Alemu et al.

European Journal
European Journal of Medical Research (2024) 29:452
https://doi.org/10.1186/s40001-024-02026-9 of Medical Research

RESEARCH Open Access

Time to death and its determinant factors

of stroke patients at Gambella General Hospital,
Gambella, Ethiopia
Chekol Alemu1* , Habitamu Wudu1, Bizuayehu Bogale2, Zerihun Getachew3 and Abebe Nega4

Abstract
Background and purpose A stroke or a cerebrovascular accident is a common cause of death and a leading cause
of long-term, severe disability in both developed and developing countries. The most recent global burden of dis-
ease report states that there were 11.9 million new cases of stroke worldwide; stroke accounts for nearly 1 in 8 deaths
globally (12%, 6.5 million deaths) and claims a life every 5 s, making it the second most common cause of death
worldwide. The goal of the study was to identify the most important factors influencing stroke patients’ time to death
at Gambella General Hospital.
Methods Data was gathered from patient files in a hospital using a retrospective study methodology, spanning
the period from September 2018 to September 2020. R 3.4.0 statistical software and STATA version 14.2 were used
for data entry and analysis. The survival time was compared using the log-rank tests and the Kaplan–Meier survival
curve. The fitness of the Cox proportional hazard model was examined.
Results The final model that was fitted was the log-logistic AFT model. A statistically significant correlation
was defined as having a p value of less than 0.05 and the accelerated factor (γ) with its 95% confidence interval
was employed. Eight days was the total median death time (95% CI 6–10). Significant predictors for shortened mortal-
ity time were age (γ = 0.94; 95% CI (0.0.920–0.980), hypertension (γ = 0.63; 95% CI (0.605–0.660), and baseline compli-
cations (γ = 0.24; 95% CI (0.223–0.256).
Conclusions The shortened timing of death was significantly predicted by age, hypertension, and baseline compli-
cations. In light of the study’s findings, health administrators and caregivers should work to improve society’s overall
health.
Keywords Stroke, Survival, Time to death, Survival analysis

Introduction
*Correspondence: A blood vessel bursts or becomes blocked by a clot, fre-
Chekol Alemu
[email protected] quently leading to a stroke, which is caused by the dis-
1
Department of Statistics, College of Natural and Computational ruption of the brain’s blood flow. Damage to the brain
Sciences, Gambella University, Gambella, Ethiopia tissue results from cutting off the delivery of oxygen and
2
Monitoring, Evaluation, Accountability and Learning (MEAL) Officer,
Doctors with Africa-CUAMM, Gambella, Ethiopia nutrients. In both industrialized and developing nations,
3
Department of Rural Development and Agricultural Extension, College a stroke or cerebrovascular accident is a frequent cause of
of Agriculture and Natural Resource, Gambella University, Gambella, demise and a major factor in severe, long-term disability
Ethiopia
4
Department of Statistics, College of Natural Sciences, Jimma University, [1].
Jimma, Ethiopia According to the most recent report on the global
burden of disease, there were 11.9 million new cases of

© The Author(s) 2024. Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0
International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long
as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if
you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or
parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated
otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not
permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To
view a copy of this licence, visit http://​creat​iveco​mmons.​org/​licen​ses/​by-​nc-​nd/4.​0/.
Alemu et al. European Journal of Medical Research (2024) 29:452 Page 2 of 10

