DETERMINING THE RISK FACTORS AMONG STROKE PATIENTS ADMITTED AT

SELECTED HEALTH CARE FACILITIES OF KAKAMEGA COUNTY, KENYA

ABSTRACT

CONTEXT: The prevalence of stroke is escalating with stroke survivors facing a more
significant threat from long term complications. The burden of dependence associated with this
impairment, the potential loss of earning capacity, the need for greater social support, and
prolonged hospitalization threaten to overwhelm health and social care systems.

AIM: The study sought to determine the risk factors among stroke patients admitted at selected
health care facilities of Kakamega County, Kenya

METHODOLOGY: The study adopted a cross-sectional study design from January 2021 to
January 2022 with a total of 153 patients who had complete medical records, confirmed
diagnosis of stroke were recruited from level four and level five hospitals of Kakamega county.
Data collection on demographic characteristics, key medical history, clinical presentations,
stroke management, in-hospital events and treatment outcomes at selected health care facilities
of Kakamega County, Kenya was done through self-administered questionnaire and analysed
using Statistical Package for Social Sciences (SPSS) version 27.0.

RESUSLTS: Men were the dominant participant (71.5%) out of 153 stroke patients who took
part in the study. The patients' average age was 56.3±12.7 years. One out of every five patients
smoked, and more than half, 64 (57%), were from rural areas. People who had hemorrhagic
stroke were associated with drinking alcohol more that of ischemic stroke. When it came to co-
morbid conditions, hypertension (40.8%) was the most frequently detected risk factor among the
stroke cases studied. Other common risk factors included atrial fibrillation in 34 (21.9%),
diabetes mellitus in 33 (21.2%), and a history of heart failure in 30 percent (19.3%) of those
diagnosed. In Table 4.1, the co-morbid illnesses with the lowest prevalence rates were chronic
renal disease (4%), coronary heart disease (11%), and a history of stroke (9%).

CONCLUSION: Treatment of stroke patients was sub-optimal and almost half of the patients
had poor treatment outcomes. Availing of thrombolytic therapy, devising appropriate preventive
measures of risk factors (hypertension), and decreasing preventable complication such as
aspiration pneumonia could improve patient outcomes

KEYWORDS: Stroke, hypertension, risk factors.

INTRODUCTION

Stroke is the most prevalent cause of long-term impairment among people and the second largest
cause of mortality globally (Reddy & Yusuf, 1998). Roughly 80% of stroke-related fatalities
1
have place in poor nations (WHO, 2002).Statistics show that stroke case mortality rates are
greater in sub-Saharan Africa (SSA) than in wealthy nations (Walker et al., 2000).This could be
connected to unmanaged risk factors including hypertension and inadequate access to healthcare.
Compared to developed countries, stroke affects younger people in developing nations (Gillum,
1988). Stroke mortality in the US is on the decline overall, although it is still greater in African
Americans than in Whites (5). Several risk factors for stroke have been identified, such as age,
sex, hypertension, hyperlipidemia, cigarette smoking, alcohol consumption, diabetes mellitus,
inadequate fruit and vegetable intake, physical inactivity, obesity, atrial fibrillation, and other
heart disorders (Lemogoum et al., 2005). Hypertension is the primary factor globally linked to
increased risk of both ischemic and haemorrhagic strokes. It accounts for 90% of all strokes in
the population (O’dennell et al., 2010).

The quality of stroke care offered throughout a patient's hospital stay has a significant impact on
outcomes such as functional ability and mortality. Providing treatment for patients with acute
stroke in specialized stroke units has been demonstrated to have a significant impact on
improving the long-term prognosis of stroke (Trialists’, 2013). The majority of hospitals in sub-
Saharan Africa lack stroke units. The Stroke Quality Enhancement Research Initiative (QUERI)
is a national group focused on improving the quality of stroke care within the US Veterans
Health Administration (VHA). They have developed a chart review methodology to assess
evidence-based quality indicators for stroke care. These indicators are based on existing stroke
quality indicators from the US Joint Commission and other relevant stroke care processes
specific to VHA medical centers (Arling et al., 2012). The quality indicators comprised the
documentation of the National Institute of Health Stroke Scale (NIHSS). Thrombolysis, where
necessary, should be administered. Anti-thrombotic therapy should be initiated by the second day
of hospitalization. Lipid management should be implemented. Deep Vein Thrombosis (DVT)
prophylaxis is recommended. Anticoagulation is necessary for patients with atrial fibrillation.
Dysphagia screening should be conducted prior oral intake, Pressure ulcer evaluation, Fall risk
Evaluation, Prompt mobilization, Cessation of smoking counselling, anti-thrombotic therapy at
discharge, Rehabilitation consult and stroke education at discharge. Most of these quality
indicators are easy to assess and include processes of care that are affordable to implement (e.g.
administering aspirin or proper nursing care) hence are relevant to resource limited settings for
monitoring and improving the quality of stroke care.

