|| Checking for direct PDF access through Ovid
To determine the construct and predictive validity and sensitivity of the sub-scale items for postural changes, sitting balance and standing balance of the Postural Control and Balance for Stroke (PCBS) test over a 90-day follow-up.In the initial phase of stroke the PCBS test scores were compared with values obtained for the Barthel Index (BI) and the four neuropsychological domains most widely studied in the literature: memory; language; visuo-spatial functions; and visual inattention. The ability of the PCBS test at an early stage to predict functional status, as measured by the BI, and tendency to falls at 90 days after stroke was studied. The sensitivity of the PCBS test in evaluating postural control was studied by comparing the changes between the initial and 90 days' measures, and by examining the ability of the test to discriminate between healthy people and stroke patients. Forty stroke patients (aged 51–89 years), measured 7 and 90 days post-stroke, and 35 healthy control subjects (aged 50–90 years) participated in the study.Moderate correlations were found between the initial PCBS test and the BI (Spearman's r = 0.56–0.79) and a negative correlation between the sum variables for visual inattention and postural changes (Spearman's r = −0.39) and sitting balance (Spearman's r = −0.55). The score for postural changes predicted functional capacity (p < 0.002) and standing balance predicted falls at the three-month follow-up (p < 0.007). The PCBS test proved to be sensitive to change: the median change in the PCBS test scores during the three-month follow-up was 6.06. The control subjects mostly obtained maximum scores, indicating that the PCBS test was able to discriminate between healthy individuals and those with stroke.The results confi rm that the PCBS test has good construct validity, good ability to predict functional capacity and safe moving, and it is sensitive to changes in balance control after stroke. Copyright © 2007 John Wiley & Sons, Ltd.