Jc. Sharma et al., Prognostic value of CT scan features in acute ischaemic stroke and relationship with clinical stroke syndromes, INT J CL PR, 54(8), 2000, pp. 514-518
The aim of this prospective study was to investigate the value of CT scan i
n prognosis of acute ischaemic stroke patients as a variable additional to
clinical stroke syndromes to develop a simple classification of CT scan fea
tures to provide a practical approach to prognosticate and manage such pati
ents. One hundred and eight nine patients admitted with ischaemic stroke we
re investigated with a CT scan. CT scan features were classified into two g
roups: large infarct, LI (across more than one lobe) and non-large infarcts
(N-LI) for all other features. Patients were also studied for clinical syn
dromes, analysing results for total anterior circulation syndrome (TACS). O
utcome was measured as early post stroke Barthel index, acute phase and 3-m
onth mortality. Patients with LI had features of severe strokes and had a l
ower Barthel index (p<0.001), lower Glasgow coma score (p<0.001), more asso
ciation with incontinence (p<0.001), pyrexia (p=0.007) and dysphagia (p<0.0
01). LI patients required higher level of care in acute wards and had a hig
her length of stay (p=0.01). Both the LI and TAGS individually had a signif
icantly higher mortality (p<0.001) and similar positive predictive value, s
ensitivity and specificity for 3-month mortality. While the combined factor
of 'large infarct and TAGS' provided the highest likelihood ratio (3.1) fo
r mortality, the factor of 'large infarct or TAGS' was the most sensitive (
85%) to identify majority of patients at a risk of mortality. N-LI patients
had a better prognosis. Classification of cerebral infarcts into large and
non-large categories identifies patients who require higher level of care
in acute wards and have a higher mortality Combined factor of 'large infarc
t and/or TAGS' identifies the majority of patients at risk of 3-month morta
lity as compared to either variable taken individually CT scan features are
complimentary to clinical syndromes for managing acute stroke patients.