Czech Training Version of the Montreal Cognitive Assessment (MoCA-CZ1) for Early Identification of Alzheimer Disease

Aim: Our aim was to determine diagnostic characteristics of the Montreal cognitive assessment (MoCA) using its training version for better administration in patients with Alzheimer's disease (AD) for the first time in Czech literature. Patients and methods: We complemented the test form with instructions for the patient, and with administration and scoring instructions. The content and the scoring were preserved from the previous Czech version. A separate Czech administration instruction manual was refined. English back translation of the data collection form and the manual was checked and approved by the Canadian author who also named it the training version. We examined 30 patients previously diagnosed with mild dementia due to Alzheimer's disease according to the new NIA-AA recommendations (78 +/- 6 years, 13 +/- 3 years of education, 56% women) and 30 cognitively normal seniors (NOS) of similar age, education and gender by MoCA-CZ1. Results: The training version of MoCA-CZ1 contains additional information. We assume that this ensures reliability, and facilitates administration and scoring of the test. Total MoCA-CZ1 scores were significantly lower in patients with AD (18 +/- 5 points) than in NOS (26 +/- 3 points) (p < 0.00001). Total scores of NOS did not correlate with age, gender or education. Analysis of receiver operating curve comparing the two groups found optimal cut off score of <= 23 points with sensitivity of 87% and specificity of 84%. The area under the curve was 0.92. Conclusions: The training MoCA-CZ1 is freely available at http://www.pcp.lf3.cuni. cz/adcentrum/testy.html. The test result can contribute to early diagnosis of AD. We recommend using the cut off score of <= 23 points for the Czech training MoCA-CZ1.