==== Front Alzheimers Dement (N Y)Alzheimers Dement (N Y)Alzheimer's & Dementia : Translational Research & Clinical Interventions2352-8737Elsevier S2352-8737(18)30040-410.1016/j.trci.2018.06.011Letters to the EditorCommunity-based serious illness care for patients with dementia Noah Hannah aDaaleman Timothy P. bcKistler Christine E. Christine_Kistler@med.unc.edubc∗a University of Vermont, Burlington, VT, USAb Department of Family Medicine, University of North Carolina at Chapel Hill School of Medicine, Chapel Hill, NC, USAc Cecil G. Sheps Center for Health Services Research, University of North Carolina at Chapel Hill, Chapel Hill, NC, USA∗ Corresponding author. Tel.: 919-966-0543; Fax: 919-966-6126. Christine_Kistler@med.unc.edu12 7 2018 2018 12 7 2018 4 324 325 © 2018 The Authors2018This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). ==== Body 1 Introduction Cognitive and psychiatric disorders are prevalent in the homebound population, with dementia and depression being especially common [1]. One intervention that could potentially improve the care of those with dementia is community-based serious illness care (CBSC). CBSC programs such as home-based primary care have been shown to reduce cost and improve care for homebound patients [2]. Patients enrolled in the Veteran Affairs' Home-Based Primary Care program cost the Veteran Affairs health care system 24% less on health expenditures, compared with the 6-month period before enrollment [3]. However, CBSC programs have not been evaluated for their effect on patient outcomes in those with dementia. Because dementia patients require expensive care and suffer from frequent acute illnesses near the end of life, CBSC may affect emergency department (ED) visits and number of hospitalizations in this population differently [4]. This study describes patients in a multidisciplinary CBSC program called the “Reaching out to Enhance the Health of Adults in their Communities and Homes” (REACH) program by dementia status. Because this population may have different needs compared with patients without dementia, we examined differences in patient characteristics between those with and without dementia, including the associations between dementia and outcomes post-REACH program initiation. 2 Methods We conducted an extensive chart review for each patient enrolled in the REACH program. We collected demographic information, including serious illness diagnoses, basic activities of daily living [5], and the presence of a home health aide or physical therapy. In addition to the chart reviews, an automated data pull extracted all primary care, specialty care, hospitalizations, and emergency department utilization in 12 months before and after the initial REACH admission visit. We first ran descriptive statistics to characterize the overall REACH sample and to compare patients with and without dementia. We derived an incidence rate ratio from a Poisson regression to determine whether the presence of dementia affected how REACH admission reduced ED visits and hospitalizations. The three predictors used in the Poisson regression were the effects of a time period (in our case, pre-REACH vs. post-REACH enrollment), the effect of dementia, and the interaction of dementia across the two time-periods. To determine whether dementia significantly affected mortality, we calculated an odds ratio using logistic regression, using mortality as the outcome and dementia as a predictor variable. In our analyses, we did not adjust for any additional variables. This is a preliminary study with a small sample size, so raising the complexity of our models heightens the risk of finding spurious statistical associations [6]. 3 Results In this sample, there were 51 patients with dementia and 108 patients without. Dementia patients were older, with a mean age of 80.6 years (standard deviation ± 9.2) compared with an average age of 65.7 years (standard deviation ± 17.8) in the nondementia group (Table 1). Compared with the nondementia group, the dementia group had more African-American people and fewer white people, more chronic conditions (2.8 ± 1.6 vs. 1.9 ± 1.5) and fewer independent activities of daily living (3.3 ± 2.4 vs. 4.1 ± 2.1). Dementia patients were slightly more likely to be married. Patients with and without dementia had comparable rates of mortality, ED visits, and hospitalizations throughout the study period. REACH does not appear to differentially influence ED visits (incidence rate ratio 0.82 [95% CI 0.60–1.14]), hospitalizations (1.02 [95% CI 0.57–1.82]), or