Journal of the Neurological Sciences 427 (2021) 117535 Contents lists available at ScienceDirect Journal of the Neurological Sciences journal homepage: www.elsevier.com/locate/jns Frequency and factors associated with post-stroke seizures in a large multicenter study in West Africa Fred S. Sarfo a,*, Joshua Akinyemi b, Albert Akpalu c, Kolawole Wahab d, Joseph Yaria e, Oladimeji Adebayo e, Morenike Komolafe f, Reginald Obiako g, Lukman Owolabi h, Godwin O. Osaigbovo i, Carolyn Jenkins j, Yaw Mensah a, Godwin Ogbole k, Benedict Calys-Tagoe l, Philip Adebayo g, Lambert Appiah a, Arti Singh c, Adekunle Fakunle m, Ezinne Uvere m, Tiwari Hemant n, Olayemi Balogun o, Osi Adeleye o, Bimbo Fawale f, Adeniyi Abdulwasiu o, Luqman Ogunjimi p, Onasanya Akinola p, Oyedunni Arulogun m, Arnette Donna q, Okechukwu Ogah r, Rufus Akinyemi o,s, Bruce Ovbiagele t, Mayowa O. Owolabi s,u,**, on behalf of SIREN a Department of Medicine, Kwame Nkrumah University of Science and Technology, Kumasi, Ghana b Department of Epidemiology and Medical Statistics, College of Medicine, University of Ibadan, Nigeria c Department of Medicine, University of Ghana Medical School, Accra, Ghana d Department of Medicine, University of Ilorin Teaching Hospital, Ilorin, Nigeria e University College Hospital, Ibadan, Nigeria f Department of Medicine, Obafemi Awolowo University Teaching Hospital, Ile-Ife, Nigeria g Department of Medicine, Ahmadu Bello University, Zaria, Nigeria h Department of Medicine, Aminu Kano Teaching Hospital, Kano, Nigeria i Jos University Teaching Hospital Jos, Nigeria j Medical University of South Carolina, SC, USA k Department of Radiology, University of Ibadan, Nigeria l Department of Community Health, University of Ghana Medical School, Accra, Ghana m College of Medicine, University of Ibadan, Nigeria n University of Alabama at Birmingham, Birmingham, AL, USA o Federal Medical Centre, Abeokuta, Nigeria p Department of Pharmacology and Therapeutics, Obafemi Awolowo College of Health Sciences, Olabisi Onabanjo University, Remo Campus, Shagamu, Ogun State, Nigeria q College of Public Health, University of Kentucky, USA r University College Hospital, Ibadan, Nigeria s Center for Genomic and Precision Medicine, College of Medicine, University of Ibadan, Nigeria t Weill Institute for Neurosciences, School of Medicine, University of California San-Francisco, USA u Department of Medicine, University of Ibadan, Nigeria A R T I C L E I N F O A B S T R A C T Keywords: Background: Post-stroke seizures (PSS) are associated with significant morbidity and mortality across the globe. Seizures There is a paucity of data on PSS in Africa. Epilepsy Purpose: To assess the frequency and factors associated with PSS by stroke types across 15 hospitals in Nigeria Stroke types and Ghana. Africa Methods: We analyzed data on all stroke cases recruited into the Stroke Investigative Research and Educational Network (SIREN). We included adults aged ≥18 years with radiologically confirmed ischemic stroke (IS) or intracerebral hemorrhage (ICH). PSS were defined as acute symptomatic seizures occurring at stroke onset and/ or during acute hospitalization up until discharge. We used logistic regression to estimate adjusted odds ratios (aOR) with 95% Confidence Interval. * Correspondence to: F. S. Sarfo, Department of Medicine, Kwame Nkrumah University of Science and Technology, Private Mail Bag, Kumasi, Ghana. ** Correspondence to: M. O. Owolabi, University College Hospital Ibadan and Blossom Specialist Medical Center, Ibadan, Nigeria. E-mail addresses: stephensarfo78@gmail.com (F.S. Sarfo), mayowaowolabi@yahoo.com (M.O. Owolabi). https://doi.org/10.1016/j.jns.2021.117535 Received 19 January 2021; Received in revised form 6 June 2021; Accepted 7 June 2021 Available online 9 June 2021 0022-510X/© 2021 Elsevier B.V. All rights reserved. F.S. Sarfo et al. J o u r n a l o f t h e N e u r o l o g i c a l S c i e n c e s 427 (2021) 117535 Results: Among 3344 stroke patients, 499 (14.9%) had PSS (95% CI: 13.7–16.2%). The mean duration of admission in days for those with PSS vs no PSS was 17.4 ± 28.6 vs 15.9 ± 24.7, p = 0.72. There were 294(14.1%) PSS among 2091 ischemic strokes and 159(17.7%) among 897 with ICH, p = 0.01. The factors associated with PSS occurrence were age < 50 years, aOR of 1.59 (1.08–2.33), National Institute of Health Stroke Score (NIHSS), 1.29 (1.16–1.42) for each 5 units rise and white cell count 1.07 (1.01–1.13) for each 10^3 mm3 rise. Factors associated with PSS in ischemic were NIHSS score, aOR of 1.17 (1.04–1.31) and infarct volume of 10–30 cm3 aOR of 2.17(1.37–3.45). Among ICH, associated factors were alcohol use 5.91 (2.11–16.55) and lobar bleeds 2.22 (1.03–4.82). Conclusion: The burden of PSS among this sample of west Africans is substantial and may contribute to poor outcomes of stroke in this region. Further longitudinal studies are required to understand the impact on morbidity and mortality arising from PSS in Africa. 