Corresponding Author: Tarwoto
E-mail: [email protected]
INTRODUCTION
A stroke is a collection of
symptoms where brain damage occurs due to reduced blood supply to parts of the
brain (Nuraini, 2015). Stroke is a neurological
disease caused by blood vessel disorders in the brain, causing characteristic
symptoms (Birawa &
Amalia, 2015). According to the American Heart Association, stroke is
an attack of hidden infarction and bleeding in the brain, spinal, and retina
(Coupland, 2017). Based on the 2020 WHO report, stroke mortality ranks second
worldwide after cardiovascular disease and the second cause of
disability-adjusted for years of life (UGIA MAY HUDATAMA,
2020). In the United States in 2019 stroke is the third
leading cause of death in the world (Yaslina et al., 2019). Stroke cases worldwide are estimated to reach 50
million people, and 9 million experience severe disabilities (Marbun et al., 2016). Based on the 2018 Riskesdas of the Ministry of
Health, the incidence of stroke is 10.9%, and every year, as many as 713,783
people experience a stroke (Mongkau et al., 2022).
Someone who has a stroke can result in loss of
independence, work, and hopelessness resulting in decreased productivity and
quality of life for sufferers (Ananda & Darliana,
2017). This situation can be prevented early by identifying
and controlling stroke risk factors. Many factors cause stroke, including age,
gender, hypertension, atrial febrile, diabetes mellitus, cholesterol, obesity,
activity, smoking, and family history of stroke (Wayunah &
Saefulloh, 2017). To identify and control stroke risk factors, the
National Stroke Association (NSA) issued a stroke risk detection model, namely
the Stroke Risk Scorecard, which contains eight stroke risk factors and is
grouped into three categories, namely high risk, caution, and low risk. Each
risk factor is categorized according to the value of the risk factor criteria,
and the results are determined based on how many risk factors are in the
high-risk, caution, and low-risk groups.
Figure 1. Stroke Risk Scorecard and Risk Scorecard
Result
A diagnostic test is carried out to predict whether
these risk factors are appropriate, namely a method (tool) to determine whether
a person is at high risk of stroke or not based on the presence of signs and
symptoms in that person. Diagnostic tests use a screening test, which is a way
to determine whether a person is at risk of suffering from a disease or not (Arif, 2017). Without screening, the diagnosis of a disease can
only be enforced after signs and symptoms appear, even though a disease has
existed long before signs and symptoms appear, which can be known if we do a
screening. Screening tests include validation tests and sensitivity and
specificity tests. Test validity is the test's ability to correctly
(accurately) which individuals are sick and which are not (Siswosudarmo, 2017). The sensitivity and specificity reflect the validity
of the test. Sensitivity is the test's ability to show which individuals are
sick from the entire population who are sick (Istiqomah et al.,
2021). Specificity is the test's ability to show which
individuals are not sick from those who are not (Nusahi, 2018).
The NSA stroke risk detection model has been tested
for sensitivity, specificity, and accuracy on 231 respondents consisting of 81
post-stroke people and 150 nonstroke people; the results obtained were 74%
sensitivity, 76% specificity, and 76% accuracy. It is felt that the
sensitivity, specificity, and accuracy of the NSA stroke detection model still
need to be optimized so that further development is needed to obtain a more
practical value. Development of the NSA stroke risk detection model by adding
two stroke risk factors, namely age, and gender, and giving weight to each risk
factor. The weight of each stroke risk factor is determined based on the
study's results and then totaled the total score. The total score ranges from
10 to 51 and then determines low risk if the score is less than 20, the
moderate risk is 20-29, and the high risk is 30 or more. The result of the next
development is an Android-based application called STRIC (Stroke Risk Score
Card). A study published in 2021 in the journal Stroke and Vascular Neurology
evaluated the effectiveness of STRIC applications in the Chinese population.
The study found that STRIC had an accuracy rate of 0.746 in predicting stroke
risk in 5 years in the studied population. In addition, it shows that STRIC has
a good ability to distinguish between stroke and non-stroke groupsBased on the
above background (Chen et al., 2022). So, the purpose of this study is to ascertain
whether the STRIC application model is effectively used to detect high risk and
requires testing for effectiveness and accuracy by conducting trials in stroke
and non-stroke groups.
METHODS
This
type of research was descriptive-analytic and was carried out in two different
places, namely in the post-stroke group conducted at the Jakarta Brain Center
Hospital and the Jakarta I Ministry of Health Polytechnic for the non-stroke
group. The number of samples was determined as many as 60 people, divided into
30 samples from the post-stroke group and 30 from the non-stroke group. The method
of sampling in the stroke group was carried out using the accidental sampling
method; namely post, stroke patients who came to the RSPON outpatient care were
sampled until the number of samples was fulfilled, and in the non-stroke group,
sampling was carried out randomly. Using the STRIC application on Android,
respondents were asked to fill in data, and the results of the respondent's
stroke risk will automatically be known whether they are at low, medium, or
high risk.
