الفهرس 
INTRODUCTIONOnly 14 pages are availabe for public view
 |  | from | 113 |  |  |
| | | from | 113 | | |
Abstract
Stroke is the second leading cause of death worldwide as well as the leading cause of long-term disability. One third of new stroke patients die each year and less than half recover and regain their independence. It is important to identify risk factors and sources of stroke in order to take steps towards preventing stroke.
The aim of the current study was to assess the determinants (risk factors) of diminished CBF in patients with ischemic motor stroke, to develop a model for estimation of the CBF by non- invasive procedure in patients with ischemic motor stroke, to validate the model for prediction of CBF and to assess the value of the model in predicting the prognosis of motor function after stroke.
To fulfill this aim, a retrospective cohort study was conducted on a sample of 200 cases with ischemic stroke attending the Emergency Department, Alexandria Main University Hospital during the period from 2015 to 2017. from the total sample of 200 cases, 160 cases were chosen by simple random sample and were used to develop the model while the remaining 40 cases were used for validation of the model.
Data collection was done using a predesigned data collection sheet concerning; personal characteristics (age, sex and smoking habit), clinical characteristics (Mean BP, ICP, aspirin dose per day, initial motor power and improvement of motor function after stroke), and investigation characteristics (serum triglycerides, SSEP and CBF).
The collected data was fed to the computer using SPSS version 21 software for tabulation and analysis. Testing normal distribution, descriptive statistics were calculated which include minimum, maximum, mean, standard deviation, univariate statistical analysis for comparison of quantitative data and detecting the relation between quantitative variables were done using the suitable statistical analysis and multiple regression analysis was adopted using two models (main model and model with interaction) to detect predictors of CBF.
The results of the present study revealed the following:
Descriptive statistics:
The development sample included 160 cases with ischemic stroke. It was revealed that the mean age was 63.04± 8.911years, 68.8% were male and 22.5% were heavy smokers.
Concerning clinical characteristics, 62.5 % of cases had Mean BP >110 mm Hg, 44.4 % had ICP >20 cm H2O, 31.9 % did not take aspirin , 11.25 % showed absent initial motor power and 18.7 % did not show improvement of motor function after stroke.
Regarding investigation characteristics, 45.6% of cases had serum triglycerides >200 mg /liter and almost half of cases (54.4%) had SSEP ≤ 3 microvolt.
The following variables were statistically significant in univariate analysis according to CBF:
Age, Mean BP, ICP, aspirin dose, serum triglycerides smoking, SSEP, low grade motor power and high grade motor power (p=< .05).
Multivariate regression analysis:
The developed model (Main model) is useful for prediction of CBF as 89.8% of the variation in the CBF is explained by predictors of CBF. The model showed good prediction accuracy as the mean absolute error was 1.2481 ± 0.96729 ml /100gm / minute and mean absolute percent error was 9.31%.
Mean BP, Aspirin and SSEP had a significant positive regression coefficient, indicating that patients with higher values had higher CBF, after controlling for the other variables in the model.
Age, ICP, serum triglycerides level and smoking had a significant negative coefficient indicating that after controlling for the other variables, patients with higher values were expected to have lower CBF.
Patients with low grade and high grade motor power were expected to have CBF more than patients with absent motor power at the time of presentation after controlling for other variables.
ROC curve showed that the model of CBF prediction was statistically significant excellent tool to predict improvement of motor function after stroke with a very high area under the curve (AUC= .997). The model had an excellent sensitivity of 100% and specificity 94.9%. Its positive predictive value was 98.7%. Its negative predictive value was 100%.
from the results of this study, it is recommended:
• National screening program for population with risk factors (hypertension, dyslipidemia)
• Health education at all places, all clinic mass media, primary health care units about risk factors, availability of services for early detection and proper management.
• Apply the model for all health subjects with multiple risk factors to predict CBF and prevent the occurrence of stroke.
• Use the model for patients with history of ischemic stroke to avoid recurrence of stroke.
• More studies are needed to investigate the impact of using the model on the prevention of occurrence or recurrence of ischemic stroke.
• Encourage studies that focus on the preventive treatment of ischemic stroke.