BRAIN INJURY , 2000, VOL. 14, NO .

6, 585 ± 594

Multimodal early onset stimulation (MEOS) in

rehabilitation after brain injury

È NER and D. TERHAAG

M . LIPP ER T G R U
UniversitaÈ t zu KoÈln, KoÈ ln, Germany

(Received 9 December 1999; accepted 24 January 2000 )

Brain Inj Downloaded from informahealthcare.com by CDL-UC Santa Cruz on 10/26/14

An early and consistent administration of the correct rehabilitation programme is of crucial importance
for the restoration and improvement of cerebral function, as well as social reintegration. This has led to
the development of a multimodal onset stimulation therapy (MEOS), which the neurosurgical inten-
sive care unit administers during an early phase of rehabilitation to patients who have been in a coma
for more than 48 hours after trauma. This study, which was carried out over a period of 2 years, focuses
on 89 patients aged 16± 65 years suffering from severe brain injury. Sixteen of these (age mean: 43.6
years) fulfil the criteria required to start the MEOS. The authors were able to identify significant
changes in two of the patients’ vegetative parameters (viz. heart and respiratory frequencies), even in
cases of deep coma (GCS 3± 4). The most significant changes were caused by tactile and acoustic
stimulation. Standardized behavioural assessment turned out to be particularly advisable in cases of
medium coma. Here, too, a stimulation of the tactile and acoustic senses resulted mainly in head and
For personal use only.

eye movements.

Introd uctio n
As a result of improvements in the rescue system and progress in intensive care
therapy, an increasing number of patients have survived severe brain injury in recent
years. An early and consistent administration of the correct rehabilitation pro-
gramme is of crucial importance for the restoration and improvement of cerebral
function, as well as social reintegration.
Early rehabilitation at this point is an integrated interdisciplinary therapy, which
starts early and proceeds continuously with changing points of interest. Its aim is to
support spontaneous recovery, to reduce the risk of early and late complications,
and to make intensive use of the brain’s own rehabilitative ability and plasticity [18].
Indeed, there are medical and nursing treatments available for early rehabilitation.
The main interest of this studyÐ apart from medical therapyÐ is in correct position-
ing, early mobilization, prophylactics of contracture, pneumonia, decubital ulcer
and thrombosis, incontinence therapy and therapy of acataposis [1].
Until now, it has not been possible to make a reliable prognosis about the
recovery from the `vegetative post-traumatic state’, not even on the basis of clinical
data or electrophysiological data such as evoked potentials. Consequently, other
electrophysiological examinations like event-related potentials or the analysis of
changes in the EEG-spectre are increasingly used to detect covert reactions to

Correspondence to: Dr M. Lippert GruÈ ner, Klinik fuÈr Allgemeine Neurochirurgie, der UniversitaÈ t
zu KoÈln, GeschaÈ ftsfuÈhrender Direktor Professor Dr. med. N. Klug, Joseph-Stelzmann Str. 9, 50931
KoÈ ln, Germany. e-mail: dsg@dr-gruner.de

Brain Injury ISSN 0269± 9052 print/ ISSN 1362± 301X online # 2000 Taylor & Francis Ltd
http:// www.tandf.co.uk/ journals
 586 M. Lippert GruÈ ner and D. Terhaag

external stimuli. Until now, this diagnostic of sensory or cognitive abilities in

comatose patients or patients with reduced consciousness has proven very difficult.
Since patients appear to react to stimuli from their surroundings, a fact which is
often observed by close relatives, it is believed that, to a certain extent, even
comatose patients undergo experiences and react. Reuter et al. [2] showed that
slow cortical potentials can be used to characterize mental functions as well as assess
comatose patients’ chances of recovery.
There is no satisfactory answer in the literature to the question of whether
further specific measures can contribute to the healing process and accelerate waking
out of coma [3, 4]. Experiments carried out on animals have uncovered the poss-
ibility of changing neuronal structures and brain functions by means of external
influences like acoustic, tactile or visual stimulation [5]. Over the last few years,
Brain Inj Downloaded from informahealthcare.com by CDL-UC Santa Cruz on 10/26/14

sensory stimulation has played an increasingly important role in early rehabilitation

therapy [6± 9, 16, 17].

