One-Year Evolution of Behavioral and Psychological Symptoms of Dementia in Patients Initially Hospitalized in Cognitive Behavioral Units: The EVITAL Prospective Cohort

Abstract

Background: The 2008–2012 French Alzheimer’s Plan has provided hospital Cognitive and Behavioral Units (CBU) to improve the management of patients with productive behavioral and psychological symptoms of dementia (BPSD). Little is known concerning the behavioral outcome of these patients after discharge.

Objective: The present study investigated the long-term evolution of BPSD over one year after CBU discharge.

Methods: The EVITAL cohort included 221 participants admitted to the CBUs of 3 French hospitals. BPSD were collected using the Neuropsychiatric Inventory (NPI) at admission and 3, 6, and 12 months after hospitalization. The global NPI score evolution was assessed using a linear mixed-effect model. A four-factor model of the NPI including behavioral dyscontrol, psychosis, mood, and agitation subscores was also analyzed.

Results: Our analysis focused on 148 patients followed up during 12 months and evaluated at each visit. The global NPI score was 48.5 (SD 21.7) at baseline, 28.8 (SD 18.7) at 3-month, 23.2 (SD 16.4) at 6-month and 20.9 (SD 15.9) at 12-month follow-up. The score significantly decreased from baseline to follow-up (F = 109.3 p < 0.0001). Moreover, the decrease was observed for each NPI subscores. The Clinical Dementia Rating (CDR) scale score was significantly linked to the baseline NPI score (t = 2.76, p = 0.009). Conversely, the NPI decline was observed whatever the CDR level.

Conclusion: The present study showed a decrease in the global NPI score and all its subscores during the year following the CBU hospitalization, regardless of the initial CDR score.

Keywords

Alzheimer’s disease behavioral disorders BPSD dementia longitudinal study

INTRODUCTION

Behavioral and psychological symptoms of dementia (BPSD) are frequent complications of Alzheimer’s disease and related disorders (ADRD) [1]. These symptoms may be transient, and vary from person to person and with the type of dementia [2]. They have been shown to be present in 50 to 90% of cases and tend to worsen with cognitivedecline [3, 4].

Many studies highlighted that BPSD are associated with a higher risk of unfavorable evolution of the disease. They are associated with deterioration of quality of life [5], unplanned hospitalizations, and premature institution placement [6]. They also have major effects on caregivers, increasing the likelihood of exhaustion and depression [7, 8].

In clinical practice, the treatment of BPSD uses non-pharmacologic and pharmacologic approaches [9]. Psychotropic drugs are often prescribed with a significant risk of untoward effects. Non-pharmacologic treatments are preferred but their levels of evidence are low. Behavioral symptoms are seen as a critical health problem that needs to promote specific management strategies.

To improve the management of BPSD, the French Alzheimer Plan 2008–2012 has provided specific hospital units called Cognitive and Behavioral Units (CBUs) [9, 10]. These units are intended for patients with ADRD and productive behavioral disorders. To date, few studies have investigated the effectiveness of these new care units on BPSD. Only three studies focused on behavioral disorders evolution in patients admitted to CBUs. The first one showed a short-term decrease in BPSD from admission to discharge [11]. The second showed a favorable effect of hospitalization on BPSD at hospital discharge and two weeks after discharge [12]. More recently, Saidlitz et al. assessed the evolution of the Neuropsychiatric Inventory (NPI) score one week before hospitalization, at admission, at discharge, and two to three weeks after discharge. They showed a decrease in the NPI score from preadmission to discharge, and stability after discharge [13]. However, these short-term effects could be explained by factors like a modification of the patients’ environment. Moreover, the benefit of CBU would be limited if this decrease in behavioral symptoms does not persist after discharge. Beyond the assessment of the short-term BPSD evolution during hospitalization, it would be important to improve our knowledge about the long-term evolution of behavioral symptoms after discharge. To our knowledge, there has been no prospective clinical investigation that examined the long-term evolution of BPSD in patients hospitalizedin CBU.