METHODOLOGY

Analytic cross-sectional study design using quantitative approach was used. The study was
carried out in level 4 and 5 hospitals of Kakamega County namely; Kakamega CGH, Butere
county hospital, Malava county hospital and Lumakanda county hospital. A total of 153 patients
who had complete medical records, confirmed diagnosis of stroke were recruited from level four
and level five hospitals of Kakamega county as at January 2021 to January 2022. We included
Patients who had complete medical records, confirmed diagnosis of stroke, and admitted in the
medical ward of the selected hospital during the period of January 2021 to January 2022. We
excluded Patients with a diagnosis of transient ischemic attack or hematoma, as well as stroke
cases with incomplete medical records (not include a patient demographic, left against medical
advice, and unidentified stroke sub-type clinically or by neuro-imaging). The study used a

2
Cluster sampling technique to group the health facilities in their respective levels of health care.
Kakamega County health facilities have been grouped into four cluster levels, namely Hospitals
(level 5 and 4), Health centers (level 3), and Dispensaries (level 2). Purposive sampling
technique to select the health facilities in level five and four facilities. The hospitals were chosen
because had the minimal requirement to admit and manage stroke patients. The sample size was
calculated using the formulae:

n= (z2pq)/δ2

Where

n = Desired sample size (when population is greater than 10,000)

z = Standard Normal Deviation which was equal to 1.96 corresponding to 95% confidence
interval

p = Prevalence of the issue under study, 50% p = 0.5

q= 1-p

δ= the error of margin, taken as 0.05. Substituting the figures above in the formula.

Thus n= 1.962 x 0.5 x 0.5/0.052

n = 384

The target population was less than 10,000 the sample size was adjusted using the formula.

nf = n/ [1+ (n/N)] Where;

nf – Desired sample size (when the population was less than 10,000).

n – Sample size (when population was more than 10,000) calculated 384.

N – Average number of patients admitted with stroke in a month Thus

nf = n/1 + (n/N)

= 384

n/1 + (384/209)

= 139

10% of the sample was used to cater for non-responses.

10% of 139 sample size to cover for those who opted out from the study

=14

3
The sample size was:

139 + 14= 153

Distribution of sample size among the selected hospitals by PPS where Kakamega CGH have
146 average number of patients admitted with stroke in a month, Butere 73, Malava 55 and
Lumakanda 43. Therefore, distribution of sample size among the selected hospitals was
computed as shown below:

Kakamega CGH

146/384x153 =85

Butere

73/384x153=29

Malava

55/384x153=22
Lumakanda
43/384x153=17

An adaptation of the data collection tool was made after reviewing relevant literature in related
fields. A semi-structured questionnaire was created in order to extract information from the
medical records of eligible patients. The data collection was carried out by two nurses who had
received specialized training. The contents of the data collection tool were cross-checked against
patient medical charts in order to ensure that they were consistent with standard hospital practice
The discharge treatment outcome was classified as either a good treatment outcome or a poor
treatment outcome based on the treating physician's discharge summary notes or a record of
physical disability measurement using the modified Rankin scale, as indicated by the treating
physician (mRs). A positive treatment outcome/improvement was considered if a patient was
discharged with no significant disability (able to carry out all pre-stroke activities without
assistance from other individuals) or if the patient had a record of mRs 2 when the study began.
A poor treatment outcome was defined as a patient who was discharged with moderate to severe
disability (bedridden, incontinent, requiring continuous care, or an mRs score of 3–5) or who
died during their hospital stay (mRs score of 6) as a result of their treatment. The length of a
hospital stay was calculated as the period of time between admission to and discharge or death
from the hospital. There was a checklist used, this was to assess key elements of stroke standard
guideline as documented by health care providers. The checklist included a list of necessary
investigations carried out on admission of stroke patients. A pre-test was done on 14 patients at
Vihiga County Referral Hospital to ensure the validity and reliability of the data collection
instruments. Prior to actual data collection, a pilot study was carried out at Vihiga County
Referral Hospital.

