Proportion of stroke types in Madagascar: A tertiary-level hospital-based case series.

BACKGROUND: Like other countries in sub-Saharan Africa, Madagascar has a high burden of stroke. The Malagasy population is unique in sharing both African and Asian ancestry. The proportion of ischemic and hemorrhagic stroke types is unknown for this population. AIM: Our aim was to establish the proportion of stroke types and known risk factors for the Malagasy population.
METHODS: We conducted a single-center, tertiary-level hospital-based case series. We included all patients with a CT-imaging confirmed stroke who presented at the emergency ward of the study hospital between January 1, 2017, and November 20, 2018.
RESULTS: Of 223 patients with CT-confirmed stroke, 57.4% (128/223, 95% CI: 51-64%) had an ischemic stroke and 42.6% (95/223, 95% CI: 36-49%) had an intracranial hemorrhage. The majority (89.5%; 85/95, 95% CI: 83-96%) of intracranial hemorrhages were intracerebral; 4.2% (4/95, 95% CI: 0-8%) had a subdural hematoma, 5.3% (5/95, 95% CI: 1-10%) had a subarachnoid hemorrhage, there was one isolated intraventricular hemorrhage (1.1%; 1/95, 95% CI: -1-3%). The prevalence of hypertension among stroke patients was high (86.6%; 187/216, 95% CI: 82-91%).
CONCLUSIONS: Our study is the first to report the proportion of stroke types and known risk factors in Madagascar. We find that the proportion of hemorrhagic strokes was unexpectedly higher than that reported from other countries in sub-Saharan Africa. Our findings highlight the need for a country-specific approach to stroke prevention, treatment, and rehabilitation and provide guidance on public health resource allocation in Madagascar.

Introduction

Stroke, one of the leading causes of permanent disability and the second leading cause of death, accounts for 6 million deaths annually. Approximately 70% of these occur in low- and middle-income countries [1, 2]. Sub-Saharan Africa (SSA), home to around a fifth of the world’s population, bears a high burden of stroke with an age-standardized incidence rate of up to 316 per 100,000, an age-standardized prevalence rate of 1,283 per 100,000, and a case fatality over one month of 24% [3-5]. In addition, stroke in SSA is increasingly affecting a younger age group and causes poorer long-term outcomes than in the developed world aggravating the social and economic toll of disease [6, 7]. As the population of SSA is the fastest growing and life expectancy is increasing most rapidly of all world regions, overall stroke prevalence is steadily rising [2, 8]. Strategies to reduce stroke burden and to adequately allocate health resources are urgently needed.

Despite the rising tide of stroke, stroke research productivity is low while epidemiology and proportion of stroke types are unknown in many countries in SSA [9]. The most recent estimate of stroke burden from the Global Burden of Disease (GBD) project included only 62 studies with participants from SSA, a mere 1.5% of the 4,058 studies used for analysis [2]. Likewise, INTERSTROKE, the largest case-control study on stroke risk factors to date, included only 3.6% of participants from SSA indicating challenges to follow-up and lack of health facilities in which computed tomography (CT) scan or magnetic resonance imaging (MRI) were available [9]. The World Health Organization, United States’ National Academy of Sciences as well as the recently inaugurated African Stroke Organization urgently call for improving local stroke data in SSA [10-12].

In Madagascar, one of the world’s least developed countries with a population of 27 million [13], the estimated life-time risk of stroke is 15.5% [14]. According to the Ministry of Health, stroke is the most common reason for in-hospital death albeit less than 5% of causes of deaths in Madagascar are registered [15, 16]. To date, the proportion of stroke types has not been described for the Malagasy population.

Our aim was to characterize the proportion of stroke types in a hospital-based case series of imaging-confirmed stroke patients. In addition, we describe the prevalence of known risk factors and fatality rates.

Methods

Study setting

This study was conducted at Soavinandriana Military Hospital in Antananarivo, a 454-bed national referral hospital. A CT scanner was available 24/7. Thrombolytic therapy was not available. There was no dedicated stroke unit. Healthcare at the hospital was free for civil servants and their families.

Study design

This was a retrospective hospital-based study including all patients with a CT-imaging confirmed stroke and accessible patient files who presented at the emergency ward of the hospital between January 1, 2017, and November 20, 2018.

Case finding methods and inclusion criteria

We identified cases based on the hospital’s emergency room register, which contained a brief medical history of all patients who sought admission for an acute illness. We extracted all cases that had a neurological deficit of sudden onset including weakness, sensory loss or inattention, speech disturbances (dysarthria or aphasia), visual problems, limb ataxia and gait unsteadiness, as well as non-specific signs including dizziness, seizures, loss of consciousness, impaired cognitive function, and thunderclap headache. We retrieved the medical records of patients with at least one of these symptoms and who were subsequently admitted to the hospital. Of those, we included all patients whose medical record contained CT images of the brain showing an ischemic or hemorrhagic stroke.

Exclusion criteria

Patients whose medical records or CT images could not be retrieved from hospital archives were excluded.

Data collection and data entry

Medical records

We digitized medical records using a digital camera (EOS 550D DSLR, Canon). Data were entered into a standardized data collection form by three trained data collectors (JR, AS, RA). Data extracted from medical records included sociodemographic characteristics (sex and age), clinical characteristics (symptoms upon arrival, time of symptom onset, admission to the ER, and CT scan, duration of hospital stay, and in-hospital mortality), ultrasound findings (echocardiography and carotid duplex scan), lab tests (glycated hemoglobin, lipid profile), medical history, and risk factors (hypertension, diabetes, body mass index, tobacco- and alcohol-consumption, family history of stroke). Data quality was continuously monitored by a supervisor who trained the data collection team. Data were crosschecked and screened for double entries, out of range values, and overall consistency. We anonymized data at the data entry level to protect participants’ personal identifiable information.