The present naturalistic study evaluated the long-term evolution of BPSD among these patients with ADRD during the year following their hospitalization in a CBU. We also investigated the factors associated with the long-term evolution of BPSD among these patients.

MATERIAL AND METHODS

Setting and patients

The EVITAL cohort included 221 participants aged 50 and older, admitted to the CBUs of 3 distinct French hospitals in Lyon and Saint Etienne. To answer our study question, further inclusion criteria for participants were: patients with ADRD at dementia stage, without severe disease that can interfere with the variables of the assessment and with a caregiver in sufficient contact with the subject to be able to note behavioral changes. A complete description of the study design was given in a previouspaper [14].

Independent variables

The factors favoring BPSD such as family situation, living place (home or nursing home), previous diseases, comorbidities, etiological diagnosis of dementia and medications were investigated during hospitalization. As the modification of living place during the study could influence the evolution of BPSD, the link between living place and behavioral symptoms was assessed both at baseline and different follow-up visits. The rehospitalization rate for BPSD was measured at each follow-up visit.

Clinical dementia stage was assessed using Clinical Dementia Rating (CDR), scoring from 1 to 3.

Psychotropic drug use was collected at CBU admission and discharge, and at each follow-up interview.

Dependent variables: BPSD assessment

The NPI [15] was designed to collect information on neuropsychiatric problems from patients with brain disorders. It is often used in international studies and has been validated in French [16].

Behavioral disorders were assessed during the inclusion visit and each subsequent interview with the patients and the caregivers. The NPI currently used assesses the frequency and severity of twelve behavioral symptoms frequently observed in ADRD. A total subscale score is calculated by multiplying the frequency and severity, and a global score is generated by summing up the total scores of the individual subscales (maximum = 144).

The NPI was completed by the caregivers at home or the referent caregivers at the nursing homes.

A NPI was also completed by the CBU teams at baseline and before CBU discharge (CBU team NPI).

Data collection procedures

Inclusion visit

The inclusion visit was performed by the neurologist, geriatrician, or psychiatrist in charge of the patients following the common practice. During this visit, the data included in the present analysis were: demographic and environmental data, etiological diagnosis and stage of dementia, comorbidities and acute somatic diseases, psychotropic drugs and dementia medication, and BPSD (NPI).

Follow-up visits: A follow-up visit was performed 3 months after discharge at the patients’ residence (home or nursing home). Follow-up phone calls were made to caregivers (patients’ informal caregivers at home or referent caregivers at the nursing homes) at 6 and 12 months. The following data were collected during these follow-up interviews: demographic and environmental data, patient’s living conditions, concomitant events, medications and non-drug therapies, rehospitalizations for BPSD, and BPSD assessed by the NPI.

Validation of phone call visits: To validate the phone call data collection procedure, both phone and face interviews were made and compared at 3-month visit for a subgroup of 18 patients and caregivers.

Ethics approval

The study protocol was reviewed and approved by an ethics committee. All procedures were conducted in accordance with the Declaration of Helsinki and ICH (International Conference on Harmonization), Guideline for Good Clinical Practice.

This is a routine care research that did not require written informed consent. Patients and their caregivers were given an information sheet at their admission to the CBU. Their acceptance or refusal was noted in medical record.

Statistical analysis

A descriptive analysis was performed on all study variables using tendency and dispersion measures for quantitative variables and relative and absolute frequencies for qualitative variables. The evolution of psychotropic drugs was assessed using linear mixed-effects models that included time as a fixed effect and subject as a random effect.

The CBU team NPI scores from baseline to CBU discharge were compared using a mixed-effects model. The following analysis included the population who completed each NPI assessment. The global NPI scores from baseline to 3, 6, and 12-month follow-ups were assessed using linear mixed-effects models that included time as a fixed effect and subject as a random effect. The global NPI score was tested using a multiple degree of freedom comparison of scores. Post hoc tests compared the baseline NPI scores with NPI scores at 3-, 6-, and 12-month follow-ups. A four-factor model of the NPI was also analyzed using a similar approach according to Hollingworth recommendations. The four identified components are behavioral dyscontrol (euphoria, disinhibition, aberrant motor behavior, and sleep and appetite disturbances), psychosis (delusions and hallucinations), mood (depression, anxiety, and apathy), and agitation (aggression and irritability) [17].

