The Impact of Hospitalization on Readmission,Institutionalization,and Mortality of People with Dementia: A Systematic Review and Meta-Analysis

Abstract

Background:

People with dementia (PwD) are at a high risk of hospitalization. Hospitals are often not adequately equipped for PwD and discharges often come unexpected. Therefore, PwD are at a risk of adverse outcomes. However, information about those outcomes is rare but crucial for the development of preventive strategies.

Objectives:

To conduct a quantitative systematic review and meta-analyses on the impact of a hospitalization on readmission, institutionalization, and mortality in PwD. To identify factors associated with these outcomes.

Methods:

PubMed, CENTRAL, and ScienceDirect were searched for studies including terms for dementia, hospital, readmission, institutionalization, and mortality. Relevant were assessed by a quality criteria sheet. Results were summarized in a table. Meta-analysis was conducted with Review Manager 5.3.

Results:

The search yielded 1,108 studies; 20 fulfilled the inclusion criteria and 10 studies were eligible for meta-analyses. The incidence and relative risk (RR) of mortality (RR 1.74 CI^95 % 1.50, 2.05) and institutionalization (RR: 2.16 CI^95 % 1.31, 3.56) of PwD was significantly higher when compared to people without dementia. Results according to readmission rate were inconsistent. Factors significantly associated with the examined adverse outcomes were severity of dementia, number of medications, and deficits in daily living activities.

Conclusion:

Hospitalization of PwD lead to adverse outcomes. An improvement in the identification of and care for PwD in the acute setting as well as in after care in the community setting, especially in the interface between both settings, is required to prevent adverse outcomes in hospitalized PwD.

Keywords

Dementia institutionalization mortality prevention strategies readmission

INTRODUCTION

The prevalence of dementia is increasing due to demographic changes and higher life expectancies [1]. Worldwide, there are approximately 46.8 million people living with dementia [1]. This number is expected to reach 131.5 million in 2050 [1]. In general, people with dementia (PwD) are at a high risk of hospitalization [2, 3]. Nowadays, in hospitals approximately 40% of the patients aged 65 and above are cognitively impaired [4]. Caring for these patients is challenging for hospital staff and requires special support and flexibility [5, 6]. Hospital wards are primarily equipped for the treatment of acute diseases and often cannot provide individualized care for PwD [6]. A systematic review revealed that PwD require more hours of nursing care, face functional decline during admission and have longer hospital stays [7]. This is a threat for PwD, and economically inefficient for hospitals.

A hospitalization can be a life-changing experience for PwD [8]. The main problems involve the interface between hospital and domestic life. Caregivers of PwD indicate that hospital discharges come unexpected, preparation of care arrangements is insufficient, and there is often no contact person available after discharge. Furthermore, PwD and (if present) their caregivers are often forced to organize follow-up medical care independently [9, 10]. A high rate of medication-related problems is a risk factor for unscheduled readmission [11]. Hospitals are required to offer discharge management, including prescribing pharmaceuticals and home care following in-hospital treatment. However, there is considerable heterogeneity, and the discharge management is still far from being standardized [12]. While a general “expert standard discharge management” has been defined [13], there is no specific management guideline for PwD.

The discharge of PwD should be planned in a way that facilitates a return to their homes [9, 14]. A lack of post-acute care can result in a higher 30-day-readmission rate [15] and increased premature institutionalization rate [16]. Additionally, it should be noted that PwD are at higher risk of death following hospitalization [17]. From an economic perspective, these adverse outcomes are crucial cost drivers [18, 19]. Specifically, institutional care is nearly twice as expensive as home care. This emphasizes that a professional discharge management of PwD is extremely relevant for healthcare providers and payers alike [19].

Several studies focused on short-term outcomes (i.e., in-hospital mortality, length of stay) of PwD [20 –22]. Only a few studies reported long-term outcomes of a hospitalization of PwD, which include unscheduled readmission, premature institutionalization, and mortality. Information about those outcomes is crucial for improving the interface between hospital and domestic life. Identifying determinants for successful discharge management can help to design research instruments and develop intervention methods to improve present outcomes.

Objective

The objective of this study is to conduct a quantitative systematic review and a meta-analysis of the impact of hospitalization on long-term outcomes for PwD. The aims of the (1) systematic review are to reveal the incidence of (a) unscheduled readmission, (b) of premature institutionalization, and (c) mortality of PwD following hospital discharge as well as to (d) reveal predictors, like socio-demographic and clinical factors significantly associated with the outcomes and therefore mainly causing those outcomes in dementia diseases. Hospitalization was defined as an admission to hospital for treatment, involving a stay of at least one night. Institutionalization was defined as a discharge or movement to a nursing home or a comparable institution. The objective of the (2) meta-analyses is to calculate the relative risk of these adverse outcomes among PwD in comparison to people without dementia.

METHODS

In order to identify all relevant publications on long-term outcomes of hospitalized PwD, a systematic search was performed.