Figure
2. STRIC application
The data obtained
is then tested for sensitivity, specificity, and accuracy with the following
formula:
Sensitivity
������ =
Sensitivity
������ =
Sensitivity
������ =
Information:
a
= true positives
b
= false positives
c
= false negatives
d = true negatives
RESULTS AND DISCUSSION
Characteristics of
Respondents
Based on the gender and age of the respondents can
be seen in table 1 below.
Table 1. Frequency distribution of respondent
characteristics based on
gender
and age in the post-stroke and non-stroke groups in 2022
|
Characteristics
|
n
|
Post Strokes
|
Non-Strokes
|
|
n
|
%
|
n
|
%
|
|
Age
|
|
|
|
|
|
|
≤ 59 years
|
36
|
10
|
33,3
|
26
|
86.7
|
|
≥ 59
|
24
|
20
|
66,7
|
4
|
13,3
|
|
Gender
|
|
|
|
|
|
|
Woman
|
32
|
13
|
43,3
|
19
|
63,3
|
|
Man
|
28
|
17
|
56,7
|
11
|
36,7
|
|
Amount
|
60
|
30
|
100
|
30
|
100
|
The respondents can be seen in table 2 below based
on the risk factors for stroke.
Table 2. Frequency Distribution of Respondents'
Characteristics Based on
Blood Pressure, Heart Rate, Body Weight, Cholesterol and
Blood Sugar,
Sports
Activities, Respondents' Family History, 2022
|
Characteristics
|
n
|
Post-Strokes
|
Non-Strokes
|
|
n
|
%
|
n
|
%
|
|
Blood
pressure
|
|
|
|
|
|
|
<
140/90 mm Hg
|
30
|
6
|
20
|
24
|
80
|
|
>=
140/90mmHg
|
30
|
24
|
80
|
6
|
20
|
|
Heart
rate
|
|
|
|
|
|
|
Regular
heart rate
|
52
|
24
|
80
|
28
|
93,3
|
|
Irregular
heartbeat
|
8
|
6
|
20
|
2
|
6,7
|
|
Smoke
|
|
|
|
|
|
|
Do
not smoke
|
34
|
17
|
56,7
|
27
|
90
|
|
Smoke
|
26
|
13
|
43,3
|
3
|
10
|
|
Cholesterol
|
|
|
|
|
|
|
<200
|
38
|
16
|
53,3
|
22
|
73,3
|
|
>=200
|
22
|
14
|
46,7
|
8
|
27,7
|
|
Blood
sugar
|
|
|
|
|
|
|
<101
|
30
|
9
|
30
|
21
|
70
|
|
101-199
|
24
|
17
|
56,7
|
7
|
23,3
|
|
>=
200
|
6
|
4
|
13,3
|
2
|
6,7
|
|
Weight Status
|
|
|
|
|
|
|
ideal
BB
|
29
|
16
|
53,3
|
13
|
43,3
|
|
Overweight
and obesity
|
31
|
14
|
46,7
|
17
|
56,7
|
|
Sports Activity
|
|
|
|
|
|
|
Routine
|
12
|
2
|
6,7
|
10
|
33,3
|
|
Rarely
or never
|
48
|
28
|
93,3
|
20
|
66,7
|
|
Family History of Stroke
|
|
|
|
|
|
|
There
are not any
|
38
|
12
|
40
|
26
|
86.7
|
|
There
is
|
22
|
18
|
60
|
4
|
13,3
|
|
Amount
|
60
|
30
|
100
|
30
|
100
|
Respondents' characteristics were based on the risk
of stroke in the post-stroke and non-stroke groups.
Table 3. Frequency Distribution of Respondents
Based on Stroke Risk Level, The year 2022
|
Variable
|
n
|
Post-Strokes
|
Non-Strokes
|
|
n
|
%
|
n
|
%
|
|
Stroke Risk Level
|
|
|
|
|
|
|
Low
|
30
|
3
|
10
|
20
|
66,7
|
|
Currently
|
30
|
5
|
16,7
|
8
|
26,7
|
|
Tall
|
30
|
22
|
73,3
|
2
|
6,6
|
|
Amount
|
60
|
30
|
100
|
30
|
100
|
STRIC Application
Effectiveness Assessment
Validation and accuracy test results, validation test, consist of
sensitivity and specificity tests. Table 4 shows the validation and accuracy
results.