To identify the factors associated with NPI evolution, the variables identified as linked to BPSD in the literature (age, gender, living at home or in a nursing home, baseline number of psychotropic drugs, number of comorbidities, and CDR) and significantly associated to NPI in univariate analysis were then included as covariables in the linear mixed-effects models.

Concerning the living place, the different NPI scores were compared between patients who stayed at home during all the study period, patients who lived in a nursing home during all the study period, and patients who entered into a nursing home during the follow-up.

All models used an unstructured covariance matrix. All statistical tests were two-tailed. A p value < 0.05 was considered significant. Statistical analyses were performed with SPSS version 17 (SPSS Software, Chicago, USA).

RESULTS

Two hundred and twenty-one patients were included in the EVITAL study. A total of 148 patients attended all follow-up visits and were included in the longitudinal analysis, and 73 did not participate in each follow-up visit. Among them, fifty patients (22.6%) died during the study, of whom 13 (5.9%) died during the hospitalization in the CBU, 13 (5.9%) before the 3-month follow-up, 10 (4.5%) between the 3-month and 6-month follow-up, and 14 (6.3%) between the 6-month and 12-month follow-up; the mean age of death was 83.2 years (SD 6.8) and the median age of death was 84 years. The mean initial NPI score was 53.3 (SD 24.4) for patients who died during the study and 47.1 (SD 20.7) for the patients staying alive (p = 0.08). Fifteen patients (6.8%) were lost of follow-up during the study, and 1 (0.4%) refused to participate in the follow-up, and 7 (3.2%) patients missed at least one visit. The flow chart of the study participants is given inFig. 1.

Demographic and clinical characteristics of the patients who completed all the visits and the other patients, including age, gender, dementia etiology, living place, number of comorbidities and acute somatic diseases, psychotropic drugs, and NPI score are given and compared in Table 1. No significant difference was observed between these two groups.

Patient initial and follow-up characteristics included in the longitudinal analysis

Dementia stage was mild for 18 patients (12.2%) (CDR 1), moderate for 85 patients (58.2%), and severe for 43 patients (29.5%) (Data were missing for 2 patients). The mean number of comorbidities was 2.87 (SD 1.8) and an acute somatic disease was diagnosed during hospitalization in the CBU for 60 patients (41.1%).

A total of 52 patients (35.2%) were re-hospitalized during the year following CBU discharge, of whom 15 (10%) were re-hospitalized for BPSD. Forty-four patients (29.7%) entered into an institution after CBU discharge, and 28 (13.5%) during the year offollow-up.

The mean number of psychotropic drugs was 1.99 (SD 1.2) at baseline, 2.28 (SD 1.02) at CBU discharge, and 2.22 (SD 0.95) at 1-year follow-up (F = 4.25, p = 0.002).

NPI score evolution

Among the 148 patients who participated in each follow-up visit, the mean CBU team NPI score was 40.5 (SD 24.6) at admission and 30.1 (SD 21.5) at CBU discharge (F = 29.8, p < 0.0001).

The mean global NPI scores (SD) ranged from 47.2 (20.2) at baseline to 27.5 (18.0) at 3-month, 21.5 (15.2) at 6-month, and 20.2 (16.0) at12-month follow-ups. The NPI score significantly decreased from baseline to 12-month follow-up (F = 109.3, p < 0.0001). Post-hoc t-tests showed a significant decline from baseline to 3-month follow-up (t = –11.7, p < 0.0001), 6-month follow-up (t = –15.2, p < 0.0001), and 12-month follow-up (t = –16.0, p < 0.0001).