Imaging data acquisition and interpretation

Nonenhanced CT was performed on a multidetector CT scanner (SOMATOM Perspective CT VC40, Siemens). Images were developed on X-ray film. We visualized and digitized those images using an X-ray film viewer and a high-resolution digital camera (EOS 550D DSLR, Canon). Digitized images were read by a neuroradiologist (ES) who entered the scan results into a standardized data collection form. Data extracted from CT images included stroke subtype (ischemic or hemorrhagic), lesion side (right, left or bilateral), age (acute, < 24 hours; subacute, 1–5 days; or chronic (> 5 days), lesion size (ischemic: lacunar, < 2/3 of vascular territory, > 2/3 of vascular territory; hemorrhagic: intracerebral hemorrhage volume < 30 ml, intracerebral hemorrhage > 30 ml), lesion expansion, previous lesions and white matter lesions (categorized according to the Fazekas scale (0, no lesions; 1, punctuate lesions; 2, beginning confluence of lesions; 3, large confluent areas)) [17].

Ischemic strokes were classified by vascular territory into anterior cerebral, middle cerebral, posterior cerebral, and vertebrobasilar artery strokes as well as strokes affecting more than one vascular territory and by etiology (i.e., cardioembolism, small-vessel disease or undetermined etiology). Intracranial hemorrhages were classified by location into typical (affecting the basal ganglia, thalamus, pons, or cerebellum) and atypical (all other locations) intracerebral hemorrhage as well as subarachnoid hemorrhage, and subdural hematoma.

Definitions of risk factors

Arterial hypertension, tobacco- and alcohol-consumption, and a family history of stroke were self-reported risk factors. Overweight was recorded as a risk factor if body-mass-index (BMI) was more than 25. Diabetes was either self-reported or was newly diagnosed during hospitalization (glycated hemoglobin > 6,5%). Hyperlipidemia was defined as blood lipid levels above the upper reference threshold of the hospital’s laboratory.

Data analysis

We used descriptive statistics to summarize the data set and independent t-tests to compare metric variables and Pearson’s Chi-Square for categorical variables. Analyses were performed in SPSS (IBM SPSS Statistics, Version 25, 2017).

Ethics approval and consent to participate

This study was approved by the Institutional Review Board of Soavinandriana Hospital (067/CENHOSOA/DG/DT) on June 28, 2018. Informed consent was waived.

Results

We included a total of 223 patients with CT confirmed stroke diagnosis. Lipid profiles were available for 62.3% (139/223, 95% CI: 56–69%) of patients, 29.1% (65/223, 95% CI: 23–35%) had HbA1c measurements, whereas cardiac echo was performed in 24.2% (54/223, 95% CI: 19–30%) and carotid duplex sonography in 9.9% (22/223, 95% CI: 6–14%) of patients.

Medical history was assessed for hypertension in 96.9% (216/223, 95% CI: 95–99%), diabetes in 80.7% (180/223, 95% CI: 76–86%), and tobacco consumption in 68.6% (153/223, 95% CI: 63–75%) of patients.

Clinical and demographic characteristics

Demographic and clinical characteristics are summarized in Table 1 stratified according to stroke type. Patients with hemorrhagic strokes were younger (60 (51–67) vs. 64 (58–72) years, p = 0.003) and more likely to be male (69.5% (95% CI: 60–79%) vs. 55.5% (95% CI: 47–64%), p = 0.034) than patients with ischemic strokes. Most patients (71.8%; 160/223, CI: 66–78%) arrived later than 6 hours after symptom onset; a first CT scan was obtained 3.2 (2.2–5.7) hours after initial presentation. Fig 1 summarizes clinical characteristics upon presentation by stroke type.

Table 1

Demographic and clinical characteristics by stroke type.

	Totala	ISa	HSa	p-valueb
Total	223 (100)	128 (57.4)	95 (42.6)
Sex	223	128	95
female	86 (38.6)	57 (44.5)	29 (30.5)	0.034
male	137 (61.4)	71 (55.5)	66 (69.5)
Age	223	128	95
median age; years (IQR)	62 (54–69)	64 (58–72)	59 (51–67)	0.003
24–35	2 (0.9)	2 (1.6)	0
36–50	32 (14.3)	9 (7.0)	23 (24.2)
51–65	99 (44.4)	57 (44.5)	42 (44.2)
66–80	70 (31.4)	47 (36.7)	23 (24.2)
81–94	20 (9.0)	13 (10.2)	7 (7.4)
Time between symptom onset and arrival at ER	223	128	95
<3h	36 (16.1)	16 (12.5)	20 (21.1)
3–4.5h	15 (6.7)	9 (7.0)	6 (6.3)
4.5-6h	12 (5.4)	4 (3.1)	8 (8.4)
6-24h	86 (38.6)	52 (40.6)	34 (35.8)
24-72h	36 (16.1)	23 (18.0)	13 (13.7)
72h-7d	22 (9.9)	12 (9.4)	10 (10.5)
>7d	16 (7.2)	12 (9.4)	4 (4.2)
Time between ER admission and CT scan	223	128	95
median time; hours (IQR)	3.2 (2.2–5.7)	3.2 (2.2–6.3)	3.2 (2.3–5.4)	0.826
Length of hospital stay	164	104	60
median length; days (IQR)	12.5 (8.2–18.2)	11.8 (9.4–18.4)	15.6 (13.3–21.4)	0.868
< 7d	16 (9.8)	12 (11.5)	4 (6.7)
7d–2w	69 (42.1)	54 (51.9)	15 (25.0)
2–4w	65 (39.6)	26 (25.0)	39 (65.0)
> 4w	14 (8.5)	12 (11.5)	2 (3.3)
Death	223	128	95
yes	49 (22.0)	22 (17.2)	27 (28.4)	0.045
Length of hospitalization until death	48	22	26
median length; days (IQR)	6.9 (3.2–12.9)	12.8 (5.6–15.3)	4.6 (1.4–10.0)	0.179
< 24h	4 (8.3)	0	4 (15.4)
24h–7d	21 (43.8)	9 (40.9)	12 (46.2)
7d–2w	13 (27.1)	6 (27.3)	7 (26.9)
> 2w	10 (20.8)	7 (31.8)	3 (11.5)

