Nghiên cứu kết quả cầm máu bằng kẹp clip đơn thuần và kẹp clip kết hợp tiêm adrenalin 1 10 000 qua nội soi điều trị chảy máu do loét dạ dày tá tràng tt tiếng anh

M INISTRY OF EDUCATION AND TRAINING

M INISTRY OF DEFENSE

108 IN STITU TE OF C LINI C A L MEDI C A L AN D P HA R MAC EU TI CA L SCI EN C ES

---------- ----------

KHAI NGUYEN DAO

RESEARCH OF THE EFICIENCY OF ENDOSCOPIC
HEMOCLIP OR HEMOCLIP COMBINED WITH
INJECTION OF ADRENALIN 1/10000
IN PATIENTS WITH PEPTIC ULCER BLEEDING
Speciality: Gastroenterology
code: 62.72.01.43

SUMMARY OF MEDICAL DOCTORAL DISSERTATION

HA NOI-2020


THE THESIS WAS DONE IN: 108 INSTITUTE OF CLINICAL
MEDICA L AND PHA RMACEUTICAL SCIENCES

Scientific supervisors

1. Assoc. Prof.Dr. Khien Van Vu
2. Assoc. Prof.Dr. Thu Ho Thi Pham

Reviewer

1.
2.
3.

This thes is will be presented at Institute Council at: 108 Institute of
Clinical Medical and Pharmaceutical Sciences
Day
Month
Year

The dissertation could be found in:
1. National Library of Vietnam

2. 108 clinical medical and pharmaceutical sciences institute


1

BACKGROUND
In Vietnam, gastrointestinal bleeding (GB) is a common
complication in both internal medicine and surgery. Upper
gastrointestinal bleeding accounts for about 70-80% of the total
bleeding in the gastrointestinal tract. The frequency of IDI ranged
from 36 / 100,000 to 172 / 100,000 people.
Endoscopic CMTH treatment has so far been classified into
three main technical groups: mechanical therapy, ablation therapy,
and injection therapy.
Endoscopic clip clamping is a mechanical treatment that has
been widely used since the 1990s, with a hemostatic effect of over
95% and a high safety. In addition, the injection of hemostatic

injection with adrenalin (epinephrine) 1 / 10,000 is also used in
clinical practice, due to its good hemostatic effect, safer than other
solutions (absolute alcohol, polidocanol ...)
In our country in recent years, clip techniques and
adrenaline1/10,000 injection of hemostasis in endoscopy to treat
gastrointestinal bleeding have been mentioned through studies. The
majority of authors used clips alone or injected adrenaline 1/10,000
alone. No author has studied a systematic and large scale on the
effectiveness of comparison between clip alone and clip combined
adrenalin 1/10,000 injection for treating gastrointestinal bleeding due
to peptic ulcer. Therefore, we conduct research on this topic with the
following two objectives:
1.
Evaluation of hemostatic results with simple clip clamps and
clip clips combined with 1 / 10,000 adrenalin injection for the
treatment of gastroduodenal ulcer bleeding.
2.
Comment on the specification, safety and a number of factors
related to the outcome of treatment.


2

Chapter 1
OVERVIEW
1.1.

ADRENALIN INJECTION METHOD

1.1.1. Injection technique and dosage:

+ Injection technique: using 4 corners of needle around the
bleeding site, with the distance from the edge of the ulcer or bleeding
place is 3-4 mm, the volume of each injection depends on the
experience of the Endoscopic Doctor , injected into the lining until
swollen and slightly changed color.
+ Dose: The volume of injection solution depends very much
on the size of ulcer causing GI bleeding. The dosage can be 0.5-1ml,
or can be used from 20 ml, 30 ml, even up to 45 ml.
1.1.2. The hemostatic effect of hemostatic injection therapy
with adrenallin
Until now, adrenalin is still a hemostatic solution that is used
extensively by endoscopic therapist because of easy using and low
cost. Today, adrenalin 1/10000 injection is combined with other
methods (hemoclips, electrocoagulation ...) in the treatment of severe
bleeding, when it is not possible to intervene deeply or difficult to
intervene (posterior wall duodenal ulcer). Adrenalin injecting around
the blood vessels is flowing may stop or reduce bleeding, which
allow to reveal the bottom of the ulcer and vascular position, and
create conditions for physicians to use combination therapy such as
endoscopic hemostatic clips.