Search strategy

The search was conducted in April 2017 using the following databases: PubMed, CENTRAL (Cochrane Central Register of Controlled Trials), and ScienceDirect. The search strategy (see Appendix 1) was developed with PubMed and adapted for the databases listed above. The search terms were “dementia” and “Alzheimer’s disease”. The term “dementia” was used to include all subtypes of dementia diseases next to Alzheimer’s disease, like Vascular, Frontotemporal, Mixed, Parkinson’s and Lewy Body Dementia. A detailed search strategy separated for each specific dementia subtype would exactly lead to the same number of initially included studies. Both terms “dementia” and “Alzheimer’s disease” were combined with the term “hospital” and terms expressing the outcomes “readmission”, “institutionalization”, and “mortality”. The terms were defined using the PICO Model: the population were PwD, the intervention was defined as a hospital admission and the outcomes were the above-mentioned terms. According to the comparator, we were interested in both the incidence of each adverse outcome listed above and in the relative risk of these adverse outcomes among PwD in comparison to people without dementia. Therefore, “people without dementia” were not initially set as a comparator in the search term. Whereas, all finally included studies were used to demonstrate the incidence of each adverse outcome in PwD, studies with a control group of persons without dementia were used in the meta-analyses only. Therefore, studies were separated according to an existing or non-existing comparator group.

In addition to the search strategy, further studies were identified through cross-referencing of initially selected studies. Our search includes articles published in English or German between 1 January 2000 and 12 April 2017. This timeframe was used to include more contemporary data. Only studies published in peer-reviewed journals were included.

Data extraction

Articles were included if they described the following outcomes: 1) at least a 28-day readmission, 2) the institutionalization, or (3) mortality rates. Studies with and without a comparison group were included. Studies that described only in-hospital outcomes and studies that were not conducted in acute hospitals were excluded. Studies were also excluded if PwD represented only a small subgroup of the analyzed population. Meta-analyses and systematic reviews were rejected. If the abstract revealed that a study addressed PwD being admitted to a hospital and included at least one of the outcomes mentioned above, the article was selected for further examination. Intervention studies with PwD in the control group, and studies considering the mortality of PwD with a specific comorbid diagnosis, were included as well. Two reviewers reviewed all abstracts and the selected studies separately. In the event of discrepancy (4.11%), a third reviewer became involved.

Data analysis

Within the systematic review, we extracted all given data concerning the incidence of the focused outcomes (readmission, institutionalization, and mortality) as a ratio (percentage) from each of the included studies. At this stage, raw data were represented and data were not pooled. Basic data, including the year of publication, country of data collection, and author were collected from each selected study. Furthermore, we abstracted information on the methodology including sample size, type of study, setting, and the inclusion of a control group. To characterize the sample of each study, we recorded information about patients’ age, sex, and, if available, degree of cognitive impairment and specific dementia diagnosis. In addition, we recorded data on the length of follow-up after hospital discharge. All factors associated with the outcomes were documented as well.

Quality assessment of included studies

The study quality was documented using a quality criteria sheet (see Appendix 2) based on the Mixed Methods Appraisal Tool (MMAT) [23]. This tool was developed for systematic reviews which include different types of studies. Each study was rated particularly with regard to this systematic review, considering the research question, sample size, description of received dementia diagnosis, follow-up time and rate, risk of selection bias, loss to follow-up, response rate and comparison group. A maximum of 18 points was achievable.

Statistical analysis

To represent the relative risk of focused adverse outcomes among PwD in comparison to people without dementia, meta-analyses were conducted using a subsample of all studies with a comparator (control group: people without dementia). Meta-analyses were performed based on the PRISMA statement (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) [24] as a manual for reviews. Thus, results of the meta-analyses complete the findings (incidence of adverse outcomes in hospitalized PwD based on data of all included studies) of the systematic review revealing the relative risk of adverse outcomes in hospitalized PwD compared to non-demented controls based on studies with a comparator group only. To achieve comparability between the studies, the meta-analysis focusing on readmission was conducted using studies which examined 28- or 30-day readmission rates. Furthermore, for all outcomes, studies with a follow-up time of at least twelve months were included. Meta-analyses were conducted with the crude overall rates (data) of the selected studies and thus, solely controlled for the sample size but not for associated confounders. The results were pooled and analyzed using the random effect analysis model. Risk ratio (RR) and the 95% confidence intervals (CI) were calculated using the Mantel-Haenszel method for each outcome and were illustrated by forest plots. Meta-analyses were conducted using Review Manager (Revman) 5.3 from the Cochrane Collaboration [25].

RESULTS

Study selection

The flow chart of the screening process and reasons for exclusion are demonstrated in Fig. 1. No additional articles were found by cross-referencing. Ultimately, 20 articles fulfilled our inclusion criteria. Characteristics of the selected studies are demonstrated in Table 1.

Fig. 1

Screening flow chart. Studies were rejected, without further examination, if they fulfilled already one of the exclusion criteria.

Table 1

Characteristics and results of included studies

Risk of bias across studies

Study quality ranged from four to 14 points, the average was ten points. The quality grading of each study is represented in Table 1. Quality grading was rather low because information about the response and follow-up rates, dementia diagnosis and dropouts was frequently missing.