There was a significant decline in scores for each NPI subdomain of psychosis (F = 27.6, p < 0.0001), behavioral dyscontrol (F = 32.3, p < 0.0001), mood (F = 90.6, p < 0.0001) and agitation (F = 53.6, p < 0.0001). Post-hoc t-tests showed a significant decline in psychosis subscore from baseline to 3-month follow-up (t = –4.9, p < 0.0001), 6-month follow-up (t = –76.7, p < 0.0001), and 12-month follow-up (t = –6.8, p < 0.0001). For behavioral dyscontrol subscore, post-hoc t-tests revealed a significant decline from baseline to 3-month follow-up (t = –5.9, p < 0.0001), 6-month follow-up (t = –8.6, p < 0.0001), and 12-month follow-up (t = –9.1, p < 0.0001). Post-hoc t-tests showed a significant decline in mood subscore from baseline to 3-month follow-up (t = –10.3, p < 0.0001), 6-month follow-up (t = –12.3, p < 0.0001), and 12-month follow-up (t = –13.2, p < 0.0001). Finally, post-hoc t-tests revealed a significant decline in agitation subscore from baseline to 3-month follow-up (t = –7.4, p < 0.0001), 6-month follow-up (t = –9.6, p < 0.0001), and 12-month follow-up (t = –10.2, p < 0.0001). These results are given in Fig. 2.

The relationship between NPI scores and socio-demographic and clinical variables was measured and the results are showed in Table 2. Only CDR score was significantly linked to baseline NPI score.

A total of 23 patients stayed at home during all the study period, 72 patients entered into a nursing home during the follow-up, and 50 patients lived in a nursing home during all the study period. The NPI scores were compared between the 3 groups at each visit. No significant differences between the 3 groups were observed for caregiver NPI scores at baseline (F = 2.29, p = 0.11), 3-month (F = 1.55, p = 0.22), 6-month (F = 1.46, p = 0.24), and 12-month follow-ups (F = 1.13, p = 0.32).

Figure 3 showed the mean NPI scores according to CDR level at baseline and at each follow-up visit. The NPI scores decreased from baseline to 12-month follow-up regardless the baseline CDR score (CDR = 1: F = 14.7, p < 0.0001; CDR = 2: F = 61.1, p < 0.0001; CDR = 3: F = 37.4, p < 0.0001). The NPI scores were then compared according to the level of CDR score at each visit. At baseline visit, the mean NPI score was 39.8 (SD 20.73) for CDR 1, 45.9 (SD 20.16) for CDR 2, and 52.7 (SD 19.82) for CDR 3. At 3-month visit, the mean NPI score was 19.33 (SD 13.77) for CDR 1, 26.52 (SD 18.11) for CDR 2, and 31.40 (SD 17.65) for CDR 3. The mean NPI score was 18.06 (SD 14.39) for CDR 1, 21.34 (SD 14.36) for CDR 2, and 22.95 (SD 15.79) for CDR 3 at 6-month visit. Finally, the mean NPI score was 16.72 (SD 16.74) for CDR 1, 18.75 (SD 15.62) for CDR 2, and 22.91 (SD 15.69) for CDR 3 at 12-month visit. The difference between the NPI scales according to the CDR score was significant at baseline (F = 2.99, p = 0.05) and 3-month follow-up visit (F = 3.11, p = 0.04). No significant difference was observed at 6-month visit (F = 0.70, p = 0.50) and 12-month visit (F = 1.23, p = 0.28)(Fig. 3).

Finally, the CDR score was included as a covariate in the linear mixed-effects model. A significant interaction was found between the CDR score and the global NPI evolution (F = 2.99, p = 0.019).

A sub-analysis of the evolution of NPI factors scores was made in each CDR group. The evolution of NPI Psychosis sub-score was significant in CDR 1 (F = 4.48: p = 0.009), CDR 2 (F = 9.94; p < 0.0001), and CDR 3 subgroups (F = 5.88; p = 0.001). The evolution of Behavioral Dyscontrol sub-score was significant for CDR 1 (F = 3.70; p = 0.02), CDR 2 (F = 11.46; p < 0.0001), and CDR 3 subgroups (F = 7.22; p < 0.0001). Concerning Mood subscore evolution, it was significant in CDR 2 (F = 31.64; p < 0.0001) and CDR 3 (F = 26.9; p < 0.0001), and it was near significance in CDR 1 subgroup (F = 2.27; p = 0.09). Finally, Agitation subscore evolution was significant for CDR 1 (F = 4.47; p = 0.008), CDR 2 (F = 20.64; p < 0.0001), and CDR 3 (F = 10.30; p < 0.0001) subgroups.