IS = ischemic stroke; HS = hemorrhagic stroke

IQR = interquartile range

a Number and (%), if not indicated otherwise

b Statistical tests: chi-square test of independence

Fig 1

Symptoms of patients with CT-confirmed stroke upon presentation in the emergency room by stroke type.

Chi-square: * p<0.05; ** p<0.01; *** p<0.001.

Imaging characteristics

Of 223 patients, 57.4% (128/223, 95% CI: 51–64%) had an ischemic and 42.6% (95/223, 95% CI: 36–49%) a hemorrhagic stroke (Fig 2). Of ischemic and hemorrhagic strokes combined, 45.7% (102/223, 95% CI: 39–52%) were in the left and 41.7% (93/223, 95% CI: 35–48%) in the right hemisphere. One in eight (12.5%, 28/223, 95% CI: 8–17%) was located bilaterally.

Fig 2

Radiological classification of ischemic and hemorrhagic stroke types.

Panel A depicts the distribution to ischemic vs. hemorrhagic stroke as percentage of all strokes, panel B and C show distribution of vascular territories and stroke subtypes separately for ischemic (B) and hemorrhagic stroke (C). ACA, anterior cerebral artery; MCA, middle cerebral artery; PCA, posterior cerebral artery; ICH, intracerebral hemorrhage.

Ischemic stroke

Vascular territories affected by ischemic strokes (128/223) are depicted in Fig 2. We found strokes affecting more than two-thirds of a vascular territory in 29.7% (38/128, 95% CI: 22–38%), less than two-thirds in 50% (64/128, 95% CI: 41–59%), and lacunar strokes in 20.3% (26/128, 95% CI: 13–27%) of patients. Based on imaging, stroke etiology was undetermined in most patients (75.5%; 96/128, 95% CI: 67–83%). Imaging was suggestive of small vessel occlusion in 15.6% (20/128, 95% CI: 9–22%) and cardioembolism as the cause of stroke in 9.4% (12/128; 95% CI: 4–15%) of patients. Almost half of patients 39.4% (50/127, 95% CI: 31–48%) had lesions consistent with previous strokes. White matter lesions indicating chronic small vessel disease were present in 71.1% (91/128, 95% CI: 63–79%) of patients. Table 2 summarizes the imaging characteristics of patients with ischemic stroke.

Table 2

Imaging characteristics of patients with ischemic stroke.

	Totala
	128 (100)
Side of stroke lesion	128
left	56 (43.8)
right	56 (43.8)
both	16 (12.5)
Age of lesion	128
acute (< 24h)	73 (57.0)
subacute (24h-5d)	14 (10.9)
chronic (> 5d)	39 (30.5)
unclear	2 (1.6)
Vascular territory	128
ACA	6 (4.7)
MCA	84 (65.6)
PCA	11 (8.6)
vertebro-basilar	6 (4.7)
multiple	21 (16.4)
Size of lesion	128
lacunar	26 (20.3)
< 2/3 of territory	64 (50.0)
> 2/3 of territory	38 (29.7)
Stroke subtype	128
cardioembolism (embolic stroke)	12 (9.4)
small-vessel occlusion (lacune)	20 (15.6)
undetermined etiology	96 (75.0)
Previous stroke on CT	50/127 (39.4)
White matter lesions	128
Fazekas 0	37 (28.9)
Fazekas 1	35 (27.3)
Fazekas 2	26 (20.3)
Fazekas 3	30 (23.4)

ACA = anterior cerebral artery; MCA = middle cerebral artery; PCA = posterior cerebral artery

a Number and % if not indicated otherwise

Intracranial hemorrhage

Of 95 intracranial hemorrhages on CT imaging, 89.5% (85/95, 95% CI: 83–96%) were classified as intracerebral hemorrhage, 5.3% (5/95, 95% CI: 1–10%) as subarachnoid hemorrhage, 4.2% (4/95, 95% CI: 0–8%) as subdural hemorrhage, and one patient (1.1%; 1/95, 95% CI: -1-3%) had an isolated intraventricular hemorrhage without intraparenchymal hemorrhage. Most intracerebral hemorrhages (76,5%; 65/85, 95% CI: 67–86%) were in the basal ganglia, thalamus, pons, or cerebellum, which are typical for a hypertensive etiology. Large hemorrhages with volumes >30ml occurred in 28.0% (26/90, 95% CI: 19–38%) and intraventricular hemorrhages in 41.5% (39/94, 95% CI: 31–52%) of patients. Table 3 summarizes the imaging characteristics of patients with hemorrhagic stroke.