Unscheduled readmission

Six studies [15 , 26–30] examined unscheduled readmission of PwD. In general, 28- or 30-day readmission rates varied between 10% and 16% [15, 30]. Long-term readmission rates were examined in three studies with follow-up times of three months or more. The range of readmission rates spanned 35% to 65% [27 –29]. Four studies [15 , 30] examined readmission of PwD in comparison to people without dementia, demonstrating inconsistent results: in the study of Harvey et al., PwD were admitted to the hospital due to an injury [26]. The study revealed that PwD had a significantly lower risk (p < 0.1) for unscheduled readmission within 28 days when compared to patients without dementia (16% versus 17.5%) [26]. Daiello et al. [15] compared the 30-day readmission rate of PwD in several general hospitals. The study revealed a significantly higher risk of readmission for PwD (23.4% versus 19.24%, p < 0.01) [15]. Tropea et al. [30] reported a 28-day readmission rate of PwD of 9.8%, as compared to 8.6% for people without dementia. Briggs et al. [29] examined readmission after twelve months, revealing a higher rate for PwD as well (65% versus 52%). The higher rate of readmission for PwD in those studies was associated with a worse nutritional status, a higher co-morbidity and a higher deterioration in activities of daily living (ADL) [28].

The meta-analysis pooled the data of three studies [15 , 31], providing a 28-day or 30-day readmission rate of PwD in comparison to people without dementia. Overall, the readmission rate of PwD was 15.9% (10,809/ 67,849) compared to 15.7% (59,417/ 379,499) in controls and thus, did not differ significantly (see Fig. 2). As such, this meta-analysis demonstrated that generally PwD did not have a higher relative risk of readmission (RR: 1.08 CI^95 % 0.88, 1.32). Nevertheless, there was a significant heterogeneity (p < 0.01) between selected studies, due to different sample sizes (i.e., n = 58083 versus n = 3908).

Fig. 2

Meta-analysis of readmission rate 28- or 30 days.

Institutionalization

In total, 14 studies reported long-term institutionalization rates after discharge of hospitalized PwD. One of those could not be included due to the fact that the living situation of the PwD prior to hospital admission was not documented [15]. The remaining 13 studies differed tremendously with respect to the length of follow-up time and study quality.

Across studies, reported institutionalization rates ranged widely from 9.2% to 73.7% [16 , 32–38]. All studies demonstrated a higher institutionalization rate of PwD compared to controls. For example, McCusker et al. [39] reported an institutionalization rate of 26.0% for PwD after one year, versus 19.0% in controls. Callahan and colleagues [16] revealed that the conditional probability of PwD for becoming institutionalized after hospitalization was 33.8% versus 11.8% in control group [16]. Factors associated with a premature institutionalization of PwD in this study were a higher age, male sex, a higher number of comorbidities (i.e., depression and delirium), and poorer nutritional status, a higher number of behavioral problems, a higher stage of dementia and a higher dependency in activities of daily living [28]. Those factors co-exist with dementia diseases and were significantly associated with a higher rate for institutionalization in PwD. Furthermore, Dramé et al. reported that inability to use the toilet was significantly associated (hazard ratio (HR): 2.48, p < 0.01) with a higher chance of institutionalization as well. Furthermore, patients living alone had a higher chance of institutionalization as well (HR: 1.52, p = 0.01) [38].

The results of the meta-analysis confirmed these findings (see Fig. 3). In total, three studies [16 , 35] were included. The average institutionalization rate of PwD was 33.3% (1061/3189) versus 8.5% (818/6937) for controls. The analysis revealed that PwD had a significantly higher risk (RR: 2.16 CI^95 % 1.31, 3.56) of institutionalization. There was no significant heterogeneity between included studies (p = 0.10). Within the sensitivity analysis, all studies (regardless of their follow-up time) were included, leading to significant heterogeneity (p < 0.01) between studies. Therefore, results cannot be interpreted with certainty (see Supplementary Figure 1).

Fig. 3

Meta-analysis of institutionalization rate 12 months after hospital discharge.

Mortality

13 studies were included in the analysis of the mortality rate of hospitalized PwD. The quality grading of the studies and the length of follow-up differed tremendously (see Table 1).

The mortality rate of PwD six to 60 months after hospital discharge ranged from 17.8% to 83.7% [17 , 40–44]. Five studies [17 , 44] evaluated mortality rates for PwD and controls. Briggs et al. [29] reported in a prospective observational study that the mortality rate of PwD within 6 months after hospital discharge was 25.0%, versus 16.0% in controls. Kimata et al. [40] conducted a study with PwD after acute myocardial infarction and revealed that the mortality rate after an average of eight months was 24.2% for PwD versus 14.6% for people without dementia. However, these findings did not persist after adjustment for several confounders. Furthermore, Tarazona-Santabalbina et al. [37] demonstrated a one-year mortality rate of 37.9% of PwD admitted due to hip-fracture compared to 21.0% in controls. A study in a geriatric hospital in Switzerland conducted by Zekry et al. [44] reported a long-term mortality rate of 63.7% for PwD after 60 months versus 55.8% in patients without dementia. According to Gage et al. [36], a higher number of medications at baseline was significantly associated with a higher mortality of PwD (3.58 medications in survivors versus 4.91 in deaths, p = 0.064). Additionally, the admission out of a nursing home was related to a higher mortality: 71.4% of PwD admitted from a nursing home died versus 31.6% from domiciliary home (p = 0.012) [36]. Van de Vorst et al. reported that male PwD had a higher one-year mortality-risk (RR: 1.26, CI^95 % 1.23–1.28) than female patients. Furthermore, our findings indicate that a higher dementia severity or poorer cognitive function, co-morbidities (i.e., depression), dehydration at admission and dependence in ADL could be relevant as well [28 , 44] (see Table 1).

