ORLANDO – A modified Mini-Mental State Examination that asks a patient to recite the days of the week backwards instead of spelling a word backwards detects dementia with fewer false positives, according to a prospective study of older patients with varying degrees of literacy. The Mini-Mental State Exam “is highly dependent on patient education,” Razia Hafiz, MD, said at the annual meeting of the American Geriatrics Society. For example, clinicians who administer the Mini-Mental State Exam (MMSE) to screen for dementia typically ask people to spell W-O-R-L-D backward, or to start at 100 and count backward in units of seven (100, 93, 86, and so on). Dr. Hafiz and her colleagues got better results when they asked 222 participants to recite the days of the week backward. “We are from Eastern North Carolina, where we have a lot of illiterate and low-literacy people. We were falsely classifying a lot of people [as having] dementia,” said Dr. Hafiz, a geriatrics fellow at East Carolina University in Greenville, N.C. They enrolled 28 illiterate patients (13%), 46 semiliterate patients (20%), and 148 literate patients (67%) at an academic outpatient geriatric center, a family medicine center, and a hospital. All participants in the open study had complaints of memory loss or a diagnosis of dementia at baseline. Each participant also completed the Mini-Cog instrument as a control. The Mini-Cog test indicates dementia if a participant cannot recall any of three items, is interpreted as normal if the individual can recall all three, and relies on the accuracy of a clock drawing when the participant recalls only one or two items. A majority, 71%, met the Mini-Cog criteria for dementia. Patients with dementia scored higher on the modified MMSE than the traditional MMSE, regardless of literacy level, education, sex, or race. Mean scores were 19.39 on the MMSE and 20.88 on the modified MMSE. On a scale of 0-5 regarding correlation with the Mini-Cog, this group scored a mean of 0.90 on the serial 7 test, 2.24 on the spelling task, and 3.88 on the weekdays exercise. In addition, participants without dementia all correctly recited the days of the week backward (a score of 5.00) and scored higher on this exercise, compared with the counting (3.85) or serial 7 task (2.02). Of the 111 participants with less than a high school education, both the Mini-Cog and modified MMSE identified 86 with dementia. In contrast, the MMSE identified 95 with dementia, suggesting some were false positives, Dr. Hafiz said. Using a modified MMSE score cutoff of 27 or less correlated better with the Mini-Cog than did the MMSE among participants who were illiterate, who were semiliterate, or who did not complete high school. With this cutoff, the modified MMSE had a higher sensitivity and specificity for all participants, she said. The study included 162 women (73%) and 60 men. The mean age of the participants was 80 years. The cohort was 51% white and 49% African American. Education level was evenly divided between 111 participants who graduated from high school and 111 others who did not. Use of the Mini-Cog as a surrogate for dementia instead of a formal diagnostic assessment was a potential limitation of the study, Dr. Hafiz said. The next step is to make sure that the modified MMSE is applicable over time to monitor the dementia disease process, she said. Damian McNamara is with the Miami bureau of Elsevier Global Medical News.