DISCUSSION

Short-term BPSD evolution

Our results point out a positive and significant effect of CBU hospitalization on the BPSD intensity after discharge. These data confirm the short-term effectiveness of these hospitalization units as established in previous studies [11 , 18].

Several factors may explain the short-term effectiveness of CBUs in caring for patients with BPSD. CBU staff is composed of trained nurses, a psychologist, a geriatrician, a neurologist, and a psychiatrist that allow a multidisciplinary and specific management of BPSD. Clinical analysis of behavioral disorders requires a comprehensive assessment to identify different etiologies such as delirium, geriatric conditions, psychological distress, psychiatric diseases, drug iatrogenic diseases, or inadequate environment [19]. About this latter point, it could have been predicted that the long-term BPSD reduction could be due to nursing-home entering. Interestingly, no significant difference was observed between the different NPI scores according to the patients living place. This result suggests that the stability of the BPSD decrease could be at least partially linked to the patients’ management in CBUs, and not only due to a modification of their living conditions.

In our study, almost 43% of the patients suffered from somatic diseases, whose treatment probably contributed to reduce behavioral disorders. Furthermore, a systematic psychiatric examination was performed to identify BPSD associated to a possible psychiatric disease, and to better select psychotropic drugs. Particular attention was also paid to drug-related iatrogeny. Prescriptions were systematically reviewed to remove potentially iatrogenic drugs, especially those with anticholinergic effects.

Another point of interest concerns the effectiveness of CBU hospitalization according to the stage of severity of dementia. At baseline, CDR levels were significantly associated with the NPI global score, with more BPSD in advanced stage of the disease. This higher prevalence of behavioral disorders in severe disease has been widely described[20, 21].

Long-term BPSD evolution

Our results point out a positive and significant effect of hospitalization on the BPSD intensity at 3, 6, and 12 months. Previous studies have shown that behavioral disorders worsen with disease progression. The REAL.FR study results pointed out that patients with moderate AD showed a significant increase in NPI hyperactivity and apathy after one-year follow-up [22] and 4-year follow-up [21]. In a two-year longitudinal prospective study of 199 demented patients, Aalten et al. reported that hyperactive symptoms increased during the follow-up, while affective symptoms remained stable [23]. A temporary beneficial effect of CBU hospitalization could have been expected but was not observed in our present results. The most striking result was a long-term effect of CBU hospitalization, characterized by a sustained decrease in BPSD after 3 months and up to one year after discharge. It is worth to note that behavioral disorders at baseline were severe with an average score close to 50 on the global NPI scale, against 20 one year later.

After one-year follow up, there was a significant decrease in BPSD regardless the initial severity. In addition, the comparison of the global NPI score according to the disease stage found no more difference, suggesting a better long-term effect of CBU care in severe form of illness.

BPSD are heterogeneous, combining different symptoms which can be classified into subgroups. Hollingworth [17] distinguished four subgroups: psychosis, behavioral dyscontrol, agitation, and mood disorder. At baseline, patients exhibited high levels of mood disorder, behavioral dyscontrol, and agitation, whereas psychosis symptoms were less common. Agitation, irritability and behavioral dyscontrol are known to strongly impede the interpersonal abilities of patients and are difficult to manage by caregivers [24, 25]. These disruptive symptoms are frequent reasons for hospitalization in CBUs. In this study, we observed an overall improvement in BPSD at discharge and during follow-up, with no significant difference between BDSP subgroups. Nevertheless, even if the effect was not significant, mood disorder seemed to improve faster than the other subgroups. These results confirm previous studies that showed an overall and nonspecific effect of CBU hospitalization on BPSD.