The meta-analysis revealed that the mortality rate of PwD was more than 1.5 times higher (33.8%, 218/646) than the mortality rate of non-demented controls (19.87%, 238/1189) (see Fig. 4). Thus, PwD had almost twice the risk (RR: 1.74, CI^95 % 1.50, 2.05) of death after hospitalization compared to patients without dementia. There was no significant heterogeneity between included studies (p = 0.58). Within the sensitivity analysis, all studies (regardless of their follow-up time) were included, leading to significant heterogeneity (p < 0.01) between studies. Therefore, results cannot be interpreted with certainty (see Supplementary Figure 2).

Fig. 4

Meta-analysis of mortality rate 12 months after discharge.

DISCUSSION

This systematic review included 20 studies that examined long-term outcomes of PwD after hospitalization. Overall, the number of empirical studies with PwD after hospitalization was low. Nevertheless, PwD had a significantly higher incidence of institutionalization and mortality after hospital discharge than did people without dementia. Results according to readmission rate are inconsistent. Factors associated with the examined adverse outcomes included a higher stage of dementia, an increased number of medications and a higher deterioration in ADL. Nutritional status, male sex, and living alone prior to admission were relevant as well.

Unscheduled readmission

Our findings indicate that readmission rates of PwD differ considerably. Results demonstrated a plausible correlation between follow-up times and evaluated readmission rates for PwD— the longer the time interval, the higher the readmission rate. However, Sganga et al. [45] conducted a study examining 1-year readmission of the elderly with a subgroup of cognitively mild to severely impaired patients (MMSE score <24). The study revealed that mentally well patients were readmitted more frequently than were cognitively impaired patients (odds ratio (OR) 1.76, CI^95 % 1.04–2.83) [45]. There are several potential reasons for this finding. For example, causes for readmission could potentially include pain after surgery or wound healing problems [46]. The progression of dementia disease is accompanied by gradual deterioration in specific cognitive functions, such as the ability to speak (aphasia) and awareness (agnosia) [47, 48]. Thus, it is possible that the readmission rate of PwD is lower due to the insufficient ability to express their mood and pain level (e.g., after surgery). Although PwD share the same age-related morbidity as other older people, they do not have the same access to pain relief as their cognitively-unimpaired counterparts [49]. Malara et al. [50] stated that self-report of cognitively impaired patients is not sufficient to assess pain. Furthermore, Sampson et al. [51] reported that PwD who died during a stay in an acute hospital received fewer palliative medications (including analgesics) than did controls without dementia. In addition, Miu et al. [52] showed that pain of PwD in nursing homes is often under-detected and not treated adequately. Furthermore, our findings revealed that PwD are frequently discharged to a nursing home. Therefore, it is possible that care in residential facilities prevents unscheduled readmission because there the medical staff can prevent outcomes, such as wound infections, more effectively than can informal caregivers. However, these hypotheses are currently unconfirmed.

Most studies examined 28-day or 30-day readmission rates. As case consolidation is applied for patients readmitted within 28 or 30 days, the increased prevalence reported in these studies could potentially be attributed to hospital efforts to avoid additional compensation or fees. We found only one study that examined long-term readmission. Therefore, there is no evidence of a general trend until now. Thus, further studies are needed to evaluate long-term readmission in comparison to people without dementia.

Institutionalization

Our findings indicate that the institutionalization of PwD was higher in comparison to people without dementia. However, institutionalization rates of PwD differed enormously between the examined studies, depending predominantly on follow-up time: the longer the time interval monitored, the higher the institutionalization rate. Premature institutionalization does not seem to be related to the type of hospital (e.g., general hospital, emergency department). Considering the institutionalization of PwD, one should acknowledge that they are generally more often institutionalized even if they have not been admitted to a hospital previously. For example, Greiner et al. [53] reported that community-dwelling PwD are at higher risk of nursing home placement as compared to controls (OR 1.30, CI^95 % 1.18–1.43). In a systematic review, Luppa and colleagues [54] described that PwD had a 16-fold higher chance to be institutionalized (OR 16.70) than people without dementia. Nevertheless, our findings of factors associated with an institutionalization after hospitalization indicate that the presence of dementia has a decisive impact on the discharge destination of PwD.