3

1.5.

HEMOCLIP METHOD

1.5.1. History and technical principles
The hemoclips technique was first used in 1975 by Hayashi

T et al. in Japan for patients with gastrointestinal bleeding causing
peptic ulcer. Since then, this technique has been used in many
countries around the world and brought good effectiveness in
hemostasis and reducing the rate of patients having to surgery. This
technique is also used to hemostasis in surgery, for patients who
undergo radiotherapy with bleeding complications, or tissue markers,
clamps to close small holes related to submucosal dissection.
The principle of clips method is based on the mechanism of
using metal clips to clamp blood vessels or tissues to pinch blood
vessels, which can stop bleeding, but do not cause inflammation or
tissue damage such as fiber injection or heat. It is usually indicated
for acute bleeding ulcer or damage such as Mallory-Weiss,
Dieulafoy .. when the lesion is small, soft.
1.5.2. Effective hemostasis of clips with endoscopic.
There are many randomized studies about hemostatic effects
of hemoclip or hemoclip combined with other treatments such as
injection (adrenalin injection, polidocanol, absolute alcohol, isotonic
salt ...), heat (photo-frezee, electrodynamics, heat frezee...).
At the beginning of the birth of this hemoclip technique,
there were some studies using hemoclip to evaluate the effectiveness
of hemostasis in patients with peptic ulcer.
To learn about the hemostatic effect of clips, there have been
a number of studies comparing the hemostatic effect of clips and
other methods. Table 1.12 presents the combined therapeutic effect


4

compared to momotherapy of patients with peptic ulcer and
gastrointestinal bleeding.

Table 1.12. Effective hemostasis of clips with other method
Author

Treatment

n

Hemost
asis (%)

Recurrent
gastrointes
tinal
bleeding
(%)

Op eration
(%)

Fatal
(%)

Chou- cs
/2003
Chung-cs
/1999

Clips
Fiber in jection
Clips

Fiber
in jection
(HSE)
Clip
+
Fiber
injection
Clips
Fiber
in jection
(Ethnol)
Clip
+
Fiber
injection
Clips
Fiber
in jection
(Poli)
Clips+
Fiber
injection
Fiber
in jection
(Epi)
Clips
Heat frezee
Clips
Heat frezee +
Fiber in jection

39
40
41
41
42

100
98
98
95
98

10
28
2
15
10

5
13
5
15
2

3
5
2
2
2

42
42
42

100
100
100

10
14
17

0
0
0

7
2
2

31
30

97
97

6
13

3
3

3
0

52
53

98
92

4
21

0
9

2
0

56
57
26
21

89
86
100
95

2
21
15
24

4
7
12
5

4
4
0
10

Shimodacs/2003

Ljubiciccs/2004
Lo-cs/2006

Cipollettacs/2001
Salt zmancs/2005

Nowadays, studies often treat the combination of different
methods, or the control comparison between hemoclip with other
methods.


5

Chapter 2
OBJECTIVES AND RESEARCH METHODS
2.1.
RESEARCH SUBJECTS
We studied 150 patients with peptic ulcer bleeding caused
by ulcers were gone to emergency department , treated at the
Internal Medicine Department of the 108 Central Military Hospital
from January 2013 to March 2017. The patient was divided into 2
random groups:
+ Group I: hemoclip (n = 75)
+ Group II: hemoclip + adrenalin(1/10000) injection (n = 75)
2.1.1. Inclusion criteria
Patients included in the study include the following criteria:
+ Symptomactic patients of peptic ulcer bleeding include
vomiting of blood and / or black stools.
+ Endoscopic images of lesions has hemostasis indication
based on Forrest's classification. Endoscopic therapeutic
interventions were performed for Forrest IA patients to Forrest IIB.
2.1.2. Exclusion criteria.
- Patients with GI bleeding due to portal hypertension.
- Patients with CI bleeding complications of stomach cancer
- Patients with contraindications for gastroduodenoscopy.
- Patients do not agree to participate in the study.
- Patients with coagulopathy
- Aneurysms lesions, Mallory-Weiss syndrome
2.2.
RESEARCH METHODS
2.2.1. Study design
- Research method: is a method of prospective, descriptive

monitoring, vertical monitoring, randomized grouping, conducting
intervention, and comparing treatment results of 2 groups.