stroke worldwide. Nearly 1 in 8 deaths (12%, 6.5 million receiving adequate rehabilitation (cure) services. This a
deaths) were attributed to stroke, making it the second series implication in terms of saving the life of patients
world wide’s most common cause of death [2]. Every especially in poorly developed societies where hemor-
2 s, a stroke occurs somewhere in the world due to the rhage strokes which are characterized by sever neuro-
increasing burden of the disease [3]. By the end of 2030, logic presentation are very much prevalent [13]. A study
it is predicted that stroke will have increased to 23 mil- which was conducted in Addis Ababa city, types and
lion new cases and 7.8 million deaths per year in the associated factor of stroke at selected public referral hos-
absence of a strong global public health response [4]. pitals in Addis Ababa; Ethiopia by [10] and another study
There are no methodologically sound stroke studies in conducted in Addis Ababa city, prevalence, nursing man-
Sub-Saharan Africa, including Ethiopia [5]. Additionally, agements and patients outcomes among stroke patients
earlier studies on stroke in Ethiopia and the rest of Africa admitted to Tikur Anbessa Specialized Hospital, Addis
were primarily descriptive summaries of stroke kinds, Ababa, Ethiopia using logistic regression model, by [14].
subtypes, patient risk profiles, and risk factor magnitude. Logistic regression has been used extensively in studies
In Ethiopia although, admission to the hospitals due to on the prevalence and risk factors for stroke, and the Cox
stroke is increased time to time. According to the latest proportional hazards model has also been used in several
data published in 2017, stroke deaths in Ethiopia reached studies using mortality as the endpoint [10, 14]. However
39,571 or 6.23% of total deaths [6]. Stroke deaths account Logistic regression does not account for the censoring
for 89.82 deaths per 100,000 people when age is taken observations. Even though a semi-parametric estimate
into consideration [6]. In high-income nations, the esti- provides more flexibility, a parametric estimate is more
mated age-adjusted incidence rate in 2010 was 138.9 per powerful provided the baseline hazard’s form is known
100,000 person-years, while in low- and middle-income in advance. There is limited evidence regarding the deter-
countries, it was 182.6 per 100,000 person-years [7]. minants of time to death of a stroke in the current study
Studies from 61 low-income individuals found an area.
increase in hemorrhagic and ischemic strokes of 22%
(from 5 to 30%) and 6% (from −7 to 18%), respectively
[8]. Even though the exact emergency burden of stroke
Objectives
The objectives of this were: (1) to estimate median sur-
in Ethiopia is not known, it has been estimated to be
vival time; (2) to identify determinant factors associated
increasing and stroke accounts for 2.5% of all hospital
with stroke-related death; and (3) to compare the survival
admissions and 13.7% of medical admissions [9]. Stroke
probability of stroke patients among different levels of
is a common and serious condition, and treatments for it
determinant factors.
have a small effect on overall health. As of 2008, it’s been
stated that the incidence of stroke in emerging nations
has overtaken that in industrialized nations [10]. Methods
In the twenty-first century, there has been a 42% Study design
decrease in stroke incidence in the high-income coun- The retrospective study design was gathered for patients
tries; stroke incidence in the low- to middle-income in the medical ward [15].
countries has increased by more than 100% [11]. The
trend is generalized because studies suggest that the geo-
graphical variations in stroke incidence and prevalence Study setting
are small. While the geographical variation of stroke The study was conducted from retrospective records at
incidence is small worldwide, the burden of stroke shows Gambella General Hospital; Gambella Peoples National
larger geographical variation. Unfortunately, most stroke Regional State, Southwestern part of Ethiopia from the
burden is carried by the low- to middle-income coun- 1st of January 2018 to the 1st of January 2022 among
tries [12]. Despite the alarming threat of stroke as a major patients who were admitted by stroke.
public health problem in Ethiopia, stroke epidemiology is
not well-studied in Ethiopia. Participants
In Ethiopia, stroke is a frequent cause of mortality Stroke patients were registered and admitted to the
and morbidity from non-communicable diseases. It has intensive care unit ward of Gambella General Hospital
been shown to be the most common neurological con- during the required period and patients for whom data
dition seen in Ethiopia. Like other developing countries for variables of interest was complete were included while
resources for stroke care and rehabilitation are defi- out of the interval period and patients with insufficient
cient (poor) in Ethiopia. Patients with stroke are often information about one of the factor variables either in the
poorly managed and discharged from hospital without registration book or in the card that was not included.
Alemu et al. European Journal of Medical Research (2024) 29:452 Page 3 of 10

