The Turkish version of the cognitive assessment scale for stroke patients (CASP-TR): a reliability and validity

Abstract

ABSTRACT Background Stroke, a prevalent neurological condition globally, is associated with a high mortality and often affects cognitive function. The Cognitive Assessment Scale for Stroke Patients (CASP) was developed to quickly identify and measure primary cognitive deficits, focusing particularly on stroke survivors. Within the Turkish context, there is a limited availability of validated cognitive screening tests for use after stroke. Aims The study aimed to translate CASP into Turkish (CASP-TR) and validate it in stroke patients. Methods and procedures After the translation of the CASP items into culturally appropriate Turkish versions, a pilot testing of the pre-final version was administered to post-stroke patients and healthy controls by two raters. Following completion of the final version of CASP-TR, 40 stroke patients were consecutively enrolled in validation process. Three blinded raters conducted all assessments with intervals of 1- or 2 days between evaluation sessions. The accuracy of CASP-TR was assessed using sensitivity, specificity, predictive values, likelihood ratios, and area under curve. CASP-TR performance in post-stroke patients with and without cognitive dysfunction was compared using the Mini-Mental State Examination (MMSE). CASP-TR data obtained by two raters were compared based on the presence of aphasia. Outcomes and results The study assessed inter-rater agreement using intraclass correlation coefficient (ICC), yielding a score of 0.92 (95% CI 0.85–0.96) between raters. Across subtests, moderate to good agreement was observed in most domains. Correlation analyses showed a strong positive relationship (r = 0.80) between mean CASP-TR and MMSE scores. CASP-TR and MMSE scores were stratified based on the presence of aphasia. CASP-TR scores from both raters were notably lower in the aphasia group compared to the non-aphasia group (p < 0.001). Similarly, MMSE scores were significantly lower in the aphasia group (p < 0.001). The patients were also classified into cognitive dysfunction and non-cognitive dysfunction groups based on MMSE scores. No statistically significant score differences were found between examiners in the cognitive dysfunction group (p = 0.267) or non-cognitive dysfunction group (p = 0.156) (p > 0.05). CASP-TR demonstrated a sensitivity of 82.4%, specificity of 78.6%, positive predictive value of 87.5%, negative predictive value of 68.7%, a positive likelihood ratio of 3.77, a negative likelihood ratio of 0.25, and an area under curve of 0.88. Conclusions This study proposes that CASP-TR is a valuable cognitive screening tool for assessing cognitive deficits in stroke patients, and it also demonstrates sensitivity to cognitive deficits in stroke patients with aphasia. The thorough validation within the acute post-stroke Turkish population demonstrates remarkably good accuracy and very good Area Under Curve (AUC) values.