Prognosis of ischaemic stroke - a multivariate analysis

The initial acute phase response predicts long-term stroke recovery

Indicators of an acute phase response (APR) in acute ischemic stroke have been shown to correlate with infarct size and predict stroke recurrence. In this study, we examined how well the APR indicators predicted long-term stroke recovery compared with standard clinical predictors of recovery. Plasma levels of interleukin-6 (IL-6), fibrinogen, white blood cells (WBCs), and serum albumin were measured within 4+/-2 days of onset in 131 stroke patients who were free of apparent infections. Standard clinical predictors included initial National Institutes of Health Stroke Scale (NIHSS), infarct size on computed tomography (CT), and Glasgow scale. The individual correlations with 6-month Glasgow outcome were IL-6, 0.42; fibrinogen, 0.24; WBC, 0.35; albumin, 0.47; NIHSS, 0.53; infarct size, 0.19; and initial Glasgow, 0.57. (all P<.005). Multiple regression analysis yielded an adjusted R(2) of .31 for the APR indicators, compared with .38 for the clinical variables. These results indicate that the initial APR is highly correlated with 6-month stroke recovery and that this correlation approaches that observed with standard clinical predictors.

Prognosis of tuberculous meningitis: a multivariate analysis

Twenty-six clinical, laboratory and CT scan features in 49 patients with tuberculous meningitis (TBM) were analysed to identify the parameters of prognostic significance. Employing univariate analysis, seven variables were found to correlate with three month outcome; these variables included the stage of TBM on admission, focal weakness, Glasgow coma scale (GCS), incontinence, infarction, hydrocephalus and shunt surgery. Analysing all 26 variables, a conditional logistic regression model was derived. In this model stage of TBM, age, focal weakness, cranial nerve palsy and hydrocephalus were found to be contributing to the three month outcome of TBM patients as assessed by the Barthel index.