Variables Operational definitions

The response variable was the time to death of patients An event is defined as the occurrence of the death of
measured in days. The survival time of the outcome stroke patients at the time of follow-up in the hospital.
of interest (death in this study) is the duration of time Censored are patients who were referred to other hos-
considered from the start of anti-stroke treatment pitals, and discharged from the hospital, or patients who
to the date of the patient’s death or censoring (i.e., didn’t develop the event and lost to follow-up.
1 = death and 0 = censoring) and the independent vari- An incomplete patient chart refers to charts that have
ables are place of residence (residence) (i.e., urban & no date of admission and major variables.
rural), sex of patients (sex), hypertension (i.e., presence Baselines complications refer to the most common
or absence), cardiac disease (presence or absence), age complications of stroke are Brain edema, or swelling of
of patients in a year, baseline complication (presence or the brain, and pneumonia.
absence), types of stroke(ischemic, hemorrhagic, Both), Stroke subtypes Strokes can be classified into two main
diabetes mellitus (presence or absence). types: ischemic (caused by a clot in a blood vessel in the
brain), or hemorrhagic (caused by a bleed in the brain).
Survival time defined as the time from the date of
Study size admission to recovery from pneumonia, determined for
The study included all stroke patients who had a follow- each participant.
up between 1st of January 2018 and 1st of January 2022
and satisfied the inclusion criteria. In this study, the
sample size was 203, since we have 259 stroke patients Statistical methods
in the study setting. But from these, only 203 stroke Data processing and analysis
patients satisfy the inclusion criteria. The data was entered into Epidata version 3.1. Statistical
analysis consists of descriptive data analysis and survival
Data collection tool and procedure model fitting to make an inference by non-parametric
Data collection was carried out by one trained nurse model, semi-parametric Cox proportional hazard mod-
and one internal medicine resident. It was secondary els, and parametric survival (AFT) accelerated failure
data that was recorded on registration charts and cards time models. All inferences were conducted at a 5% sig-
via nurses, laboratory technicians, medical doctors, and nificance level using R version 3.4.0 &STATA 14.2 were
clinicians. statistical software packages used for analysis. Kaplan
The hospital’s registry is used to extract data from Meier curves and log-rank tests were used to compare
stroke patients’ initial date of admission up to the date the survival experience of different categories of patients
of the patient’s death or censored during the study [16, 17].
period, the medical ward registration chart and the Univariable Cox-proportional hazards regression
patient’s identification cards were used to select the model was fitted for each predictor. Those variables hav-
variables in the study by trained clinicians. The cards ing a p ≤ 0.25 in the bivariable analysis were selected [18,
were prepared by the Federal Ministry of Health to be 19]. Moreover, further variable selection was undertaken
uniformly used by clinicians to early identify and docu- using stepwise backward variable selection proportional-
ment clinical and laboratory variables. Thus, the data ity hazard assumption was tested using Schoenfeld resid-
were collected from patient follow-up records based on uals tests, graphically using the log–log plot of survival,
the variable in the study. The necessary history used for and time-dependent test using (tvc) command [20–22].
the study was taken from the patient and/or caregivers Log logistic AFT model was employed as the final model
by the language they understood. To ensure the qual- for the study. Acceleration factor (γ ) with its 95% con-
ity of data, the data abstraction tool was developed in fidence interval and p values was used to measure the
English, translated to the local language (Amharic), and strength of association and to identify the statistically sig-
back-translated into English to check its consistency. nificant result. p < 0.05 was considered a statistically sig-
Data on socio-demographic characteristics, and clinical nificant association [17, 23, 24].
characteristics of patients, including risk factors, clini-
cal presentation, subtypes of stroke, and time to death
of patients owing to stroke, were collected using the Results
data collecting form. Of a total of 259 patient medical records, 56 were
excluded because of incompleteness, and the remaining
203 were included in the final analysis. Therefore this
study included a total of 203 stroke patients fulfilling the
Alemu et al. European Journal of Medical Research (2024) 29:452 Page 4 of 10

Table 1 Descriptive summary of stroke patient data set at GGH

from 2018 to 2022
Variable Category Event (status)
Censored (%) Died (%)

Sex Male 84 (55.26) 29 (56.86)

Female 68 (44.74) 22 (43.14)
Stroke type Ischemic 87 (57.24) 25 (49.02)
Hemorrhagic 40 (26.32) 23 (45.10)
Both 25 (16.45) 3 (5.88)
Residence Rural 91 (59.87) 30 (58.82)
Urban 61 (40.13) 21 (41.18)
Baseline complication Yes 95 (62.50) 37 (72.55)
No 57 (37.50) 14 (27.45)
Hypertension Yes 109 (71.71) 37 (72.55)
No 43 (28.29) 14 (27.45)
Diabetes mellitus Yes 43 (28.29) 19 (37.25)
No 109 (71.71) 32 (62.75) Fig. 1 Discharge status of stroke patient’s data set at GGH from 2018
Cardiac disease Yes 65 (42.76) 29 (56.86) to 2022
No 87 (57.24) 22 (41.14)
Continuous variable Mean S.D
Age 54.49 21.08
(62.75%) were dead. Of the patients included in the study,
132 (65.02%) had baseline complications of which 37
(72.55%) were dead, and 71 (34.98%) were free from base-
inclusion criteria in Gambella General Hospital. Sum- line complications of which 14 (27.45%) were dead. The
mary results for covariates included in this study are pre- table below shows that the mean age of the respondents
sented in Table 1. was 54.49 years, with a standard deviation of 21.08 years.
As shown in Table 1, a total of 259 stroke patients were Among those patients with stroke, 152 (74.87%) were
treated in the hospital GGH during the study period censored and 51 (25.13%) died, as shown in Fig. 1.
from January 2018 to January 2022. Of the total popula-
tion, this study included 203 stroke patients for whom Non‑parametric survival analysis
data for variables of interest are complete. Of all 203 The estimated value of the survivor function patients
stroke patients, 152 (74.9%) were censored or not experi- decreases at an increasing rate from the time of ori-
enced the event and 51 (25.1%) died. The study included gin until 10 days and remains constant after 15 days, as
113 male patients of which 56.86% were dead. Among 90 shown in Fig. 2.
female patients in the study, 43.14% died. Of the patients Figure 3 depicts the probability of survival of patients
included in the study, 146 were Hypertensive patients based on hypertension, demonstrating that patients
of which 72.55% died, and 57 were free from Hyperten- without hypertension had a longer survival time than
sion of which 27.45% were dead. There were 82 (40.39%) those with hypertension this indicates that patients with-
urban patients of which 21 (41.18%) died, 121 (59.61%) out hypertension have better survival experience than
of them were rural patients of which 30 (58.82%) were patients with hypertension.
dead. Of the patients included in the study, 94 (46.31%)
were Cardiac disease patients of which 29 (56.86%) were Log‑rank test
dead, 109 (53.69%) were free from Cardiac disease of The log-rank test was used to compare survival time
which 22 (43.14%) were dead. There were 112 (55.17%) between categories of different predictors. Based on this
ischemic-type patients of which 25 (49.02%) were dead, test, survival time among different groups of predictors
63 (31.03%) of them were hemorrhagic patients of which such as the presence of baseline complication of patients,
23 (45.10%) were dead and 28 (13.79%) of them were both hypertension and cardiac disease were significantly dif-
type stroke patients of which 3 (5.88%) were dead. Of the ferent in survival time at a 5% level of significance and
patients included in the study, 62 (30.54%) were Diabe- estimated median time to death stroke patients for all
tes mellitus patients of which 19 (37.25%) were dead, 141 observations were 8 days with 95% CI [6, 10]. The median
(69.46%) were free from Diabetes mellitus of which 32 death time of patients due to stroke varied among various
Alemu et al. European Journal of Medical Research (2024) 29:452 Page 5 of 10

