Gerocognitive Examination Research Articles

Scoring Accuracy of the Self‐Administered Gerocognitive Examination by Primary Care Providers

AbstractBackgroundPrimary Care Providers (PCPs) are critical for case finding of Mild Cognitive Impairment to provide management, referrals, and treatments. Conducting cognitive assessments are time consuming and difficult to incorporate into time‐limited visits. The Self‐Administered Gerocognitive Examination (SAGE) is a reliable and valid, pen and paper, cognitive assessment tool easily incorporated into a PCP visit without time required for staff to administer. Accurate scoring by PCP or staff is critical to interpreting the degree of impairments and any changes over time.MethodsPatients over 64 years‐old without a diagnosed cognitive disorder, presented for a non‐acute care visit with their PCP and were given a SAGE test during their appointment. The SAGE was administered as part of the provider’s standardized method for screening for cognitive impairment. Detailed scoring instructions were provided to the PCP office before subject enrollment. Differences in SAGE scoring between the PCP office and the Ohio State University Memory Disorders Research Center (MDRC) were evaluated. Paired t test and 95% confidence interval were used to compare the mean SAGE scores.ResultsOf the 110 potential recruits, 93 SAGE tests were graded by the PCP office and by MDRC. Scoring discrepancies, the MDRC scored SAGE minus PCP office scored SAGE, ranged from ‐7 points to +3 points with a total of 80% of the SAGE tests resulting in a scoring error by the PCP office; 74% of the total score discrepancies were within ± 2 points. The mean difference was ‐1.49 with a 95% CI (‐1.82, ‐1.17) and p<0.0001 with PCPs scores tending about 1.5 points higher. The largest discrepancies between the scores occurred with the problem‐solving question (40%) and the 3‐D figure question (29%).ConclusionsThe PCP office provided an incorrect score (typically higher) 80% of the time which downplays cognitive deficits. When scoring mistakes are made, accuracy is diminished, impacting the interpretation of the test results and perceived changes over time. PCP offices need to emphasize accurate scoring or consider utilizing the digital format of SAGE, called eSAGE or BrainTest® (available at https://braintest.com). It can send HIPPA compliant automatically scored results to the PCP office.

Identifying mixed Lewy body and Alzheimer’s disease pathologies using functional neuroimaging

AbstractBackgroundLewy body dementia (LBD) and Alzheimer’s disease (AD) are the two most common neurodegenerative causes of dementia. Autopsy studies report many cases with these clinical diagnoses having mixed pathology with alpha‐synuclein, tau and amyloid proteinopathies. There is limited research in trying to distinguish these potential mixed pathology conditions prior to death. We used functional neuroimaging methods including FDG PET, amyloid PET, and DaTscan SPECT to evaluate clinically diagnosed subjects with dementia with Lewy bodies (DLB), AD and PD to investigate how often potential mixed pathologies could be present.MethodEighteensubjects (all Caucasian, male/female ratio 15:3) so far were recruited from Ohio State University Memory Disorder and Movement Disorder Clinics in this currently enrolling study. Subjects were diagnosed based on standard clinical criteria for DLB, PD and AD by fellowship trained cognitive and movement disorders faculty. Cognitive testing including Mini‐Mental State Examination (MMSE) and Self‐Administered Gerocognitive Examination (SAGE) were performed. Neuroimaging evaluations performed by neuroradiology faculty were blinded to clinical diagnosis.ResultNine subjects with clinical diagnosis of Lewy body disease (5 DLB and 4 PD) and nine subjects with clinical diagnosis AD were compared. The Lewy body subjects were 71.3 ± 7.4 years old, MMSE 22.3 ± 5.4, and SAGE 9.7 ± 4.8. AD subjects were 67.1 ± 9.6 years old, MMSE 18.6 ± 5.2, and SAGE 9.0 ± 5.2. One subject (11%) of the clinical AD cases showed mixed LBD and AD features, one subject (11%) of the clinical AD cases was negative for amyloid with normal DaTscan uptake, two subjects (40%) of the clinical DLB cases showed mixed LBD and AD features, one subject (20%) of the clinical DLB cases was positive for amyloid with normal DaTscan uptake, and all four subjects with clinical PD showed neuroimaging findings consistent with PD.ConclusionBased on functional neuroimaging, 21% of clinically diagnosed AD or DLB subjects had evidence of potential mixed pathologies and 14% were misdiagnosed. Although the sample size is small, functional neuroimaging studies revealed possibly incomplete or incorrect diagnoses in about one third of subjects clinically diagnosed by subspecialists using standard clinical diagnostic criteria. Larger studies are required to confirm these findings.