Since dementia is frequently underdiagnosed [55], it is possible that PwD are institutionalized after hospitalization because dementia was first detected in the hospital. Occasionally, hospital admission reveals self-neglect of PwD, a scenario potentially leading to nursing home admission [56]. Furthermore, hospitals are often not particularly dementia friendly: the new environment, difficulties in communication, lack of special units for PwD, and frequent change of caregivers can deteriorate the state of dementia [5]. Bail et al. [57] revealed that PwD suffer more frequently from preventable complications (i.e., urinary tract infections, pneumonia, delirium) than do patients without dementia. PwD are also at a high risk of in-hospital falls [58], which also may lead to institutionalization.

Literature reviews indicate that the risk of institutionalization depends on the state of dementia: Briggs et al. [29] reported a mean standardized MMSE score of 23 in the dementia group and an institutionalization rate of solely 15% after 12 months. In comparison, Dramé et al. [59] reported a mean MMSE score of 15 and an institutionalization rate of 40%. Tarazona-Santabalbina et al. [37] examined institutionalization rates in relation to cognitive status: 18.3% of patients with mild dementia, 13.2% of patients with moderate dementia, and 40% of patients with severe dementia were institutionalized directly after discharge. Differences in dementia state in the included studies could also explain the large range of reported institutionalization rates. Thus, an evaluation of the dementia severity in a hospital could give information about the needs of PwD after hospitalization.

Furthermore, it is important to try and maintain the personal capabilities of PwD, as worsening in the capability to perform the daily life was associated with institutionalization [28]. The high rate of nursing home placements indicates that improvements in care for PwD in general hospitals could reduce both, premature institutionalization and related costs.

Mortality

Since dementia is an illness of the elderly, the risk of death is already high due to the advanced age. Most included studies conducted their examinations with participants averaging some 80 years of age (see Table 1). According to the WHO, life expectancy in 2015 was 76.8 years in Europe and 76.9 years in the Americas [60]. In addition, dementia as a disease is related to increased mortality [61]. The review revealed that all included studies with a comparison group showed a yet higher mortality of PwD after hospitalization as compared to patients without dementia.

Former studies examining mortality of hospitalized PwD showed similar findings. A review based on hospital administrative data showed that in-hospital mortality, complications during treatment and readmission were all higher for PwD than for the comparison group [62].

Implications for clinicians and policymakers

Our findings indicated that the more advanced the dementia disease, the higher the risk of an adverse outcome. To initiate effective interventions aiming to avoid those outcomes, a group of high-risk patients should be detected at the time of hospital admission. These are PwD with reduced nutritional status, male sex, high number of prescribed medications, and high deficits in ADLs [63]. Early detection of these patients by screening and giving them special attention could improve their outcome after discharge. Furthermore, our findings indicate that an improvement in care for PwD is necessary: hospitalization should be prevented as long as possible, and, if necessary, be organized in a specialized dementia sensitive environment with trained staff. For example, Fabris et al. [32] indicated that PwD benefit from a specific home care program, a scenario which could prevent or delay admission to a general hospital.

Limitations and future research

Studies were performed in different countries all over the world. As a result, there is large heterogeneity of settings, study designs, methods, and results. Discrepancies in health care systems exist and limit comparison between studies. Therefore, only studies published in English and German were included. The restriction to the databases PubMed, CENTRAL, and ScienceDirect (and the non-consideration of Embase) might have caused a bias of identified and finally included studies. Furthermore, advances in medical care have improved the interface between hospital and primary care. Therefore, a restriction of the literature search to the last 17 years seems to be logical and an advantage to reveal the effect of a hospital admission in the current healthcare systems and their treatment and caring options for PwD. However, a more comprehensive search should be an issue in further research to assess changes in adverse outcomes over time. The relative lack of relevant publications led us to include different types of studies. Thus, study samples significantly differ according to socio-demographic and clinical factors, which in particular, bias and limits the validity of presented incidence rate for each focused adverse outcome and the calculated relative risk of PwD compared to non-demented controls. Because of the absence of more studies, it was unfeasible to demonstrate incidence rates or results of meta-analyses separated, for example for stages of dementia severity. We tried to discuss the extracted results in relation to the given socio-demographic and clinical differences between the included studies, demonstrating that adverse outcomes are more likely in patients with a higher comorbidity or higher functional impairment. Furthermore and despite those limitations, demonstrated results indicated a consistent trend of adverse outcomes in PwD after hospitalization. However, according to the existing bias of patients’ parameters, further research is needed to reveal the impact of different socio-demographic and clinical factors of hospitalized PwD on readmission, institutionalization and mortality in more detail.

As both the number of included studies and the quality of those studies were roughly low to moderate, further research is needed. In particular, additional studies with a long follow-up time reporting response rate and drop outs in detail, conducted in such a way as to reduce selection bias, are necessary to confirm our findings and increase robustness of the quantitative estimates. Furthermore, relevant risk factors must be appropriately addressed to avoid premature institutionalization, unscheduled readmission or death of PwD. More intervention studies should examine the benefit of wards specialized on PwD.