Table 3

Imaging characteristics of patients with hemorrhagic stroke.

	Totala
	95 (100)
Side of stroke lesion	95
left	46 (48.4)
right	37 (38.9)
both	12 (12.6)
Age of lesion	95
acute (< 24h)	90 (94.7)
subacute (24h-5d)	5 (5.3)
chronic (> 5d)	0
unclear	0
Intracerebral hemorrhage	85/95 (89.5)
Location of origin	85
typical	65 (76.5)
atypical	20 (23.5)
Intraventricular hemorrhage	39/94 (41.5)
Infratentorial origin of hemorrhage	8/95 (8.4)
ICH volume >30ml	26/90 (28.9)
Subarachnoid hemorrhage	5/95 (5.3)
Subdural hemorrhage	4/95 (4.2)

ICH = intracerebral hemorrhage

a Number and (%), if not indicated otherwise

Ultrasound imaging

S1 Table in S1 File summarizes echocardiographic and carotid duplex sonography findings by stroke type.

In-hospital mortality

Forty-nine patients died during the hospital stay (22.0%; 49/223, 95% CI: 17–28%). In-hospital mortality differed significantly between ischemic stroke (17.2%; 22/128, 95% CI: 11–24%) and hemorrhagic stroke (28.4%; 27/95, 95% CI: 19–38%, p = 0.045, Table 1). Most in-hospital deaths were directly attributed to the brain lesion (63.3%; 31/49, 95% CI: 49–77%). Secondary causes of death included chest infection, respiratory failure and cardiac infarction.

Risk factors

Hypertension was the most prevalent risk factor across all groups 187/216 (86.6%; 187/216, 95% CI: 82–91%), followed by active alcohol consumption 38.8% (59 /152, 95% CI: 31–47%), tobacco consumption 29.4% (45/153, 95% CI: 22–37%), and diabetes 17.8% (32/180, 95% CI: 12–23%). Alcohol consumption was more common among patients with a hemorrhagic stroke (64.0% (48/75, 95% CI: 53–75%) vs. 45.5% (35/77, 95% CI: 34–57%), p = 0.022); diabetes was more common among patients with ischemic stroke (23.6% (25 /106, 95% CI: 15–32%) vs. 9.5% (7/74, 95% CI: 3–16%), p = 0.015). Blood lipid profiles revealed dyslipidemia in 30.2% (42/139, 95% CI: 23–38%) of patients. One-third 34.7% (25/72, 95% CI: 24–46%) were overweight (body-mass-index >25). Table 4 summarizes contributing risk factors among patients with ischemic or hemorrhagic stroke.

Table 4

Contributing risk factors among patients with ischemic or hemorrhagic stroke.

	Totala	ISa	HSa	p-valueb
Total	223 (100)	128 (57.4)	95 (42.6)
Stroke family history	42	24	18
	11/42 (26.2)	3/24 (12.5)	8/18 (44.4)	0.020
Tobacco	153	85	68
non-smoker	71 (46.4)	38 (44.7)	33 (48.5)	0.411
active smoker	45 (29.4)	23 (27.1)	22 (32.4)
former smoker	37 (24.2)	24 (28.2)	13 (19.1)
Alcohol	152	77	75
no consumption	69 (45.4)	42 (54.5)	27 (36.0)	0.012
active consumption	59 (38.8)	21 (27.3)	38 (50.7)
former consumption	24 (15.8)	14 (18.2)	10 (13.3)
Hypertension	216	127	89
	187 (86.6)	113 (89.0)	74 (83.1)	0.216
Diabetes	180	106	74
	32 (17.8)	25 (23.6)	7 (9.5)	0.015
BMI	72	43	29
median; BMI (IQR)	23.5 (20.3–25.8)	24.2 (19.5–25.7)	22.0 (20.6–26.5)	0.570
<18.5	9 (12.5)	8 (18.6)	1 (3.4)
18.5–25.0	38 (52.8)	19 (44.2)	19 (65.5)
25.0–30.0	18 (25.0)	10 (23.3)	8 (27.6)
30.0–35.0	6 (8.3)	5 (11.6)	1 (3.4)
>35.0	1 (1.4)	1 (2.3)	0
HbA1c	65	45	20
normal	32 (49.2)	19 (42.2)	13 (65.0)	0.090
increased	33 (50.8)	26 (57.8)	7 (35.0)
Total cholesterol	157	96	61
normal	134 (85.4)	81 (84.4)	53 (86.9)	0.681
decreased	9 (5.7)	5 (5.2)	4 (6.6)
increased	14 (8.9)	10 (10.4)	4 (6.6)
LDL cholesterol	139	83	56
normal	97 (69.8)	55 (66.3)	42 (75.0)	0.185
increased	42 (30.2)	28 (33.7)	14 (25.0)
HDL cholesterol	142	84	58
normal	63 (44.4)	35 (41.7)	28 (48.3)	0.834
decreased	79 (55.6)	49 (58.3)	30 (51.7)
Triglycerides	156	96	60
normal	133 (85.3)	83 (86.5)	50 (83.3)	0.585
decreased	3 (1.9)	1 (1.0)	2 (3.3)
increased	20 (12.8)	12 (12.5)	8 (13.3)

IS = ischemic stroke; HS = hemorrhagic stroke

BMI = body mass index (kg/m2); HbA1c = glycated hemoglobin; LDL = low-density lipoprotein; HDL = high-density lipoprotein

IQR = interquartile range

a Number and % if not indicated otherwise

b Statistical tests: chi-square test of independence

Discussion

Our study describes the proportion of stroke types in Madagascar using a hospital-based, imaging-confirmed series of cases. The conspicuous strengths of our study were the inclusion of stroke patients irrespective of an individual’s financial situation to reduce the risk of selection bias and use of a standardized image analysis protocol.