The statistical package for social science (SPSS) version 27 was used to analyse the data. The
socio-demographic variables and clinical results between stroke subtypes were described using

4
 descriptive statistics such as proportions, means, and standard deviations. The chi-square test or
Fisher's exact test were used to compare categorical variables, and the t-test was used to compare
continuous variables. For multivariate analysis, variables with a p-value of less than 0.25 in
binary logistic regression analysis were selected. The researchers employed multivariable
logistic regression to find independent predictors of poor clinical outcomes in the hospital. It was
determined that predictors with a probability value less than 0.05 and a confidence interval that
did not contain 1 were statistically significant

Ethical approval and permission for the study was obtained from Masinde Muliro University
Institutional Research and Ethics Committee (IERC). Data collection permission for the study
was sought from the County government of Kakamega and the National Commission of Science
Technology and Innovation (NACOSTI). Participation in the study was voluntary, and all
participants taking part in the study signed an informed consent form for participation.
Information on patients remained confidential and was used only for the study. Safety of the
collected data was guaranteed by using a computer with a password known by just the researcher
as well as storage in cupboards under lock and key. Before the study, permission to conduct
research was sought through a letter of approval from the university ethical review board and the
health care facilities involve.

RESULTS

Men were the dominant participant (71.5%) out of 153 stroke patients who took part in the study.
The patients' average age was 56.3±12.7 years. One out of every five patients smoked, and more
than half, 64 (57%), were from rural areas. People who had hemorrhagic stroke were associated
with drinking alcohol more that of ischemic stroke. When it came to co-morbid conditions,
hypertension (40.8%) was the most frequently detected risk factor among the stroke cases
studied. Other common risk factors included atrial fibrillation in 34 (21.9%), diabetes mellitus in
33 (21.2%), and a history of heart failure in 30 percent (19.3%) of those diagnosed. In Table 4.1,
the co-morbid illnesses with the lowest prevalence rates were chronic renal disease (4%),
coronary heart disease (11%), and a history of stroke (9%).

Table 1: Risk factors and socio-behavioral characteristics among stroke patients among
stroke patients admitted at selected health care facilities of Kakamega County, Kenya
Variables Ischemic (n = 112) Hemorrhagic Total (n = 153) p-value
(n = 41)
Age (in years) ‡ 57.2±11.5 53.6±13.6 56.3±12.2 0.128
Sex (male), n (%) 79 (70.4) 30 (74.6) 110 (71.5) 0.607
Residence (rural) 64 (57.0) 21 (52.7) 86 (55.9) 0.681
Smoking, n (%) 17 (14.9) 10(25.8) 29 (18.6) 0.172
Alcohol, n (%) 15(11.4) 12 (29.7) 27 (17.3) 0.009†
Co-morbid
conditions
Hypertension, n (%) 41 (36.5) 21 (52.7) 65 (40.8) 0.072

5
Diabetes, n (%) 20 (17.8) 12 (30.7) 33 (21.2) 0.087
AF, n (%) 23 (20.4) 10(25.8) 34 (21.9) 0.406
Pervious stroke, n 7 (6.1) 2 9 (5.5) 0.615*
Heart
(%) failure, n (%) 22 (18.6) 8 (20.9) 30(19.3) 0.734
CAD, n (%) 8 (2.0) 5 12 0.734
CKD, n (%) 6 (5.3) 6 (14.6) 14 (8.8) 0.110*
Note; Abbreviations: AF atrial fibrillation, CAD, coronary artery disease, CKD, chronic kidney
disease.