1. Introduction 3. Definition of terms Stroke is a major cause of symptomatic seizures among older adults 3.1. Post-stroke seizures [1–3]. It is estimated that between 5 and 15% of stroke patients develop seizures within two years of stroke onset [4]. The pathogenesis of early Post stroke seizures (PSS) were defined as acute symptomatic sei- post-stroke seizures (PSS) following an ischemic stroke is putatively zures occurring after an acute stroke following ILAE recommendations linked to a lowering of seizure threshold secondary to local ionic shifts, [21,22]. Early PSS was classified as symptomatic seizures occurring the release of excitotoxic neurotransmitters and the presence of global within 7 days of stroke onset while late PSS was defined as symptomatic hypoperfusion with cortical hyperexcitability [2]. The mechanisms for seizures occurring after 7 days of stroke onset. For this study, seizures post-stroke seizures in intracerebral hemorrhage involve direct stimu- were diagnosed clinically and were classified as acute symptomatic PSS latory effects of blood degradation products on neural tissues and based on medical history from a witness (often a family member) of focal extracellular glutamate toxicity [7,8]. or generalized seizures at the time of presentation for admission with Sub-Saharan Africa is currently at the epicenter of a stroke epidemic stroke or by clinically documented seizures during hospitalization for characterized by a younger age of onset and very poor short- and long- acute stroke. No electroencephalographic studies were performed to term outcomes from mortality and post-stroke morbidity [5–14]. There confirm diagnosis of seizures and we did not record seizures into focal- are no reports from large scale multi-center studies on the burden of onset or generalized onset for the purposes of this report. post-stroke seizures except a few single center studies [15–17]. Furthermore, the delineation of factors associated with occurrence of 3.2. Vascular risk factors of stroke post-stroke seizures according to stroke types within the sub-Saharan African context remains to be elucidated. We therefore present data We collected basic demographic and lifestyle data including, socio- on the frequency and factors associated with post-stroke seizures by the economic status, cardiovascular risk profile, dietary patterns, routine primary stroke types from the Stroke Investigative Research and Edu- physical activity, stress, depression, cigarette smoking, and alcohol use cation Networks (SIREN) study. The SIREN study is the largest study on using a validated INTERSTROKE instrument [23]. We have reported stroke in Africa to date involving 15 sites in northern and southern belts these definitions in our previous publications [9,24]. of Nigeria and Ghana. 2. Methods 3.3. Statistical analysis 2.1. Study design We compared demographic and vascular risk factor data among stroke cases who reported with post-stroke seizures versus those without The study protocol has been previously published [18]. In brief, post-stroke seizures using Student’s t-test for parametrically distributed stroke cases were consecutively consenting adults aged ≥18 years with continuous data and Chi-squared tests for categorical data. We assessed clinical stroke presenting within 8 days of current symptom onset or ‘last factors associated with post-stroke seizures among stroke cases and by seen without deficit’. We confirmed all stroke diagnosis using either CT stroke types (ischemic and hemorrhagic strokes) using a multivariable or MRI scan typically within 10 days of symptom onset. Ethical approval logistic regression model. Covariates which were included in the was obtained from all study sites and informed consent was obtained multivariate logistic models were selected if they achieved a p-value of from all subjects [18]. In unconscious or aphasic patients, consent was <0.10 in bivariate analyses. Sensitivity analyses for factors associated obtained from next of kin. with early PSS and late PSS were also performed. All statistical tests of hypotheses were two-sided. Statistical analyses were performed with 2.2. Stroke phenotyping Stata MP version 14. Stroke diagnosis and phenotyping were based on clinical evaluation 4. Results and brain neuroimaging (CT or MRI), ECG, transthoracic echocardiog- raphy, and carotid Doppler ultrasound performed according to stan- 4.1. Characteristics of participants with post-stroke seizures dardized protocols (SOP) at each site. Presumed etiological sub-types of ischemic stroke were defined etiologically using the A-S-C-O-D classifi- We enrolled 3344 patients meeting study criteria of an acute stroke. cation into A: Atherosclerosis, S: Small-vessel occlusion, C: Cardiac pa- The frequency of post stroke seizures was 499 (14.9%; 95% CI: thology, O: Other causes and D: dissection [ 19] and intracerebral 13.7–16.2%). Among those with PSS, 382 (76.6%) had early PSS and hemorrhage was classified etiologically into Structural, Medication- 117 (23.4%) with late PSS. The mean age of those with post-stroke related, Amyloid angiopathy, Systemic/other disease, Hypertension seizures of 58.3 ± 15.3 years was significantly lower than 60.2 ± and Undetermined causes (SMASH-U) [20]. 