Among our study population, the majority had ischemic strokes (128/223; 57.4%) predominantly in the middle cerebral artery territory followed by lacunar strokes. Hemorrhagic strokes accounted for 95/223 (42.6%) of cases and the majority occurred in locations typical for hypertensive intracerebral hemorrhage. Almost 90% of stroke patients had hypertension, underscoring the importance of implementing effective prevention strategies to reduce stroke burden. Around a third of ischemic and hemorrhagic strokes were severe, affecting more than 2/3 of a vascular territory or exceeding a bleeding volume of 30ml indicating high levels of functional disability and emphasizing the need for stroke rehabilitation. The median age of our study population was 62 years and patients were more likely to be male (61.4%), exemplifying the substantial economic impact of stroke in Madagascar by affecting a relatively young and productive population.

Scientific literature on stroke in Madagascar is scarce. The Pubmed/MEDLINE database contains only six publications using the search terms “stroke” and “Madagascar”, three of which are hospital-based case series. None included an unselected sample of imaging-confirmed stroke patients. Razafindrasata et al. included only patients with acute motor deficits; 150 of 227 patients had CT imaging, 45% of those were hemorrhagic strokes [18]. Rasaholiarison et al. included only patients with lacunar strokes; all 83 patients had CT imaging, 67% of those were hemorrhagic strokes [19]. Stenumgård et al. report clinical characteristics, socio-demographic factors, and outcomes in 30 consecutive stroke patients but only 3 of those had a CT [20]. Remaining publications on stroke in Madagascar are case reports [21-23].

The Stroke Investigative Research and Educational Network (SIREN) study, the largest study on the proportion of stroke types and associated risk factors in SSA to date, included 2,118 consecutive case-control pairs from Ghana and Nigeria [24]. INTERSTROKE, an international case-control study included 973 stroke patients from Mozambique, Nigeria, South Africa, Sudan, and Uganda [25]. Compared to results from SIREN and INTERSTROKE, stroke patients in Madagascar were older (59.0 and 58.7 vs. 62.1 years) and had a higher likelihood of hemorrhagic stroke (32% and 30.2% vs. 42.6%); hypertension was the most common risk factor in all studies. Hypertension, the most important risk factor for stroke, is common in Madagascar but not more common than in other countries in SSA. Using previous guideline recommendations, 27.0% and 29.7% of rural and urban populations in Madagascar have hypertension defined as blood pressure readings greater than 140/90 mm Hg [26]. This is similar to the prevalence of hypertension found in rural and peri-urban populations in Uganda, South Africa, Tanzania, and Nigeria [27]. Taken together, the higher rate of hemorrhagic strokes in our study might be caused by other modifiable risk factors or genetic predisposition to intracerebral hemorrhage [28, 29].

Compared to other hospital-based case series in SSA ranging from 25.9 to 41.1% of all stroke patients [30-32], the fatality rate in our study (22.0%) was low. This indicates either successful treatment or selection bias which might have been caused excluding patients who died in the emergency room before being admitted to hospital.

Our study has limitations. First, the results of our single-center study in an urban setting might not reflect the true community burden of stroke. However, access to healthcare including CT scans was free of charge at the study hospital for patients reducing selection bias otherwise introduced by a households’ ability to pay [33]. In addition, prompt imaging and comprehensive investigation would not have been feasible in a community-setting. Second, the study hospital being a tertiary-level referral hospital might have introduced a selection bias towards more severe cases of stroke, which might explain the relatively high mortality rate. Third, by including CT-confirmed strokes only, minor strokes for which a CT scan might not have been performed or posterior circulation strokes, for which CT imaging is known to be less sensitive, might be underrepresented. Fourth, while medical records including demographic and clinical data were available for all patients, they were not equally thorough and self-reported medical history may lead to an underestimation of negative findings. On the other hand, patients’ medical history was consistently assessed for relevant risk factors, like hypertension, in 96.9% of records. Fifth, clinical data on stroke severity was limited and no standardized data on stroke severity was available. Nevertheless, we used imaging characteristics to determine stroke severity. Last, our study was not designed to assess stroke prevalence or incidence, as the details of the source population from which the study hospital draws its patients from were unknown.

In conclusion, our study results contribute to determining the clinical outcome and prognosis of stroke patients in Madagascar and provide guidance on public health resource allocation for stroke prevention, treatment, and rehabilitation. In addition, our results may encourage community-based stroke surveillance studies to be conducted in Madagascar.

Ultrasound imaging by stroke type.

(PDF)

Click here for additional data file.

7 Jul 2022

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Ismail Ibrahim Ismail, MD

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. Thank you for stating the following in the Competing Interests section:

"I have read the journal´s policy and the authors of this manuscript have the following competing interests: ME received funding from DFG under Germany's Excellence Strategy – EXC-2049 – 390688087, BMBF, DZNE, DZHK, EU, Corona Foundation, and Fondation Leducq. ME reports grants from Bayer and fees paid to Charité from AstraZeneca, Bayer, Boehringer Ingelheim, BMS, Daiichi Sankyo, Amgen, GSK, Sanofi, Covidien, Novartis, Pfizer, all outside the submitted work."