Fig. 2 K–M plots of the survival function of stroke patients at GGH from 2018 to 2022

Fig. 3 K–M survival plot by hypertension of stroke patients

categories of predictors. For example, the median death Survival of the patients is significantly related to sex,
time of patients who had cardiac disease was 6 days and age, diabetes mellitus, and hypertension at a 25% level
those who had no cardiac disease were 10 days. The of significance were selected as candidate potential vari-
median times to death of patients with past baseline ables. In the 2nd step, all selected predictors in 1st step
complication and without baseline complication were 10 were fitted in the proportional hazard model and candi-
and 7 days, respectively, as shown in Table 2. date predictors at a 10% level of significance were chosen
using the backward selection method, variables duration,
history of ARTI, insurance status, and clinical presenta-
The univariable and multivariable analysis result tion during admission were selected as candidate poten-
The 1st step in the model-building process is univariable tial variables.
analysis. Predictors which had an association at a p value All selected predictors were fitted in the proportional
of 0.25 in univariable Cox regression were included in hazard model and candidate predictors at a 10% level of
multivariable Cox regression. significance were chosen using the backward selection
Alemu et al. European Journal of Medical Research (2024) 29:452 Page 6 of 10

Table 2 Median time to death and log-rank test by predictors of stroke patients
Variable Category Median death time (95% CI) Log-rank p value
X2 value (df)

Residence Rural 5 (4–6) 0.16 (1) 0.68

Urban 4 (4–6)
Cardiac disease No 10 (11–16) 6.23 (1) 0.035*
Yes 6 (6–15)
Hypertension No 6 (4–5) 5.4 (1) 0.0175*
Yes 5 (5–6)
Baseline complication No 10 (11–16) 5.34 (1) 0.010*
Yes 7 (9–17)
Stroke type Ischemic 9 (6–13) 0.951 (2) 0.826
hemorrhagic 6 (4–10)
Both 8 (6–12)
Diabetes mellitus No 5 (4–6) 0.00 (1) 0.9976
Yes 5 (4–6)
Sex Male 9 (5–14) 0.5 (1) 0.21
Female 7 (6–9)
Total observations 203 8 (6–10)
*
indicates that the comparison difference was significant at a 5% significance level and df is the degree of freedom