14.1 years for those without seizures, p = 0.006. The characteristics of those with PSS are compared with those without PSS in Table 1. Those with PSS were less likely to earn monthly income > $100 (49.8% vs 2 F.S. Sarfo et al. J o u r n a l o f t h e N e u r o l o g i c a l S c i e n c e s 427 (2021) 117535 Table 1 Comparison of demographic and clinical characteristics of stroke cases with Post-stroke seizures versus those with no Post-stroke seizures. Variable All Stroke type Ischemic Stroke Intracerebral hemorrhage No Post-stroke Post-stroke P-value No Post-stroke Post-stroke P-value No Post-stroke Post-stroke P-value seizures seizures seizures seizures seizures seizures N = 2845 n = 499 N = 1797 n = 294 N = 738 n = 159 Country, Ghana, n (%) 974 (34.2) 161 (32.3) 0.391 582 (32.4) 86 (29.3) 0.285 371 (50.3) 72 (45.3) 0.254 Gender, Male, n (%) 1578 (55.5) 292 (58.5) 0.208 939 (52.3) 159 (54.1) 0.567 463 (62.7) 109 (68.6) 0.166 Age, mean ± SD 60.2 ± 14.1 58.3 ± 15.3 0.006 62.4 ± 13.8 61.3 ± 15.0 0.195 54.7 ± 13.2 51.7 ± 13.7 0.010 <30 37 (1.3) 18 (3.6) 0.001 20 (1.1) 14 (4.8) <0.001 10 (1.4) 4 (2.5) 0.017 30–49 584 (20.6) 117 (23.5) 285 (15.9) 39 (13.3) 241 (32.7) 71 (44.7) 50–69 1433 (50.4) 232 (46.7) 894 (49.8) 143 (48.6) 383 (51.9) 67 (42.1) ≥ 70 787 (27.7) 130 (26.2) 596 (33.2) 98 (33.3) 104 (14.1) 17 (10.7) Domicile Rural, n (%) 272 (9.6) 40 (8.1) 0.508 176 (9.8) 23 (7.8) 0.288 63 (8.6) 14 (8.9) 0.832 Semi-urban, n (%) 828 (29.2) 152 (30.6) 516 (28.8) 96 (32.7) 208 (28.3) 41 (26.0) Urban, n (%) 1735 (61.2) 305 (61.4) 1101 (61.4) 175 (59.5) 463 (63.1) 103 (65.2) Monthly Income >$100, n (%) 1551 (54.9) 247 (49.8) 0.036 989 (55.5) 150 (51.0) 0.153 418 (56.8) 81 (51.9) 0.266 Education, (some) n (%) 2310 (81.4) 405 (81.2) 0.890 1417 (79.1) 226 (76.9) 0.382 640 (86.8) 138 (86.8) 0.988 Hypertension, n (%) 2739 (96.3) 462 (92.6) <0.001 1721 (95.8) 264 (89.8) <0.001 725 (98.4) 155 (97.5) 0.443 Dyslipidemia, n (%) 2402 (84.4) 402 (81.1) 0.059 1570 (87.4) 253 (86.4) 0.628 588 (79.7) 117 (73.6) 0.089 Diabetes, n (%) 1068 (37.6) 205 (41.2) 0.126 740 (41.2) 133 (45.2) 0.191 209 (28.4) 46 (28.9) 0.885 Cardiac Disease, n (%) 352 (12.4) 43 (8.7) 0.017 266 (14.8) 32 (10.9) 0.076 54 (7.3) 9 (5.7) 0.456 HDL-Cholesterol, mg/dl, 47.7 ± 19.3 48.7 ± 20.3 0.316 46.4 ± 18.4 45.4 ± 17.9 0.457 52.1 ± 20.5 55.3 ± 22.9 0.100 mean ± SD HDL-Cholesterol ≤18.54 mg/ 78 (2.7) 12 (2.4) 0.668 48 (2.7) 10 (3.4) 0.480 17 (2.3) 2 (1.3) 0.406 dl, n (%) LDL-Cholesterol, mg/dl, mean 121.8 ± 51.1 116.1 ± 53.3 0.039 121.6 ± 50.8 116.1 ± 53.5 0.131 127.8 ± 51.5 120.9 ± 53.9 0.160 ± SD LDL-Cholesterol ≥61.2 mg/dl, 2203 (90.6) 338 (86.2) 0.007 1393 (90.4) 202 (87.1) 0.116 626 (93.3) 119 (88.8) 0.071 n (%) LDL/HDL ratio, mean ± SD 3.0 ± 1.9 2.7 ± 1.7 0.034 3.0 ± 1.9 2.9 ± 1.8 0.433 2.8 ± 2.0 2.5 ± 1.5 0.089 LDL/HDL ratio > 2.96, n (%) 925 (38.4) 130 (33.4) 0.063 607 (39.7) 88 (38.3) 0.683 235 (35.4) 36 (26.9) 0.057 LDL/HDL ratio by thirds: ≤ 2.00, n (%) 760 (31.5) 143 (36.8) 0.079 456 (29.8) 79 (34.4) 0.351 226 (34.0) 53 (39.6) 0.159 2.01–2.96, n (%) 726 (30.1) 116 (29.8) 466 (30.5) 63 (27.4) 203 (30.6) 45 (33.6) ≥ 2.97, n (%) 926 (38.4) 130 (33.4) 608 (39.7) 88 (38.3) 235 (35.4) 36 (26.9) Total Cholesterol, mmol/l, 192.0 ± 57.7 186.3 ± 58.4 0.065 191.3 ± 57.9 183.3 ± 58.9 0.051 200.0 ± 57.5 196.3 ± 57.5 0.496 mean ± SD Total Cholesterol ≥93.6 mg/ 2415 (98.0) 383 (96.0) 0.014 1536 (98.1) 224 (95.7) 0.022 666 (98.2) 133 (97.1) 0.376 dl, n (%) Triglyceride, mg/dl, mean ± 126.2 ± 83.9 120.9 ± 71.6 0.231 130.1 ± 85.5 122.3 ± 67.2 0.185 122.4 ± 84.2 122.2 ± 81.2 0.978 SD Triglyceride ≥30.6 mg/dl, n 2447 (99.5) 395 (99.5) 0.950 1559 (99.7) 234 (100.0) 0.439 668 (99.0) 134 (99.3) 0.751 (%) Waist-to-hip Ratio, mean ± SD 0.9 ± 0.1 0.9 ± 0.1 0.217 0.9 ± 0.1 0.9 ± 0.1 0.105 0.9 ± 0.1 0.9 ± 0.1 0.473 Waist-to-hip Ratio raised, n 2207 (83.0) 387 (82.7) 0.883 1420 (84.5) 236 (86.1) 0.493 552 (79.9) 120 (79.0) 0.795 (%) Waist-to-hip Ratio by thirds: ≤ 0.90, n (%) 703 (26.4) 118 (25.2) 0.294 426 (25.3) 61 (22.3) 0.055 202 (29.2) 42 (27.6) 0.772 0.91–0.96, n (%) 941 (35.4) 183 (39.1) 591 (35.2) 117 (42.7) 251 (36.3) 53 (34.9) ≥0.97+, n (%) 1017 (38.2) 167 (35.7) 664 (39.5) 96 (35.0) 238 (34.4) 57 (37.5) WHR**, Lowest vs highest 1017 (59.1) 167 (58.6) 0.866 664 (60.9) 96 (61.2) 0.956 238 (54.1) 57 (57.6) 0.529 thirds, n (%) WHR**, 1st vs 2nd + 3rd 1958 (73.6) 350 (74.8) 0.585 1255 (74.7) 213 (77.7) 0.274 489 (70.8) 110 (72.4) 0.693 thirds, n (%) BMI*** (kg/m2), mean ± SD 26.8 ± 5.3 26.7 ± 5.6 0.867 26.9 ± 5.4 26.5 ± 5.2 0.230 26.3 ± 4.9 27.2 ± 6.3 0.066 BMI*** > 30 kg/m2, n (%) 491 (21.3) 87 (21.9) 0.812 332 (22.5) 55 (23.0) 0.867 103 (17.6) 22 (18.2) 0.880 Physical Activity (some 2658 (95.1) 460 (95.0) 0.957 1680 (95.2) 271 (94.4) 0.553 699 (95.6) 149 (96.1) 0.777 activity), n (%) Tobacco use in past 12 102 (3.7) 20 (4.1) 0.620 54 (3.1) 7 (2.5) 0.581 41 (5.6) 12 (7.6) 0.331 months, n (%) Tobacco (any use), n (%) 272 (9.6) 51 (10.4) 0.616 170 (9.5) 33 (11.4) 0.325 81 (11.0) 16 (10.2) 0.758 Alcohol (current user), n (%) 478 (16.9) 91 (18.4) 0.397 262 (14.6) 34 (11.7) 0.181 182 (24.8) 49 (31.0) 0.104 Alcohol (any use), n (%) 926 (32.7) 172 (34.8) 0.350 545 (30.5) 88 (30.2) 0.943 311 (42.3) 70 (44.3) 0.646 Alcohol use categories: Never Use, n (%) 1908 (76.5) 322 (74.9) 0.129 1245 (78.4) 203 (81.2) 0.442 424 (68.2) 88 (62.9) 0.020 Ever Low Use, n (%) 521 (20.9) 89 (20.7) 308 (19.4) 44 (17.6) 172 (27.7) 38 (27.1) Ever High Use, n (%) 66 (2.7) 19 (4.4) 35 (2.2) 3 (1.2) 26 (4.2) 14 (10.0) Stress, n (%) 525 (19.9) 101 (22.7) 0.178 321 (19.3) 59 (22.4) 0.234 156 (22.8) 30 (20.8) 0.606 Depression, n (%) 203 (7.3) 40 (8.3) 0.479 130 (7.5) 25 (8.7) 0.451 60 (8.3) 11 (7.1) 0.595 Family history of CVD, n (%) 1096 (38.5) 169 (33.9) 0.048 689 (38.3) 102 (34.7) 0.232 328 (44.4) 57 (35.9) 0.047 Adding salt at table, n (%) 