Please confirm that this does not alter your adherence to all PLOS ONE policies on sharing data and materials, by including the following statement: "This does not alter our adherence to  PLOS ONE policies on sharing data and materials.” (as detailed online in our guide for authors http://journals.plos.org/plosone/s/competing-interests).  If there are restrictions on sharing of data and/or materials, please state these. Please note that we cannot proceed with consideration of your article until this information has been declared.

Please include your updated Competing Interests statement in your cover letter; we will change the online submission form on your behalf.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: No

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: No

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: No

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

Reviewer #3: No

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: I do not think that this study represents the true stroke figures. The authors included patients with CT-confirmed stroke. A significant proportion of stroke patients would have a small infarction in MRI, which would not appear in CT. I think the authors should have included stroke patients meeting the old WHO definition of stroke, irrespective of the neuroimaging: acute-onset focal neurological deficits lasting more than 24hours. This would increase the proportion of ischemic stroke patients. Please consider adding this data to your study. Moreover, I think a significant proportion of stroke patients in SSA would not present to the hospital, which would explain the relatively higher prevalence of hemorrhagic stroke in this study.

"Sub-Saharan Africa (SSA), home to around a fifth of the world’s population, bears a high burden of stroke with an age-standardized incidence rate of 160 per 100,000" Actually, this is not a high incidence. The incidence of stroke in Germany is around 250/100,000 per year

"As the population of SSA is the 73 fastest growing and fastest ageing of all world regions" Please clarify this sentence.

In table S1: Please explain what do you mean by the following terminologies: instable plaque, thrombus, reduced blood flow velocity

Minor mistake: table 1 is described as figure 1

Reviewer #2: Very interesting well- written article that discussed the proportion of stroke types among Malagasy population

I have some comments that might help improve this nice article.

Line 164: (BMI) was more than 25 kg/m2, being an index it should not have units

Table 1: Length of hospital stay (days) (164, 104, 60) is a bit confusing → (days) should be (no of patients ) Then (days) should be in the next rows, same applied for Length of hospitalization until death

Table 2: side of lesion was bilateral in 12.5% of cases however 70% of cases has Fazekas I -III which means that 70 % of patients have bilateral lesions. So, what do you mean here by bilateral lesions?

You did not mention the causes of death in your patients (chest infection , respiratory failure , pulmonary embolism, expanding hge/ inf etc). This is very important to know because your mortality rate is relatively high 22%.

What about the other important stroke risk factors like ischemic heart disease, oral contraceptive pills, migraine, atrial fibrillation, anticoagulants ?

You should mention in the limitation section that being a tertiary-level hospital this may cause selection bias as you will receive more of complicated and severe cases and this might explain the relatively higher morality rate among your cohort.

Reviewer #3: This single center retrospective study using CT-confirmed strokes only, could miss a lot of ischemic strokes like minor strokes , posterior circulation strokes and others for which a CT scan might not have been less useful

Even demographic and clinical data were not equally thorough and self-reported medical history may lead to an underestimation of negative findings.

clinical data on stroke severity was limited and no standardized data on stroke severity was available.

All above mentioned might shed great doubt on the message of the study

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: Yes: Ahmed Elhfnawy

Reviewer #2: No

Reviewer #3: Yes: Ossama yassin mansour

**********

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

14 Aug 2022

Dear Dr Chenette,

We would like to thank you and the reviewers for the constructive and insightful comments on our previous submission. By incorporating the editor’s and reviewers’ suggestions, we believe that we were able to significantly improve the quality of this manuscript. Please find on the following pages a point-by-point response to each of the comments.

Reviewer #1:

1. I do not think that this study represents the true stroke figures. The authors included patients with CT-confirmed stroke. A significant proportion of stroke patients would have a small infarction in MRI, which would not appear in CT. I think the authors should have included stroke patients meeting the old WHO definition of stroke, irrespective of the neuroimaging: acute-onset focal neurological deficits lasting more than 24hours. This would increase the proportion of ischemic stroke patients. Please consider adding this data to your study.

Response: We thank the reviewer for pointing this out and fully agree with this comment. As non-contrast head CT is known to have low sensitivity for detecting ischemic strokes overall and performs even worse in patients with posterior fossa stroke, it would have been preferable to include patients based on the old WHO definition of stroke. However, the overall accuracy and completeness of the available medical records was poor. Most records did not specifically mention the suspicion or diagnosis of a stroke and did not contain standardized information on stroke severity or the clinical course during admission. Thus, using the old WHO definition of stroke would have drastically reduced our sample size and the overall reliability of results. Therefore, we identified cases based on the hospital’s emergency room register, from which we extracted all cases that had a neurological deficit of sudden onset. Of those, we included all patients whose medical record contained CT images of the brain showing an ischemic or hemorrhagic stroke. A particular strength of our study was that CT findings were not obtained from reports, but all images were digitized and read by an experienced neuroradiologist who entered the scan results into a standardized data collection form prior to analysis.

To address the reviewer’s concern the following sentence is included in the limitation section of the manuscript (page 20, line 327-329):

“Third, by including CT-confirmed strokes only, minor strokes for which a CT scan might not have been performed or posterior circulation strokes, for which CT imaging is known to be less sensitive, might be underrepresented.”

2. Moreover, I think a significant proportion of stroke patients in SSA would not present to the hospital, which would explain the relatively higher prevalence of hemorrhagic stroke in this study.

