.

E.O. Jørgensen, S. Holm / Resuscitation 36 (1998) 111–122

119

Fig. 5. (Continued)

4. Discussion The present report establishes that orderly brain recovery can be identified during the period of unconsciousness by a fixed-order return of interrelated neurological signs and EEG configurations; and, after

awakening by a fixed-order return of motor, sensory and mental faculties. A similar systematic description of the natural history of global brain ischaemia is not available in the literature. However, Bokonjic [25], who studied 27 patients retrospectively following circulatory arrest or suicidal hanging, noted that one patient, who

E.O. Jørgensen, S. Holm / Resuscitation 36 (1998) 111–122

120

Table 3 Neurological signs related to EEG in 115 patients regaining consciousness Electrocortical configurations ICA I

CCA d II

12

I

CCA t II

CCA a

I

II

I

II

30

17

26

34

38

68

9

All cranial nerve reflexes

+ −

8 2a

23 2

16 1

24

34

23

67

8

Decorticate posturing

+ −

9 3

12 14

1 16

3 21

1 28

64

8

Stereotypic reactivity

− +

8 4

21 5

14 2

23 1

10 19

12 10

4 60

5 3

Defensive reactivity

+ −

3 13

24

19 10

10 12

60 4

3 5

N

The presence (+) or absence (−) of neurological signs at the time of recording of ICA (intermittent cortical activity) or CCA (continuous cortical activity). a Information not available in two patients. Similar discrepancies between the total number of patients given at top of the column and the figures in this and other columns indicate missing data. Table 4 Recovery after awakening Postawakening recovery Severe disability

Group I (N= 78) Group II

Moderate disability

No disability

Persistent

Transient

Persistent

Transient

31 15

47 22

21 8

26 14

26 14

The numbers of patients with persistent or transient disability after awakening are shown

had ‘‘little or no recordable cortical activity’’ in the EEG shortly after the ischaemic event, presented decerebrate posturing; these are joint features which in the present study defined the phase of cephalic reacti6ity. Another of his patients, whose EEG ‘‘in addition to long periods of almost complete absence of electrical activity, showed runs of regular 1 – 3 Hz activity of 80 – 200 mV mainly in the frontal and temporal regions’’, had ‘‘spasticity with tonic neck reflexes’’, a description which fits ours for the phase of intermittent cortical acti6ity. Finally, Bokonjic divided the course of post-awakening recovery into an initial period of disorientation, progressing to impaired memory which preceded complete restitution of mental faculties. Our work adds the description of a characteristic sequential return of elementary faculties after awakening, defining the phases of se6ere-, moderate-, and slight/no disability. We identified patients with abnormal recovery courses by their incomplete recoveries and abrupt or gradual losses of function. The gradual deterioration somewhat resembled the course of events observed in patients with expanding supratentorial lesions. How-

ever, such lesions are usually characterized by focal neurological deficits as distinct from the non-focal findings following global brain ischaemia. Post-resuscitation convulsive phenomena like those depicted herein have been reported to reflect injury to the brain structures [6]. In our studies we found no consistent correlation between fits or seizures and the EEG configurations. The term ‘coma’ is extensively used and has been defined as sleeplike, unarousable unresponsiveness without awareness of self or environment, specifically qualified by the lack of spontaneous or provoked eye opening [11]; patients with ‘open eyes’ were accordingly considered awake but not necessarily aware; if awake, but unaware they were said to be in a ‘vegetative state’ [11,16,17,19,26]. However, at 12 h following resuscitation we found ‘open eyes’ in half the patients considered neither awake nor aware. ‘Open eyes’, in our opinion, merely signified an alive brain stem as this sign has also been found in anencephalic infants [27] and decerebrate cats [28]. We differentiated the awake, but severely disabled individu-

E.O. Jørgensen, S. Holm / Resuscitation 36 (1998) 111–122

121

Fig. 6. Salient features of postischaemic brain recovery. The steps of reco6ery during unconsciousness: I, phase of exclusive presence of cranial nerve reflexes; II, phase of cephalic reactivity; III, phase of intermittent cortical activity; IV, phase of continuous cortical activity. The steps of reco6ery after awakening: V, phase of severe disability; VI, phase of moderate disability; VII, phase of slight or no disability.