P a tien ts a nd m ethod s
This paper reports on the multimodal stimulation carried out on comatose patients
at a neurosurgical intensive care unit. A stimulation therapy was developed in co-
operation with the Neurosurgical Department and Department of Juvenile
For personal use only.

Psychiatry of the University of Cologne. The neurosurgical intensive care unit

administers it at an early rehabilitation phase to patients who have been in a
coma for more than 48 hours after trauma (table 1).
Stimulation therapy was applied until the maximal values, shown in table 2, on
the three subscales of the Glasgow Coma Scale were reached. The present study,
which was carried out over a period of 2 years, focuses on 89 patients aged 16± 65
years suffering from severe brain injury. Sixteen of these (age mean 43.6 years)
qualified for MEOS, which lasted for an average of 9.81 days (range 1± 30 days).
The patients received early rehabilitation treatment at the neurochirurgical intensive
care unit during an average period of time of 27.9 days. The average Glasgow Coma
Score (GCS) at the beginning of the stimulation therapy was 6.56 (range 3± 9).

Table 1. Criteria taken into consideration at the beginning of stimulation

therapy

Including criteria
. severe head trauma with coma for at least 48 hours (GCS < 8)
. cardiopulmonal stability
. normal intracranial pressure
. no mechanical ventilation
. no sedation
. no severe infections

Table 2. Maximal values on the three subscales of the Glasgow Coma Scale

Maximal values of the GCS

. reaction to painful stimuli: undirected defending movements (4 out of 6 points)
. verbal reaction: unrecognizable sounds (2 out of 5 points)
. eye-opening: until spontaneous eye opening (4 out of 4 points)
 MEOS in rehabilitation after brain injury 587

Multim o d a l ea rly ons et s tim ula tio n (MEO S )

The stimulation therapy consists of acoustic, tactile, olfactory, gustatory and
kinesthic procedures (table 5), administered daily in two units of 1 hour each
following a well determined pattern (tables 3 and 4).
Special demands have to be made concerning restriction of frequency and
intensity of sensory stimulation in order to avoid straining the reduced possibilities
of the injured brain. Controlled stimulation therapy should include low noise levels
and adequate intervals between stimulation and medical and nursing activities.
Furthermore, the patient’ s notion of time should be supported by alternating phases
of activity with intervals without therapy. Rather than following a static pattern,
stimulation units are based on dialogue answers and the actual function level
achieved in the diverse sections.
Brain Inj Downloaded from informahealthcare.com by CDL-UC Santa Cruz on 10/26/14

Table 3. Stimulation block A

Facial Motor
Block 1 play Verbal Eye Head activity

Baseline (10 minutes before beginning)

Orofacial Stim. Finger
(10 minutes) Brush
Disturbances:
For personal use only.

Pause (10 minutes)

Acoustics (10 minutes)
Disturbances:
Pause (10 minutes)
Visual (10 minutes)
Disturbances:
Pause (10 minutes)
Smell (5 minutes)
Disturbances:
Pause (10 minutes)
Taste (5 minutes)
Disturbances:
Baseline (10 minutes later)

Table 4. Stimulation block B

Facial Motor
Block 2 play Verbal Eye Head activity

Baseline (10 minutes before beginning)

Tactile Stim. Manual
(10 minutes) Brush
Disturbances:
Pause (10 minutes)
Proprioceptive and kinaesthetic Stim.
(10 minutes)
Disturbances:
Pause (10 minutes)
Thermic Stimuli Cold
(10 minutes) Heat
Disturbances:
Baseline (10 minutes later)
 588 M. Lippert GruÈ ner and D. Terhaag