188 (6.8) 45 (9.5) 0.035 109 (6.3) 20 (7.3) 0.529 65 (9.0) 19 (12.4) 0.193 Adding salt at table categories: Never/rarely, n (%) 1913 (69.4) 334 (70.8) 0.029 1230 (70.8) 191 (69.5) 0.803 481 (66.6) 116 (75.8) 0.002 Occasionally, n (%) 655 (23.8) 93 (19.7) 399 (23.0) 64 (23.3) 176 (24.4) 18 (11.8) (continued on next page) 3 F.S. Sarfo et al. J o u r n a l o f t h e N e u r o l o g i c a l S c i e n c e s 427 (2021) 117535 Table 1 (continued ) Variable All Stroke type Ischemic Stroke Intracerebral hemorrhage No Post-stroke Post-stroke P-value No Post-stroke Post-stroke P-value No Post-stroke Post-stroke P-value seizures seizures seizures seizures seizures seizures N = 2845 n = 499 N = 1797 n = 294 N = 738 n = 159 Very often, n (%) 188 (6.8) 45 (9.5) 109 (6.3) 20 (7.3) 65 (9.0) 19 (12.4) Green vegetable consumption, 1940 (73.4) 329 (74.1) 0.758 1222 (73.6) 193 (75.4) 0.547 504 (72.3) 101 (69.2) 0.445 n (%) Whole grains consumption, n 2225 (83.5) 375 (84.1) 0.743 1399 (83.8) 215 (83.3) 0.859 588 (83.8) 125 (86.2) 0.462 (%) Legumes consumption, n (%) 1791 (67.7) 299 (68.0) 0.912 1117 (67.5) 170 (66.9) 0.848 477 (68.1) 99 (69.2) 0.781 Fruit consumption, n (%) 2259 (85.3) 374 (83.9) 0.447 1410 (84.9) 217 (83.8) 0.630 587 (84.1) 118 (81.9) 0.524 Sugar consumption or 771 (29.6) 145 (33.3) 0.122 474 (29.0) 77 (30.7) 0.580 216 (31.2) 54 (37.8) 0.125 otherwise, (%) Meat consumption or 2285 (85.6) 381 (85.0) 0.752 1409 (84.4) 214 (82.6) 0.474 603 (85.4) 128 (87.7) 0.476 otherwise, (%) Fish consumption or 2474 (93.1) 393 (88.5) 0.001 1532 (92.1) 225 (87.6) 0.014 663 (94.2) 131 (90.3) 0.088 otherwise, (%) Serum sodium 137.6 ± 9.1 138.2 ± 11.6 0.278 137.5 ± 8.6 137.8 ± 13.4 0.642 137.9 ± 10.3 139.2 ± 9.2 0.241 White blood cell count, mean 16.9 ± 2.9 21. 5 ± 3.6 0.017 17.1 ± 3.0 19.3 ± 3.0 0.420 16.1 ± 2.7 23.8 ± 4.3 0.015 ± SD NIHSS score, mean ± SD 12.7 ± 8.6 15.2 ± 9.3 <0.001 12.1 ± 8.3 14.3 ± 9.2 <0.001 14.3 ± 8.8 16.7 ± 9.3 0.008 Location of lesions Lobar 149 (20.2) 57 (35.9) <0.001 Non-lobar 589 (79.8) 102 (64.2) Volume of lesions ≤10cm3 1414 (64.5) 203 (50.6) <0.001 1109 (74.0) 158 (62.7) 298 (43.6) 44 (29.9) 10.1-30 cm3 452 (20.6) 109 (27.2) 209 (14.0) 53 (21.0) 0.001 240 (35.1) 56 (38.1) 0.003 > 30cm3 326 (14.9) 89 (22.2) 180 (12.0) 41 (16.3) 146 (21.4) 47 (32.0) Duration of admission in days, 15.9 ± 24.7 17.4 ± 28.6 0.716 15.9 ± 24.9 19.1 ± 35.4 0.819 15.0 ± 22.8 14.0 ± 13.3 0.529 mean ± SD NIHSS – National Institute of Health Stroke Score. 54.9%), to be hypertensive (92.6% vs 96.3%), to have cardiac disease 4.3. Factors associated with occurrence of post-stroke seizures (8.7% vs 12.4%) and to consume fish regularly (88.5% vs 93.1). How- ever, the white cell count and National Institute of Health Stroke Score In Table 2, we show 12 potential factors associated with occurrence were significantly higher among those with post-stroke seizures than of post-stroke seizures overall in bivariate analysis. Upon adjusting for those without. Also 22.2% with post-stroke seizure had a lesion volume potential confounders, the adjusted odds ratio (95% CI) of three factors of >30cm3 compared with 14.9% among those without seizures. The which remained independently associated with post-stroke seizures mean duration of admission days for those with PSS vs no PSS was 17.4 were: age < 50 years 1.59 (1.08–2.33), NIHSS score at presentation 1.29 ± 28.6 vs 15.9 ± 24.7, p = 0.72. (Table 1). (1.16–1.42) for each 5 units rise and white cell count 1.07 (1.01–1.13) for each 10^3 mm3 rise. Two factors were independently associated with 4.2. Characteristics of participants with post-stroke seizures by stroke type post-stroke seizures among those with ischemic stroke were NIHSS score and lesion volume of >30cm3 (see Table 3). Among patients with ICH, There were 294 (14.1%) post-stroke seizures found among 2091 post-stroke seizures were associated with alcohol use 5.91 (2.11–16.55) patients with ischemic stroke and 159 (17.7%) among 897 with intra- and lobar bleeds 2.22 (1.03–4.82) (Table 4). Factors associated with cerebral hemorrhage, p = 0.01. Data on stroke type information were early PSS were stroke severity and white blood count (Table S1 in missing for 356 participants. Among 1864 patients with Oxfordshire supplementary information). Late-onset PSS was associated with stroke Community Stroke Project (OCSP) classification data, 778 (41.7%) had severity as shown in supplementary information Table S2. LACI, 635 (34.1%) had PACI, 268 (14.4%) had TACI and 183 (9.8%) had POCI. Proportions with post-stroke seizures by OCSP classification of 5. Discussion ischemic strokes were 18.7% for those with TACI, 16.9% for POCI, 14.1% for LACI and 11.8% among those with PACI (p = 0.039), Fig. 1A. In this large multi-center study across 15 hospitals in Ghana and Among 1386 ischemic stroke patients with data on etiology, 54.5% Nigeria, we found the frequency of post-stroke seizures to be 14.9% had small vessel occlusion (SVO), 26.3% had large artery atherosclerotic (95% CI of 13.7–16.2%). The frequency of post-stroke seizures was disease (LAA), 16.1% had cardio-embolic stroke (CE) and 3.1% had significantly higher among those with spontaneous intracerebral hem- other causes. In decreasing order, 17.3% with LAA, 14.4% with SVO, orrhage at 17.7% compared with 14.1% among those with ischemic 13.4% with CE and 7.0% with other