Nghiên cứu đặc điểm lâm sàng, cận lâm sàng và một số yếu tố liên quan của bệnh nhân nhồi máu não cấp trên lều tiểu não có thông khí cơ học tt tiếng anh

MINISTRY OF EDUCATION AND TRAINING MINISTRY OF NATIONAL DEFENSE

THE THESIS WAS DONE AT 108 INSTITUTE OF CLINICAL
MEDICAL AND PHARMACEUTICAL SCIENCES

108 INSTI TUTE O F CLINICAL MEDICAL AND
PHARMACEUTICAL SCIENCES

Scientific instructors:
1. Prof. PhD. Nguyen Van Thong
TRAN THI OANH

2. PhD. Nguyen Hong Quan

Reviewer 1:
STUDYING CLINICAL, SUBCLINICAL CHARACTERISTICS

Reviewer 2:

AND SOME RELAVANT FACTORS OF PATI ENTS WITH ACUTE
CEREBRAL INFARCTIONABOVETHE CEREBELLUM TENT

Reviewer 2:

REQUIRI NGMECHANICAL VENTI LATION

This thesis will be presented at Institute Council at:
Speciality: Neurology
Code: 62720147

Day

Month

Year

The thesis can be found at:
SUMMARY OF MEDICAL DOCTORAL THESIS
1. National library
2. Library of 108 Institute of clinical medical pharmacological sciences

HA NOI – 2019


1

2

INTRODUCTION

THE NEWCONTRIBUTIONOF THE THESIS

Stroke is one of the leading causes of death and disability in adults, with
about 80-85% of cerebral infarction. Patients with severe cerebral
infarction often have consciousness disorders, loss of ability to protect the
airway, sputum congestion, causing respiratory failure. Intubation and
mechanical ventilation for these patients is needed to support breathing to
protect the airway and ensure adequate oxygen supply to brain cells.
Although the proportion of patients with right ventilated cerebral infarction
is not high (10-16%), the prognosis is very poor. All patients have severe

1. The thesis topic has scientific, practical and topical significance to
contribute to show some clinical, subclinical and imagingcharacteristics of
patients with acute cerebral infarction above the cerebellum tent requiring
mechanical ventilation.
2. Determining a number of factors related to mechanicalventilation of
patients with acute cerebral infarction above thecerebellum tent, some
prognostic factors of mortality and prognosis ofgood functional state mRS
0-3 at the time of 1 year. This will helpclinicians predict and prescribe
timely treatment intervention.

clinical circumstances, complex evolutions, need many pos itive treatments
but high mortality. The death rate in the hospital is 35-75%. The patients

THE S TRUCTURE OF THE THESIS

who survived are mostly with severe neurological sequelae and dependent.

The thesisconsists of130pages: 2pagesof introduction, 36pages of
overview, 13pages of subjects and methods, 33pages ofresearchresults,

There have been many studies in the world and in the country about
patients with cerebral infarction in general but there have not been many
studies on patients with mechanical ventilation with cerebral infarction

43pages of discussion, 2pages of conclusionsand 1 page of suggestion, 40
tables,13 charts, 9 images, 1 figures and 148 references.

about factors related to ventilation indications and prognostic factors in
these patients.

Chapter 1 - OVERVIEW

Therefore, we conducted the subject: "Studying clinical, subclinical
characteristicand some relevant factors of patients with acute cerebral

1.1. Physiology of cerebral infarction

infarction above the cerebellum tent requiring mechanical ventilation"

A cerebral infarction occurs when the amount of brain blood falls below
18–20 ml / 100g brain / m inute, the centre of the infarction is the necrotic
area with a blood flow of 10-15ml/100g brain/minute, around this area

with two objectives as follows:
1. Description of clinical, subclinical characteristics and some factors
related to mechanical ventilation of patients with acute cerebral
infarction above the cerebellar tent.
2. Identify some of prognostic factors of patients with acute cerebral
infarction above the cerebellum tent requiring mechanical ventilation.

(Penumbra area) has a blood flow of 20-25ml/100g brain / minute,
although brain cells are still alive but inactive. The area cells die over a few
hours and are different for every patient. This is the window time for
reperfusion treatment interventions. Treatment measures to save this area.
1.2. Edema in cerebral infarction


3

4

Cerebral edema in a large cerebral infarction causes increased
intracranial pressure, which can lead to a brain hernia, aggravate
neurological deficiencies and high mortality if left untreated. The clinical
development of cerebral edema in p atients with massive cerebral infarction
can be divided into 3 levels: fulminant (within 24-36 hours), slowly (over
several days), or initial acute course then descending (about a week).
Cytotoxic Edema: Once clogged, there is a stop of oxygen exchange in the
damaged area which leads to the cell losing energy, losing the function of
the transport membrane, the ion pump stops working, Na + from outside
spills into the cell, dragging water causes the cell to swell causing cytotoxic
edema. This type of edema does not respond to anti-edematous drugs
according to the osmotic mechanism.
Vasogenic Edema: Occurs 4 to 12 hours after embolization, due to a
profound change in the endothelial lining of the capillaries, stagnation of
glycogen in stellar cells, causing bulging star cells, breaking the tight bonds
between intracellular cells tissue and between endothelial cells and stellar
cells leads to blood barrier brain (BBB), the fluid from the lumen of the
artery is released causing brain edema. Brain edema b ecomes the strongest
on the third to fifth day and is reduced after one to two weeks. This type of
edema responds to anti-edematous drugs according to the osmotic
mechanism.
As recommended by the American Heart Association/American Stroke
Association in 2014, the signs predict malignant cerebral edema and poor
prognosis on cranial CT include increased mid-cerebral artery photon, dot
sign on film within 6 hours, infarction of one-third or more of the midcerebral artery blood supply region, or midline shift push of 5 mm or more
on the cranial CT scan in the f irst 2 days is also associated with increased
nerve damage and death early in the acute phase. The American Heart
Association/American Stroke Association (2014) recommends serial CT
scans during the first 48 hours of stroke to assess the risk of malignant

brain edema.