6

- Sample size: Choose convenient 150 random pateints from
January 2013 to March 2017 was divided into 2 groupss. In group I,
there were 75 patients patients received hemoclip , group II had 75
patients treated with hemocip combined with adrenalin(1/10000)
injection
2.2.2. The technique of endoscopy hemostasis by clip
+ Step 1: Reveal ulcers
As soon as the ulcer is detected, it is necessary to assess the location,
condition of the ulcer, bleeding level immediate ly and bring the
scanner to the most favorable position to stop bleeding.
Pump , c lean the ulcer, if blood c lot remains on the ulcer, clots must
be removed ,firstly, wash with strong pressure, if blood clot does not
turn out, tools used to take out.
+ Step 2: Insert the device that has inserted the clip through
the biopsy channel into the lesion position.
+ Step 3: Pull out the handle to open the clip. If les ion of
the stomach, the clip will be opened more convenient. However, in
order to increase the convenience of opening the clip, the person
performing the procedure must perform the endoscopy in accordance
with the regulations (do not twist the machine, cross-bend ...). If
lesson of the duodenum, it is necessary to expose the surrounding
lesions, so that the clip can be opened smoothly.
+ Step 4: Principle of clip : c lamp directly into blood vessels
at the ulcer causing bleeding. Clips are paired perpendicular to the

blood vessels. The main physician has just pushed the clip into the
clamping position, while sucking the vapor to pull the damage into
the clip. when the correct position has been determined, the
secondary physician coordinates losely with the main physician to
clam the clip. In case the blood vessel is still flowing, it should be
clamped above and below the position of the blood vessel that is


7

following. This technique is applicable to all ulcers causing GI
bleeding. Depending on the size of the ulcer, it can be clamped by 110 clips. Results are achieved when the blood flow from the ulcer is
not visible.
+ Ulcer causing bleeding is classificated in Forrest IIB : In case
old clot is easy to turn out, proceed to clamp clip immediately. In the
case new clots is hard to cling, use pliers or snares to remove the
clot, after taking the clot while combining the washing and clamping
clips. Need to monitor closely hemodynamics (pulse, blood pressure,
oxygen pressure ...) during the procedure.
+ In case blood continues to flow, stop bleeding. If unsuccessful,
blood continues to flow, consultation should be switched to other
methodic treatment (vascular intervention or surgery)
2.2.3. Hemostatic technique with adrenalin injection
+ Prepare injection solution:
- Adrenalin( 1mg / 1ml) tube.
- Physiological saline (Nacl 0.9%).
- Use 10ml syringe, take 1ml of adrenaline solution, then
take 9ml of saline solution to get 10 ml of adrenaline diluted 1 /
10,000.
+ Technique of Adrenalin injection :

First, use an injection containing 10ml of adrenalin solution
to inject around the ulcer causing bleeding
With Forrest IA and IB ulcers, inject around the bleeding site
until the mucosa swells with volumes for a injections ranges 0.5-3 ml
With Forrest IIA ulcer, inject 4 positions around the ulcer
with dose is 0.5 - 3 ml.
With Forrest FIIB ulcer having blood clot, inject to remove
the clot, inject around the location of blood clot, the dose of each
injection is 0.5 - 5 ml until the mucosa swells. As a result, blood c lot


8

may pop out and convenient for washing and removing the clot.
Then, the ulcer is performed a washing pump to reveal , if blood
clots remain on the ulcer, the blood c lot must be removed, firstly,The
ulcer is flushed with strong pressure, if the blood clot does not turn
out, snares , pliers or three prong pliers were used to take out the
objects.
Claming clip is performed additionally after injecting
adrenalin. If it fails, the blood continues to flow, requiring
Consultation to another method of treatment is required (vascular
intervention, or surgery).
2.2.4. Identify results
2.2.4.1. Criteria for first hemostasis (immediate hemostasis)
The first period is evaluated by endoscopic images right after
the end of the procedure, with pictures:
+ Blood does not flow: After the procedure is performed, the
results are successful immediate ly. They include the image of blood
from the ulcer does not flow, wash and clean the bottom ulcer,

follow after that, blood bleeding is not see.
+ Blood flow does not hold: the results is failure after
performing the procedure as well as the blood still flow, physician
must coordinate other endoscopic interventions such as adrenalin
1/1000 injection. After the combination treatment is performed,
physician must check again if the ulcer is no longer bleeding,
treatment has been successful.
+ Blood flow without holding, vascular intervention or
surical method: After using other endoscopic interventions, the blood
continues to flow, vascular intervention or surgical surgery must be
performed immediately
2.2.4.2. Evaluate res ults after hemostasis:
Based on clinical and investigation:


The first hemostatic result:


9

+ Success: when the first colonoscopy was successful, the
clinical condition of patients is stable, there is not vomiting blood,
defecation is yellow stool.
+ Failure: when the first endoscopy must use other
hemostatic measures or surgical transfer or in case the patient has
recurrent bleeding with clinical signs :vomit blood, go outside the
black stool, pulse fast, low blood pressure and endoscopy immage
with recurrent bleeding ulcer must be intervened several times.


The second hemostatic result:

Evaluate the results of hemostasis when the patient has
recurrent bleeding, with the same criteria as the first hemostasis
assessment.
2.2.4.3. Evaluation of general hemostatic results
As a general assessment of the results through hemostasis:
+ Good: Only one intervention has to be conducted, the
results are successful immediately, the ulcer is not bleeding, clinical
monitoring is stable, stool is yellow.
+ Average: Patients with recurrent bleeding, but through
successive hemostasis, clinica l monitoring is better , without
bleeding, defecation.
+ Poor: Hemostatic procedure failed, other endoscopic
intervention must be used or the patient must switch to vascular
intervention or surgery, or the patient has severe fatalities, the
procedure is not achieved.
2.2.4.4. Criteria of Technical and safety evaluation
- Evaluation of technical complications: perforation when
rough intervention, seeing tools through ulcers or exposed lateral
lesions, necrosis.


10

- Assess the condition of pulse, blood pressure before and
after intervention
2.2.4.5. Comparison of effective hemostasis: clip alone and clip
combinated with adrenalin 1/10000 injection
+ Compare the successful hemostatic rate between the two

methods after performing the procedure.
+ Compare the rate of recurrent bleeding between the two
methods.
+ Compare hemostatic results of two methods when patients
leave the hospital.
+ Compare the amount of blood transmitted to the patient
between the two methods.
+ Compare length of hospital stay between two methods.
+ Technically: Compare the average number of clips used
for each technique. The average total of adrenalin solution injections
and the total mean volume (ml) of adrenalin 1 / 10,000 solution for
the combined treatment group (group II)
2.2.5. Using a scale to assess the risk of recurrent GI bleeding,
the risk of death after treatment.
According to the guidelines for treatment for pateint with GI
bleeding which not due to by portal hypertension, there are 2
commonly used scales: Full Rockall scale and Blatchford scale
2.2.6. Data processing
Data is processed according to the medical statistical method
with the help of 20.0 SPSS software and processed at the Institute of
Preventive Medicine Training and Public Health, Hanoi Medical
Univers ity.


11

Chapter III
RESULTS
3.1. GEN ERAL

CHARACTERIS TICS

OF

THE

RES EARCH

GROUP

Average age, rate of male and female, history of
gastrointestinal bleeding (first, second, over 2 times), comorbidities
and drug history between two different groups were not statistically
significance (p> 0.05). Clinical symptoms (abdominal pain, vomiting
blood, go outside black stool), classify the level of blood loss in
clinical (severe, moderate, mild), the number of peptic ulcer (1 drive,
2 drives, on 2 drives), position of peptic ulcer (stomach, duodenum),
stomach-duodenal ulcer size (<1 cm, 1-2 cm,> 2 cm), different
between 2 groups were not statistically significance (p> 0.05).
3.2.
COMPARE OF TREATMENT RESULTS OF 2
GROUPS
3.2.1. Evaluate first hemostatic res ults (1st)
Table 3.11. first hemostatic res ults of two groups
Hemostatic results
Plus
Group I Group II
n(%)
n (%)
n (%)

Success
Failure
Plus

143
(95,3%)
7 (4,7%)
150
(100%)

68
(90,7%)
7 (9,3%)
75
(100%)

P

75 (100%)
0 (0%)
75 (100%)

0,013b

b: Fisher's Exact Test
Comment:
- Hemostatic effect of group II: 75/75 patients (100%)
- Successful hemostatic effect of group I: 68/75 patients
(90.7%).