Dementia is associated with several other clinical factors that could cause adverse outcomes [29, 64]. Therefore, frailty could as well have an effect on long-term adverse outcomes. Briggs et al. [29] found, that PwD with frailty had not a higher risk of mortality or readmission than PwD without frailty. Nevertheless, further studies including frailty as a covariate that possibly influences the results are necessary. Additionally, the studies included in this review did examine the outcomes of patients with all different types of dementia. Many studies did not examine the type of dementia, e.g., Dramé et al. [38] including PwD with a diagnosis of dementia confirmed by DSM-IV criteria, regardless of the original etiology of the syndrome. Nevertheless, Zekry et al. [44] reported that PwD with vascular dementia had a higher risk of mortality comparing to PwD with Alzheimer’s disease. Thus, further studies should examine the outcomes focusing on different types of dementia. On the other hand, different types of dementia are difficult to discriminate and patients often show mixed forms [64, 65].

Conclusion

Overall, there are currently only few studies evaluating long-term outcomes of PwD after hospitalization. Pertinent studies indicate that PwD are at a high risk of institutionalization and mortality. Major factors associated with the examined outcomes include an advanced stage of dementia, a higher number of medications and deterioration in ADL. The chances of PwD returning to their previous lives after hospital admission are quite low. Thus, according to demonstrated results, it could be concluded that unscheduled hospital admission should be prevented, if possible. But, PwD have many co-morbidities and thus a higher need for medical treatments and care. Therefore, prevention is not achievable in any case. In addition, sometimes is a hospital admission worth to avoid long-term adverse outcomes in PwD. Therefore, better all-round identification and care of PwD, both in the community and acute setting is required. Furthermore, the interface between both settings, starting with PwD discharge, should be improved.

DISCLOSURE STATEMENT

Authors’ disclosures available online (https://www.j-alz.com/manuscript-disclosures/17-1128r3).

Footnotes

The supplementary material is available in the electronic version of this article: .

Appendices

Appendix 1 – PubMed example search strategy

Database: PubMed >2000 to 12th of April 2017

dementia [Title/Abstract] OR alzheimer*[Title/ Abstract] Filters: English German 122102.

dementia [Title/Abstract] OR alzheimer*[Title/ Abstract] AND trial [Title/Abstract] OR study [Title/Abstract]) Filters: English German 50278.

dementia [Title/Abstract] OR alzheimer*[Title/ Abstract] AND trial [Title/Abstract] OR study [Title/Abstract]) AND hospital Filters: English German 11664

dementia [Title/Abstract] OR alzheimer*[Title/ Abstract] AND study [Title/Abstract] OR trial [Title/Abstract] AND hospital AND “Residential Facilities”[Mesh]) OR “Mortality”[Mesh] OR “Institutionalization”[Mesh] OR rehospitalisation OR rehospitalization OR re-hospitalisation OR re-hospitalization OR readmission 775

Appendix 2 – quality criteria sheet

References

Prince

, Comas-Herrera

, Knapp

, Guerchet

, Karagiannidou

(2016) World Alzheimer Report 2016: Improving healthcare for people living with dementia. Coverage, Quality and costs now and in the future. http://www.alz.co.uk/research/WorldAlzheimerReport2016.pdf. Accessed on November 20, 2017.

Phelan

, Borson

, Grothaus

, Balch

, Larson

(2012) Association between incident dementia and risk of hospitalization. JAMA 307, 165-172.

Natalwala

, Potluri

, Uppal

, Heun

(2008) Reasons for hospital admissions in dementia patients in Birmingham, UK, during 2002-2007. Dement Geriatr Cogn Disord 26, 499-505.

Schäufele

(2016) General Hospital Study – GHoSt. Zusammenfassung einer repräsentativen Studie zukognitiven Störungen und Demenz in den Allgemeinkrankenhäusern von Baden-Württemberg und Bayern,http://www.bosch-stiftung.de/content/language1/downloads/Studie_Demenz_im_Akutkrankenhaus.pdf. Accessed on November 19. 2017.

Dewing

, Dijk

(2014) What is the current state of care for older people with dementia in general hospitals? A literature review. Dementia 15, 106-124.

Moyle

, Olorenshaw

, Wallis

, Borbasi

(2008) Best practice for the management of older people with dementiain the acute care setting: A review of the literature. Int J Older People Nurs 3, 121-130.

Mukadam

, Sampson

(2011) A systematic review of the prevalence, associations and outcomes of dementia in older general hospital inpatients. Int Psychogeriatr 23, 344-355.

Pinkert

, Holle

(2012) People with dementia in acute hospitals. Z Gerontol Geriatr 45, 728-734.

Chenoweth

, Kable

, Pond

(2015) Research in hospital discharge procedures addresses gaps in care continuity in the community, but leaves gaping holes for people with dementia: A review of the literature. Australas J Ageing 34, 9-14.

10.

Langer

, Ewers

(2013) “It’s not the old life that we live anymore…”–counselling of relatives in nurse discharge planning. Pflege 26, 311-320.

11.

Leendertse

, Egberts

ACG

, Stoker

, van

den Bemt PMLA

(2008) Frequency of and risk factors for preventable medication-related hospital admissions in the Netherlands. Arch Intern Med 168, 1890-1896.

12.