1.3. Indications and role of mechanical ventilation in patients with

cerebral stroke
The most common causes of hypoxemia in brain stroke patients may be
due to partial obstruction of the airways due to sputum stagnation,
respiratory depression and hypoventilation, choking pneumonia and
collapse. In these cases, mechanical ventilation helps improve blood
oxygenation, maintain oxygenation to the brain and reduce intracranial
pressure, but excessive ventilation should be avoided. SpO2 target> 94%
and pCO2 35 - 40 mmHg. In patients stroke with impaired consciousness,
or signs of brain stem dysfunction, decreased oropharyngeal motion and
airway reflex loss are at high risk of choking pneumonia. Intubation for this
patient is necessary to protect the airway and prevent choke complications.
Some patients have coma, disorders of breathing, have apnea, intubation
and mechanical ventilation to ensure respiration, ensure oxygen to the brain
and body to prevent brain edema progression.
The American Heart Association/Stroke Association 2014recommends
for mechanical ventilation in the treatment of acute cerebral infarction:
Intubation may be considered for patients with decreased levels of
consciousness resulting in poor oxygenation or impaired control of
secretions.
1.4. Hyperventilation and role of pCO2 in treatment intracranial pressure
Reducing pCO2 is known as a cerebral artery contraction that reduces
cerebral blood flow leading to a reduction in intracranial pressure, mainly
due to changes in pH around the blood vessels. The effect of reducing
cerebral blood flow is temporary, after 4 hours brain blood flow has been
restored 90%. In addition, a rapid increase in pCO2 causes a decrease in the
pH around the blood vessels, causing vasodilation to increase brain blood

volume and increase intracranial pressure ("rebound hyperemia"). Use
hyperventilation should only be used short in cases of life-threatening
increase in intracranial pressure, pending surgical intervention. pCO2
should be normalized as soon as possible.


5

6

Chapter 2 - SUBJECTS AND METHODS

Clinical variables: gender, age, medical history, time of admission, pulse,
blood pressure, temperature, level of consciousness at admission on
Glasgow scale, NIHSS score, degree of paralysis, language disorder,
sensory disorders, urinaryincontinence, pupil abnormalities, light reflexes,
head-eye deviation, progression of symptoms, related mechanical
ventilation complications.
Subclinical variables: hematology, biochemistry, coagulation, arterial blood
gases.
Imaging variables: CT, CTA, DSA: parenchymal, artery damage, midline shift.
Variables of treatment outcome: death, live, mRS at discharge, 1 year.
2.2.4. Research contents
Patients were divided into 2 groups: MVgroup and non MVgroup. The
patients were divided into two groups, the MVgroup and the non MV group.
MVis indicated when at least one of the following criteria: Glasgow ≤ 8, loss
of reflexes protects airway causing mucus congestion, patients with
consciousness disorders, stimulation must use safety drugs strong spirit causes
respiratory depression, patients with respiratory failure, circulatory failure.
Describe the clinical and paraclinical features with analysis and comparison

between two groups of MV and non MV group to highlight c linical and
subclinical characteristics of patients requiring MV.
Identify factors related to MV, factors related to prognosis of death at
hospital discharge and mRS 0-3 at 1 year. The supposedly relevant
variables are included in univariate analysis and logistic multivariate
regressions to find meaningful prognostic factors.
2.3. Data analysis
Data processing using SPSS 16.0 software.
Description of clinical, subclinical, imaging features: neurological signs
on onset, on admission and during hospitalization, intubation designation,
subclinical characteristics, imaging, complications during MV, treatments
and outcome.

2.1. Studying subjects
Severe cerebral infarction patients above the cerebellum tent were
treated at Strokecenter-Central MilitaryHospital108from 9/2013 – 6/2017.
2.1.1. Criteria for selecting a patient
The patients was diagnosed as stroke according to the World Health
Organization (1989) stroke definition, arriving at the hospital 72 hours
prior to the onset of cerebral infarction. Images of hemispherical infarction
on CT/MRI/ Severe nerve damage with NIHSS≥15 score (if the patient was
hospitalized prematurely, the damage on the first CT was unknown, the
patient would be diagnosed for a second time on CT. Patients were divided
into 2 groups: mechanical ventilation group and non mechanical ventilation
group.
2.1.2. Exclusion criteria
History of stroke with mRS score> 2 points, patients with severe
medical conditions such as liver failure, severe kidney failure, cancer,
COPD,…
2.2. Research methods:

2.2.1. Study design:Progressive, description, follow-up study
2.2.2.Sample size:
Sample size is determined by formula:
p (1-p)
n = Sample size to study
2
2
n = Z (1-α/2 ) -------Z (1-α/2) : At the probability level 95% (Z =1,96)
2
d
d: The desired accuracy (d = 0,05)
p: Estimated ratio, the rate ofcerebral
infarctionpatients
requiring
mechanical
ventilationin previous studies, p= 0,11.
→ Based on the above formula, the estimated patient sample sizeis 150.
In the period of taking data from 9/2013 – 6/2017, we collected 166
patients including 84 ventilated severe cerebral infarction patients and 82
severe cerebral infarction patients without mechanical ventilation.
2.2.3. Research variables


7

8

Analys is of related factors: Chi-square test of qualitative or quantitative
variables with clustering. Statistically significant variables in Chi-square
test were included in univariate regression analys is. Variables related to

MVand mortality in univariate analysis with significance level p <0,05
were included in multivariate regression analysis to identify independent
prognostic variables.
2.4. Diagramresearch

Table 3.4. Neurological symptoms onset

ACUTE ISCHEMIC STRO KE
- Images of HI on CT/MRI
- NIHSS≥15if the damage on the first CT is
unknown, will be diagnosed for 2nd on CT
(n = 166 patients)

ISC HEMIC
STROKE without

MV
(n = 82patients)

ISC HEMIC
STROKE with MV

(n = 84patients)

Non MV(n=82)
p
n
%
n
%

Conscious disorders onset
67
79,8
31
37,8 <0,001
Lips/ coma does not say
53
63,1
41
50
>0,05
Hemiplegia in the face
83
98,8
82
100
>0,05
Hemiplegia
84
100
82
100
>0,05
Headache
9
9,5
5
8,0
>0,05
Dizzy

8
9,5
7
9,1
>0,05
Vomiting /nausea
14
16,7
3
3,7
<0,05
Urinary incontinence
56
66,7
16
19,5 <0,001
Comment: Disturbances of consciousness, vomiting/nausea, and urinary
incontinence were statistically different with p <0,05.
 Neurologic symptoms at hospital arrival
Table 3.5. Neurologic symptoms at hospital arrival
Symptoms

Symptoms
1. Descripti on of cli nical ,
subcl ini cal chara cteristics and
so me factor related to MV

Clinicalcharac
teristics

Subclinical
charac ter istics

Related factors
of MV

2. Identi f y some of prognosti c
f actors in pati ents ischem ic
stroke with MV

Prognostic factors
of mortality

Related factors of
mRS 0-3 at 1 year

Chapter 3 – RESULTS
3.1. Clinical, subclinical characteristics and some factors related to MV in
patients with acute cerebral infarction above the cerebellar tent.
3.1.1. Clinical symptoms
 Neurological symptoms onset

Average Glasgow score
Average NIHSS score
Glasgow point at
admission≤10
NIHSS point at admission>20
Head-eye deviation
Dilated pupils admission
Language d isorders

Urinary incontinence
admission
Severe paralysis
(Muscle strength 0/5-1/5)
Average strength of arm
Average strength of leg

MV(n=84)

MV(n=84) Non - MV(n=82)
p
n
%
n
%
10,31 ± 2,02
11,84 ± 1,95
<0,001
22,82 ± 5,39
19,90 ± 3,73
<0,001
51

60,7

27

32,9

<0,001

52
36
10
83

61,9
42,9
11,9
98,8

34
16
1
82

41,5
19,5
1,2
100

< 0,01
<0,001
< 0,01
> 0,05

62

73,8

4

4,9

<0,001

78

92,9

64

78

< 0,01

0,62 ± 0,94
0,87 ± 1,05

< 0,05
< 0,01

0,32 ± 0,64
0,45 ± 0,67


9

10

Comment: The Glasgow average score was lower for the MV group than
for the non MV group. The average NIHSS score for MV was higher than
for the non MV group. Signs of head-eye deviation met 42,9% in the MV
group, higher than the MV group in 19,5%. Severe paralysis ½ people in 2
groups are 92,9% MV group and 78% MV group.
 Some characteristics related to mechanical ventilation
Table 3.10. Indication for intubation

Sum
34
40,5
≤ 3 days
13
38,2
> 3 days
21
61,8
Mean duration fromintubationtotracheostomy (days)
3,74 ± 1,21
Mean duration of MV
≤ 3 days
31
36,9
4 – 7 days
46
54,8
≥ 8 days
7
8,3
Average MV t ime (days)

4,40 ± 2,28
Comment: The rate of intubation was mainly in the first and second day
after admission. The rate of tracheostomy early 38,2%. The group with MV
time of 4-7 days had the highest rate of 54,8%.

Indication for intubation
Nerve

n

%

Glasgow ≤ 8

18

21,4

Protect airway

23

27,4

96,4

40
3

47,6

3,6

3,6

Progession of symtoms
Respiratory failure, circulatory failure

Sum
84
100
100
Comment: Only 3,6% indicated intubation due to respiratory failure,
circulatory failure. 96,4% indicated intubation related to nerve.