Table 5. Modes of stimulation

Acoustic stimulation:
Acoustic stimulation, which is applied by the nurse, the patient’s relatives and other closely related
individuals, includes such methods as referring to events of the social sphere in casual conversation,
or reading aloud literature which the patient knows and likes. It can be carried out directly by one
of the individuals named above or indirectly by playing tapes prepared for this purpose. Apart from
speech, this mode of stimulation can include favourite music or high or low tones and different
sounds.
Tactile stimulation:
This mode of stimulation should focus on the qualities of touch, pressure and temperature, mainly
installed by physical therapy treatments: different kinds of massage and cooling applications of the
extremities and different warming applications of the body. This leads to stimulation, therapy and
prophylactic effects.
Olfactory stimulation:
Brain Inj Downloaded from informahealthcare.com by CDL-UC Santa Cruz on 10/26/14

The patients are exposed to a series of familiar smells, both pleasant and unpleasant. Artificial or
natural essences in small glass containers are given in alternation.
Gustatory stimulation:
Different kinds of food and flavours are placed on the previously moistened tongue with a cotton
bud.
Visual stimulation:
The most important means of applying this mode of stimulation involve coloured lights and
variations of brightness following a day-and-night-rhythm. If the eyes are open, this can be done
by moving bright coloured objects in front of the eyes. If the eyes are closed, this kind of
stimulation is performed by exposing the closed eyes to coloured lights, before opening them
mechanically, with care being taken to avoid dazzling.
For personal use only.

Kinaesthetic and proprioceptive stimulation:

This mode of stimulation should involve vestibular and proprioceptive stimuli, which go further
than the normal nursing and physiotherapeutic measures. This is all the more so since they involve
changes in position of the body and the extremities in horizontal and vertical direction.

Table 6. Documentation parameters of behaviour observation

. Documentation parameters
. eye-opening
. eye movements
. eye fixations
. head movements
. mimic responses
. aimed and non-aimed motor reactions
. articulations

As mentioned, the so-called dialogue answers to the stimuli only take the form
of vegetative symptoms in the coma patient. Such answers are monitored during the
whole stimulation phase by means of feedback system recording and continuous
registration of heart- and respiratory frequency, blood pressure and galvanic skin
response, including the possibility of direct graphical reproduction. These electro-
physiological data were recorded and analysed with the Paron-biofeedback device
and the VITAPORT system developed at Cologne University. In addition, the
VITAPORT system permits the registration of EEG and EMG.
The observation of the patient’ s behaviour grows increasingly important as coma
depth reduces (table 6). A standardized behavioural assessment protocol was devel-
oped during the first stages of the examinations on the basis of the KRS, which also
 MEOS in rehabilitation after brain injury 589

focuses on vegetative changes (table 7). The patient’s most frequent reactions were
hierarchically ordered into mimic reactions, vocal utterings, arousability/ attentive-
ness, motor reactions and vegetative changes.
The clinical observation of the patients under stimulation showed that changes
in the vegetative parameters (for example: changes of the heart and respiratory
frequencies) precede any visible behavioural change. Still, it was necessary to
prove that these changes differ from the baseline without stimulation. Therefore,
a 10 minute baseline was recorded before stimulation and every 10 minutes
between the stimulation sessions (A-B-A design).

Res ults
Brain Inj Downloaded from informahealthcare.com by CDL-UC Santa Cruz on 10/26/14

Significant changes were identified in the vegetative parameters (heart and respira-
tory frequencies) of the patients, even those in deep coma (GCS 3± 4). Figures 1 and
2 show the monitoring of a multimodal stimulation therapy administered to a 38-
year old patient, a therapy which was started on the 4th post-traumatic day. At this
time, the GCS reached 4 points. The first stimulation block did not lead to any

Table 7. Coma observation scale (COS)

(1) Mimical reactions

For personal use only.