method. Variables of sex, age, and hypertension were Table 3 AIC, BIC and log-likelihood of the candidate parametric
selected as candidate potential variables. models
All selected variables at a 10% level of significance in Distribution AIC BIC Log-likelihood
the second step and the non-significant variable in the
univariate analysis at a 25% level of significance were Exponential 578.10 607.54 −281.049
modeled together using the forward selection method Weibull 231.58 264.70 −106.79
the following predictors were selected at a 10% level of Log-normal 204.42 237.54 −93.21
significance. Log-logistic 143.58 160.70 −68.79
Age, presence of baseline complication, hypertension,
and diabetes-mellitus were statistically significant at a 5%
significance level and those predictors were selected as
log-logistic AFT model has (AIC = 143.58, BIC = 160.70)
the final model. It is the best model compared to forward
which is selected as a good model to fit the survival time
and backward selection methods since it has the smallest
of stroke patients data than other accelerated failure time
value of AIC.
model such as exponential, Weibull and lognormal as a
Using different methods predictors of age and hyper-
baseline distribution.
tension violate the proportional hazard assumption.
The final model results are shown as follows according
Thus, we doubt the accuracy of the PH assumption and
to Table 4 under the log-logistic AFT model. Hyperten-
consider the AFT model for this data set.
sion, baseline complication, and age of stroke patients
were significant at a 5% significance level. An accelera-
Accelerated failure time (AFT) model tion factor greater than one (positive coefficient) indi-
When PH assumptions were not satisfied, the paramet- cates extending the time to death while an acceleration
ric AFT model should be used instead of the Cox model factor less than one (negative coefficient) indicates short-
[25]. ened time to death. The output of the final log-logistic
Multivariable analysis of exponential, Weibull, log- AFT model is presented in Table 4. This output showed
normal and log-logistic parametric models was done Stroke patients with hypertensive, with baseline com-
using all significant predictors in the final multivariable plications and patients who were older had significantly
Cox PH model at a 5% level of significance. To compare shortened survival times. The estimated acceleration
the efficiency of different models AIC and BIC was used. factor for patients with hypertension is 0.63 with (95%
A model having the minimum AIC and BIC value was CI 0.605 0.660). The confidence interval for the accelera-
selected as a good model. Accordingly, from Table 3, tion factor did not include one and the p value is small
Alemu et al. European Journal of Medical Research (2024) 29:452 Page 7 of 10

Table 4 Summary result of the final Log-logistic AFT model of stroke patients
Variable Category 
β SE Sig γ 95% CI (γ )

Baseline complication No (ref )

Yes −1.43 0.035 0.0023* 0.24 (0.223 0.256)
Hypertension No (ref )
Yes −0.46 0.022 0.003* 0.63 (0.605 0.660)
Diabetes mellitus No (ref )
Yes 0.068 0.036 0.59 1.071 (0.998 1.149)
Age −0.06 0.012 0.001* 0.94 (0.92 0.98)
Constant 0.534 0.0502 0.000 1.705 (1.545 1.881)
AIC = 155.6, BIC = 188.7, gamma = 1/p = 0.142
P = Shape parameter,γ Indicates Acceleration factor; 95% CI (γ ): 95% confidence interval for acceleration factor, S.E standard error for the coefficient, and * indicates
significance at 5% significance level

the Log-logistic baseline distribution plot and Cox-Snell

residual plot make approximately a straight line through
the origin than the rest AFT models. So this plot suggests
that the Log-logistic AFT model is appropriate. Also, the
likelihood ratio test in Table 5 shows that the model is
significant and the log-likelihood values of the null model
and the full model indicate that the model had a signifi-
cant improvement after the covariates were added to the
model.

Discussion
Stroke, also known as a cerebrovascular accident, is a
prominent cause of severe, long-term impairment in
Fig. 4 Log-logistic baseline distributions plot of stroke patients
both industrialized and developing nations. For the effec-
at GGH from 2018 to 2022 tive management of stroke patients and the develop-
ment of a stroke preventive strategy, the time to death
and the factors that determine it are crucial. The goal of
this study was to pinpoint the variables that affected how
(p = 0.003). This indicates hypertensive patients have less long it took stroke victims at Gambella General Hospital
survival time than patients who are not hypertensive. to pass away. A total of 203 patients were enrolled in the
Similarly acceleration factor for patients with baseline study to determine the associated factors of time to death
complication was 0.24 with (95% CI 0.223 0.256) the γ CI for stroke patients; of those patients, 74.9 were censored
did not include one and the p value is small (p = 0.0023). or did not experience the event, and 25.1% died. This
This implied the expected survival time of stroke patients study agrees with the study conducted by [26], that 27.2%
decreased by 76% for patients with baseline complication perished, while 72.8% were censored or did not witness
as compared to patients who have no baseline complica- the tragedy. People with hypertension, baseline complica-
tion (reference), finally holding other factors constant in tions, and older ages were greater at risk for stroke than
the model. Finally holding other factors constant in the person’s without hypertension, baseline complications,
model, for the age of the stroke patients for 1 year change and younger ages. The average time for all patients was
in the age of patients the log of survival time is decreased 6 with a standard deviation of 3.2 this study agrees with a
by 0.06. study conducted by [26].
Survival models that were parametric, semi-paramet-
Model diagnostic ric, and nonparametric were all used in this investiga-
To check whether the fitted model adequately describes tion. Based on the Kaplan–Meier estimate approach, a
the data or not two graphical methods and the Likeli- non-parametric method is utilized to compare the dif-
hood ratio test were used Adequacy of Parametric Base- ferences between each categorical covariate. The Cox PH
lines plot and CoxSnell residual plot. From Figs. 4 and 5, model was used to fit a semi-parametric survival analysis.
Alemu et al. European Journal of Medical Research (2024) 29:452 Page 8 of 10