causes, respectively had post-stroke strokes. The prevalence of post-stroke seizures in the present study was seizures, p = 0.23, by chi-squared test for trend (Fig. 1B). Among 801 quite high and is comparable with a figure of 17.9% (14.6–21.8%) found patients with ICH, 90.8% had hypertensive ICH, 4.0% had structural in an Indian study [25]. Otherwise, most of the previous studies have lesions, 2.7% had undetermined causes, 1.5% had cerebral amyloid reported much lower prevalence of post-stroke seizures ranging from associated bleeds, and 0.5% each had either medication-related or sys- 3.0% in Taiwan [26], 3.9% to 6.3% from three Italian studies [27–29], temic disease-related bleeds. The frequencies of post-stroke seizure 4.2% in Denmark [30], 4.1% from the US [31], 8.9% from an interna- occurrence by etiology of ICH in decreasing order were 50.0% for those tional collaborative study involving tertiary medical centers in Canada, with medication-related ICH, 41.7% in amyloid-related bleeds, 18.2% in Australia, Israel and Italy [32] and one study from Egypt reported 9.3% bleeds of undetermined causes, 17.5% in hypertensive-related bleeds [15]. While differences in study designs and cohort characteristics may and 9.4% with structural lesions such as aneurysms and AV malforma- underlie these differences in prevalence observed across studies, a key tions, p = 0.09 (Fig. 2). Data on ICH stroke subtype information were reason could be differences in the time window for defining early post- missing for 96 participants. stroke seizures which has varied between 1 and 30 days in various 4 F.S. Sarfo et al. J o u r n a l o f t h e N e u r o l o g i c a l S c i e n c e s 427 (2021) 117535 Fig. 1. Frequency distribution of post-stroke seizures by etiology of Intracere- bral hemorrhage in West Africa. studies. In our study, the observation window for identifying PSS was within the period of stroke onset until discharge from hospital or death. The average duration of hospitalization for those with PSS of 17 days was not significantly different from a mean of 16 days for those without PSS. Admittedly, there is potential for the differential duration of hos- pitalization to influence our ability to identify PSS. For instance, those with mild stroke might have be discharged earlier than those with severe stroke and therefore less likely to have seizures identified. However, we did not find a significant association between duration of hospitalization and risk of PSS. Intracerebral hemorrhage, cerebral infarction with hemorrhagic transformation, stroke severity and alcoholism are factors associated with early post-stroke seizures from meta-analytic data [33]. Late-onset seizures were associated with cortical involvement and stroke severity [33]. In our study, stroke severity was independently associated with post-stroke seizure with each 5 units rise in the NIHSS score corre- sponding to a 29% higher odd of seizures (95% CI: 16–42%). Acute stroke patients <50 years old had a 59% higher odds of a post-stroke seizures than those who were older. This higher proclivity for post- stroke seizures occurrence among young west Africans may be explained by the relative preponderance of intracerebral hemorrhage in this age group [9,34]. In sensitivity analyses, the odds ratio of post- stroke seizure among those <50 years with ICH was 1.82 (95% CI: 0.94–3.51) compared with 1.12 (0.69–1.82) among those with ischemic stroke. Leukocytosis was also independently associated in a graded Fig. 2. Frequency distribution of post-stroke seizures by etiology of Ischemic manner with post-stroke seizures, with 7% higher odds for each 10^3 rise Stroke in West Africa. (1A) by Oxfordshire Community Stroke Project classifi- in white blood cell count at presentation. It is uncertain whether cation; (1B) by ASCO classification. leukocytosis is simply a marker of stroke severity [35], is a result of post- 5 F.S. Sarfo et al. J o u r n a l o f t h e N e u r o l o g i c a l S c i e n c e s 427 (2021) 117535 Table 2 Table 4 Multivariable logistic regression analysis for factors associated Post-stroke Multivariable logistic regression analysis for factors associated Post-stroke sei- seizures. zures among patients with intracerebral hemorrhage. Unadjusted P-value Adjusted OR P-value Unadjusted OR P-value Adjusted OR P- OR (95% CI) (95% CI) (95% CI) (95% CI) value Age < 50 years 1.33 0.009 1.46 0.060 Age < 50 1.73 0.002 1.82 (0.93–3.55) 0.080 (1.07–1.66) (0.98–2.16 (1.22–2.45) Income <100$ 1.23 0.036 1.06 0.735 Dyslipidemia 0.71 0.091 0.99 (0.41–2.36) 0.979 (1.01–1.48) (0.75–1.51) (0.48–1.06) Hypertension 0.48 <0.001 0.66 0.339 BMI (continuous) 1.03 0.069 0.96 (0.88–1.03) 0.266 (0.32–0.71) (0.28–1.55) (1.00–1.07) Dyslipidemia 0.79 0.059 0.82 0.455 Alcohol use (Ever 2.59 0.007 5.79 0.001 (0.62–1.01) (0.49–1.38) High use) (1.30–5.17) (1.99–16.89) Cardiac disease 0.67 0.018 0.95 0.870 White blood count 1.07 0.018 1.02 (0.89–1.17) 0.770 (0.48–0.93) (0.52–1.74) (1.01–1.14) Family history of CVD 0.82 0.048 0.95 0.771 NIHSS each 5 units 1.16 0.009 1.19 (0.98–1.45) 0.066 (0.67–0.99) (0.67–135) higher (1.04–1.30) Salt (very often) 1.44 0.036 1.51 0.118 Lobar bleed 2.21 <0.001 2.23 (1.02–4.87) 0.044 (1.02–2.02) (0.90–2.52) (1.52–3.20) Fish consumption 0.57 0.001 0.73 0.322 Lesion volume (0.41–0.79) (0.38–1.37) <10.0 cm3 1.00 1.00 Hemorrhagic