Assessing cognitive function in cancer patients

Goals of work: Some patients with cancer suffer cognitive impairment after chemotherapy. A brief, sensitive instrument is required to screen patients for cognitive impairment. In this study, we compare three possible screening instruments. Materials and methods: Cancer patients (n=31) who had received adjuvant chemotherapy within 2 years underwent cognitive evaluation on three occasions. Fluent English speakers (n=20) completed the High Sensitivity Cognitive Screen (HSCS), the computer-based Headminder and CogHealth, and the Functional Assessment of Cancer Therapy-cognitive function (FACT-COG) questionnaire. Patients not fluent in English (NESB) (n=11) performed CogHealth and Headminder. Results: The patients were aged 31-65 years, and 94% were women with breast cancer. At baseline, 6 of 20 (30%) had moderate-severe cognitive impairment on HSCS, 17 of 31 (55%) had scores greater than one standard deviation (SD) below the mean for normative data in one to two of three domains on Headminder, and on CogHealth, 8 of 31 (26%) were greater than one SD below the mean in at least two of six domains. A large practice effect was seen for the HSCS, with moderate-severe cognitive impairment decreasing from 30 to 5% between the first and second assessment. On FACT-COG, 9 of 19 patients (47%) rated their cognition as greater than one SD below that of a pre-chemotherapy breast cancer control group, with impact on quality of life greater than one SD below for 10 of 19 (53%) patients. No correlation was seen between objective cognitive testing and the FACT-COG. Conclusions: CogHealth and Headminder were suitable for NESB patients. The HSCS is not recommended for longitudinal studies with short intervals between testing due to practice effect. There is poor correlation between the patients' perception of their cognitive impairment and the objective tests.