mortality (OR 1.42 [95% CI 0.68–2.98]) based on whether a patient has dementia.Table 1 Patient characteristics, n = 159 Patient characteristics Dementia patients, N = 51 All other patients, N = 108 Total, N = 159 Age, mean (SD) 80.6 (9.2) 65.7 (17.8) 70.4 (17.0) Female 34 (67%) 55 (51%) 89 (56%) Race  White 26 (51%) 74 (69%) 100 (63%)  African-American 23 (45%) 29 (27%) 52 (33%)  Other 3 (6%) 4 (4%) 6 (4%) Patient is married 26 (51%) 41 (38%) 67 (42%) Number chronic health conditions, mean (SD) 2.8 (1.6) 1.9 (1.5) 2.2 (1.5) Independent ADLs, mean (SD) 3.3 (2.4) 4.1 (2.1) 3.9 (2.2) Home health aide present before REACH 25 (49%) 32 (30%) 57 (36%) Physical therapy/occupational therapy/nursing/social work present before REACH 17 (33%) 43 (40%) 60 (38%) Mortality within study period 17 (33%) 28 (26%) 45 (29%) Twelve months before REACH enrollment, mean (SD)  Visits to ED 2.6 (2.6) 2.8 (4.9) 2.8 (4.3)  Hospitalizations 1.1 (1.7) 1.5 (2.2) 1.4 (2.1) Twelve months after REACH enrollment, mean (SD)  Visits to ED 2.2 (4.1) 2.1 (3.6) 2.1 (3.7)  Hospitalizations 0.6 (1.1) 0.9 (1.6) 0.8 (1.5) Pre-REACH and post-REACH change in, mean (SD)  Visits to ED 0.4 (4.4) 0.7 (4.6) 0.6 (4.5)  Hospitalizations 0.5 (1.9) 0.6 (2.0) 0.6 (1.9) Abbreviations: ADLs, activities of daily living; ED, emergency department; SD, standard deviation; REACH, Reaching out to Enhance the Health of Adults in their Communities and Homes. 4 Discussion The dementia patients in REACH differed from nondementia patients on a variety of characteristics. Dementia patients were more likely to be older, female, and African-American. These patients also had a higher number of chronic conditions and were more likely to receive assistance from a home health aide. However, none of these differences appeared to change the REACH program's effect on patients with dementia. Our results agree with what is known about dementia in the larger population. It is a disease of aging [7], impairs activities of daily living [8], requires aides for assistance [9], and is more prevalent among African-Americans than white Americans [10]. Importantly, in the REACH sample, having a diagnosis of dementia did not appear to affect mortality, nor did it augment or decrease the effect of REACH on hospitalizations or ED visits. These results are limited to a single home-based care program in central North Carolina, with patients enrolled in a discrete time period, and measures hospitalization and emergency department utilization over only a 12-month before and after program enrollment time period. Subsequent work could also investigate whether particular components of CBSC programs are especially effective for patients with dementia. Many opportunities exist for evaluating and improving home-based care, a promising intervention for improving the health of homebound patients with dementia. Acknowledgments Funding: This research was funded by the Duke Endowment. ==== Refs References 1 Qiu W.Q. Dean M. Liu T. George L. Gann M. Cohen J. Physical and mental health of homebound older adults: an overlooked population J Am Geriatr Soc 58 2010 2423 2428 21070195 2 Edes T. Kinosian B. Vuckovic N.H. Nichols L.O. Becker M.M. Hossain M. Better access, quality, and cost for clinically complex veterans with home-based primary care JAGS 62 2014 1954 1961 3 Beales J.L. Edes T. Veteran's affairs home based primary care Clin Geriatr Med 25 2009 149 154 viii–ix 19217499 4 Sachs G.A. Shega J.W. Cox-Hayley D. Barriers to excellent end-of-life care for patients with dementia J Gen Intern Med 19 2004 1057 1063 15482560 5 Katz S. Downs T.D. Cash H.R. Grotz R.C. Progress in development of the index of ADL Gerontologist 10 1970 20 30 5420677 6 Babyak M.A. What you see may not be what you get: A brief, nontechnical introduction to overfitting in regression-type models Psychosomatic Med 66 2004 411 421 7 Types of Dementia. Alzheimer's Association Available at: http://www.alz.org/dementia/types-of-dementia.asp 2017 8 Ciro C.A. Maximizing ADL performance to facilitate aging in place for people with dementia Nurs Clin N Am 49 2014 157 169 9 Scherer R.K. Scarmeas N. Brandt J. Blacker D. Albert M.S. Stern Y. The relation of patient dependence to home health aide use in Alzheimer's disease J Gerontol A Biol Sci Med Sci 63 2008 1005 1009 18840808 10 Potter G.G. Plassman B.L. Burke J.R. Kabeto M.U. Langa K.M. Llewellyn D.J. Cognitive performance and informant reports in the diagnosis of cognitive impairment and dementia in African Americans and whites Alzheimers Dement 5 2009 445 453 19896583