2.1 Study design
An institution based prospective, cross-sectional study was conducted at Tikur Anbessa Specialized Hospital (TASH) between July and December, 2016.
2.2 Study Setting
TASH is located in Addis Ababa, the capital of Ethiopia. It is the largest and the first tertiary level referral university teaching hospital with an estimated bed capacity of over 700. It has an emergency unit, an intensive care unit, a neurology ward, a clinical laboratory, and a radiology department with a Computerize Tomography (CT) and Magnetic Resonance Imaging (MRI) services. There is an integrated clinical service composed of highly trained neuro-radiologists, neurologists and emergency care physicians who serve for the entire nation.
2.3 Study population
The study population was all patients presenting to the emergency department with neurologic deficits suggestive of acute stroke and who fulfilled the inclusion criteria of age greater than 14 years, presentation within 72 hours of symptom onset, and confirmation of stroke diagnosis using emergency brain CT scan. A non-fasting capillary blood sample was used to estimate the admission blood glucose (BG) using a calibrated glucometer. Exclusion criteria were: age less than 14 years, patients with clinical course consistent with transient ischemic attack (TIA), trauma related cerebral hemorrhage, subarachnoid hemorrhage, prior neurologic deficit, prior stroke, unknown duration from symptom onset and who declined to give consent.
2.4 Sample size and sampling technique
The sample size was calculated using the single population proportion formula n=Z2*pq/d2 = 164 with estimated population proportion of 30%[10], α= 5% and d= 5%. During the study period, all first-ever acute stroke patients who met the inclusion criteria were enrolled consecutively.
2.5 Baseline data collection
The socio-demographic and clinical data were collected using a structured questionnaire that was pre-tested in 8 acute stroke patients having weakness on the non-dominant side and we got a complete response. We attempted to mitigate selection bias by randomizing participant selection using the same inclusion criteria detailed above. The National Institute of Health Stroke Scale (NIHSS) was used to measure the stroke severity. NIHSS is a 15-item neurologic examination stroke scale that sums the scores from individual elements to provide an overall stroke impairment score. It is a well-validated and commonly used assessment tool to evaluate the baseline neurologic deficit in acute stroke patients. The NIHSS score was dichotomized into: mild stroke (NIHSS score ≤ 14) and severe stroke (NIHSS score >14)[11].
In all patients, BG was documented and hyperglycemia was defined as BG level ≥140 mg/dl and normoglycemia as BG <140mg/dl[12]. DM was diagnosed if the patient had a pre-existing diagnosis of DM or was taking anti-diabetics drugs before stroke onset. If the BG was >200 mg/dl and a repeat Fasting blood glucose (FBG) test showed ≥126 mg/dl, then patients were classified as newly diagnosed DM based on ADA diagnostic criteria[13].
2.6 Thirty-day outcome
The modified Rankin Scale (mRS) was used for the assessment of functional outcome at 30-day after the index stroke onset, which may coincide with the time of hospital discharge and for those who were discharged earlier, we followed them by telephone using standardized interview protocol. The mRS is six-point ordinal scale ranging from 0 (no symptoms) to 6 (death) measuring the degree of disability or dependence in everyday life, including instrumental and basic activities of daily living (ADL). The functional mRS scoring was subcategorized into; good outcome (independence in ADL, mRS score (0-2), and poor outcomes (dependence in ADL, mRS score 3-5). A mRS score of 6 signifies a patient death (30-day mortality)[13].
2.7 Statistical analysis
Data were analyzed using the Statistical Package for Social Sciences version 20.0 software package (SPSS Inc, Chicago, IL, USA). Descriptive statistics were summarized using the frequency and proportion table for categorical data and mean (standard deviation) or median (interquartile range) for continuous data. The Pearson’s Chi square test were used to compare noncontinuous data. For dichotomous data, multiple logistic regression models were used to determine univariate and multivariate associations between 30-day outcomes and patient related variables including; demographic factors; prior history of DM; type of stroke based on neuroimaging findings; stroke severity by NIHSS and BG category. Adjusted odds ratios with 95% confidence interval (CI) were estimated. Level of statistical significance was set at p < 0.05.
2.8 Ethics approval and consent to participate
Ethics approval was obtained from the Institutional Review Board (IRB) of College of Health Science in Addis Ababa University. A written consent was obtained from all patients when clinically appropriate. For patients with depressed consciousness or severe aphasia, written consent was obtained from their care takers. For participants aged 14-18 years, an assent form was obtained from their parents or legal guardians. Confidentiality and anonymity of all patient data were kept throughout the study. The study was performed in accordance with the Declaration of Helsinki.