Chart 3.7. Time of intubation
Table 3.11. Characteristics in mechanical ventilation
Characteristics
n
%
Average intubation time from admission (days)
1,64 ± 0,91
Intubation in the first day
49
58,3
Intubation for the first 2 days
69
82,1
Successful extubation
50
59,5

Time of
tracheostomy

Table 3.12. Arterial blood gas averages the points
The first day after MV
Day 3 after MV
Factors
(n= 84)
(n= 58)
pH
7,436 ± 0,057
7,439 ± 0,048
pCO2
36,8 ± 8,9
37,9 ± 6,75
pO2
132,3 ± 54,3
112,8 ± 37,15
HCO3
24,59 ± 4,08
25,99 ± 3,29
Comment: Blood g as factors at the time of day 1 after MV and day 3 after
MV had reached the goal.
Table 3.15. Complications related to MV
Complications
n
Pneumonia
30
Reflux

21
Gastrointestinal bleeding
5
Canyn around bleeding
1
Re endotracheal intubation
4

%
35,7
25
5,95
1,2
4,8

Comment: Common complications: pneumonia 35,7%, reflux 25%


11

12

3.1.2. Image characteristics
Table 3.19. Characteristics of images on CT scan first
MV
Non -MV
Sum
Image characteristics
(n=84)
(n=82)

(n=166)
n
%
n
%
n
%
Hypodensity
60
71,4
47
57,3
107
64,5
Early brain imaging
5
6
3
3,7
8
4,8
Unknown damage
19
22,62
32
39,0
51
30,7
p
>0,05

Comment: 64,5% of patients had hypodensity at the first CT scan on
admission, 30,7% had no damage of images.
Table 3.22. Image of edema brain on CT scan
MV
Non -MV
Sum
(n=84)
(n=82)
(n=166)
Image of edema brain
None
Only blurry brain groove

n
0
14

%
0
16,7

n
27
21

%
32,9
25,6

n

27
35

%
16,3
21,1

Blurry brain groove and
8
9,5
17
20,7
25
15,1
ventricular collapse
Midline shift
62 73,8
17
20,7
79
47,6
Sum
84
100
82
100
166
100
p
<0,05

Comment: Image of edema brain differs between the two groups
Table 3.23. Midline shift classification on CT scan
MV
Non - MV
Midline shift
(n=62)
(n = 17)
n
%
n
%
Degree 1
9
14,5
8
41,7
Midline shift
Degree 2
16
25,8
6
35,3
classification
Degree 3
37
59,7
3
17,6
p
<0,01

Average midline shift(mm)
10,04 ± 4,69
5,25 ± 3,43
p
<0,001
Comment: Midline shift difference between the 2 groups p<0,001.

3.1.3. Some factors related to mechanical ventilation in patients with
acute cerebral infarction above the cerebellar tent.
Results of univariate analysis including 12 clinical and subclinical
variables with statistical significance related to MV were included in the
multivariate regression analysis.
Table 3.30. Factors related to mechanical ventilation in multivariate analysis
Factors
OR
95%CI
p
Ages > 60

0,378

0,137 - 1,045

0,061

Conscious disorders onset

5,097

1,752 - 14,832

0,003

Vomiting /nausea onset

6,586

1,138 - 38,131

0,035

Urinary incontinenceonset

8,027

2,628 - 24,518

0,000

Glasgow point at admission≤10

0,888

0,298 - 2,639

0,830

NIHSS point at admission> 20

0,790

0,260 - 2,397

0,677

Head-eye deviation

1,992

0,661 - 6,002

0,221

Dilated pupils admission

7,699

0,443 - 133,935

0,161

Temperature admission> 37,50 C

5,228

0,929 - 29,416

0,061

Pules > 90l/p

1,700

0,641 - 4,508

0,287

Leukocytes > 10G/l

3,212

1,149 - 8,982

0,026

Midline shift > 5mm

13,511

4,392 - 41,560

0,000

Comment:In the multivariate analysis, 5 variables were statistically
significant: onset consciousness consciousness, vomiting/nausea onset,
urinary incontinenceonset, leukocyte > 10G / l, midline shift > 5mm.
3.2. Study some prognostic factors in patients with cerebral infarction
in upper cerebellum tent with mechanical ventilation
3.2.1. Clinical outcome

13

Chart3.12. Functional status upon discharge
Comment: In the MV group, no patients had a level of mRS 1-2. Mortality
(mRS 6) 34,5%. In the group with no MV, mRS 4-5 was 68,3%.
3.2.2. Some of factors related to prognosis of death in patients with
cerebral infarction in the cerebellum tent with MV
Table 3.32. Some clinical factors related to clinical outcome at discharge in
patients with cerebral infarction abovethe cerebellum tent with MV
Clinical outcome at discharge
Factors
p
Dead (n=29) Alive (n=55)
Conscious disorders onset
22 (75,9)
45 (81,8)
0,518
Vomiting /nausea onset
5 (17,2)
9 (16,4)
0,918
Urinary incontinenceonset
25 (86,2)
31 (56,4)
0,006
Average Glasgow score
10,31 ±2,12
10,29 ±2,01 0,967
Glasgow point at admission≤10

18 (62,1)
33 (60)
0,854
Average NIHSS score
22,76 ± 5,65 22,84 ± 5,34 0,951
NIHSS point at admission> 20
13 (44,8)
36 (65,5)
0,356
Decreased Glasgow ≥ 2 at intubation
18 (62,1)
25 (45,5)
0,173
Glasgow point ≤8 at intubation
18 (62,1)
28 (50,9)
0,364
Intubation in the first day
17 (58,6)
32 (58,2)
1,000
Aggravation in the first 48 hours
22 (75,9)
41 (75,4)
1,000
Head-eye deviation
13 (44,8)
23 (41,8)
0,791