4 points: aimed mimic reaction to stimulus

3 points: non-aimed mimic reaction to stimulus
2 points: non-aimed spontaneous mimic reaction
1 point: orofacial spontaneous automatisms
(2) Verbal reactions
5 points: at least one intelligible sentence
4 points: at least one intelligible word
3 points: unintelligible verbal reactions to stimulus
2 points: unintelligible spontaneous verbal reactions
1 point: spontaneous swallowing, coughing
0 points: none
(3) Arousability/ attentiveness
6 points: following a source of stimulation
5 points: lingering on stimulus (longer than 5 seconds)
4 points: turning to stimulus
3 points: opening of the eyes, movement of the eye and eyelids after stimulation
2 points: spontaneous opening of the eyes
1 point: spontaneous movement of the eyes and eyelids
0 points: none
(4) Motor activity
5 points: motor reaction on request
4 points: aimed motor reaction to stimulus
3 points: non-aimed motor reaction to stimulus
2 points: non-aimed spontaneous motor reaction
1 point: extensor/ flexor synergia
0 points: none
(5) Changes in the vegetative state
4 points: reproducible in combination with stimulus
3 points: inconstant in combination with stimulus
2 points: spontaneous and in combination with stimulus
1 point: spontaneous
0 points: none
 590 M. Lippert GruÈ ner and D. Terhaag
Brain Inj Downloaded from informahealthcare.com by CDL-UC Santa Cruz on 10/26/14

Figure 1. The block A stimulation of D.A., a 38-year old patient, 4 days after he sustained severe brain injury
(GSC 4) does not cause any observable stimulation-dependent changes in the heart frequency (highest line),
respiratory frequency, galvanic skin response.
For personal use only.

Figure 2. The block B stimulation of D.A., a 38-year old patient, 4 days after he sustained severe brain injury
(GSC 4) does cause observable stimulation-dependent changes in the heart frequency (highest line).

observable changes in the vegetative functions (figure 1). However, the vegetative
functions of the heart frequency, in particular, did show significant changes during
the second stimulation block, especially during tactile stimulation (figure 2). The
most important changes were found after tactile and acoustic stimulation.
Standardized behavioural assessment turned out to be particularly advisable in
cases of medium coma (GCS 5± 6). Again, a stimulation of the tactile and acoustic
senses resulted mainly in head and eye movements.
 MEOS in rehabilitation after brain injury 591

Thanks to the development of the Cologne VITAPORT system, the authors

were able to measure surface EMG and continuous EEG. Besides therapeutic
stimulation, which was performed according to the ABA-design, an apparative
experimental model was developed. The different channels were stimulated with
a fixed duration and sequence over a period of 24 hours, during which time the
vegetative and electrophysiological parameters were registered (figure 3). Two EEG
channels were recorded with an additional supraorbital electrode in order to detect
ocular artefacts. Changes in vegetative and electrophysiologic parameters during
stimulation (figure 4) were compared with the spontaneous registration. The infor-
mation about the changes in the EEG activity during the day enables one to look
for `therapeutical windows’, i.e. phases of the patient’s weakness that can most
adequately be used for therapy planning.
Brain Inj Downloaded from informahealthcare.com by CDL-UC Santa Cruz on 10/26/14

Two years later, 14 patients were reassessed, whilst two patients were prevented
by structural problems from attending the examination. One of the patients ex-
amined remained in a vegetative state (GOS 2), two suffered severe neurological
and neuropsychological deficiencies and were dependent on care (GOS 3), six had a
GOS of 4 and suffered from severe functional deficiencies, though they did perform
the tasks of daily life on their own, and three had a GOS 5. Two of these had been
able to return to their old jobs.
For personal use only.

Figure 3. Representation of the visible changes in the EEG spectrum during a 20-hour examination of a 24-year
old patient, 10 days after he sustained severe brain injury.
 592 M. Lippert GruÈ ner and D. Terhaag
Brain Inj Downloaded from informahealthcare.com by CDL-UC Santa Cruz on 10/26/14
For personal use only.

Figure 4. Change in the EEG spectrum as a result of vibration stimulation.