12

3.2.2. General hemostatic results.
Table 3.15. General hemostatic res ults
General
Hemostatic

Plus

Group I

Group II

n (%)

n (%)

n (%)

Good

143 (95,3%)

68 (90,7%)

75 (100%)

Medium

5 (3,3%)

5 (6,7%)

0 (0,0%)

Bad

2 (1,3%)

2 (2,7%)

0 (0%)

Plus

150 (100%)

75 (100%)

75 (100%)

res ults

P

0,025a

a: Chi-Square Tests
Comment: the good hemostatic effect of group II was 100%,

significantly higher (p <0.05) than the good hemostatic effect in
group I (90.7%)
3.2.3. Time in hospital

Chart 3.2. Average treatment time
Comment: Using Independent-Samples T Test, the results
showed that the average treatment time of group I (7.7 ± 6.5)
was higher than that of group II (5.8 ± 2.6), the difference
was statistically significant with p = 0.025.


13

3.3.

HEMOSTATIC
RESULTS
CLASSIFICATION

FOR

FORREST

3.3.1. Hemostatic results with Forres t I level (A + B)
Table 3.21. Hemostatic res ults with Forrest I
The first

Group I

Group II

hemostatic res ult

n (%)

n (%)

Success

23 (82,1%)

41 (100%)

Failure

5 (17,9%)

0

Total

28 (100%)

41 (100%)

p

0,009b

b: Fisher's Exact Test

Comment: Hemostasis results for Forrest I in group II
reached: 100%, in group I was 82.1%, the difference was statistically
significant with p <0.05.
3.4.
SPECIFICATION AND SAFETY
3.4.1. Comment on specification and safety
Table 3.24. Comment on specification and safety
Targets of the research
Number of clips used

Group I

Group II

p

2,07 ± 0,827

2,19 ± 1,087

0,448c

Number of ad renaline inject ions

3,73 ± 1,018

Vo lu me of adrenalin (ml)

7,49 ± 2,056

Catastrophe

No

No

c: Independent-Samples T Test
Comment: The number of clips used between group I and
group II was not statistically significant (p> 0.05). The number of
injections and volume of adrenalin solution 1 / 10,000 in group II
was respectively 3.73 ± 1.0 (nose) and 7.49 ± 2.0 (ml).


14

3.5.

RELATIONSHIP BETWEEN RATING SCALE AND
TREATMENT RESULTS

3.5.1. The scale of Rockall and Blatchfort with injured position
Table 3.28. Relation between the scale of Rockall, Blatchford and injured position

Rating scale

Injured position

Pe

Stomach

Duodenum

Clinical Rockall

1,68 ± 1,04

1,10 ± 1,15

0,002

Full Rockall

3,72 ± 1,13

3,10 ± 1,15

0,002

Blatchford

10,44 ± 2,82

8,71 ± 3,64

0,015

e. Mann-Withney test
Comment: Average score of the scales of Clinical Rockall , Full
Rockall and Blatchford increased higher was statistically significant

in patients with stomach ulcer versus duodenal ulcer (p <0.05).
3.5.2. Prognosis of recurrent gastrointestinal blee ding
Table 3.33. Prognostic factors assess the risk of recurrent gastrointestinal bleeding

Prognostic factor

Acreage

Cut-off

Confidence

for
recurrent bleeding

under the
curve

thres hold

intervals 95%
(Min-Max)

Age

0,585

≥ 60

0,376 - 0,793

Ulcer size

0,583

≥ 2,0

0,176 - 0,991

Full Rockall score

0,529

≥ 3,5

0,210 - 0,849

Batchfort score

0,662

≥ 11,5

0,410 – 0,913

Comments: Factors had directly proportional to the risk of
recurrent gastrointestinal bleeding include: Age, size of ulcer,
Rockall and Blatchford scores with cut-off points: 60 years, 2.0 cm,
3.5 points and 11.5 points.