Fig. 5 Cox–Snell residuals plots of log-logistic baseline distribution

Table 5 Likelihood ratio and significance of the Log- logistic AFT modifiable risk factor for stroke and hypertension was
model the most frequent co-morbidity that occurred in 50.6% of
Loglik(intercept only) Loglik(model) Chisq DF Sig all stroke patients [28].
Age was found to be a key determinant in this study
−273.12 −68.79 408.66 5 0.000 when determining the time until a stroke patient died;
as patients aged, their chances of survival reduced. This
result is consistent with another investigation in the liter-
Schoenfeld residuals and the Cox PH model’s assump- ature that found that becoming older is the primary, non-
tions were tested graphically, and both were shown to be modifiable driver of stroke risk [28]. For this study, there
false. The researcher then proposed a parametric AFT are no significant differences between rural and urban
survival model as a substitute for the Cox PH model to patients with time to death. In contrast, another study
suit the pneumonia data from Gambella General Hospi- conducted in Tanzania showed there were significant dif-
tal. For the Stroke patient dataset at Gambella General ferences between rural and urban populations with time
Hospital, the researcher fit AFT models using several to death [29].
baseline distribution patterns. The baseline distributions
used in this study were Exponential, Weibull, Log-nor- Conclusion
mal, and Log-logistic. The log-logistic AFT model was This study used the survival time of Stroke patients’ data-
selected as a better AFT model than Weibull, Exponen- set of those patients who started their Stroke treatment
tial, and log-normal models based on comparison crite- from 1st January 2018 to 1st January 2022 years to deter-
ria with smaller AIC and BIC values. The overall median mine the determinant factors of time to death of Stroke
time from stroke patients was 8 days (mean = 6 days; patients in Gambella General Hospital. Out of the total
standard deviation = 3 days). This study is almost consist- 203 stroke patients who started Stroke treatments, about
ent with the Research conducted by [26]. 25.1% died at the end of the study. The estimated median
Age, baseline complications, and hypertension were survival time of stroke patients was 8 days.
statistically significant predictors of the survival status To determine the associated factors of survival time
of stroke in this study. This study is consistent with the of stroke patients, the Cox PH model was used and the
study conducted by [26] Hypertension is a highly predic- PH assumption was checked by graphical, Schoenfeld
tor of death of stroke patients. This is consistent in the residual plot and global test. Then, AFT model was fitted
literature, multiple studies have identified hypertension because the assumption of the Cox proportional model
as the leading risk factor for stroke in SSAs [27]. Other was violated. Different AFT models using different base-
studies have found that hypertension is an important line distributions were applied. Among them using AIC
Alemu et al. European Journal of Medical Research (2024) 29:452 Page 9 of 10