Cross‐sectional association between the area deprivation index (ADI) and cognitive impairment in community‐dwelling older veterans

AbstractBackgroundThe evidence that lower socio‐economic status (SES) may be associated with late‐onset dementia is inconsistent. Differences may be related to how SES was assessed. The area deprivation index (ADI) is a validated, composite measure of neighborhood socioeconomic status representing a geographic area level of social deprivation. Previous research has shown an association between ADI and cognitive impairment, but less is known about the Veteran population. The study aim was to determine the cross‐sectional association between the area deprivation index (ADI) and cognitive impairment in a Veteran population.MethodThis is a cross‐sectional study among community‐dwelling Veterans aged ≥50 years enrolled in a VA primary care clinic from July 2019‐May 2020. Patients were mailed questionnaires including sociodemographic, the Self‐Administered Gerocognitive Examination (SAGE). Clinical information was extracted from the electronic health records (EHR). To assess frailty, we used a 31‐item VA Frailty Index (VA‐FI) generated from EHR data matched to the study date. We calculated the ADI from 17 socioeconomic indicators available from the US Census. ADIs were generated for census tracts corresponding to veterans’ addresses. Higher ADI values are consistent with increased deprivation. After adjusting for age, gender, marital status, race, ethnicity, body mass index and frailty, odds ratios (ORs) and 95% confidence intervals (CIs) were calculated using binomial logistic regression models with cognitive impairment (MCI and dementia) as the outcome variable and ADI as the independent variable.ResultParticipants were 1060 Veterans (response rate of 20.14%), mean age of 68.39 (SD=8.52) years, 59.0% Caucasian, 69.2% non‐Hispanic, 95.8% male, 12.0% screened positive for MCI and 15.6% for dementia. Among ADI quintile groups, patients in the highest group had a significantly higher proportion of cognitive impairment (40.59%) when compared to lower groups, p<.001. No significant differences were found between quintiles 1 to 4. The highest ADI quintile was the only group associated with cognitive impairment, adjusted OR: 1.916 (95% CI:1.187‐3.093), p=.008.ConclusionThis study shows a cross‐sectional association between ADI and dementia in community‐dwelling older Veterans. Further investigation of the links between neighborhood social deprivation and cognitive impairment may assist in the development of strategies that lower the incidence of cognitive impairment in these communities.

Results from the self‐administered gerocognitive examination (SAGE) are not associated with misreporters of energy intake among multiethnic older adults

Examine association between misreporting on 24‐hr dietary recalls (24HDR) and SAGE, a screening tool to identify mild to severe cognitive impairment (MSCI). Adults (n=265) from a substudy of the Multiethnic Cohort (MEC) completed 3 telephone 24HDR. The sample represented the 5 ethnic groups in MEC (African American (AA), Hawaiian, Japanese, Latino, non‐Hispanic white (NHW)). SAGE scores defined likely to be normal (17–22) or MSCI (¡Ü16). Estimated Energy Requirement (EER) was computed using low activity. The ratio of reported energy intake (rEI) to EER identified under (<0.75), acceptable (0.75–1.25), or overreporters (>;1.25). The primary outcome of reporting status (under or acceptable/over) was examined using logistic regression adjusted for confounders.The sample included 131 men and 134 women, ages 55–80y. The majority (80%) were characterized as unimpaired. The proportion of underreporters in the sample was 35%. Adjusted for sex, ethnicity, and age, there was no significant association between cognitive status and misreporting. Independent of cognitive status, AA were more likely (p=.046) and Japanese were less likely (p=.023) to underreport compared to NHW.In this sample of older adults, no association between cognitive impairment and misreporting of rEI based on dietary recalls was found.Support for this work was made possible by the National Cancer Institute NIH/NCI 4R37 CA 54281.