Sensitivity������� =
Sensitivity������� =
Accuracy �������� =
PPV ����������������� =
NPV ���������������� = 
.
Table 4. Results of sensitivity, specificity, and
accuracy tests
on stroke and non-stroke post respondents in 2022
|
Variable
|
Results
|
Total
|
|
|
High
Risk
|
Low/Moderate
Risk
|
|
|
Stroke Post
|
22
|
8
|
30
|
|
Non-Strokes
|
2
|
28
|
30
|
|
|
24
|
36
|
60
|
|
Sensitivity = 91.67
|
Accuracy = 83.33
|
NPV = 93.33
|
|
Specificity = 77.78
|
PPV = 73.33
|
|
Results
Explanation
The STRIC application is an application developed
from previous research conducted by researchers. Unlike the NSA version of
stroke risk detection, the application of STRIC in determining stroke risk is
through the weight or score of each risk factor. The NSA version of stroke risk
detection to determine the degree or level of stroke grouped into three parts,
namely low risk, warning, and high risk, with each containing eight stroke risk
factors, namely blood pressure, atrial fibrillation, smoking, cholesterol,
diabetes mellitus, activity, obesity, and family risk of stroke. High-risk criteria if the
high-risk section has more than or equal to 3 risk factors. Determination of
warning if the risk factors in the warning section have warning risk factors as
many as 4-6 risk factors and determining low risk if in the low-risk column if
there are 6-8 risk factors.
In detecting stroke risk,
the NSA version does not accommodate age and gender factors. The results showed
that male sex factors were 1.4 times more than female. At the age of > 59
years, have four times more risk. Thus, the factors of age and gender become
the factors that must be considered as part of the assessment of stroke risk
factors. Even so, the issue of gender still has many differences in certain
locations or countries. For example, the results of the analysis (Feigin et
al., 2017) state that the lifetime risk
of stroke is higher for women than men, with a 1 in 4 stroke risk for women
after the age of 25; however, the incidence of stroke is higher in women than
men for those <30 years, while the stroke risk rate is higher in men than
women during midlife (Vyas et
al., 2021). The age factor contributes
to the occurrence of stroke; this is related to the presence of atheroma, which
starts with damage to blood vessels with increasing age. The effect of gender
on the incidence of stroke is influenced by the presence of the hormone
estrogen, which protects blood vessels by inhibiting arteriosclerosis. Thus it
is suspected that men have a higher risk of stroke (Nuraini,
2015).
Another problem in
determining risk criteria is high risk if you have > three high-risk
factors, warning risk if there are 4-6 risk factors, and low risk if you have
6-8 risk factors. There are eight risk factors in total, so if at high risk,
there are two risk factors; moderate risk factors have three risks and low-risk
factors have five factors. Thus, no one is included in that category. Thus, it is necessary to develop a more valid stroke
risk detection model with high effectiveness and validity.
Sensitivity, specificity, and accuracy tests were
carried out to determine the ability of the STRIC application to detect stroke
risk. The sensitivity value of the STRIC application is 91.67%, which means
that the ability to predict stroke risk for those who have had a stroke is very
high. The specificity value of the STRIC application is 77.78%; this means that
the ability to measure the STRIC application to provide non-stroke (negative)
results is 77.78% or very good. The STRIC application is very valid for
measuring stroke risk from these two measurements. A test's validity shows a
test's ability (accuracy) to get a value that matches the actual reality. In
the world of health, reality is divided into 2, sick and not sick. So that the
assessment of validity includes two aspects: the accuracy of assessing diseased
conditions is called sensitivity, and the accuracy of assessing non-painful
conditions is called specificity. In other words, sensitivity is the ability of
a test to state positively about those who are sick, while specificity is the
ability of a test to state negatively about those who are not sick (Putra et al., 2016). Thus, the validity of the STRIC application can
assess the positive condition of a stroke for those who have a stroke and can
say no stroke for those who do not have a stroke
The accuracy value of the STRIC application is
83.33%, which means that the ability of the STRIC application to detect the
risk of stroke for all subjects tested correctly is 83.33%. A stroke risk
detection tool requires a high level of sensitivity to guarantee that the tool
is used and strengthens the suspicion that there is a stroke risk for subjects
who have had a stroke. The sensitivity, specificity, and accuracy values
obtained in the STRIC application show that this application is very effective
in detecting the risk of stroke.