als from those neither awake nor aware (including those with ‘open eyes’), by their ability to obey simple orders or to answer simple questions, and by their orientating eye movements. Although there are many reports of EEG findings after global brain ischaemia [3 – 6,8,29], none has specified the mode or the sequence of electrocortical recovery. Our work helps to clarify the post-resuscitation EEG recovery course. The frontal or diffuse continuous alpha activity, herein called ‘pre-awakening alpha’, is indistinguishable from the ‘alpha coma pattern’ found in patients with structural brain lesions [30 – 34]. This pattern has been reported as an exceptionally rare and usually ominous finding in victims of circulatory arrest [35,36]. However, the fact that we found ‘pre-awakening alpha’ activity in 31 patients of whom 25 regained consciousness with nine progressing to complete recovery, documents that this pattern emerges as a facet of the natural electrocortical recovery from circulatory arrest. The above description of the natural history of neurological recovery from circulatory arrest may prove useful as a base for control data for future clinical studies of brain resuscitation including assessment of

interventions to prevent or revert structural brain injuries. References [1] Safar P, Bircher NG. Cardiopulmonary Cerebral Resuscitation. World Federation of Societies of Anaesthesiologists. London: Saunders, 1988. [2] Rogers MC, Kirsch JR. Current concepts in brain resuscitation. J Am Med Assoc 1989;261:3143– 7. [3] Hochaday JM, Potts F, Epstein E, Bonazzi A, Schwab RS. EEG changes in acute cerebral anoxia from cardiac or respiratory arrest. Electroencephalogr Clin Neurophysiol 1965;18:575–86. [4] Prior PF, Volavka J. An attempt to assess the prognostic value of the EEG after cardiac arrest. Electroencephalogr Clin Neurophysiol 1968;27:333 – 7. [5] Binnie CD, Prior PF, Lloyd DSL, Scott DF, Margerison JH. Electroencephalographic prediction of fatal anoxic brain damage after resuscitation from cardiac arrest. Br Med J 1970;4:265–8. [6] Prior PF. The EEG in Acute Cerebral Anoxia. Amsterdam: Excerpta Medica, 1973. [7] Lemni H, Hubbert CH, Faris AA. The electroencephalogram after resuscitation of cardiocirculatory arrest, J Neurol Neurosurg Psychol 1973;36:997 – 1002. [8] Møller M, Holm B, Sindrup E, Lyager Nielsen B. Electroencephalographic prediction of anoxic brain damage after resuscitation from cardiac arrest in patients with acute myocardial infarction. Acta Med Scand 1978;203:31 – 7.

