Study Design and Population:
This cohort study was conducted at a tertiary public teaching hospital, Muhimbili University of Health and Allied Sciences Academic Medical Center (MAMC), in Dar es Salaam, Tanzania. MAMC offers super-specialized medical care for all specialties and receives referrals from public and private hospitals from all over the country as well as walk in patients. We consecutively enrolled consenting participants who were admitted at MAMC with a clinical diagnosis of first ever stroke based on the World Health Organization (WHO) definition for stroke (11). Participants or their next of kin were required to provide written informed consent and had to be ≥18 years at the time of consent prior to enrollment. Study participants were prospectively enrolled between June 2018 to January 2019 and each participant was followed up for outcomes to 30 days from admission.
Data collection:
An interviewer based structured questionnaire was administered to all study participants or their caregivers if the participant was unable to communicate. The questionnaire captured sociodemographic information, mobile numbers, previous stroke risk factors such as history of hypertension, DM, smoking, alcohol consumption, cardiac disease, and HIV infection. Other information collected included: use of medications for hypertension, diabetes, HIV, illicit drugs and use of hormonal contraception for females. The date of onset of stroke symptoms and date of arrival at the hospital were also recorded. Cigarette smoking and alcohol consumption was categorized as ever smoked and taken alcohol in life or never smoked or taken alcohol respectively. Current smokers/current alcohol consumers were defined as cigarette smoking/alcohol consumption within the last 12 months respectively.
All participants had their waist and hip circumference measured using a tape measure and recorded in centimeters. Pulse was checked for rate and rhythm and blood pressure (BP) was measured using a standard digital BP machine, AD Medical Inc. Three BP readings were collected spaced 5 minutes apart, while the participant was at rest and an average BP was computed. Hypertension was defined as a systolic blood pressure (SBP) ≥140 mmHg or diastolic blood pressure (DBP) ≥90 mmHg. Examination also included measuring temperature, precordial and neck carotid auscultation. All examination findings were recorded in pre-specified case report forms.
We aseptically collected 15mls of venous blood from each study participant: 5mls were analyzed for random total cholesterol, triglycerides, low density and high density lipoproteins using BIO- SYSTEMS machine, 5 mls were analyzed for complete blood count using HEMOLYZER 3 PRO machine and 5mls were analyzed for sickling test. Sickling test was performed using sodium metabisulphite and slides were viewed using Olympus microscope.
Capillary fingertip blood samples were collected from each participant to check for random blood glucose (RBG) levels and HIV rapid testing using a glucometer GLUCOPLUSTM and SD Bioline respectively. Consent for HIV testing was obtained during the initial consent process prior to enrolment. A fasting blood glucose (FBG) sample was collected the following morning for participants with (RBG) levels of ≥11.1 mmol/l. DM diagnosis was defined as a RBG reading of ≥ 11.1 mmol/l or a FBG reading of ≥ 7 mmol/l. For participants who were HIV reactive to SD Bioline, were tested using Unigold Biotech.
Non-contrast brain computer tomography (NCCT) using GE Healthcare Optima or magnetic resonance imaging (MRI) with GE SIGNA CREATOR were performed on study participants at the MAMC radiology department and images interpreted by a trained radiologist.
Transthoracic echocardiography (ECHO) using GE Medical Systems was performed by a trained cardiologist and interpretation was based on European Society of Cardiology/American Society of Echocardiography guidelines (12). Left ventricular muscle mass was assessed using a four chamber view at the end of diastole. A septal thickness > 10mm and > 11mm was considered left ventricular hypertrophy (LVH) for females and males respectively. Left atrium (LA) size measurements were performed in m-mode at the end of systole, a diameter >3.8cm and >4.0cm for females and males respectively was regarded as LA enlargement. Mitral stenosis (MS) was defined as mitral valve area of ≤ 1.5cm2 in short axis view and mean pressure gradient of ≥ 5mmHg using continuous Wave Doppler. A 12-lead electrocardiogram (ECG) using Bionet machine was performed by the principle investigator on the study participants to look for evidence of atrial fibrillation.
Stroke severity was assessed using the National Institute of Health Stroke Scale (NIHSS) (11). A score of 1 – 4 was defined as minor stroke, 5 – 15 moderate stroke, 15 – 20 moderately severe and 21 – 42 severe stroke. Stroke outcomes were categorized using the Modified Rankin Scale (MRS) (11) at 24 hours, 72 hours, 7 days, 14 days and at 30 days from admission, with scores ranging from 0 (no symptoms) to 6 (death). Death of date was recorded at each point and time to event was computed using date difference between the date of last contact (date of death or date of last follow up) and date of admission.
Data analysis:
Data were transferred from the questionnaires in full and entered into SPSS version 20.0 for analysis. The proportions of stroke by age were calculated with 95% confidence intervals. Continuous variables were summarized and presented as means and standard deviation (SD) or medians with Interquartile Range (IQR). Stroke risk factors by age were summarized as proportions and comparisons were made using Pearson’s Chi square test or Fisher’s exact test. Kaplan Meier analysis was applied to compare survival probabilities by age groups from date of admission to 30 days. The associations between various patient characteristics with fatality were examined using the Cox-proportional hazards model. Hazard ratios (HR), 95% confidence intervals (CI) and corresponding p values were obtained from the models adjusting for potential confounders. Variables with p value <0.2 in the univariate analysis were included in the multivariate analysis model and significance level was set as a p value of <0.05.