15

3.5.2.1. Relationship between age thres hold and the risk of
recurrent gastrointestinal blee ding.
Table 3.34. Relationship between age threshold and the risk of
recurrent gastrointestinal bleeding
Age

Recurrent bleeding

Total

Yes : n (%)

No: n (%)

< 60

1 (33,3%)

90 (61,2%)

91 (60,7%)

≥ 60

2 (66,7%)

57 (38,8%)

59 (39,3%)

Total

3 (100%)

147 (100%)

150 (100%)

Comment: With the age threshold above 60 is the factor
prognosis risk of recurrent gastrointestinal bleeding, with sensitivity,
specificity: 66.7% and 61.2% respectively.
3.5.2.2. Relationship of ulcer size> 2.0 cm with recurrent
gastrointestinal blee ding.
Table 3.35. Relative ulcer size with risk of recurrent gastrointestinal
bleeding
Ulcer size

Recurrent bleeding

Total

Yes : n (%)

No: n (%)

< 2,0 cm

2 (66,7%)

140 (95,2%)

142 (94,7%)

≥ 2,0 cm

1 (33,3%)

7 (4,8%)

8 (5,3%)

Total

3 (100%)

147 (100%)

150 (100%)

Comment: When the ulcer size > 2.0 cm is the factor
prognosis risk of recurrent gastrointestinal bleeding with sensitivity,
specificity: 33.3% and 95.2%, respectively.


16

3.5.2.3. Relationship between Full Rockall and the risk of

recurrent gastrointestinal blee ding.
Table 3.36. Relationship between Full Rockall and the risk of
recurrent gastrointestinal bleeding
Rockall

Recurrent bleeding

Total

score

Yes : n (%)

Yes : n (%)

< 3,5 điểm

1 (33,3%)

87 (60,0%)

88 (59,5%)

≥ 3,5 điểm

2 (66,7%)

58 (40,0%)

60 (40,5%)

Total

3 (100%)

145 (100%)

148 (100%)

Comment: With a full Rockall threshold ≥ 3.5 points is the
factor prognosis risk of recurrent gastrointestinal bleeding, with
sensitivity (66.7%), specificity (60.0%).
3.5.2.4. Relationship betwee n Blatchford points and the risk of
recurrent gastrointestinal blee ding.
Table 3.37. Relationship between Blatchford points and the risk of
recurrent gastrointestinal bleeding.
Blatchford
points

Recurrent bleeding

Total

Yes : n (%)

Yes : n (%)

< 11,5 điểm

1 (33,3%)

91 (66,9%)

92(66,2%)

≥ 11,5 điểm

2 (66,7%)

45 (33,1%)

47 (33,8%)

Total

3 (100%)

136 (100%)

139 (100%)

Comment: With Blatchford threshold points ≥ 11.5 pointss
the factor prognosis risk of recurrent gastrointestinal bleeding, with
sensitivity (66.7%), specificity (66.9%)


17

Chapter IV
DISCUSSION

4.1.

GENERAL CHARACTERISTICS
The study results show that the average age in group I and
group II is not statistically significant (p> 0.05). The study results
also show that the average age of both groups is 54.56 ± 17.4 (years).
The proportion of male patients with the disease was higher (66.0%)
than the female patient (34.0%). The overall male to female ratio is:
1.94. The gastrointestinal bleeding rate in the first-time in both
groups are high (group I: 69.3%), group II (73.3%). History of
medication and associated diseases in our study shows that the
difference between Group I and Group II is not statistically
significant. Assessing the level of blood loss on the light, moderate
and severe sieve level accounted for respectively: 14.7%, 56.0%,
29.3%. Regarding the number of ulcers in the study, most patients
have an ulcer. There was no difference in the degree of blood loss on
the sieve and the number of ulcers between group I and group II (p>
0.05). According to Forrest classification, the research results
showed: the ratio of IA, IB, IIA and IIB Forrest are: 11.3%; 34.7%;
32.7% and 21.3%, respectively. The difference between the two
groups was statistica lly significant (p <0.05).
4.2.
COMPARISON
OF
TREATMENT
RESULTS
BETWEEN 2 GROUPS
4.2.1. Comparison of the initial hemostatic effective by
endoscopic.
Research results (table 3.11) presented the initial hemostatic

effect for patients with gastrointestinal bleeding due to peptic ulcers:
the hemostatic effect of group II reached: 75/75 patients (100%), the
hemostatic effect of group I is 68/75 patients (90.7%). So, the
combination therapy (hemoclips + adrenalin injection 1/10,000) in