stroke vs 1.32 0.011 1.14 0.521 10.1 - 30cm3 1.58 0.037 1.34 (0.61–2.96) 0.464 ischemic stroke as (1.07–1.62) (0.76–1.70) (1.03–2.43) referent >30.0 cm3 2.18 0.001 1.70 (0.69–4.22) 0.251 NIHSS score as a 1.16 <0.001 1.29 <0.001 (1.38–3.44) continuous variable (1.09–1.24) (1.16–1.42) SMASH-U per each 5 units Amyloid 12.08 0.006 7.43 0.279 higher angiopathy (2.05–71.11) (0.20–280.28) White blood cell count 1.04 0.022 1.06 0.039 Others (referent) 1.00 continuous variable (1.01–1.08) (1.01–1.12) Duration of 0.99 0.632 1.00 (0.99–1.01) 0.676 Volume of lesions admission for (0.98–1.01) ≤10cm3 1.00 1.00 stroke in days 10.1-30 cm3 1.68 <0.001 1.33 0.192 (1.30–2.17) (0.87–2.04) > 30cm3 1.90 <0.001 1.03 0.924 Indeed, cerebral inflammation is a prime etiological factor for icto- (1.44–2.51) (0.61–1.72) genesis and epileptogenesis via alterations in ionic channel sensitivity, Duration of admission 1.00 0.259 1.00 0.309 for stroke in days (0.98–1.01 (0.99–1.01) neurotransmitter uptake and glia-associated modulation of the extra- cellular electrical milieu as elegantly reviewed by Shimada et al. [36] There were also differences in risk factor profile for post-stroke seizure by stroke type. We observed among patients with intracerebral Table 3 Multivariable logistic regression analysis for factors associated Post-stroke sei- hemorrhage that alcohol use and lobar bleeds were significantly asso- zures among patients with ischemic strokes. ciated with post-stroke seizures. However, while binge drinking may predispose to ICH with attendant PSS, alcohol withdrawal seizures after Unadjusted OR P-value Adjusted OR P- ICH in alcoholic may conversely predispose to PSS. Thus an assessment (95% CI) (95% CI) value of serum alcohol levels or specific questions to patients/relatives may Age < 50 1.07 0.662 – help distinguish between these two possible causes of PSS among (0.78–1.48) Hypertension 0.39 <0.001 0.60 0.230 alcohol users with ICH for clinical management purposes. It is intriguing (0.25–0.60) (0.26–1.38) to note that, unlike the situation in high-income countries, we found a Cardiac disease 0.70 0.078 0.84 0.536 low frequency of medication-related ICH, perhaps as a consequence of (0.48–1.04) (0.47–1.47) low utilization rates of these pharmacological agents for primary or Total cholesterol 1.00 0.051 1.00 0.385 secondary CVD risk reduction in our settings [37,38]. Among ischemic (continuous) (0.99–1.00) (0.99–1.00) WHR (tertiles) strokes however, stroke severity and larger infarct volumes presented a 0.91–0.96, n (%) 1.38 0.057 1.59 0.073 higher risk for occurrence of seizures. This observation aligns with the (0.99–1.93) (0.96–2.66) higher frequency of seizures among those with total anterior circulation ≥0.97+, n (%) 1.01 0.956 1.24 0.428 infarcts using the OCSP classification observed in our study and found by (0.72–1.42) (0.73–2.12) Fish consumption 0.60 0.015 0.65 0.198 others [39]. Furthermore, those with large artery atherosclerotic disease (0.40–0.91) (0.33–1.25) with its tendency to cause larger territorial infarcts had the highest NIHSS, each 5 units 1.15 0.001 1.21 0.001 frequency of seizures among ischemic stroke clinical sub-types. A sur- higher (1.06–1.25) (1.08–1.37) prisingly high proportion (≈14%) of ischemic stroke subjects with small Lesion volume 3 vessel occlusive disease had post-stroke seizures. This observation re-<10.0 cm 1.00 1.00 10.1 - 30cm3 1.78 0.001 1.90 0.010 quires further studies because a significant majority of lacunar infarcts (1.26–2.51) (1.16–3.11) occur in the basal ganglia, an area with lower predilection for seizure >30 cm3 1.60 0.015 1.05 0.872 generation. Further on-going analysis may throw more light on the (1.10–2.33) (0.57–1.96) unique determinants of seizures occurrence in the context of basal Duration of admission 1.00 0.082 1.00 0.266 for stroke in days (0.99–1.01) (0.99–1.01) ganglionic bleeds and lacunar strokes in the African context. convulsive leukocytosis or reflective of infections occurring after PSS 5.1. Implications such as aspiration pneumonitis. It is tempting to speculate that disrup- tion of the blood brain barrier and invasion of leukocytes into the site of Given that approximately 15% of patients presenting with acute cerebral injury after a stroke may incite abnormal neuronal firing. stroke for hospitalization have concomitant seizures, there is the need 6 F.S. Sarfo et al. J o u r n a l o f t h e N e u r o l o g i c a l S c i e n c e s 427 (2021) 117535 for a heightened index for suspicion by clinicians for its screening and [3] G. Wang, H. Jia, C. Chen, S. Lang, X. Liu, C. Xia, et al., Analysis of risk factors for identification to treatment. International guidelines from the US and first seizure after stroke in Chinese patients, Biomed. Res. Int. (2013) 702–871. [4] S.T. Herman, Early post-stroke seizures: is it time for prospective treatment trials? Europe do not recommend primary or secondary anti-epileptic drug Neurology 7 (2011) 776–778. (AED) treatment for prophylaxis for post-stroke acute symptomatic [5] R. Walker, D. Whiting, N. Unwin, F. Mugusi, M. Swai, Stroke incidence in rural and seizures although in clinical practice clinicians often do so. The use of urban Tanzania: a prospective, community-based study, Lancet Neurol. 