Prediction score for post-stroke cognitive impairment (PSCI) after acute ischemic stroke

Abstract

It was estimated that patients with ischemic stroke and post-stroke cognitive impairment (PSCI) have been increasing. In addition, this PSCI is often late diagnosed when it has already developed into post-stroke dementia. Only a few studies have developed a scoring system of predictor factors cognitive impairment (CI) for post-acute ischemic stroke in Indonesia. This study aimed to develop a scoring system of predictor factors of CI for post-stroke ischemic patients. The patients included were >18 years old diagnosed with acute ischemic stroke who underwent mini-mental state examination (MMSE) and clock drawing test (CDT) examination on day-30 at Bethesda Hospital Yogyakarta. It was retrospective cohort study design and samples were obtained from the stroke registry and medical records. Patients who had a history of CI and incomplete medical records were excluded. The results of MSSE and CDT at day-30 were the outcomes of this study. To evaluate the relationship between the independent variable and the dependent variable, chi-squared tests were perforemd followed by multivariate logistic regression analysis with Hosmer-Lemeshow tests with backward likelihood-ratio (LR) method and by assessing the final area under the curve (AUC) model. The final model was transformed into a scoring system to determine the value of probability prediction of PSCI, the optimal cut-off point, the sensitivity value and specificity value of the cognitive impairment scoring system at day-30 after acute ischemic stroke. A total of 140 subjects were included in the study with an average age of 62.8 years, 86 (61.4%) males and 54 (38.6%) females. Ninety-one subjects (65%) experienced post-stroke CI. The multivariate analysis showed age >70 years, education level ≤6 years, modified ranking score (mRS) >3 at diagnosis, Barthel index score ≤4 at diagnosis, the number of multiple lesions and the location of lesion in the cortex were independent predictor factors affecting CI 30 days after acute ischemic stroke. The developed predictor score obtained AUC discrimination value of 82.6% (95%CI:0.757-0.896) and calibration value of p>0.366. The scoring system had a value range of 0-7, and with a cut-off ≥1, it had a sensitivity value of 86.8% and a specificity value of 59.2%. It can be concluded that the predictor score has a good performance in predicting the occurrence of PSCI at day-30 after acute ischemic stroke.