Response: We fully agree with the comment that in general, only a minority of patients after stroke can be expected to present to a hospital in SSA. This might in particular affect more severe stroke cases. However, financial constraints are one of the main reasons for patients to forgo or delay care in SSA [1, 2]. Thus, a conspicuous strength of our study was the inclusion of stroke patients irrespective of an individual’s financial situation which is unique in Madagascar and likely reduced this selection bias.

3. "Sub-Saharan Africa (SSA), home to around a fifth of the world’s population, bears a high burden of stroke with an age-standardized incidence rate of 160 per 100,000" Actually, this is not a high incidence. The incidence of stroke in Germany is around 250/100,000 per year.

Response: We thank the reviewer for spotting this error and apologize for our oversight. The age-standardized incidence rate is up to 316 per 100,000 for Sub-Saharan Africa. We have amended the sentence as follows (page 4, line 67-69):

"Sub-Saharan Africa (SSA), home to around a fifth of the world’s population, bears a high burden of stroke with an age-standardized incidence rate of up to 316 per 100,000 [3]."

4. "As the population of SSA is the fastest growing and fastest ageing of all world regions..." Please clarify this sentence.

Response: We apologize for the lack of clarity. Sub-Saharan Africa (SSA) has a very high fertility rate of 4.7 births per woman compared to 2.4 births worldwide [4]). At the same time, life expectancy is increasing more rapidly in SSA than anywhere else [5]. To improve clarity, we have amended the sentence as follows (page 4, line 72-74):

“As the population of SSA is the fastest growing and life expectancy is increasing most rapidly of all world regions...”

5. In table S1: Please explain what do you mean by the following terminologies: instable plaque, thrombus, reduced blood flow velocity.

Response: We agree with the reviewer that these terms were insufficiently defined. We amended the legend of Supplementary Table 1 as follows:

“Supplementary Table S1. Carotid duplex sonography findings by stroke type:

b unstable atherosclerotic lesion at risk of rupture

c blood clot in carotid artery at risk of embolization

d post-stenotic reduced blood flow velocity in internal carotid artery”

6. Minor mistake: table 1 is described as figure 1.

Response: We agree with the reviewer that the journal’s requirement to insert a figure legend right after the paragraph in which the figure is cited can be confusing especially when a table and a figure legend are inserted together. We have changed the order of figure and table to improve readability of the manuscript.

Reviewer #2: Very interesting well- written article that discussed the proportion of stroke types among Malagasy population. I have some comments that might help improve this nice article.

1. Line 164: (BMI) was more than 25 kg/m2, being an index it should not have units.

Response: We thank the reviewer for spotting this. We deleted the unit.

2. Table 1: Length of hospital stay (days) (164, 104, 60) is a bit confusing → (days) should be (no of patients). Then (days) should be in the next rows, same applied for Length of hospitalization until death.

Response: We apologize for the inconsistency and have adjusted the rows in the Table 1 as suggested by the reviewer for improved clarity and readability.

3. Table 2: side of lesion was bilateral in 12.5% of cases however 70% of cases has Fazekas I -III which means that 70 % of patients have bilateral lesions. So, what do you mean here by bilateral lesions?

Response: We thank the reviewer for the opportunity to clarify this inaccuracy. The lesions we described in the first section of Table 2 refer to the stroke lesions. We found bilateral stroke lesions in 12.5% of cases. In addition to rating the stroke on CT, we included information on white matter lesions using the Fazekas Scale to provide additional information on chronic small vessel disease and cardiovascular risk [7]. We have amended Table 2 accordingly to clarify that we describe only stroke lesions in the first section of the table.

4. You did not mention the causes of death in your patients (chest infection, respiratory failure, pulmonary embolism, expanding hge/ inf etc). This is very important to know because your mortality rate is relatively high 22%.

Response: We fully agree with the reviewer and are grateful for the opportunity to provide more details about the causes of death in the study population. Unfortunately, data quality on the secondary causes was poor and did not allow to distinguish between different causes of death. We included the following paragraph in our results section (page 15, line 249-252):

“Most in-hospital deaths were directly attributed to the brain lesion (63.3%; 31/49, 95% CI: 49-77%). Secondary causes of death included chest infection, respiratory failure, and cardiac infarction.”

5. What about the other important stroke risk factors like ischemic heart disease, oral contraceptive pills, migraine, atrial fibrillation, anticoagulants?

Response: We thank the reviewer for this comment. Unfortunately, details about these risk factors were documented very inconsistently in the available medical records. For most cases it remained unclear whether the information was missing, a risk factor had been ruled out or if screening had not been performed at all. Therefore, we decided to refrain from reporting on risk factors for which data was inaccurate or incomplete.

6. You should mention in the limitation section that being a tertiary-level hospital this may cause selection bias as you will receive more of complicated and severe cases and this might explain the relatively higher mortality rate among your cohort.

Response: We fully agree with the reviewer that the selection of our study hospital might have contributed to a higher number of severe cases of stroke. We have included the following sentence in the limitations section of the manuscript (page 20 line 325-327):

“Second, the study hospital being a tertiary-level referral hospital might have introduced a selection bias towards more severe cases of stroke, which might explain the relatively high mortality rate. Third...”

Reviewer #3:

This single center retrospective study using CT-confirmed strokes only, could miss a lot of ischemic strokes like minor strokes, posterior circulation strokes and others for which a CT scan might not have been less useful.

Even demographic and clinical data were not equally thorough and self-reported medical history may lead to an underestimation of negative findings.