18

group II is better than group I (p = 0.013) in hemostasis of patients
with peptic ulcer and gastrointestinal bleeding from IA to IIB Forrest.
Sung JJY et al. collected over 15 research studies with 1156
patients with gastrointestinal bleeding due to peptic ulcer divided
into 4 different treatment groups: hemoclips group (n = 390),
combination group with hemoclips and hemostatic injection (n =
242), hemostatic injection group (n = 359), e lectro-frezee group with
and without hemostasis injection. Research results show:
+ The hemostatic effect of hemoclip group (86.5%) is higher
than the hemostatic injection group (75.4%) (RR 1.14, 95% CI 1.001,30). Reduce the surgical risk for the hemoclip group compared to
the injection group
+ The hemostatic effect of hemoclip + hemostatic injection
group (88.5%) is higher than the hemostatic injection group (78.1%)
(RR 1.13, 95% CI 1.03-1, 23), reducing the surgical risk for
hemoclip + hemostatic injection group compared to the injection
group.
Chung IK el at. performed endoscopic hemostasis treatment
for 124 patients with gastrointestinal bleeding due to peptic ulcer.
The patient was divided into 3 groups: hemoclip group (n = 41),
adrenaline injection group (n = 41) and the combination group:
hemoclip + adrenalin injection (n = 41). The results showed that the
initial hemostatic effect of the hemoclip group, the adrenalin

injection group and the combination group were: 97.6%; 95.1% and
97.6%, respectively. However, the incidence of recurrent
gastrointestinal bleeding was different, its’ hemoclip group,
adrenalin injection group and combination group were: 2.4%; 14.6%
and 9.5%, respectively. The surgery rate of hemoclip group,
adrenalin injection group and combination group accounted for 4.9%;
14.6% and 2.3%, respectively. The hemostatic rate for 3 groups were:
95.1%; 85.4% and 95.2%, respectively. All complications are in the
adrenalin injection group.


19

Gevers AM et al. studied the hemostatic effect in two groups:
the hemoclip group (n = 35) and the combination group using
hemoclip + adrenalin injection (n = 32). The results showed that: The
rate of early recurrent gastrointestinal bleeding of hemoclip group
was 13/35 patients (37.14%) and those of combination group was
8/32 patients (25.0%). Similarly, the rate of hemostatic failure is
12/35 patients (34.3%) in hemoclip group and 8/32 patients (25.0%)
in combination group.
4.2.2. Res ults of hemostatic res ults betwee n 2 endoscopic
groups
The hemostatic result after hemoclip treatment (group I)
compared with hemoclip + adrenaline injection group (group II) is
the main goal in this study.
Long-term hemostatic results are assessed by clinical criteria
that patients are stabilized blood pressure and blood pressure index,
yellow stools, and increase RBC, hemoglobin, hematocrit. Regarding
general hemostatic results (table 3.15) in hemoclip group, there were

90.7% good results, 6.7% fair results and 2.7% poor results, group II
(hemoclip + adrenaline injection) had 100% good results, the
difference between the two groups was statistically significant with p
= 0.025.
These results shown that combination therapy will achieve
better results about initial hemostatic effect and overall clinical
efficacy. A collection of different studies around the world found
that the combination method should be applied for case of complex
damages and difficultly intervene and the combination of methods
will increase the effectiveness of treatment and reduce complications.
4.2.3. Time in hospital of 2 groups.
Time in hospital is also a indirect parameter assessing the
effectiveness of each treatment. In our study, the average number of


20

days of treatment in group I was: 7.7 ± 6.5; those in group II: 5.8 ±
2.6, the difference is statistically significant (p <0.05).
4.3.
HEMOSTATIC
RESULTS
WITH
FORREST
CLASSIFICATION
In our study, the assessment of endoscopic hemostatic
effects with I Forrest (Table 3.21). The results showed that the
hemostatic effect of group II (combination group: adrenalin injection
+ hemoclip) reached 100%. Meanwhile, the successful effect of
hemostasis of group I was: 23/28 (82.1%), 5/28 patients (17.9%)