Kuske

, Graf

, Hartig

, Quasdorf

, Vollmar

, Bartholomeyczik

(2014) Dementia considered? Safety-relevant communication between health care settings: A systematic review. J Public Health 22, 383-393.

13.

Deutsches

Netzwerk für Qualitätsentwicklung in der Pflege

(2009) Auszug aus der abschließenden Veröffentlichung Expertenstandard Entlassungs management in der Pflege Expertenstandard Entlassungsmanagement in der Pflege. http://www.dnqp.de/fileadmin/HSOS/Homepages/DNQP/Dateien/Expertenstandards/Entlassungsmanagement_in_der_Pflege/Entlassung_Akt_Auszug.pdf. Accessed on November 18, 2017.

14.

Kable

, Chenoweth

, Pond

, Hullick

(2015) Health professional perspectives on systems failures in transitional care for patients with dementia and their carers: A qualitative descriptive study. BMC Health Serv Res 15, 567.

15.

Daiello

, Gardner

, Epstein-Lubow

, Butterfield

, Gravenstein

(2014) Association of dementia with early rehospitalization among Medicare beneficiaries. Arch Gerontol Geriatr 59, 162-168.

16.

Callahan

, Tu

, Unroe

, LaMantia

, Stump

, Clark

(2015) Transitions in care in a nationally representative sample of older Americans with dementia. J Am Geriatr Soc 63, 1495-1502.

17.

van

de Vorst IE

, Vaartjes

, Geerlings

, Bots

, Koek

(2015) Prognosis of patients with dementia: Results from a prospective nationwide registry linkage study in the Netherlands. BMJ Open 5, e008897.

18.

Annear

, Tierney

, Vickers

, Palmer

(2016) Counting the cost of dementia-related hospital admissions: A regional investigation. Australas J Ageing 35, E32-E35.

19.

Wübker

, Zwakhalen

SMG

, Challis

, Suhonen

, Karlsson

, Zabalegui

, Soto

, Saks

, Sauerland

(2015) Costs of care for people with dementia just before and after nursing home placement: Primary data from eight European countries. Eur J Health Econ 16, 689-707.

20.

Motzek

, Junge

, Marquardt

(2017) Einfluss der demenz auf verweildauer und erlöse im akutkrankenhaus. Z Gerontol Geriatr 50, 59-59.

21.

Beauchet

, Launay

, Annweiler

, Remondiere

, de

Decker L

(2014) Geriatric inclusive art and risk of in-hospital mortality ininpatients with dementia: Results from a quasi-experimental study. J Am Geriatr Soc 62, 573-575.

22.

Heßler

, Schaufele

, Hendlmeier

, Junge

, Leonhardt

, Weber

, Bickel

(2018) Behavioural and psychological symptoms in general hospital patients with dementia, distress for nursing staff and complications in care: Results of the General Hospital Study. Epidemiol Psychiatr Sci 27, 278-287.

23.

Pluye

, Gagnon

, Griffiths

, Johnson-Lafleur

(2009) A scoring system for appraising mixed methods research, and concomitantly appraising qualitative, quantitative and mixed methods primary studies in Mixed Studies Reviews. Int J Nurs Stud 46, 529-546.

24.

Moher

, Liberati

, Tetzlaff

, Altman

, Group

(2009) Preferred reporting items for systematic reviews and meta-analyses: The PRISMA Statement. PLOS Med 6, e1000097.

25.

The Cochrane Collaboration Review Manager (RevMan) [Computer program].Version 5.2. Copenhagen: The Nordic Cochrane Centre, 2014.

26.

Harvey

, Mitchell

, Brodaty

, Draper

, Close

(2016) Theinfluence of dementia on injury-related hospitalisations andoutcomes in older adults. Injury 47, 226-234.

27.

Goldberg

, Bradshaw

, Kearney

, Russell

, Whittamore

, Foster

PER

, Mamza

, Gladman

JRF

, Jones

, Lewis

, Porock

, Harwood

(2013) Care in specialist medical and mental health unit compared with standard care for older people with cognitive impairment admitted to general hospital: Randomised controlled trial (NIHR TEAM trial). BMJ 347, f4132.

28.

Bradshaw

, Goldberg

, Lewis

, Whittamore

, Gladman

JRF

, Jones

, Harwood

(2013) Six-month outcomes following an emergency hospital admission for older adults with co-morbid mental health problems indicate complexity of care needs. Age Ageing 42, 582-588.

29.

Briggs

, Dyer

, Nabeel

, Collins

, Doherty

, Coughlan

, O’Neill

, Kennelly

(2017) Dementia in the acute hospital: The prevalence and clinical outcomes of acutely unwell patients with dementia. QJM 110, 33-37.

30.

Tropea

, LoGiudice

, Liew

, Gorelik

, Brand

(2017) Poorer outcomes and greater healthcare costs for hospitalised older people with dementia and delirium: A retrospective cohort study. Int J Geriatr Psychiatry 32, 539-547.

31.

Mitchell

, Harvey

, Brodaty

, Draper

, Close

(2016) Hipfracture and the influence of dementia on health outcomes and accessto hospital-based rehabilitation for older individuals. Disabil Rehabil 38, 2286-2295.

