The Northwick Park Examination of Cognition: A brief cognitive assessment tool for use in acute stroke services

Abstract

Background/Aims: Assessment of cognitive impairment following stroke forms an important part of diagnosis, treatment and rehabilitation planning. However, none of the available and widely used tools were developed specifically for use in stroke services. Most screening tools were developed for dementia, and consequently are biased toward an evaluation of memory function, provide inadequate assessment of executive function and are mainly verbally administered, limiting their utility in aphasic patients. Methods: One hundred and sixty-six stroke patients admitted to a hyper acute stroke unit and a control sample of 100 healthy participants completed the Northwick Park Examination of Cognition (NPEC). The NPEC includes 22 sub-tests in the domains of reasoning, episodic memory, language, perception and attention/executive function. Multiple input (verbal, visual) and output (spoken, written, gesture) modalities increase accessibility to patients with various deficits/lesion locations. Findings: Mean time from stroke to assessment was 5.6 days (SD=7.9). Seventy five percent of patients gained impaired scores (mild, moderate or severe impairment), which was evident at the group level on all subtests. Left and right cortical stroke patients differed significantly (P<0.05) on specific verbal (R>L) and spatial attention (L>R) subtests. Sensitivity and specificity for the detection of cognitive impairment (sensitivity=0.90; specificity=0.80; area-under-curve [AUC]=0.93) were equivalent or superior to data reported for established cognitive screening tools (AUC=0.53–0.89). Patients were disproportionately impaired on high vs low attentional demand cancellation tasks (P<0.0001). Conclusions: The NPEC is brief, freely available and has good sensitivity and specificity for differentiating stroke patients from controls in terms of cognitive functioning. The inclusion of timed executive function measures and comparable verbal and non-verbal sub-tests permits characterisation of cognitive dysfunction in different stroke subtypes.