14
Dilated pupils admission
21 (72,4)
19 (34,5)
0,001
Loss of light reflection
24 (82,8)
10 (18,2)
0,000
0
Temperature admission>37,5 C
11 (37,9)
6 (10,9)
0,003
AverageSBPadmission
142,79±33,55 145,56 ±21,93 0,65
AverageDBPadmission
85,93 ± 15,47 85,58 ± 14,39 0,918
Averagepulesadmission
89,72 ± 22,22 87,58 ± 19,84 0,653
Pneumonia
7 (24,1)
23 (41,8)
0,108
Comment: Factors with statistical significance: urinary incontinence onset,
dilated pupils admission, loss of light refraction, temperature admission
0
>37,5 C
Table3.33. Some subclinical factors related to clinical outcome at hospital
discharge in patients with cerebral infarction in the cerebellum tent with MV

Clinical outcome at discharge
Factors
p
Dead (n=29) Alive(n=55)
Leukocytes > 10G/l
18 (62,1)
39 (70,9)
0,409
Blood g lucose >11,1 mmol/l
5 (17,2)
5 (9,1)
0,303
pCO2 on the first MV < 35mmHg
12 (41,4)
17 (30,9)
0,337
Midline shift >5mm
23 (79,3)
30 (54,5)
0,025
Table 3.34. Some factors are associated with mortality prognosis in
univariate regression analysis
Factors
OR
95%CI
p
Urinary incontinenceonset
4,839
1,483 - 15,784 0,009
0

Temperature admission>37,5 C
4,991
1,609 - 15,480 0,005
Dilated pupils admission
5,500
1,998 - 15,139 0,001
Loss of light reflection
16,063 5,290 - 48,778 0,000
Midline shift >5mm
3,194
1,125 - 9,070
0,029
Comment: The factors in the table are all related to statistically significant
mortality outcomes in univariate analys is.


15

16

Table 3.35. Some factors related to mortality prognosis in multivariate
logistic regression analysis

Chapter 4 – DISCUSSION

Factors
OR
95%CI
p
Urinary incontinenceonset

4,326
1,062 - 17,617
0,041
0
Temperature admission>37,5 C
3,087
0,636 -14,967
0,162
Dilated pupils admission
1,149
0,204 -6,481
0,875
Loss of light reflection
22,426 2,324 - 216,392
0,007
Midline shift >5mm
0,819
0,172 -3,899
0,802
Comment: When analyzing multivariate logistic regression, the factors
associated with mortality outcome were statistically significant: urinary
incontinenceonset, loss of light reflection.
3.2.3.Some factors related to mRS 0-3 good function at 1 year
- There are 55 patients discharged. At 1 year after discharge, there were 7
patients losing follow-up, 11/48 patients died, 20,3% mRS 0-3.
Table 3.39. Some prognostic factors of good functional mRS 0-3 at 1 year
in univariate regression analysis
Factors
OR
95%CI

p
Ages > 60
0,102
0,024 - 0,438
0,002
Pneumonia
0,229
0,055 - 0,957
0,043
Comment:Factors with a negative predictive effect on good functional mRS
0-3 at 1 year include: ages > 60, pneumonia
Table 3.40. Some prognostic factors of good functional mRS 0-3 at 1 year
in multivariate regression analysis
Significant variables related to the good functional mRS 0-3 at 1 year in
univariate analysis were included in the multivariate regression analys is.
Factors
OR
95%CI
p
Ages > 60
0,091
0,019 - 0,427
0,002
Pneumonia
0,192
0,038 - 0,962
0,045
Comment: When analyzing multivariate logistic regression, the negative
predictive factors giving good results of mRS 0-3 recovery at 1 year of
statistical significance include: ages > 60, pneumonia.

4.1. Clinical, subclinical characteristics and some factors related to MV
in patients with cerebral infarction above the cerebellum tent
4.1.1. Clinical characteristics, imagings
- Conscious disorders onset
The study results showed that 79,8% of patients with MV had
consciousness disorder onset compared with 37,8% in the group without
MV(p<0,001). Santoli (2001) reported 69% of patients had consciousness
disorder onset. In Gupta’s study (2014), 60% of stroke patients had onset in
MV group compared with 12% in non-MVgroup (p <0,05). In major
cerebral infarction, which causes widespread cerebral infarction, some
cases of early cognitive impairment due to the influence of the neural
activation n etwork in the lower part of the hypothalamus by damage to the
hemisphere.
- Vomiting/nausea onset
Vomiting/nausea is an uncommon manifestation of cerebral infarction.
In the study, the group with MVhad the rate of vomiting / nausea was
16,7%, different from the group with MVwithout 3,7% (p <0,05). Gupta
(2014) showed signs of vomiting in 43,3% of patients with MV and 14,3%
with non-MV(p <0,05). Signs of vomiting more likely may be due to
studies conducted in both patients with cerebral infarction and cerebral
hemorrhage.
-Urinary incontinence onset
The study results showed that urinary incontinencein the MVgroup was
66,7% and this sign in the MV group was 17,1%. Doan Thi Huyen (2009)
studied a group of large brain infarcts with the ratio of urinary incontinence
was 66,67%. In the study of Nguyen Van Tuyen (2013) also recorded the
rate of urinary incontinencein patients with MVwas 97,01% compared to
the group without MV2,08% (p <0,05).