9 (2010) 786–792. carbamazepine and phenytoin is rife but more recent studies support the [6] M.C. Ezejimofor, O.A. Uthman, O. Maduka, et al., Stroke survivors in Nigeria: a use of levetiracetam, lamotrigine and gabapentine for treatment of post- door-to-door prevalence survey from the Niger Delta region, J. Neurol. Sci. 372 stroke seizures [40–42]. A clearer resolution of the potential role of short (2017 January 15) 262–269. or longer-term use of AEDs in the management of post-stroke seizures [7] F.S. Sarfo, D.O. Awuah, C. Nkyi, J. Akassi, O.K. Opare-Sem, B. Ovbiagele, Recent patterns and predictors of neurological mortality among hospitalized patients in remains to be resolved. Prospective studies are also urgently required to Central Ghana, J. Neurol. Sci. 363 (2016) 217–224. assess the impact of post-stroke seizures on short- and long-term out- [8] R.W. Walker, A. Jusabani, E. Aris, et al., Post-stroke case fatality within an incident comes of strokes, in particular the subsequent risk of development of population in rural Tanzania, J. Neurol. Neurosurg. Psychiatry 82 (9) (2011) 1001–1005. post-stroke epilepsy in this region. A recent study from Ghana among [9] F.S. Sarfo, B. Ovbiagele, M. Gebregziabher, K. Wahab, R. Akinyemi, A. Akpalu, et 1101 stroke survivors reported a post-stroke epilepsy prevalence of al., Stroke among young West Africans: evidence from the SIREN (Stroke 11.4% with male sex, cortical infarcts, elevate blood pressure as pre- Investigative Research and Educational Network) large multisite case-control study, Stroke. 49 (5) (2018 May) 1116–1122. dictive factors while antihypertensive use was protective against post- [10] A. Akpalu, M. Gebregziabher, B. Ovbiagele, F. Sarfo, H. Iheonye, R. Akinyemi, et stroke epilepsy [43]. This finding perhaps highlights the need for pre- al., Differential impact of risk factors on stroke occurrence among men versus diction models for personalized care for post-stroke seizure women in West Africa, Stroke. 50 (4) (2019 Apr) 820–827. [11] F.S. Sarfo, M. Nichols, S. Qanungo, A. Teklehaimanot, A. Singh, N. Mensah, et al., management. Stroke-related stigma among West Africans: patterns and predictors, J. Neurol. Sci. 375 (2017) 270–274. [12] F.S. Sarfo, C. Jenkins, A. Singh, M. Owolabi, A. Ojagbemi, N. Adusei, et al., Post- 5.2. Limitations stroke depression in Ghana: characteristics and correlates, J. Neurol. Sci. 379 (2017) 261–265. This study is limited by non-confirmation of clinically reported sei- [13] F.S. Sarfo, M. Agbenorku, S. Adamu, V. Obese, P. Berchie, B. Ovbiagele, PINGS Study investigators. The dynamics of poststroke depression among Ghanaians, zures with electroencephalographic studies and detailed evaluation of J. Neurol. Sci. 405 (2019 Oct 15) 116410. potential causes of seizures in the acute stroke setting. The burden of [14] F.S. Sarfo, J. Akassi, S. Adamu, V. Obese, B. Ovbiagele, Burden and predictors of pre-stroke epilepsy risk among our study population was not ascertained poststroke cognitive impairment in a sample of Ghanaian stroke survivors, as was documentation of seizures types (focal vs generalized), number of J. Stroke Cerebrovasc. Dis. 26 (11) (2017) 2553–2562. [15] N. Shehta, R.M. Fahmi, B.M. Ramadan, E.M. Emad, A.F. Elsaid, Early post-stroke seizures (recurrent or single), and types of seizure treatments offered. seizures in a sample of Egyptian patients with first-ever stroke, Neurol. India 66 None of patients in this series had thrombolytic therapy which has been (2018) 1031–1035. shown to be associated with PSS at a frequency of 1 in every 15 ischemic [16] C. Napon, A. Dabilgou, J. Kyelem, J. Kabore, Post-stroke epilepsy in Burkina Faso (West Africa), J. Neurol. Sci. 368 (2016 Sep 15) 47–48. stroke patients receiving reperfusion treatment [44]. Varying pro- [17] I.F. Aiwansoba, O.W. Chukwuyem, Early post-acute stroke seizures: clinical profile portions of the study participants did not have data on stroke type or and outcome in a Nigerian stroke unit, Ann. Afr. Med. 13 (1) (2014 Jan-Mar) subtypes and had to be excluded from secondary data analysis. Finally, 11–15. [18] A. Akpalu, F.S. Sarfo, B. Ovbiagele, R. Akinyemi, M. Gebregziabher, R. Obiako, et causal associations between seizures and stroke cannot be drawn due to al., Phenotyping stroke in sub-Saharan Africa: Stroke Investigative Research and the cross-sectional design of the study. Education Network (SIREN) Phenomics protocol, Neuroepidemiology 45 (2015) 73–82. [19] P. Amarenco, J. Bogousslavsky, L.R. Caplan, G.A. Donnan, M.G. Hennerici, New 5.3. Conclusion approach to stroke subtyping: the A-S-C-O (phenotypic) classification of stroke, Cerebrovasc. Dis. 27 (2009) 502–508. [20] A. Meretoja, D. Strbian, J. Putaala, S. Curtze, E. Haapaniemi, S. Mustanoja, et al., The burden of post-stroke seizures among this sample of west Afri- SMASH-U: a proposal for etiologic classification of intracerebral hemorrhage, cans is substantial and may contribute to poor outcomes of stroke in this Stroke 43 (2012) 2592–2597. region. Further longitudinal studies are required to understand the [21] M. Holtkamp, E. Beghi, F. Benninger, R. Kalviainen, R. Rocamora, H. Christensen, impact on morbidity and mortality arising from post-stroke seizures in et al., European Stroke Organisation. European Stroke Organisation guidelines for the management of post-stroke seizures and epilepsy, Eur. Stroke J. 2 (2017) Africa. 