E.O. Jørgensen, S. Holm / Resuscitation 36 (1998) 111–122

122

[9] Willoughby JO, Leach BG. Relation of neurological findings after cardiac arrest to outcome. Br Med J 1974;3:437–9. [10] Plum F, Caronna JJ. Can one predict outcome of medical coma? Outcome of severe damage to the CNS. Ciba Symposium. New York: Elsevier, 1975:121–139. [11] Bates D, Caronna JJ, Cartlidge NEF, Knill-Jones RP, Levy DE, Shaw DA, Plum F. A prospective study of nontraumatic coma: Methods and results in 310 patients. Ann Neurol 1979;2:211 – 20. [12] Snyder BD, Ramirez Lassepas M, Lippert DM. Neurologic status and prognosis after cardiopulmonary resuscitation: I. A retrospective study. Neurology 1977;27:807–11. [13] Caronna JJ, Finkelstein S. Neurological syndromes after cardiac arrest. Stroke 1978;9:517–20. [14] Earnest MP, Breckinbridge JC, Yarnell PR, Oliva PB. Quality of survival after out-of-hospital cardiac arrest. Neurology 1979;29:56 – 60. [15] Dougherty JH, Rawlenson D, Levy DE, et al. Hypoxic ischaemic brain injury in the vegetative state. Neurology 1979;29:591 – 6. [16] Snyder BD, Loewenson RB, Gumnit RJ, Hauser WA, Leppik IE, Ramirez-Lassepas M. Neurologic prognosis after cardiopulmonary arrest: II. Level of consciousness. Neurology 1980;30:52 – 8. [17] Levy DE, Bates D, Coronna JJ, Cartlidge NEF, Knill-Jones RP, Lapinski RH, Singer BH, Shaw MB, Plum F. Prognosis of non-traumatic coma. Ann Intern Med 1981;94:293–301. [18] Longstreth WT, Diehr P, Inue TS. Prediction of awakening after out-of-hospital cardiac arrest. New Engl J Med 1983;308:1378– 82. [19] Levy DE, Caronna JJ, Singer B II, et al. Predicting outcome from hypoxic-ischaemic coma. J Am Med Assoc 1985;253:1420– 6. [20] Jørgensen EO. Requirements for recording the EEG at high sensitivity in suspected brain death. Electroencephalogr Clin Neurophysiol 1974;36:65–9. [21] Jørgensen EO, Malchow-Møller A. Natural history of global and critical brain ischaemia. Part I. EEG and neurological signs during the first year after cardiopulmonary resuscitation in patients subsequently regaining consciousness. Resuscitation 1981;9:133 – 53.

.

[22] Jørgensen EO, Malchow-Møller A. Natural history of global and critical brain ischaemia. Part II. EEG and neurological signs in patients remaining unconscious after cardiopulmonary resuscitation. Resuscitation 1981;9:155 – 74. [23] Pallis C. ABC of brain stem death. Collection of articles published in Br Med J, The Devonshire Press, Torquay Jan, 1983. [24] Jørgensen EO. Aortocervical, internal carotid and vertebral angiography in total cerebral infarction. Acta Radiol 1973;14:369– 78. [25] Bokonjic N. Stagnant anoxia and carbon monoxide poisoning. Electroencephalogr Clin Neurophysiol 1963; Suppl 21. [26] Jeannett B, Plum F. Persistent vegetative state after brain damage. Lancet 1972;I:734 – 57. [27] Gamper E. Bau und Leistungen eines menschlichen mittelhirnwesens (archinencephalie mit encephalocele). A Ges Neurol Psychiat 1926;102:154 – 235. [28] Villablanca J. Behavioural and polygraphic study of ‘sleep’ and ‘wakefulness’ in chronic decerebrate cats. Electroencephalogr Clin Neurophysiol 1966;21:562 – 77. [29] Gro¨nquist YKJ, Soldon TH, Faulconer A Jr. Cerebral anoxia during anaesthesia: prognostic significance of electroencephalic changes. Ann Chir Gynaec Fenn 1952;4:149 – 52. [30] Lundervold A, Hauge T, Jo¨ken AC. Unusual EEG in unconscious patients with brain stem atrophy. Electroencephalogr Clin Neurophysiol 1956;8:665 – 70. [31] Bickford RG, Klass DW. In: Caveness WF, Walker AE, editors. Head Injuries. Philadelphia: Lippincott, 1966. [32] Naquet R, Vigouroux RP, Choux M, Barand C, Charmant JH. EEG in recent cranial trauma in a reanimating service. Rev Neurol 1967;117:512 – 9. [33] Westmoreland FF, Klass DW, Sharbrough FW, Reagan TJ. Alpha coma. Electroencephalographic, clinical, pathologic, and etiologic correlations. Arch Neurol 1975;32:713 – 820. [34] Chase TN, Moretti L, Prensky AL. Clinical and electroencephalographic manifestations of vascular lesions in the pons. Neurology 1968;18:358 – 68. [35] Grindal AB, Suter C. ‘Alpha pattern coma’ in high voltage electrical injury. Electroencephalogr Clin Neurophysiol 1975;38:521 – 6. [36] Alving J, Møller M, Sindrup E, Lyager Nielsen B. Alpha pattern coma following cerebral anoxia. Electroencephalogr Clin Neurophysiol 1979;47:95 – 101.