and BIC, the Log-logistic AFT model is a better-fitted the national level, but the patients came from differ-
survival time of Stroke patients’ dataset than other AFT ent regions of the country. As the data is gathered from
baseline distributions. the treatment card of patients of the study has limited
The best model to fit the data to explain the survival number of variables considered as risk factors for the
time of the Stroke patient dataset in Gambella General survival time of stroke patients.
Hospital was the Log-logistic AFT model, which was
Abbreviations
revealed using the graphical technique and Cox-Snell AIC Akaike information criterion
residuals plots. AFT Accelerated failure time
In Gambella General Hospital, the results of a Log- BIC Bayesian information criterion
CVA Cerebrovascular accident dalys
logistic AFT model revealed that age, hypertension and CI Confidence interval
baseline complication were found to be determinant DALYs Disability-adjusted life-years
factors of the survival status of stroke patients. Patients GGH Gambella General Hospital
GBD Global burden of disease
without hypertension and baseline complications had HR Hazard ratio
considerably longer survival time (higher survival expe- PH Proportional hazards
rience). While 1-year increases in age (older age) short- SNNPR Southern Nations Nationalities and Peoples region
SSA Sub-Saharan Africa
ened the survival time by 0.94 times. The health giver S.D Standard deviation
to be planned and awareness about the risk factors of S.E Standard error
stroke, and the benefit of regular medical checkups and TPAs Tissue plasminogen activators
TR Time ratio
treatment follow up should be given to the community. WHO World Health Organization
Strategies for screening and management of hyperten-
sion, age and baseline complication should be given pri- Acknowledgements
We would like to acknowledge Gambella university office of research
ority as they are the most prevalent determinant factors directorate for their sponsorship and financial support for this study and the
identified. Identifying and managing early stroke compli- Gambella General Hospital Health staff in Gambella to undertake this study
cations are important for the prevention of early stroke with their cooperation and permission in using the data.
related mortality. To prevent strokes we should focus on Author contributions
reducing vascular risk factors such as high blood pres- Chekol Alemu was involved in this study from the data acquisition, inception
sure stroke patients. to design, data cleaning, data analysis, and interpretation and drafting and
revising of the manuscript. Habitamu Wudu, Bizuayehu Bogale, Zerihun
Based on this study, the following recommendations Getachew and Abebe Nega were involved in principal supervision, interpreta-
are forwarded for policy makers and the responsible bod- tion, data analysis, and revising the final manuscript. All authors read and
ies: age, baseline complication and hypertension were approved the final manuscript.
significant factors and need to be considered when plan- Funding
ning and developing policies against stroke to increase The only funder for the study was Gambella University. The funding body did
patient’s survival time. Additionally, special attention not have any role in study design, data collection, data analysis, interpretation
of data, or in writing the manuscript.
should be given for old age patients in order to prolonged
death timing. Availability of data and materials
Based on the finding of the study the following rec- The datasets used and/or analyzed during the current study are available from
the corresponding author upon reasonable request.
ommendations were made for ministry of health, the
community at large, Gambella General Hospital and
Declarations
researcher. Community outreach program has to be
planned and awareness about risk factors of stroke, ben- Ethics approval and consent to participate
efit of regular medical checkup and treatment follow up All methods are performed according to the relevant regulations and guide-
lines of the journal. The ethical clearance approval letter was obtained from
should be given to the community and periodic follow up the Gambella University Institutional Review Board research directorate ethical
and adherence to the treatment of determinant hyperten- approval committee (with reference number GURPGC/201/2015). The struc-
sion, baseline complication can minimize the chance of tured questionnaire was developed by the researcher and the secondary data
from patients’ charts or log-book were collected by well-experienced health
getting stroke. workers from Gambella General Hospital, Gambella, Ethiopia. The study was
conducted without individual informed consent obtained from all subjects
and their literate legal guardian because of the secondary nature of the data.
Limitation of study All methods were performed per the Declarations of Helsinki.
The following were some of the study’s limitations.
Consent for publication
There is a lack of published literature in the country Not applicable. No person’s details, images, or videos are being used in this
regarding the survival time of stroke disease, with ref- study.
erences to the outcomes of other countries. This study
Competing interests
used Gambella General Hospital data from a single The authors declare that they have no competing interests.
hospital, which does not represent the prevalence at
Alemu et al. European Journal of Medical Research (2024) 29:452 Page 10 of 10