Clinical data on stroke severity was limited and no standardized data on stroke severity was available.

All above mentioned might shed great doubt on the message of the study.

Response: We thank the reviewer for this comment. As also outlined in our responses to Reviewer 1, point 1 above we fully agree that it would have been preferable to include patients based on the clinical definition of stroke as well as to include more demographic and clinical data on stroke severity. However, overall accuracy and completeness of the available medical records for this retrospective analysis was poor. The points have been addressed in the limitations section of the manuscript with our initial submission (page 20, line 310-325):

Our study has limitations. First, the results of our single-center study in an urban setting might not reflect the true community burden of stroke. However, access to healthcare including CT scans was free of charge at the study hospital for patients reducing selection bias otherwise introduced by a households' ability to pay. In addition, prompt imaging and comprehensive investigation would not have been feasible in a community-setting. Second, by including CT-confirmed strokes only, minor strokes for which a CT scan might not have been performed or posterior circulation strokes, for which CT imaging is known to be less sensitive, might be underrepresented. Third, while medical records including demographic and clinical data were available for all patients, they were not equally thorough and self-reported medical history may lead to an underestimation of negative findings. On the other hand, patients’ medical history was consistently assessed for relevant risk factors, like hypertension, in 96.9% of records. Fourth, clinical data on stroke severity was limited and no standardized data on stroke severity was available. Nevertheless, we used imaging characteristics to determine stroke severity. Last, our study was not designed to assess stroke prevalence or incidence, as the details of the source population from which the study hospital draws its patients from were unknown.

Conducting research in a resource-restricted public health setting harbors additional challenges and limitations due to an overall lack of resources and poor data quality. In this manuscript, we draw from multiple data sources to best describe the proportion of stroke types in a hospital-based case-series in Madagascar’s capital Antananarivo, the only hospital in the country where patients can obtain a CT scan regardless of an individual’s financial situation. Despite a high burden of disease, scientific literature on stroke in Madagascar is scarce; there is a mere one-digit number of stroke studies from Madagascar. This lack of research draws an inconclusive picture, hampers efficient public health resource allocation in Madagascar, and adds to overall health inequity. There is an urgent need for more and better quality stroke data in SSA.

References:

1. McLane HC, Berkowitz AL, Patenaude BN, McKenzie ED, Wolper E, Wahlster S, et al. Availability, accessibility, and affordability of neurodiagnostic tests in 37 countries. Neurology. 2015;85(18):1614-22.

2. Swindle RN, David. Barriers to Accessing Medical Care in Sub-Saharan Africa in Early Stages of COVID-19 Pandemic. Poverty and Equity Notes. Washington, DC. ©: World Bank; 2021.

3. Akinyemi RO, Ovbiagele B, Adeniji OA, Sarfo FS, Abd-Allah F, Adoukonou T, et al. Stroke in Africa: profile, progress, prospects and priorities. Nature Reviews Neurology. 2021;17(10):634-56.

4. World Bank. Fertility rate, total (births per woman): The World Bank Group; 2022 [Available from: https://data.worldbank.org/indicator/SP.DYN.TFRT.IN.

5. Wan He IA, Dzifa Adjaye-Gbewonyo. Africa Aging: 2020. U.S. Government Printing Office, Washington D.C. : U.S. Census Bureau; 2020.

6. Reutern G-Mv, Goertler M-W, Bornstein NM, Sette MD, Evans DH, Goertler M-W, et al. Grading Carotid Stenosis Using Ultrasonic Methods. Stroke. 2012;43(3):916-21.

7. Sharma R SS, Cascella M. White Matter Lesions. Treasure Island (FL): StatPearls Publishing; 2022. Available from: https://www.ncbi.nlm.nih.gov/books/NBK562167/.

Submitted filename: Response to Reviewers.docx

Click here for additional data file.

19 Sep 2022

26 Sep 2022

Reviewer #1:

Please remove the terminologies: "unstable plaque" and "thrombus" from table S1. Actually, unstable plaque means plaque neovascularity or floating thrombus. I don't know if this is what the authors mean. I cannot understand what the authors mean by thrombus?

Response: We thank the reviewer for pointing this out and apologize for the imprecise terminology in the previous version. We acknowledge that by non-contrast enhanced duplex ultrasound the diagnostic accuracy to assess the stability of a plaque and the sensitivity to detect intraluminal thrombi is limited. We report the findings as recorded by the Malagasy physician performing the ultrasound examination. To better describe the findings, we added details of the classification used to assess stability and clarified the term “thrombus”.

The legend for “unstable plaque” now reads as:

“unstable atherosclerotic lesion at risk of rupture described as “echolucent” or “predominantly echolucent” corresponding to types 1 and 2 in the Gray-Weale classification [1].”

We omitted the term “thrombus” and specified to “intraluminal thrombus” which is also reflected in the legend now to read as:

“intraluminal carotid artery thrombus at risk of embolization”

Submitted filename: Response to Reviewers_rev2.pdf

Click here for additional data file.

2 Oct 2022

Proportion of stroke types in Madagascar: a tertiary-level hospital-based case series

PONE-D-22-13670R2

Dear Dr. Julius Emmrich

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Ismail Ibrahim Ismail, MD

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

5 Oct 2022

PONE-D-22-13670R2

Proportion of stroke types in Madagascar: a tertiary-level hospital-based case series

Dear Dr. Emmrich:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Ismail Ibrahim Ismail

Academic Editor

PLOS ONE