Cs. Anderson et al., VALIDATION OF A CLINICAL CLASSIFICATION FOR SUBTYPES OF ACUTE CEREBRAL INFARCTION, Journal of Neurology, Neurosurgery and Psychiatry, 57(10), 1994, pp. 1173-1179
The validity of a clinical classification system was assessed for subt
ypes of cerebral infarction for use in clinical trials of putative str
oke therapies and clinical decision making in a population based strok
e register (n = 536) compiled in Perth, Western Australia in 1989-90.
The Perth Community Stroke Project (PCSS) used definitions and methodo
logy similar to the Oxfordshire Community Stroke Project (OCSP) where
the classification system was developed. In the PCSS, 421 cases of cer
ebral infarction and primary intracerebral haemorrhage (PICH), confirm
ed by brain imaging or necropsy, were classified into the subtypes tot
al anterior circulation syndrome (TACS), partial anterior circulation
syndrome (PACS), lacunar syndrome (LACS), and posterior circulation sy
ndrome (POCS). In this relatively unselected population, relying exclu
sively on LACS for a diagnosis of PICH had a very low sensitivity (6%)
and positive predictive value (3%). Comparison of the frequencies and
outcomes (at one year after the onset of symptoms) for each subgroup
of first ever cerebral infarction in the PCSS (n 248) with the OCSP (n
= 543) registers showed uniformity only for LACI. For example, there
were 27% of cases of TACI in the PCSS compared with 17% in the OCSP (d
ifference = 10%; 95% confidence interval (95% CI) 4% to 16%) and 15% o
f cases in the PCSS compared with 24% in the OCSP were POCI (differenc
e = 9%; 95% CI 3% to 15%). Case fatalities and long term handicap acro
ss the subgroups were not significantly different between studies, but
the frequencies of recurrent stroke were significantly greater for PO
CI in the OCSP compared with the PCSS. Although this classification sy
stem defines subtypes of stroke with different outcomes, simple clinic
al measures-level of consciousness, paresis, disability, and incontine
nce at onset-are more powerful predictors of death or dependency at on
e year. It is concluded that simple clinical measures that reflect the
severity of the neurological deficit should complement this classific
ation system in clinical trials and practice.