Long-term effectiveness of CBUs is probably due to several reasons. The environmental and relationship factors play a crucial role in the development of BPSD. During hospitalization, an important place is given to family support, in terms of information, education, and psychological support. Concerning patients living in the community, it can be assumed that improvement measured after hospitalization was partially due to the better tolerance of caregivers to behavioral disorder. In other words, caregivers could have learned better strategies to cope and manage with BPSD. Moreover, after discharge, the patients hospitalized in CBUs had more regular and frequent medical evaluations. In case of severe BPSD, caregivers can seek the help of a trained professional more quickly. This opportunity to facilitate access to professionals is reassuring and can also limit the risk of subsequent hospitalizations. Furthermore, knowledge of the patient’s individual situation observed during its hospitalization in a CBU can be transferred to the patient’s family or nursing home team [26] in order to help them take better care of their patient after CBU discharge. This point may contribute to the stability of BPSD after CBU discharge.

Interestingly, the rehospitalization rate was lower in our study than in previous publications. Baudemont and Saidlitz reported 12.8 and 20% of rehospitalizations, respectively, within the 3 months following CBU discharge [13, 27]. Unfortunately, these studies did not specify the reasons for rehospitalization, but it can be supposed that the major reason was an increase in BPSD.

The death rate of 22.6% was relatively high in our study, this point should be discussed. The mortality rate was compared to the literature data. A French study reported similar mortality rate of 23% at a mean age of 86.7 years in patients hospitalized in hospital short-term units [28]. For a similar rate of death the mean age of death of the patients included in our study was 3 years lower [28]. However, a number of patients included in EVITAL cohort had severe-stage dementia (CDR 2 or 3), and the observed mortality rate was coherent with the severity of the disease. BPSD could have been considered as a mortality risk factor in our study, but no significant difference was observed on NPI score between patients who died and those completing the follow-up. Furthermore, it would be interesting to study the factors related to death in EVITAL cohort.

In our study, the mean number of psychotropic drugs increased from baseline (1.99) to CBU discharge (2.32), and remain stable during the year of follow-up. Similar results were reported by Koskas, but not by Saidlitz who showed a diminution of psychotropic drug prescriptions after CBU hospitalization. These differences could be explained by different behavioral profiles between the studies, as our patients exhibited more severe behavioral disturbances with a mean NPI score of 47 at admission against 39 in Saidlitz’s study. Although there is a moderate but significant increase in psychotropic drugs, the mean number of psychotropic drugs at hospital discharge was lower than in other studies. Moreover, it could be interesting to more precisely assess the different psychotropic drugs evolution with qualitative pharmacological approach.

Limitations

The present study may have several limitations. First, some patients who were living at home before their admission to the CBU moved to a new place of residence, e.g., in an institution, after discharge. NPI was assessed by two different persons for these patients. However, the inter-rater reliability of the NPI has been established [15]. Besides, to deal with this limitation, subgroup analyses were made separately for patients living at home from baseline to one-year follow-up and for patients institutionalized from baseline to the end of the study.

Secondly, the group of patients included in the study was not compared with a control group of patients not hospitalized in CBUs. However, it would be very difficult to identify patients with the same clinical characteristics in non-specific care units. Our patients might have been compared to other types of patients, leading to biased findings. In spite of this, our data were compared to the literature data, all showing an increase in or at least stability of the BPSD in similar types of patients. On the opposite, we observed a consistent and stable decrease in BPSD in our patients. These findings were seen both for patients with mild, moderate or severe stage of the disease, and for the different subgroups ofBPSD.

In conclusion, our naturalistic study is the first prospective cohort to investigate the long-term evolution of BPSD over one year after CBU discharge. A stable decrease in BPSD was seen both for patients with mild, moderate, or severe stage of the disease, and for the different subgroups of BPSD. As the management of patients with dementia and BPSD is currently a major challenge for the patients and their families, these data are very encouraging. In the future, it will be important to identify more precisely the factors associated with this BPSD reduction in order to improve patient management in CBUs and after hospitalization discharge.