‡Expressed as mean and standard deviations (SD); *Fisher’s exact test, †Statistically significant
at p- value <0.05

DISCUSSION

According to the findings of the current study, males were more likely than females to suffer a
stroke than females. The majority of stroke patients, as opposed to those who suffered from
hemorrhagic stroke, presented with ischemic stroke. High blood pressure was shown to be the
most common risk factor in the current investigation. The most prevalent type of stroke subtype
diagnosed in this study was ischemic stroke, which was the most common subtype overall. This
finding is consistent with multiple previous studies, which found that ischemic strokes were
more common than hemorrhagic strokes in the general population. On the other hand, there have
been studies conducted in Ethiopia that have found an equal prevalence of both stroke subtypes
as well as a high prevalence of hemorrhagic stroke in the population. This variation could be
related to disparities in stroke diagnosis (clinical versus neuroimaging), differences in socio-
economics and risk factors at the population level, or a combination of these factors. The
prevalence of hemorrhagic stroke is increasing in Sub-Saharan Africa, despite the fact that
ischemic stroke continues to be the most common type of stroke presentation in the region.
Hemorrhagic stroke is associated with a significant risk of fatality. People in most of Sub-
Saharan Africa have different risk factors for stroke (like high blood pressure), environmental
factors, study design (community-based vs. institutionalized), and clinical diagnosis that aren't
the same.

In the current investigation, hypertension was shown to be the most common concomitant illness
among those who had suffered a stroke (42.2%). Compared to other research, this finding was
consistent with the finding that hypertension is the most common risk factor for stroke around
the world. 2 When it comes to stroke patients in Sub-Saharan Africa, hypertension is still
underdiagnosed or poorly managed among those who are receiving treatment, which contributes
to poor treatment outcomes. Because of low awareness, restricted access to healthcare, and a lack
of a healthy lifestyle, it is possible that this tendency may continue. All of these things can help
cut down on the burden of cerebrovascular disease. Preventing, diagnosing, and treating
hypertension, as well as raising public awareness of the condition, can all play a role. The health
seeking behaviors is usually reflected by the value citizens’ place on health. Myths and beliefs
play a key role in the health seeking behaviours of a person. A comparison study indicated more
stroke incidences in urban areas compared to the rural areas. This was associated to the health

6
seeking behaviours. Many individuals in the urban centers were seeking medical attention in
private pharmacies and hence delaying care (Gichana, 2015). In the study, taking herbal
medicine, visiting traditional healers and missing scheduled appointments delayed initiation of
treatment. A minority also engaged in lifestyle practices such as alcohol consumption or cigarette
smoking that may contribute to stroke condition. A previous study in KNH by Gichana, (2015)
showed that 18.2% were cigarettes smokers. Majority of the smokers later suffered stroke.

Majority have comorbid conditions such as hypertension, diabetes or heart diseases. A very
small proportion presented within the required 12 hours after onset of stroke which could have
worsened the outcome of clinical management of their condition. In a study to determine the
outcomes of early vs. delayed admissions to a neurological department of Central India Institute
of Medical Sciences, Hundred and four patients admitted in the facility were grouped as an early
referral (within 24 hours from onset of symptoms) and late referrals (after 24 hours from onset of
the symptoms). Long-term outcomes were determined in the delayed and early admission.
Outcomes of Death, dependency, coma, disability, treatment burden depression and recovery in
both groups were analysed. The analysis was performed to determine the mortality rate of stroke
in the hospital for 12 months, and a comparison was drawn. Analysis showed better prognosis
with a high recovery rate of 90 % with a level of improvement in early admissions compared to
delayed admission with a recovery rate of 23% (Kaddumukasa et al., 2017). Similarly, the ratio
of dependency was lower in early admission 6% compared to delayed admission 18%. Less
than a third in this study were within acceptable range of BMI. Obesity increases the risk for
high blood pressure, diabetes, stroke and high blood cholesterol. Once considered a problem in
developed countries, obesity is on the rise in developing countries (Temu et al., 2021). Four
million of the Kenyan population is obese (Mkuu et al., 2018). Being obese and overweight
increases risk of stroke (Mkuu et al., 2018). Obesity has a significant effect on stroke directly or
through predisposing conditions such as diabetes and hypertension leading to stroke (Mkuu et
al., 2018).

