Factors affecting the results of the clock drawing test in elderly patients hospitalized for physical rehabilitation.

The Clock Drawing Test (CDT) is a recognized and accepted instrument for the early diagnosis of dementia in the elderly. In a prospective study we evaluated the association between the results of this test and a broad range of clinical, functional and sociodemographic variables. The study was conducted on elderly patients hospitalized for rehabilitation following stroke or hip fracture (HF) in the geriatric ward of a university hospital in southern Israel. The administration of the CDT and its scoring system were adapted from Sunderland et al. and Wolfe-Klein et al. The study was conducted on all 425 elderly patients who were hospitalized during the study period and who were capable of completing the test. Stepwise multiple regression was used to evaluate the association between the results of the CDT and the other variables. The mean CDT score (+/- SD) for the entire study population was 7.8 +/- 2.5 and 145 patients (34%) had scores of 6 or below. Of the 41 variables that were tested, significant associations with the CDT were found for the following four variables only: the Folstein minimental test (beta = 0.447, p < 0.0001), the cognition value from the admission FIM (beta = 0.252, p < 0.0001), years of education (beta = 0.183, p = 0.0001), and the patient's age (beta = -0.075, p = 0.037). The total variance of the CDT explained by these four variables (Adjusted R2) was 0.554. We conclude that in the study population there was a significant proportion of patients with low CDT scores. This score, in this population, is influenced in particular by two other measures of cognitive function and by the formal level of education, together with a weaker effect of age.