Footnotes

ACKNOWLEDGMENTS

This study was funded by the French Ministry of Health, via Rhône-Alpes Regional Health Agency.

The authors would like to acknowledge the care team of CBU for the contribution to patients’ inclusion in EVITAL study.

Authors’ disclosures available online ().

References

Finkel

, Costa e Silva

, Cohen

, Miller

, Sartorius

(1996) Behavioral and psychological signs and symptoms of dementia: A consensus statement on current knowledge and implications for research and treatment. Int Psychogeriatr 8(Suppl 3), 497–500.

Lyketsos

, Steinberg

, Tschanz

, Norton

, Steffens

, Breitner

(2000) Mental and behavioral disturbances in dementia: Findings from the Cache County Study on Memory in Aging. Am J Psychiatry 157, 708–714.

Robert

, Verhey

, Byrne

, Hurt

, De Deyn

, Nobili

, Riello

, Rodriguez

, Frisoni

, Tsolaki

, Kyriazopoulou

, Bullock

, Burns

, Vellas

(2005) Grouping for behavioral and psychological symptoms in dementia: Clinical and biological aspects. Consensus paper of the European Alzheimer disease consortium. Eur Psychiatry 20, 490–496.

Benoit

, Brocker

, Clement

, Cnockaert

, Hinault

, Nourashemi

, Pancrazi

, Portet

, Robert

, Thomas

, Verny

(2005) Les symptomes psychologiques et comportementaux de la demence: Description et prise en charge. Rev Neurol (Paris) 161, 357–366.

Gonzalez-Salvador

, Arango

, Lyketsos

, Barba

(1999) The stress and psychological morbidity of the Alzheimer patient caregiver. Int J Geriatr Psychiatry 14, 701–710.

Lyketsos

, Lopez

, Jones

, Fitzpatrick

, Breitner

, DeKosky

(2002) Prevalence of neuropsychiatric symptoms in dementia and mild cognitive impairment: Results from the cardiovascular health study. JAMA 288, 1475–1483.

Donaldson

, Tarrier

, Burns

(1997) The impact of the symptoms of dementia on caregivers. Br J Psychiatry 170, 62–68.

Sink

, Covinsky

, Barnes

, Newcomer

, Yaffe

(2006) Caregiver characteristics are associated with neuropsychiatric symptoms of dementia. J Am Geriatr Soc 54, 796–803.

Noblet-Dick

, Balandier

, Demoures

, Drunat

, Strubel

, Voisin

(2013) Description of cognitive-behavioral specialized units in France: Results of a national investigation. Geriatr Psychol Neuropsychiatr Vieil 11, 151–156.

10.

Pongan

ED-C

, Laurent

, Vermorel

, Gonthier

, Krolak Salmon

, Rouch

(2014) Organisation and functioning of cognitive-behavioral units in Rhône-Alpes region one year after opening;. Neuro Psycho Griatrie 14, 319–325.

11.

Koskas

, Feugeas

, Saad

, Belqadi

, Daraux

, Drunat

(2011) Management of behavioral and psychological symptoms of dementia in a dedicated psychogeriatric unit: A pilot experience. Alzheimer Dis Assoc Disord 25, 184–186.

12.

Delphin-Combe

, Roubaud

, Martin-Gaujard

, Fortin

M-E

, Rouch

, Krolak-Salmon

(2012) Efficacité d’une Unité Cognitivo-Comportementale sur les symptômes psychologiques et comportementaux des démences. 11ème Réunion francophone sur la maladie d’Alzheimer et les syndromes apparentés, Toulouse.

13.

Saidlitz

, Sourdet

, Voisin

, Vellas

(2016) Management of behavioural symptoms of dementia in a specialized unit care. Psychogeriatrics. doi: 10.1111/psyg.12193

14.