Limitations of the Study

The county has many health care facilities with many providers of health but the research was
conducted in a few selected hospitals from the county; therefore, the outcome of research cannot
be generalized to a large population. Lack of follow up in prediction of long-term outcome for
stroke patients after discharge. This was a hospital-based study that did not involve follow up to
patients who were discharged to determine the long-term outcomes. The convenient sampling
method that was used in selecting the health care facilities may have brought up a selection bias.
In addition, stroke management performance and stroke outcome were used as a measure of the
actual performance of the care givers, future studies should include other dependent variables
related to stroke management performance

CONCLUSION

One out of every five patients smoked, and more than half, were from rural areas. Ischemic
stroke, was the most often seen subtype in this investigation. Stroke patients were in their mid-

7
50s when they were admitted, with a male preponderance. One of the most common comorbid
conditions observed was hypertension (high blood pressure). Both ischemic stroke and
hemorrhagic stroke patients utilized aspirin and atorvastatin as the most prevalent drugs, whereas
ACEIs were the most widely used antihypertensive medications in patients with both forms of
stroke. A considerable number of stroke patients had poor treatment results as a result of their
illness. Older age, heart failure, a lower level of consciousness on admission, and aspiration
pneumonia were all found to be strongly related to a poor prognosis. All of these things could
improve the health of patients: the use of thrombolytic treatment, the development of appropriate
risk factor prevention strategies, and the reduction of preventable complications like aspiration
pneumonia. The stroke Standard guideline for management of CVA is an important factor for
health care workers in obtaining optimal level of performance.

REFERENCES

1. Reddy, K. S., & Yusuf, S. (1998). Emerging epidemic of cardiovascular disease in

2. World Health Organization. (2002). The world health report 2002: reducing risks,
promoting healthy life. World Health Organization.

3. Walker, R. W., McLarty, D. G., Kitange, H. M., Whiting, D., Masuki, G., Mtasiwa, D.
M., ... & Alberti, K. M. (2000). Stroke mortality in urban and rural Tanzania. The
Lancet, 355(9216), 1684-1687.

4. Gillum, R. F. (1988). Stroke in blacks. Stroke, 19(1), 1-9.

5. Mozaffarian, D., Benjamin, E. J., Go, A. S., Arnett, D. K., Blaha, M. J., Cushman, M., ...
& Turner, M. B. (2015). Heart disease and stroke statistics—2015 update: a report from
the American Heart Association. circulation, 131(4), e29-e322.

6. Lemogoum, D., Degaute, J. P., & Bovet, P. (2005). Stroke prevention, treatment, and
rehabilitation in sub-saharan Africa. American journal of preventive medicine, 29(5), 95-
101.

7. O'donnell, M. J., Xavier, D., Liu, L., Zhang, H., Chin, S. L., Rao-Melacini, P., ... &
Yusuf, S. (2010). Risk factors for ischaemic and intracerebral haemorrhagic stroke in 22
countries (the INTERSTROKE study): a case-control study. The Lancet, 376(9735), 112-
123.

8. Trialists’Collaboration, S. U. (2013). Organised inpatient (stroke unit) care for

8
9. Arling, G., Reeves, M., Ross, J., Williams, L. S., Keyhani, S., Chumbler, N., ... &
Bravata, D. M. (2012). Estimating and reporting on the quality of inpatient stroke care by
Veterans Health Administration Medical Centers. Circulation: Cardiovascular Quality
and Outcomes, 5(1), 44-51.

10. Gichana, J. K. (2015). Social determinants of stroke among stroke patients attending
medical outpatient clinic at Kenyatta National Hospital (Doctoral dissertation, University
of Nairobi).

11. Kaddumukasa, M., Kayima, J., Nakibuuka, J., Blixen, C., Welter, E., Katabira, E., &
Sajatovic, M. (2017). Modifiable lifestyle risk factors for stroke among a high risk
hypertensive population in Greater Kampala, Uganda; a cross-sectional study. BMC
Research Notes, 10, 1-6.

12. Temu, T. M., Macharia, P., Mtui, J., Mwangi, M., Ngungi, P. W., Wanjalla, C., ... &
Kibachio, J. (2021). Obesity and risk for hypertension and diabetes among Kenyan
adults: Results from a national survey. Medicine, 100(40), e27484.

13. Mkuu, R. S., Epnere, K., & Chowdhury, M. A. B. (2018). Peer reviewed: prevalence and
predictors of overweight and obesity among Kenyan women. Preventing chronic
disease, 15.