17

18

- Glasgow points at admission
The mean Glasgow score was 10,31 ± 2,02 in the MVgroup and 11,84 ±
1,95 in the non - MVgroup (p <0,001). 20,2% of MVgroup patients had a
score of Glasgow ≤ 8 on admission. In both groups, the proportion of
patients with a score of Glasgow 9-12 was predominant (63,1% in theMV
and 59,8% in the non-MVgroup). The score of Glasgow admission in
Mengi's study (2018) was 11,5 ± 2,78.
- The degree of nerve damage conform the NIHSS scale
The NIHSS score is an indicator of the severity of nerve damage. The
average NIHSS admission to hospital in the MV group (22,82 ± 5,39) was
higher than that in the non-MV group (19,90 ± 3,73), p <0,001. In Santoli’s
study (2001), the average NIHSS score was 21,12 ± 5.
- Head-eye deviation
Head-eye deviationusually appear after a large cerebral infarction in the
tent, either due to lesions around the 8th region of the pre-motor region of
the upper frontal lobe or due to severe brain damage brain edema leading to
brain compression and appearance this sign. Head-eye deviation is a
serious prognosis factor in stroke patients. Research results showed that
29,4% of patients showed signs of head-eye deviation, and met a higher
rate in the MVgroup was 42,9% compared with 19,5% in the non- MV
group. In the Gupta study (2014), head-eye deviation in the MV group was
17,3% and in the non-MV groupwas 10,9%.
- Signs of mydriasis
In the MVgroup, 11,9% of patients showed signs of mydriasis, in the
non -MVgroup, this rate was 1,2% (p <0.01). When there was aggravation

of neurological d eficiencies, the rate of mortality was 47,6% of patients in
MVgroup. In the Gupta’s study (2014), the group of patients with MVwho
had MVhad the rate of mydrias is abnormalities of 17,3% and the group of
MV with 10,9% (p = 0,25). In the Gujjar’s study(1998), the rate of
mydriasis abnormalities was 16% in mechanicallyventilated cerebral
infarction patients.

- Indication of intubation
The study results showed that 96,4% of intubation and MVrelated to
nerves (including Glasgow score ≤ 8 was 21,4%, loss of airway protection,
risk of sputum congestion was 27,4% and nerve progression was 47,6%).
Traditionally, some authors have indicated intubation and MVwhen
impaired consciousness with a score of Glasgow ≤8, or consciousness
preserved but impaired oropharyngeal function,congestion of phlegm. The
study results showed that 45,2% of patients with a score of Glasgow 9-10
when intubated had neurologic progression, or loss of airway protective
reflexes.
According to Nguyen Hong Quan (2012), indication ofintubation
related to nerves was 85,6%. In published studies, intubation indicated was
mostly related to nerves: Gujjar’s study (1998) was 82%; Schielke's study
(2005)was 71%; Berrouschot’s study (2000) was 90%; Milhaud'sstudy
(2004) was 86%.
- The time of MV
The average MV time was 4,40 ± 2,28 days, the group of MVpatients 47 days accounted for the highest proportion (54,8%). In Berrouschot’s
study (2000), averageMV time was 172 ± 182 hours (7,17 ± 7,58 days),
Santoli's study (2001), averageMV time was 8,6 ± 8,8 days. Popat's study
(2018) averageMV time was 6,5 ± 5,9 days. In massive cerebral infarction,
adverse neurologic events usually occur within the first 48 hours, especially
malignant brain edema, requiring MV. At the end of the period of severe
cerebral edema, the patient has adequate self-breathing, can stop MV, avoid

prolonged MV, limiting the complications of MV.
- Characteristics of arterial blood gases
The pO2 index on the first day:132,3 ± 54,3 and 112,8 ± 37,15 on day
3. This was appropriate because the patients in the study had good
pulmonary ventilation and 96,4%of intubation was related to nerves.
Arterial b lood gas on the first day after MV pH 7,436 ± 0,057, pCO2 36,8
± 8,9 mmHg, on day 3 MVpH 7,439 ± 0,048, pCO2 37,9 ± 6,75 mmHg.


19

20

This result was similar to the result of arterial blood gas in Pelosi’s study
(2011), pH 7,4 ± 0,1 and pCO2 37 ± 13 mmHg. In the practice of MV in
patients with brain stroke or traumatic brain injury, the level of pCO2 is
recommended to be maintained at 35-40 mmHg, in case of severe cerebral
edema, hyperventilation may be achieve at pCO2 level of 30 -35 mmHg
during short.
- Complications related to MV were recorded: pneumonia 35,7%,
gastroesophageal reflux 25,0%, gastrointestinal bleeding 5,95%. These are
often the complications of MV patients in general in intensive care units.
The rate of pneumonia in the study of Nguyen Van Tuyen (2013) was
recorded as 13,43%, in the study of Catalino (2018) was 20,8%.
- Images
On the first CT scanner at hospital arrival, 64,5% of patients had images
of parenchymal reduction (MVgroup 71,4%, non -MVgroup 57,3%),
30,7% of patients hadunknown damage. In the MVgroup, 100% of the
patients had mass effect levels, of which 73,8% had midline shift. The rate
of midline shiftin the non - MVgroup was 20,7%. Gupta’s study (2014): the

rate of midline shiftin the MVgroup was 53,3% while the non-MV group
was 0,8%. In the MVgroup, in 62 patients who had midline shift, the rate of
midline shift> 5mm accounted for 85,5%, midline shift level 3 (> 10mm)
accounted for 59,7%.
In our study, midline shift> 5mm was a factor related to MVin
univariate analysis (OR = 13,867; p = 0,000) and multivariate (OR =
13,511; p =0,000). The midline shift> 5mm was also a predictor of
mortality in univariate analysis (OR = 3,194; p = 0,029).
In clinical practice, doctors predict an increase in intracranial pressure
or degree of brain compression indirectly through visual compression of
the mediastinum or medial medial compression. The midline shift on CT
scannerwas one of the most obvious evidence of cerebral edema.
4.1.2. Factors related to MV in patients with cerebral infarction in the
cerebellum tent