were ineffective. The difference is statistically significant (p = 0.009).
The results showed that combination therapy was more effective than
momotherapy treatment of gastrointestinal bleeding I Forrest.
4.4.
CHARACTERISTICS AND SAFETY OF TECHNICAL
In our study (Table 3.24), the number of clips used between
Group I and Group II is 2.07 ± 0.827, 2.19 ± 1.087, the difference is
not statistica lly significant (p> 0,05). The number of injections and
the volume of 1 / 10,000 adrenaline solution in Group II were: 3.73 ±
1.0 (nasal) and 7.49 ± 2.0 (ml), respectively.
Hemostasis with hemoclip or hemoclip with adrenalin
injection is one of the most commonly used treatments for patients
with gastrointestinal bleeding. It is an uncomplicated technique and
high hemostatic effect.
In our study, no complications occurred in any patient in
both groups.
4.5.
SOME FACTORS RELATED TO HEMOSTATIC
RESULTS
4.5.1. Relation betwee n clinical Rockall scale, Blatchfort
Our study showed clinical Rockall score, full Rockall,
Blatchfort: 1.28 ± 1.14; 3.30 ± 1,18 and 9,27 ± 3,48, respectively.
There is no difference in the average clinical Rockall score, full


21

Rockall, Blatchfort between Group I and II (p> 0.05). So, based on
the full Rockall scale, the patients in both groups have average scores,
but still have to follow up. Similarly, the average score of Blatchfort

scale for 2 groups is above 6 points and treatment interventions are
required.
4.5.2. Factors

related with the

prognosis of recurrent

gastrointestinal blee ding
The results from Table 3.33 show that: ulcer size is most
closely related to the risk of recurrent bleeding, with a 95%
confidence interval, max area is 0.991; followed by two Blatchford
and Rockall scales, max area is 0.913 and 0.849 respectively. The
risk assessment threshold of gastrointestinal bleeding with age, ulcer
size, full Rockall score, and Blatchford are: ≥ 60, ≥ 2.0, ≥ 3.5, ≥ 11.5.
So, 4 factors (age, ulcer size, score of full Rockall scale and
Blatchfort) play roles in predicting risk of recurrent gastrointestinal
bleeding.
Based on these thresholds, we calculated the sensitivity,
specificity for each specific parameter, presented specifically in
tables 3.34 to 3.37.
In summary: upper gastrointestinal bleeding causing peptic
ulcer is an emergency disease encountered both internal and external
medical. Today, thanks to better understanding of the mechanism of
gastrointestinal bleeding causing peptic ulcer, it contributed to limit
this complication. With modern and common facilities, endoscopic
hemoclip combined with adrenalin injection are one of the combined
measures being applied by many countries, contributing to successful
hemostasis, reducing blood flow rate, reducing surgical rate and
mortality rate.

22

CONCLUSION
The study on 150 patients with gastrointestinal bleeding was
divided into two groups: Group I: hemoclips (n = 75), group II:
injections of adrenaline 1/10,000 + hemoclips (n = 75), the following
conclusions are obtained:
1. Evaluate the hemostatic res ults of two groups
- Initial hemostatic effect of 2 groups
Successful hemostatic effect of group I reached: 68/75
patients (90.7%), those of group II: 75/75 patients (100%), the
difference was statistically significant (p = 0.013 ).
- Overall hemostatic effect
The overall hemostatic effect with Good, fair and poor levels
in group I were: 90.7%, 6.7% and 2.7%, respectively. The hemostatic
effect with good, fair and poor levels in group II were 100%; 0% and
0%. The difference is statistically significant (p = 0.025).
- Hemostatic effective according to Forres t classification of 2
groups
+ With I Forrest (A + B): The hemostatic result in group II
reached 100%, those in group I reached 82.1%, the difference was
statistically significant with p = 0.009.
- Time in hospital of 2 groups:
+ The average time in hospital of group I is: 7.7 ± 6.5 (days),
that of group II is: 5.8 ± 2.6 (days), the difference is significant with
p = 0.025.
2. Technical characteristics, safety and some factors related to
the treatment res ults

- Number of injections and volume of adrenalin solution
1/10,000 used in group II were 3.73 ± 1,018 (injections) and 7,49 ±
2,056 (ml) respectively.


23

- The number of hemoclips used in group I is: 2.07 ± 0.827,
that in group II is: 2.19 ± 1.087, the difference is not statistically
significant (p = 0.448).
- No complications, no fatal patients when performing the
procedure in groups I and II.
- Clinical Rockall score, full Rockall and Blatchford score
increased significantly in patients with gastric ulcer versus duodenal
ulcer (p <0.05).
- Factors that correlate positively with risk of recurrent
gastrointestinal bleeding include: Age, size of ulcer, Rockall and
Blatchford scores with cutoffs: 60 years old, 2.0 cm, 3.5 points and
11.5 points.