Pongan

, Freulon

, Delphin-Combe

, Dibie-Racoupeau

, Martin-Gaujard

, Federico

, Waissi

, Richard

, Jacqueline

, Fabre

, Trombert-Paviot

, Krolak-Salmon

, Laurent

, Rouch

(2014) Initial and long-term evaluation of patients with Alzheimer’s after hospitalization in cognitive and behavioural units: The EVITAL study design. BMC Psychiatr 14, 308.

15.

Cummings

, Mega

, Gray

, Rosenberg-Thompson

, Carusi

, Gornbein

(1994) The Neuropsychiatric Inventory: Comprehensive assessment of psychopathology in dementia. Neurology 44, 2308–2314.

16.

Robert

, Médecin

, Vincent

, Staccini

, Cattelin

, Goni

(1998) L’inventaire Neuropsychiatrique: Validation de la version Française d’un instrument destiné á évaluer les troubles du comportement chez les sujets déments.e Gérontologique. L’ann’e 5, 63–87.

17.

Hollingworth

, Hamshere

, Moskvina

, Dowzell

, Moore

, Foy

, Archer

, Lynch

, Lovestone

, Brayne

, Rubinsztein

, Lawlor

, Gill

, Owen

, Williams

(2006) Four components describe behavioral symptoms in 1,120 individuals with late-onset Alzheimer’s disease. J Am Geriatr Soc 54, 1348–1354.

18.

Delphin-Combe

, Roubaud

, Martin-Gaujard

, Fortin

, Rouch

, Krolak-Salmon

(2013) Effectiveness of a cognitive-behavioral unit on behoral and psychological symptoms of dementia. Rev Neurol (Paris) 169, 490–494.

19.

Ford

(2014) Neuropsychiatric aspects of dementia. Maturitas 79, 209–215.

20.

Benoit

, Brocker

, Clement

, Cnockaert

, Hinault

, Nourashemi

, Pancrazi

, Portet

, Robert

, Thomas

, Verny

(2005) Behavioral and psychological symptoms in dementia: Description and management. Rev Neurol (Paris) 161, 357–366.

21.

Gonfrier

, Andrieu

, Renaud

, Vellas

, Robert

(2012) Course of neuropsychiatric symptoms during a 4-year follow up in the REAL-FR cohort. J Nutr Health Aging 16, 134–137.

22.

Benoit

, Staccini

, Brocker

, Benhamidat

, Bertogliati

, Lechowski

, Tortrat

, Robert

(2003) Behavioral and psychologic symptoms in Alzheimer’s disease: Results of the REAL.FR study. Rev Med Interne 24(Suppl 3), 319s–324s.

23.

Aalten

, Jolles

, de Vugt

, Verhey

(2007) The influence of neuropsychological functioning on neuropsychiatric problems in dementia. J Neuropsychiatry Clin Neurosci 19, 50–56.

24.

Torrisi

, De Cola

, Marra

, De Luca

, Bramanti

, Calabro

(2016) Neuropsychiatric symptoms in dementia may predict caregiver burden: A Sicilian exploratory study. Psychogeriatrics. doi: 10.1111/psyg.12197

25.

Feast

, Moniz-Cook

, Stoner

, Charlesworth

, Orrell

(2016) A systematic review of the relationship between behavioral and psychological symptoms (BPSD) and caregiver well-being. Int Psychogeriatr 28, 1761–1774.

26.

Strubel

(2015) The specificities of care in cognitive-behavior units. Soins Gerontol 20, 37–42.

27.

Baudemont

, Merlet

, du Boisgueheneuc

, Liuu

, Tartarin

, Ragot

, Paccalin

(2012) Cognitive-behavioral specialized units: Predictive factors of readmissions within three months. Geriatr Psychol Neuropsychiatr Vieil 10, 277–283.

28.

Duport

, Boussac-Zarebska

, Develay

, Grémy

(2013) Maladie d’Alzheimer et démences apparentées: Taux d’ALD, de patients hospitalisés et de mortalité en France métropolitaine, 2007 et 2010. Bull Epidémiol Hebd 30, 361–368.