Factors related to MV in multivariate regression analysis include:
conscious
disorders
onset,
nausea/vomiting
onset,
urinary
incontinenceonset, leukocyte > 10G/land midline shift > 5mm.
- Conscious disorders onset
Research results show that conscious disorders onsetwas a factor related
to MV(OR = 5,097; p = 0,003). The onset cognitive disorder associated
with MV was a factor mentioned in some studies. According to
Berrouschot (2000), consciousness disorders was related to mechanical
ventilation (OR = 2,37; p = 0,0002). In Gupta’s study (2014), the
prevalence of conscious disorders onsetin the MV group was 60%

compared to 12% in the nonMVgroup (p <0,05); and loss of consciousness
onset associated with MV(OR = 13,345, p = 0,022).
- Vomiting/nausea onset
Our results show that vomiting is related to MVin multivariate
regression analysis (OR = 6,586; p = 0,035). Gupta’s study (2014) showed
signs of vomiting in 43,3% of patients on MVgroup and 14,3% in non
MVgroup (p <0,05) and signs of vomiting in patients group died 56,2% vs
28,6% in the surviving patient group.
- Urinary incontinence onset
Urinary incontinenceonset of stroke is a factor related to MVin multiple
variables (OR= 8,027; p = 0,000). In Gupta’s study (2014), the prevalence
of urinary incontinencein MVgroup was 90% compared to 31,8% in non
MVgroup (p <0,05). It is possible that Gupta's study included patients with
cerebral infarction and cerebral hemorrhage, so the incidence of urinary
incontinencewas also higher. In a study of moderate and large brain
haemorrhage in a tent, Nguyen Van Tuyen (2013) also recorded the rate of
urinary incontinencein MVgroup was 97,01% compared to those
nonMVgroup 52,08% (p <0,05).
- Leukocytes> 10G/l
When invest igating factors related to MV, leukocytes> 10G/l are
relevant in multivariate analysis (OR = 2,708; p = 0,046). In the study of


21

22

Tran Ngoc Tai (2005), the group of patients with cerebral infarction
cerebral death died had a higher leukocyte index than the group of
surviving patients (12,2 G/l and 11,3 G/l, there were no statistically

significant difference). The proportion of patients with leukocyte>10G /l
was higher in the group of mechanical ischemic cerebral infarction (67,9%
and 48,8% p <0,05).
- Midline shift > 5mm
In our study the group of MVpatients were all very severe. Signs of
intraventricular tamponade and medial compaction indirectly reflect the
degree of cerebral edema on cranial computerized tomography. Research
results, midline shift > 5mm are factors related to MVdesignation in
multivariate analysis (OR = 13,511; p = 0,000).
4.2. A number of prognosis factors for patients with cerebral infarction
above the cerebellum tent have MV
4.2.1. About a some of prognostic factors of mortality
The prognostic factors of mortality in multivariate regression analysis are:
urinary incontinence onset, loss of light reflection.
- Urinary incontinence onset
The results of the study showed that the prevalence of urinary
incontinence onsetin patients who died was 86,2% compared to 56,4% in
patients alived (p = 0,006). Urinary incontinence onsetof stroke were
prognostic factors of mortality in univariate analysis (OR = 4,839; p =
0,009) and multivariate (OR = 4,326, p = 0,041). Gupta's (2014) study
showed that the rate of urinary incontinence onsetin p atients who died was
96,9% compared to 82,1% of the alived patients (p = 0,088). Li et al.
(2018) reported that urinary incontinence was more in the mRS 4-6 group
than in the mRS 0-3 group (27,9% and 7,1%; p <0,001).
- Lose light reflection
Patients with hemispherical infarction have unilateral or bilateral
irradiation loss as a result of severe cerebral hypoxia or tenteral hernias that
press on nerve III in the brain stem. Often in these cases, consciousness
will deteriorate rapidly. There should be indications of timely emergency

interventions, including MV. In our study, the proportion of patients
showing signs of mydriasis loss with light met 75,9% in the death group
and 16,4% in the alived patient group (p = 0.001). Lose light reflectionis a
predictor of death in multivariate regression analysis (OR = 22,426; p =
0,007).
Some authors have also shown an association between signs of pupil
loss of reflex and severe prognosis and death in mechanically ventilated
stroke patients. Steiner's study (1997) recorded the loss of pupil reflexes
with light has significant prognosis of death at 2 months. Bushnell (1999)
found that losing pupil reflexes to light was an independent 30-day
prognosis factor. Santoli (2001) also noted that irradiation loss to light is
also a prognostic factor of death at 1 year (RR = 1,57; p = 0,021). Gupta’s
study (2014) loss of light reflection when intubation was associated with
mortality prognosis (p = 0,015).
4.2.2. Some factors related to functional status at 1 year tim e.
Prognostic factors for good recovery of mRS 0-3 in multivariate
regression analysis include: ages > 60 years, pneumonia.
- Ages> 60
Advanced age has been reported to adversely affect the outcome of
treatment and rehabilitation in patients with extensive brain infarction in
many studies. Research results at 1 year, among patients> 60 years of age
good recovery rate of mRS 0-3 is 7,7%, while in patients ≤ 60 years old,
the recovery rate of mRS 0-3 is 36,8% (p = 0,001). Age above 60 has
negative predictive meaning with good recovery results mRS 0-3 at 1 y ear
in univariate regression analysis (OR=0,1102; p = 0,002) and multivariate
(OR = 0,091; p = 0,002).
In Huang's study (2012) assessed 6-month recovery of patients with
extensive cerebral infarction, in patients> 60 years the recovery rate of
mRS 0-3 met 36,4% while the recovery rate inferior to mRS 4-6 was 63,6%
(p = 0,036). In Foerch's study (2004), age over 75 years had a negative