CONCLUSION
The application of STRIC stroke risk resulted
in a sensitivity value of 91.67%, specificity of 77.78%, and accuracy of
83.33%. Thus, the benefit of this good sensitivity is being able to identify
individuals with high risk precisely and accurately, so that appropriate and
effective precautions can be provided. Thus, applying STRIC can help prevent
stroke and improve the prognosis for high-risk individuals. However, STRIC
sensitivity is not absolute and there can still be cases that go undetected as
high-risk even though they have certain risk factors. Therefore, it is
necessary to conduct periodic evaluations and seek additional information
related to risk factors in individuals who are not detected by STRIC to ensure
that all high-risk individuals have been correctly identified.
REFERENCES
Ananda, Z., & Darliana, D. (2017).
Kecemasan dengan Kualitas Hidup pada Pasien Stroke. Jurnal Ilmiah Mahasiswa
Fakultas Keperawatan, 2 (3), 1-10.
Arif, M. (2017). Hubungan pelaksanaan screening test menelan
dengan kejadian disfagia pada pasien baru yang menderita stroke akut. Jurnal
Kesehatan Perintis, 4 (2), 61�67, https://doi.org/10.33653/jkp.v4i2.230
Birawa, A. B. P., & Amalia, L.
(2015). Stroke
pada usia muda. Cermin Dunia Kedokteran, 42 (10), 736�739.
Chen, X.-G., Shi, S.-Y., Hu, L., Chen, Y., Sun, H.-W., Zhou,
L., Lu, Z.-B., Wang, H., Wang, X.-S., & Yu, J. (2022). Longitudinal changes
in the hypothalamic�pituitary�adrenal axis and sympathetic nervous system are
related to the prognosis of stroke. Frontiers in Neurology, 13.
Doi: 10.3389/fneur.2022.946593
Feigin, V. L., Norrving, B., & Mensah, G. A. (2017).
Global burden of stroke. Circulation Research, 120 (3), 439�448, https://doi.org/10.1161/CIRCRESAHA.116.308413
Istiqomah, L., Irfani, F. N., &
Shafriani, N. R. (2021). Literature
Review: Tingkat Efektivitas Penggunaan Rapid Test Antibodi Metode
Immunocromatography Sebagai Skrining Awal Pasien Covid-19.
Marbun, A. S., Juanita, J., &
Ariani, Y. (2016). Hubungan antara stres
dan gaya hidup dengan kualitas hidup pasien stroke. Jurnal Keperawatan
Sriwijaya, 3 (1), 1-10.
Mongkau, L., Langi, F. L. F. G., & Kalesaran, A. F. C.
(2022). Studi
Ekologi Prevalensi Diabetes Melitus Dengan Stroke Di Indonesia. PREPOTIF: Jurnal Kesehatan Masyarakat, 6 (2), 1156�1162, https://doi.org/10.31004/prepotif.v6i2.4027
Nuraini, B. (2015). Risk factors of hypertension. Jurnal
Majority, 4 (5).
Nusahi, B. S. (2018). Sensitivitas
Dan Spesifisitas Metode Immunochromatography (Ict-Tb) Terhadap GeneXpert MTB Pada
Tersangka Penderita TB. Universitas Muhammadiyah Semarang.
Putra, I. W. G. A. E., Epid, M.,
Sutarga, I. M., Kardiwinata, M. P., KM, S., Suariyani, N. L. P.,
IntHlth&Dev, M., Septarini, N. W., Subrata, D. I. M., & Erg, M. (n.d.).
Program Studi Kesehatan
Masyarakat Fakultas Kedokteran Universitas Udayana Denpasar, 2016.
Siswosudarmo, R. (2017). Tes diagnostik (Diagnostic test). J.
Metodol. Penelit, 12.
Ugia May Hudatama, U. M. A. Y. H.
(2020). Penerapan ROM (range
of motion) pada asuhan keperawatan pasien stroke dengan gangguan kebutuhan
mobilitas fisik jarong sungai sariak, kec. Baso kabupaten agam kota bukittinggi
tahun 2020. Universitas
Perintis Indonesia.
Vyas, M. V, Silver, F. L., Austin, P. C., Yu, A. Y. X.,
Pequeno, P., Fang, J., Laupacis, A., & Kapral, M. K. (2021). Stroke
incidence by sex across the lifespan. Stroke, 52 (2), 447�451.
Wayunah, W., & Saefulloh, M. (2017). Analisis faktor yang
berhubungan dengan kejadian stroke di rsud indramayu. Jurnal Pendidikan
Keperawatan Indonesia, 2 (2), 65�76. https://doi.org/10.17509/jpki.v2i2.4741
Yaslina, Y., Murni, L., & Najwa, L. (2019). Hubungan
Karakteristik Individu Dan Dukungan Sosial Dengan Perilaku Pencegahan Stroke
Pada Masyarakat Diwilayah Kerja Puskesmas Gulai Bancah. Prosiding Seminar
Kesehatan Perintis, 2 (1), 86.
