Intra-individual variability in women with breast cancer.

Studies assessing cognitive functioning in women treated for breast cancer have used primarily standardized neuropsychological tests and examined accuracy and/or reaction time as outcome measures: they have been inconsistent in identifying the cognitive domains affected and the severity of deficits. In other contexts of neural development and disorders, measures of Intra-individual variability (IIV) have proven useful in identifying subtleties in performance deficits that are not captured by measures of central tendency. This article presents proof of concept that assessing IIV may also increase understanding of the cognitive effects of cancer treatment. We analyzed mean accuracy and reaction time, as well as IIV from 65 women with breast cancer and 28 age and education matched controls who performed the Conner's Continuous Performance Test, a "Go-NoGo" task. Although there were no significant differences between groups using measures of central tendency, there was a group × inter-stimulus interval (ISI) interaction for IIV Dispersion (p < .001). Patient Dispersion was more variable at shorter ISI than controls and less variable at long ISI, suggesting greater sensitivity to presentation speed. Interpretation of IIV differences requires further investigation. Our results suggest that future studies would benefit from designs that allow analysis of IIV measures in studies assessing cognition in cancer survivors.