predictive effect for survival at 6 months after stroke with MV (OR = 0,1; p
= 0,004).


23

24

- Pneumonia
The study results showed that the rate of pneumonia was 17,6% in the
mRS 0-3 recovery group and 48,4% in the 4-6 mRS poor recovery group (p
=0,035). Pneumonia had a negative predictive effect with good recovery of
mRS 0-3 at 1 year in univariate regression analysis (OR = 0,222; p = 0,043)
and multivariate (OR = 0,192; p = 0,045). This result is similar to the result
of Li (2018) when studying the factors related to the good outcome of mRS
0-3 in patients with massive cerebral infarction at 3 months recorded the
rate of pneumonia seen in the group. mRS 0-3 and mRS 4-6 group were
67,1% and 36,3% respectively; (p<0,001) and predictive negative with
good recovery results mRS 0-3 at 3 months in univariate analysis (OR =
0,42; p <0,05).

Indication for intubationdue to Glasgow point ≤8 21,4%, respiratory
protection 27,4% and neurological progress worsens 47,6%;
 58,3% patients was intubated on the first day;
+ Pneumonia 35,7%;
 Factors related to mechanical ventilation in patients with cerebral
infarction in the cerebellar tent:
- Regression analysis of multivariate logistis, the factors related to MVwith
statistical significance with p <0.05 include: consciousness disorder onset
(OR = 5,097), vomiting onset ( OR = 6,586), urinary incontinenceonset

(OR = 8,027), leukocytes>10G/l (OR = 3,212), midline shift> 5mm(OR =
13,511).
2. A number of prognostic factors in patients with cerebral infarction
above the cerebellum tent with MV
 Some prognostic factors of death:
- Regression analys is of multivariate logistis, prognostic factors of death at
the hospital with statistical significance with p <0,05 include: urinary
incontinence (OR =5,901), loss of light reflection (OR = 22,949).
 Some factors related to good functional condition mRS = 0-3 at 1 year:
- Regression analysis of multivariate logistis, the predictive negative
factors with good functional status mRS 0-3 at 1 year with statistical
significance with p <0,05 including: ages> 60 (OR =0,091), pneumonia
(OR = 0,192).

CONCLUSIONS
The study involved 166 patients with cerebral infarction include 84
MVpatients and 82 non - MVpatients who were hospitalized for 72 hours
of stroke onset, treated at the Stroke Center - Central Military Hospital 108
from 9/2013 – 6/2017. The research results are as follows:
1. Clinical, subclinical characteristics and some factors related to MV
in patients with cerebral infarction above cerebellum tent.
 Clinical and subclinical characteristics:
+ Conscious disorders onset: 79,8%.
+ Average Glasgow score at admission 10,31 ± 2,02;
+ Average NIHSS score admission2,82 ± 5,39.
+ Severe hemiplegia (92,9%), head-eye deviation(42,9%),vomiting
onset (16,7%), urinary incontinenceonset (66,7 %); admission temperature
above 37,5C (20,2%).
 75% of neurological progress worsens in 48 hours.
Middle artery occlusion 43,9%.

+ Midline shift73,8%; 85,5% of patients had midline shift> 5mm, an
average ofmidline shift10,04 ± 4,69mm.

RECOMMENDATIONS
1. Factors related to MVnoted in the results of the study can be considered
for the design of mechanical ventilation in patients with cerebral infarction
in cerebellar tent.
2. A number of factors related to prognosis of death recorded in the study
can help clinicians have more information about prognosis of patients,
indicating better treatment options.


PUBLISHED ARTICLES RELATED TO THE THESIS

1. Tran Thi Oanh, Nguyen Van Thong, Nguyen Hong Quan (2015),

Study on clinical and imaging features of large cerebral infarction
patients above cerebellar tent with mechanical ventilation. Journal of
Medicine and Pharmacy Clinical Studies 108, volume 10 - Issue 9/2015,
pages 147-153.
2. Tran Thi Oanh, Nguyen Van Thong, Nguyen Hong Quan, Nguyen
Van Tuyen (2018), Study on some factors related to mechanical
ventilation in patients with severe hemispherical infarction patientsJournal of Medicine and Pharmacy Clinical Studies 108- 2018, Episode 13
- Special Numbers, pages 176 - 182.
3. Tran Thi Oanh, Nguyen Van Thong, Nguyen Hong Quan, Nguyen
Van Tuyen (2018), Study some prognosis factors in hemispherical
cerebral infarction patients with mechanica l ventilation- Vietnam
Journal of Medicine 471 - Special Issue - October, pp 152 -158.