D is c us s io n
The therapeutic effect and the scientific basis of the stimulation therapy under
discussion is controversial. LeWinn and Diamanesku [10] reported that eight out
of 16 patients who had received daily multi-sensory stimulation and passive move-
ments of the limb completely recovered from a deep coma. Also, the remaining
eight patients eventually recovered from their comas: four of them were able to live
largely independent lives at home. Similarly, Pierce et al. [11] showed that an
intensive multi-sensory stimulation led to significant improvements in comatose
patients. Wood [12] emphasized the importance of a so-called `sensory regulation’,
which consists of a regulation of background noise levels and introduces silent
periods and breaks between the therapy sessions. It aims at reaching the highest
level of patient vigilance possible during purposeful stimuli-presentation. Hall et al.
[13] noticed that a directed, person-related stimulation with individual meaningful
stimuli seemed to evoke more significant reactions than undirected stimulation.
The use of behavioural parameters for the assessment of these reactions is a
problem, because in most settings the person performing this task is also the one
carrying out stimulation. For this reason, the evaluation of electrophysiologic par-
ameters is crucial. Fast changes can be detected in a more objective way.
 MEOS in rehabilitation after brain injury 593

Pfurtscheller et al. [14] detected significant changes in the EEGs of some comatose
patients (GCS 4± 6) after 10 minutes of tactile or visual stimulation. These results
suggest that EEG frequency analysis may be put to good effect in determining to
what extent comatose patients can react. Besides the EEG analysis, the measurement
of event-correlated vegetative potentials is important for the evaluation of coma
stimulation.
Like neuropsychological and remedial educational early support of children
suffering from severe brain damage, the early rehabilitation of patients having sus-
tained brain injury should include an intensive unit to restore communication.
Intensive care treatment should include offerings of dialogue based on aspects of
the patient’s life and preferences. These aspects will also be used in looking for
possible answers to this dialogue. Because the communicative ability of the deep
Brain Inj Downloaded from informahealthcare.com by CDL-UC Santa Cruz on 10/26/14

comatose patient is still deficient, these answers consist of vegetative symptoms,

which are partly spontaneous, and partly caused by stimulation. It is possible to
document these responses, e.g. by recording heart and respiratory frequency, blood
pressure, galvanic skin response, and by observing the patients’ behaviour. The
importance of the assessment of vegetative changes lies in the fact that non-visible
(`covert’) behaviour or autonomic potentials can be detected equally well as directly
eye-catching (`overto) behaviour [15].
Significant changes of vegetative parameters (heart and respiratory frequencies)
were shown in these patients, even those in deep coma (GCS 3± 4). The most
For personal use only.

significant changes were found after tactile and acoustic stimulation. Standardized
behavioural assessment made sense especially in medium coma (GCS 5± 6). Here,
also, first tactile, acoustic and proprioceptive/ kinaesthetic stimuli resulted mainly in
mimic changes and eye movements.
As expected, changes in behaviour were frequent in all modes of stimulation
in light coma stages, when the stimulus-dependent changes in the vegetative par-
ameters have grown significantly less important and the range of observable changes
in behaviour has widened.

Conc lus ion

The present results indicate that stimulation treatment should always concentrate on
the individual patient rather than follow a fixed pattern. It should be based on a
close observation of patterns of behaviour, and, at least in deep coma stages, involve
the registration of vegetative parameters. The stimulation qualities proved particu-
larly appropriate at the beginning of the stimulation treatment in both the acoustic
and the tactile domains. It is at this stage that most of the changes in the parameters
investigated could be observed. As the depth of the coma decreases, the therapy
should be extended to the remaining senses. However, since some patients also
showed reactions to the visual or proprioceptive stimuli in early stages of the
treatment, it appears advisable to carry out `diagnostic’ stimulations of all the
sense parameters at the beginning of, as well as during the treatment in order to
determine individual response patterns. Also, it could be important to identify
parameters predicting a favourable or unfavourable outcome in non-responding
comatose patients during such diagnostic stimulation. At present, the number of
observations available is still too small to provide an answer to this question, and the
follow-up is too short to determine the usefulness of this diagnostic approach.
Preliminary data seem to support the hypothesis that the absence of any response
 594 MEOS in rehabilitation after brain injury

to external stimuli is indicative of an unfavourable outcome. However, one cannot

infer a good prognosis from patients’ aptitude to give vegetative responses.

A cknowled g em ents
This study was supported by Bundesministerium fuÈ r Bildung und Forschung
(FoÈ rderkenn zeichen 01 k0 9517).