103–115. [22] R.A. Fisher, C. Acevedo, A. Arzimanoglou, A. Bogacz, J.H. Cross, C.E. Elger, et al., ILAE official report: a practical clinical definition of epilepsy, Epilepsia 55 (2014) Funding 475–482. [23] M. O’Donnell, D. Xavier, C. Diener, R. Sacco, L. Lisheng, H. Zhang, et al., Rationale The SIREN (Stroke Investigative Research and Education Networks) and design of INTERSTROKE: a global case-control study of risk factors for stroke, and SIBS Genomics studies were funded by the National Institutes of Neuroepidemiology 35 (2010) 36–44. [24] M.O. Owolabi, F. Sarfo, R. Akinyemi, M. Gebregziabher, O. Akpa, A. Akpalu, et al., Health grant U54 HG007479 and R01NS107900 under the H3Africa Dominant modifiable risk factors for stroke in Ghana and Nigeria (SIREN): a case- initiative. control study, Lancet Glob. Health 6 (4) (2018) e436–e446. [25] R.P. Goswami, P.S. Karmakar, A. Ghosh, Early seizures in first-ever acute stroke patients in India: incidence, predictive factors and impact on early outcome, Eur. J. Declaration of Competing Interest Neurol. 19 (2012) 1361–1366. [26] H.C. Liao, S.H. Chen, C.D. Yang, Y.W. Chen, Clinical profile and outcomes of early seizures in Asian patients with acute intracerebral hemorrhage, J. Acute. Med. 9 All authors have no conflicts to declare. (4) (2019) 172–177. [27] A. Pezzini, M. Grassi, E. Del Zotto, A. Giossi, I. Volonghi, P. Costa, Complications of Appendix A. Supplementary data acute stroke and the occurrence of early seizures, Cerebrovasc. Dis. 35 (2013) 444–450. [28] A. Alberti, M. Paciaroni, V. Caso, M. Venti, F. Palmerini, G. Agnelli, Early seizures Supplementary data to this article can be found online at https://doi. in patients with acute stroke: frequency, predictive factors and effect on clinical org/10.1016/j.jns.2021.117535. outcome, Vasc. Health Risk Manag. 4 (2008) 715–720. [29] E. Beghi, R.D. Alessandro, S. Beretta, D. Consoli, V. Crespi, L. Delaj, Incidence and predictors of acute symptomatic seizures after stroke, Neurology. 77 (2011) References 1785–1793. [30] J. Reith, H.S. Jorgensen, H. Nakayama, H.O. Raaschou, T.S. Olsen, Seizures in [1] W.A. Hauser, J.F. Annegers, L.T. Kurland, Incidence of epilepsy and unprovoked acute stroke: predictors and prognostic significance. The Copenhagen stroke study, seizures in Rochester, Minnesota: 1935-1984, Epilepsia 34 (1993) 453–468. Stroke 28 (1997) 1585–1589. [2] I.E. Silverman, L. Restrepo, G.C. Mathews, Poste stroke seizures, Arch. Neurol. 59 [31] D. Labovitz, W.A. Hauser, R.L. Sacco, Prevalence and predictors of early seizures (2002) 195–201. and status epilepticus after first stroke, Neurology 57 (2001) 200–206. 7 F.S. Sarfo et al. J o u r n a l o f t h e N e u r o l o g i c a l S c i e n c e s 427 (2021) 117535 [32] C.F. Bladin, A.V. Alexandrov, A. Bellavance, N. Bornstein, B. Chamber, R. Cote, et [39] J. Burn, M. Dennis, J. Bamford, P. Sandercock, D. Wade, C. Warlow, Epileptic al., Seizures after stroke: a prospective multicenter study, Arch. Neurol. 57 (2000) seizures after a first stroke: the Oxfordshire Community Stroke Project, BMJ. 315 1617–1622. (7122) (1997) 1582–1587. [33] C. Zhang, X. Wang, Y. Wang, J.G. Zhang, W. Hu, M. Ge, et al., Risk factors for post- [40] L. Sykes, E. Wood, J. Kwan, Antiepileptic drugs for the primary and secondary stroke seizures: a systematic review and meta-analysis, Epilepsy Res. 108 (10) prevention of seizures after stroke, Cochrane Database Syst. Rev. 1 (2014), (2014) 1806–1816. CD005398. [34] F.S. Sarfo, B. Ovbiagele, M. Gebregziabher, O. Akpa, A. Akpalu, K. Wahab, et al., [41] S.P. Singh, R. Sankaraneni, A.R. Antony, Evidence-based guidelines for the Unraveling the risk factors for spontaneous intracerebral hemorrhage among West management of epilepsy, Neurol. India 65 (Suppl S1) (2017) 6–11. Africans, Neurology. 94 (10) (2020 Mar 10) e998–e1012. [42] T. Glauser, E. Ben-Menachem, B. Bourgeios, A. Cnaan, C. Guerriro, R. Kalviainen, [35] P. Welsh, M. Barber, P. Langhorne, A. Rumley, G.D. Lowe, D.J. Stott, Associations et al., ILAE Subcommission on AED guidelines. Updated ILAE evidence review of of inflammatory and haemostatic biomarkers with poor outcome in acute antiepileptic drug efficacy and effectiveness as initial monotherapy for epileptic ischaemic stroke, Cerebrovasc. Dis. 27 (3) (2009) 247–253. seizures and syndromes, Epilepsia 54 (2013) 551–563. [36] T. Shimada, T. Takemiya, H. Sugiura, K. Yamagata, Role of inflammatory [43] F.S. Sarfo, J. Akassi, V. Obese, S. Adamu, M. Agbenorku, B. Ovbiagele, Prevalence mediators in pathogenesis of epilepsy. Role of Inflammatory Mediators in the and predictors of post-stroke epilepsy among Ghanaian stroke survivors, J. Neurol. Pathogenesis of Epilepsy, Mediat. Inflamm. 2014 (2014), 901902, 8 pages. Sci. 418 (2020) 117138. [37] F.S. Sarfo, S. Adamu, V. Obese, M. Agbenorku, P.A. Opare-Addo, B. Ovbiagele, [44] A. Lekoubou, J. Fox, P. Ssentongo, Incidence and association of reperfusion Atherosclerotic event risk and risk reduction therapies among Ghanaian therapies with poststroke seizures. A systematic review and meta-analysis, Stroke. hemorrhagic stroke survivors, J. Neurol. Sci. 424 (2021) 117389. 51 (2020) 2715–2723. [38] F.S. Sarfo, B. Ovbiagele, J. Akassi, G. Kyem, Baseline prescription and one-year persistence of secondary prevention drugs after an index stroke in Central Ghana, eNeurologicalSci. 6 (2017) 68–73. 8