Research Admission From Long Term Care Facility and Delirium Are Most Predictive of Death Following CDI in Hospitalized Patients Presenting Author: Laurie Archbald-Pannone, MD, MPH, University of Virginia Author(s): Laurie Archbald-Pannone, MD, MPH, Tim A. McMurry, PhD; and Richard Guerrant, MD Introduction/Objective: Although incidence & severity of Clostridium difficile infection (CDI) in hospitalized patients have increased, there are no clinical tools at diagnosis to assist the clinicians in predicting patients who are at greatest risk of death following CDI. We evaluated known and novel clinical factors to create a score to predict death following CDI. Design/Methodology: We collected demographic and clinical information (HSR-IRB# 13630) on a prospective cohort of 362 hospitalized patients diagnosed with CDI (5/10-8/11, XPERT, GeneOhm), who were followed for 12 months from diagnosis. We built a statistical model to predict death within 100 days of diagnosis. Data was split into two subsets, a training set (n¼262) used for model selection, and a test set (n¼100) used for model validation.10 clinical predictors were screened for univariate association. Significant var- iables (0.05) were considered for use in the final predictive model, selected by an all subsets algorithm using AIC as the optimality criterion. The area under the ROC curve was used to assess the model’s ability to separate the patients who died within 100 day of diagnosis with those who were alive. Results: Significant univariate factors were included in predictive model and the selected model with coefficients estimated on the training set included age, co-morbidity score, peripheral WBC, admitted from a long term care facility, diagnosed while in ICU, and delirium. The logistic regression coefficients were converted to a points scale. The points score was as effective as the raw logistic regression for differentiating patients who survive from those who did not; with areas under the ROC curve of 0.829 and 0.828, respectively. We calculated that each decade of life, a 10 point increase in WBC, and each point on Charlson co-morbidity index was worth 1 point; admission from long term care facility and ICU at diagnosis were equally predictive and each worth 4 points. Acute diagnosis of delirium was most impactful, worth 5 points. Conclusion/Discussion: The literature supports that the clinical factors of age, co-morbidity score, and peripheral WBC have some predictive power for hospitalized patients with CDI. However, we have found that the factors most predictive of death following CDI were if the patient diagnosed in ICU, admitted from long term care facility, or developed delirium during hos- pitalization. Author Disclosures: NIH/NIAID and TechLab, Inc. Effectiveness of the Brief Interview for Mental Status (BIMS) as a Screening Tool for Cognitive Impairment in the Geriatric Heart Failure Population Presenting Author: Kenneth Koncilja, MD, Cleveland Clinic Foundation Author(s): Kenneth Koncilja, MD, Elizabeth Hinds, Barbara Messinger- Rapport, MD, PhD, CMD; and Amanda Lathia, MD Introduction/Objective: Cognitive impairment in patients with heart failure is highly prevalent and often under-recognized. As a result of impaired cognition, these patients, who are followed throughout the care continuum, often have difficulty complying with fluid and salt restrictions, medication regimens, and appointments. Noncompliance due to cognitive impairment may reduce patient quality of life and may increase the risk of hospital admissions and readmissions and the risk of mortality due to heart failure. The Brief Interview for Mental Status (BIMS) is used in skilled and long term care facilities to screen for cognitive impairment. There is no accepted brief screening tool for cognitive impairment in heart failure. Design/Methodology: A cross-sectional pilot study was conducted at two nurse-run community heart failure clinics to determine the validity of the BIMS as a screening tool for cognitive impairment in the geriatric heart failure population. The subjects were English speaking adults, 55 years of age and older, with a diagnosis of heart failure and no documented history of cognitive impairment or dementia. The Montreal Cognitive Assessment (MOCA), a fairly complex test for cognitive impairment validated in a population with vascular disease, was used as the gold standard. The MOCA and the BIMS were administered to each subject. Demographic and past medical information were collected from both interview and review of the electronic medical record. Results: Out of 32 subjects enrolled, the majority were Caucasian (59%) and male (56%) with an average age of 75. Using the MOCA as the gold standard, only 4/32 subjects (12.5%) were cognitively intact, and 5/32 subjects (15.6%) were identified as having probable dementia (MOCA score of less than 19). However, 28/32 subjects (87.5%) attained a normal score on the BIMS (score of 13-15 out of 15). Only 4 of the 28 patients exhibiting cognitive impairment by MOCA standards were found to have cognitive impairment by the BIMS (sensitivity of 14.3%). Similarly, only 4 of the 28 patients (14.3%) identified as cognitively intact by the BIMS were classified as having normal cognition by the MOCA. Conclusion/Discussion: The BIMS is not a sensitive test for cognitive impairment in the geriatric heart failure population. This finding brings into question the use and reliability of the BIMS in skilled and long term care settings for elderly heart failure patients. More research is encouraged to develop effective and reliable screening tools for cognitive impairment in patients with chronic illnesses, including heart failure, in long term and post-acute care. Author Disclosures: Dr. Amanda Lathia has a Geriatric Academic Career Award (GACA), which is a career development grant from the Health Resources and Services Adminstration for interdisciplinary teaching in geriatrics. Dr. Lathia is the principle investigator for the above study and also has been primary research mentor for Dr. Koncilja and Elizabeth Hinds. Also, Elizabeth Hinds received the Medical Student Training in Aging Research (MSTAR) grant from the American Federation of Aging Research from July through September of 2013. She enrolled subjects and completed the cognitive testing for the above study. Efficacy of Brain Brightening for Enhancing Performance and EEG Alpha Rhythm in Early and Middle Stage Dementia Presenting Author: Phyllis Gaspar, PhD, RN, University of Toledo Author(s): Phyllis Gaspar, PhD, RN, Jacob Elliott, PhD, Donna Algase, PhD, RN, Murthy Gokula, MD, CMD, Saaid Arshad Siddiqui, MD, Leelasri Vanguru, MBBS, Archana Kulai, MD, Sreekiran Thotakura, MD, MPH; and Xuan Li, MPH Introduction/Objective: Purpose: A potential treatment to assist those with dementia is termed brain brightening (BB), a process of stimulation to the brain by providing Audio Visual Entrainment (AVE). AVE creates physiological changes including altered EEG activity, limbic stabilization, increases in neurotransmitters (serotonin, endorphin, melatonin), possible increase in dendritic growth, and increased cerebral blood flow. This study examines effects of BB on EEG patterns, cognitive functioning, ADL and dementia-related behavior in elders with dementia residing in ALF. Theoretical rationale: A physiological model incorporating changes that occur with AVE provides rationale for the BB intervention. The central hypothesis is that brain brightening will modulate brain wave frequency of elders with dementia toward the average of 10 cycles per second by increasing alpha and decreasing theta waves. This increase would thereby improve alertness and concentration (and subsequently ADL, motor and behavior) in elders with early and middle stage dementia. Design/Methodology: Subjects: ALF residents were contacted for consent and assent. Inclusion criteria were: age 60 years or greater; diagnosis of dementia; English speaking; chart negative for seizures, pacemaker, and cochlear implant; and Clinical Dementia Rating Scale of 0.5-2.4. Four subjects (1 sham; 3 intervention) completed participation. Two partici- pants were unable to complete sessions (headache and acute change in Poster Abstracts / JAMDA 15 (2014) B3eB28 B26