Referenc es
1. D E Y OUNG, S. and G RASS, R . B .: Coma recovery programme. Rehabilitation Nursing, 12: 121±
124, 1987.
2. R EUTER, B . M ., L INKE, D. B . and K URTHEN, M .: Cognitive processes in comatose patients?
Brain Inj Downloaded from informahealthcare.com by CDL-UC Santa Cruz on 10/26/14

A brain mapping study using P 300. Archiv fuÈ r Psychologie, 141: 155± 173, 1989.
3. K ATER, K . M .: Response of head-injured patients to sensory stimulation. Western Journal of Nursing
Research, 11: 20± 33, 1989.
4. S CHREIBER, P. and M AI , N.: U È berlegung zur spezifischen Rehabilitation bei Patienten mit
Schwerem SchaÈ del-Hirn-Trauma. Rehabilitation , 29: 238± 241, 1990.
5. R OSENZWEIG, M . R .: Animal models for effects of brain lesions and for rehabilitation. In: P.
Bach-y-Rita (editor) Recovery of Function. Theoretical Considerations for Brain Injury Rehabilitation
(Baltimore: University Park Press), pp. 127± 172, 1980.
6. A NSELL, B . J. and K EENAN, J. E.: The Western Neuro-Sensory Stimulation Profile: a tool for
assessing slow-to-recover head-injured patients. Archives of Physical Medicine Rehabilitation, 70: 104±
108, 1989.
For personal use only.

7. H ELLWICK, L. D .: Stimulation programs for coma patients. Critical Care Nurse, 8: 47± 51, 1994.
8. L A P UMA , J., S CHIEDERMAYER, D . L ., G ULVAS, A . E . and S IEGLER, M .: Talking to comatose
patients. Archives of Neurology, 45: 20± 22, 1988.
9. L IPPERT-G RüNER, M ., Q UESTER, R . and T ERHAAG, D .: FruÈ hstimulation, ein Bestandteil des
fruÈ hrehabilitativen Behandlungskonzeptes auf der neurochirurgischen Intensivstation. Die
Rehabilitation , 2: 111± 115, 1997.
10. L E W INN, E . B . and D IAMANESCU , M . D.: Environmental deprivation and enrichment in coma.
Lancet, 2: 156± 157, 1978.
11. P IERCE, J. P., L YLE, D. M ., Q UINE, S. et al.: The effectiveness of coma arousal intervention. Brain
Injury, 4: 191± 197, 1990.
12. W OOD , R . L .: Evaluating sensory regulation as a method to improve awareness in patients with
altered states of conciousness: a pilot study. Brain Injury, 6: 411± 418, 1992.
13. H ALL, M ., M ACD ONALD , S . and Y OUNG, G. C.: The effectiveness of directed multisensory
stimulation versus non-directed stimulation in comatose CHI patients: pilot study of a single
subject design. Brain Injury, 6: 435± 445, 1992.
14. P FURTSCHELLER, G., S CHWARZ, G. and G RAVENSTEIN, N .: Clinical relevance of long latency
SEP’s and VEP’ s during coma and emergence from coma. Electroencephalography and Clinical
Neurophysiology, 62: 88± 98, 1985.
15. Z IEGER, A .: FruÈ hrehabilitation komatoÈ ser Patienten auf der neurochirurgischen Intensivstation.
Zentralbl. Neurochir., 53: 92± 113, 1992.
16. M ITCHELL , S., B RADLEY, V . A ., W ELCH, J. L. et al.: Coma arousal procedure: therapeutic inter-
vention in the treatment of head injury. Brain Injury, 4: 273± 279, 1990.
17. N AIR, M . K . C., S UJ A , M ., B ABU, G. et al.: Early stimulation: trivandrum model. Indian Journal of
Pediatrics, 59: 663± 667, 1992.
18. W HYTE, J. and G LENN, M . B .: Care and rehabilitation of patient in persistent vegetative state.
Journal of Head Trauma Rehabilitation, 1: 39± 53, 1986.