20. Schoenfeld D. Partial residuals for the proportional hazards regression

model. Biometrika. 1982;69(1):239–41.
Received: 13 April 2024 Accepted: 13 August 2024 21. Lemeshow S, May S, Hosmer DW Jr. Applied survival analysis: regression
modeling of time-to-event data. Hoboken: John Wiley & Sons; 2011.
22. Marubini E, Valsecchi MG. Analysing survival data from clinical trials and
observational studies. Hoboken: John Wiley & Sons; 2004.
23. Lee ET, Wang J. Statistical methods for survival data analysis. Hoboken:
References John Wiley & Sons; 2003.
1. Johnston SC, Mendis S, Mathers CD. Global variation in stroke burden and 24. Nelson WB. Recurrent events data analysis for product repairs, disease
mortality: estimates from monitoring, surveillance, and modelling. Lancet recurrences, and other applications. Philadelphia: SIAM; 2003.
Neurol. 2009;8(4):345–54. 25. Boersma E, Kertai MD, Schouten O, Bax JJ, Noordzij P, Steyerberg EW, et al.
2. Collaborators G. Global, regional, and national incidence, prevalence, and Perioperative cardiovascular mortality in noncardiac surgery: validation of
years lived with disability for 354 diseases and injuries for 195 countries the Lee cardiac risk index. Am J Med. 2005;118(10):1134–41.
and territories, 1990–2017: a systematic analysis for the Global Burden of 26. Ababu DG, Getahun AM. Determinants of stroke mortality through
Disease Study 2017. 2018. survival models: the case of Mettu Karl Referral Hospital, Mettu, Ethiopia.
3. Powers WJ, Derdeyn CP, Biller J, Coffey CS, Hoh BL, Jauch EC, et al. 2015 Stroke Res Treat. 2022. https://​doi.​org/​10.​1155/​2022/​99851​27.
American Heart Association/American Stroke Association focused update 27. Akinyemi RO. Epidemiology of parkinsonism and Parkinson’s dis-
of the 2013 guidelines for the early management of patients with acute ease in Sub-Saharan Africa: Nigerian profile. J Neurosci Rural Pract.
ischemic stroke regarding endovascular treatment: a guideline for health- 2012;3(03):233–4.
care professionals from the American Heart Association/American Stroke 28. Asefa G, Meseret S. CT and clinical correlation of stroke diagnosis, pattern
Association. Stroke. 2015;46(10):3020–35. and clinical outcome among stroke patients visting Tikur Anbessa Hospi-
4. Abegunde DO, Mathers CD, Adam T, Ortegon M, Strong K. The burden tal. Ethiop Med J. 2010;48(2):117–22.
and costs of chronic diseases in low-income and middle-income coun- 29. Allen LA, Stevenson LW, Grady KL, Goldstein NE, Matlock DD, Arnold RM,
tries. Lancet. 2007;370(9603):1929–38. et al. Decision making in advanced heart failure: a scientific statement
5. Adeloye D, Basquill C. Estimating the prevalence and awareness from the American Heart Association. Circulation. 2012;125(15):1928–52.
rates of hypertension in Africa: a systematic analysis. PLoS ONE.
2014;9(8):e104300.
6. Schütte S, Acevedo PNM, Flahault A. Health systems around the world–a
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in pub-
comparison of existing health system rankings. J Glob Health. 2018.
lished maps and institutional affiliations.
https://​doi.​org/​10.​7189/​jogh.​08-​010407.
7. Feigin VL, Forouzanfar MH, Krishnamurthi R, Mensah GA, Connor M, Ben-
nett DA, et al. Global and regional burden of stroke during 1990–2010:
findings from the Global Burden of Disease Study 2010. Lancet.
2014;383(9913):245–55.
8. Krishnamurthi RV, Feigin VL, Forouzanfar MH, Mensah GA, Connor M,
Bennett DA, et al. Global and regional burden of first-ever ischaemic and
haemorrhagic stroke during 1990–2010: findings from the Global Burden
of Disease Study 2010. Lancet Glob Health. 2013;1(5):e259–81.
9. Deresse B, Shaweno D. Epidemiology and in-hospital outcome of stroke
in South Ethiopia. J Neurol Sci. 2015;355(1–2):138–42.
10. Ng M, Fleming T, Robinson M, Thomson B, Graetz N, Margono C, et al.
Global, regional, and national prevalence of overweight and obesity
in children and adults during 1980–2013: a systematic analysis for the
Global Burden of Disease Study 2013. The lancet. 2014;384(9945):766–81.
11. Feigin VL, Lawes CM, Bennett DA, Barker-Collo SL, Parag V. Worldwide
stroke incidence and early case fatality reported in 56 population-based
studies: a systematic review. The Lancet Neurol. 2009;8(4):355–69.
12. Ng M, Fleming T, Robinson M, Thomson B, Graetz N, Margono C, et al.
Global, regional, and national prevalence of overweight and obesity
in children and adults during 1980–2013: a systematic analysis for the
Global Burden of Disease Study 2013. The Lancet. 2014;384(9945):766–81.
13. Greffie ES, Mitiku T, Getahun S. Risk factors, clinical pattern and outcome
of stroke in a referral hospital, Northwest Ethiopia. Clin Med Res.
2015;4(6):182–8.
14. Asres AK, Cherie A, Bedada T, Gebrekidan H. Frequency, nursing manage-
ments and stroke patients’ outcomes among patients admitted to Tikur
Anbessa specialized hospital, Addis Ababa, Ethiopia a retrospective,
institution based cross-sectional study. Int J Afr Nurs Sci. 2020;13:100228.
15. Mohammed T, Mahmud S, Gintamo B, Mekuria ZN, Gizaw Z. Medica-
tion administration errors and associated factors among nurses in Addis
Ababa federal hospitals, Ethiopia: a hospital-based cross-sectional study.
BMJ Open. 2022;12(12):e066531.
16. Kaplan EL, Meier P. Nonparametric estimation from incomplete observa-
tions. J Am Stat Assoc. 1958;53(282):457–81.
17. Collett D. Modelling survival data in medical research. Boca Raton: CRC
Press; 2023.
18. Muche R. Applied Survival Analysis: Regression Modeling of Time to
Event Data. DW Hosmer, Jr., S Lemeshow. New York: John Wiley, 1999, pp.
386, US $89.95. ISBN: 0–471–15410–5. Oxford University Press; 2001.
19. Collett D. Modelling survival data. In: Collett D, editor. Modelling survival
data in medical research. Boston: Springer; 1994. p. 53–106.