32.

Fabris

, Molaschi

, Aimonino

, Ponzetto

, Maero

, Tibaldi

, Nicola

, Varetto

, Barresi

, Cavallero

, Boschis

, Plastino

, Vitale

(2004) Home care for demented subjects: New models of care and home-care allowance. Arch Gerontol Geriatr Suppl 155-162.

33.

Wancata

, Benda

, Meise

(2004) Non-cognitive symptoms of dementia – prevalence and consequences. Psychiatr Prax 31, 346-351.

34.

Seitz

, Gill

, Austin

, Bell

, Anderson

, Gruneir

, Rochon

(2016) Rehabilitation of older adults with dementia after hip fracture. J Am Geriatr Soc 64, 47-54.

35.

McCusker

, Cole

, Abrahamowicz

, Primeau

, Belzile

(2002) Delirium predicts 12-month mortality. Arch Intern Med 162, 457-463.

36.

Gage

, Cheynel

, Williams

, Mitchell

, Stinton

, Katz

, Holland

, Sheehan

(2015) Service utilisation and family support of people with dementia: A cohort study in England. Int J Geriatr Psychiatry 30, 166-177.

37.

Tarazona-Santabalbina

, Belenguer-Varea

, Rovira

Daudi E

, Salcedo

Mahiques E

, Cuesta

Peredo D

, Domenech-Pascual

, Gac

Espinola H

, Avellana

Zaragoza JA

(2015) Severity of cognitive impairment as a prognostic factor for mortality and functional recovery of geriatric patients with hip fracture. Geriatr Gerontol Int 15, 289-295.

38.

Drame

, Lang

P-O

, Jolly

, Narbey

, Mahmoudi

, Laniece

, Somme

, Gauvain

J-B

, Heitz

, Voisin

, de

Wazieres B

, Gonthier

, Ankri

, Saint-Jean

, Jeandel

, Couturier

, Blanchard

, Novella

J-L

(2012) Nursing home admission in elderly subjects with dementia: Predictive factors and future challenges. J Am Med Dir Assoc 13, 83.e17-20.

39.

McCusker

, Cole

, Dendukuri

, Belzile

, Primeau

(2001) Delirium in older medical inpatients and subsequent cognitive and functional status: A prospective study. CMAJ 165, 575-583.

40.

Kimata

, Hirakawa

, Uemura

, Kuzuya

(2008) Absence of outcome difference in elderly patients with and without dementia after acute myocardial infarction. Int Heart J 49, 533-543.

41.

Meier

, Ahronheim

, Morris

, Baskin-Lyons

, Morrison

(2001) High short-term mortality in hospitalized patients with advanced dementia: Lack of benefit of tube feeding. Arch Intern Med 161, 594-599.

42.

Pilotto

, Sancarlo

, Panza

, Paris

, D’Onofrio

, Cascavilla

, Addante

, Seripa

, Solfrizzi

, Dallapiccola

, Franceschi

, Ferrucci

(2009) The Multidimensional Prognostic Index (MPI), based on a comprehensive geriatric assessment predicts short- and long-term mortality in hospitalized older patients with dementia. J Alzheimers Dis 18, 191-199.

43.

Shaw

, Bond

, Richardson

, Dawson

, Steen

, Mckeith

(2003) Multifactorial intervention after a fall in older people. BMJ 326, 1-6.

44.

Zekry

, Herrmann

, Graf

, Giannelli

, Michel

J-P

, Gold

, Krause

K-H

(2011) Mild cognitive impairment, degenerative and vascular dementia as predictors of intra-hospital, short- and long-term mortality in the oldest old. Aging Clin Exp Res 23, 60-66.

45.

Sganga

, Landi

, Volpato

, Cherubini

, Ruggiero

, Corsonello

, Fabbietti

, Lattanzio

, Gravina

, Bernabei

, Onder

(2017) Predictors of rehospitalization among older adults: Results of the CRIME Study. Geriatr Gerontol Int 17, 1588-1592.

46.

Lee

, Daniels

, Wild

JRL

, Wilson

(2017) Readmissions after general surgery: A prospective multicenter audit. J Surg Res 209, 53-59.

47.

Blair

, Marczinski

, Davis-Faroque

, Kertesz

(2007) A longitudinal study of language decline in Alzheimer’s disease and frontotemporal dementia. J Int Neuropsychol Soc 13, 237-245.

48.

DS3-Leitlinie Demenzen. Deutsche Gesellschaft für Neurologie,http://www.dgn.org/images/red_leitlinien/LL_2016/PDFs_Download/038013_LL_Demenzen_2016.pdf. Accessed November 19, 2017.

49.

Regan

, Colling

, Tapley

(2015) Pain management: A fundamental component of dementia care. Nurs Stand 30, 43-50.

50.

Malara

, De

Biase GA

, Bettarini

, Ceravolo

, Di

Cello S

, Garo

, Praino

, Settembrini

, Sgro

, Spadea

, Rispoli

(2016) Pain assessment in elderly with behavioral and psychological symptoms of dementia. J Alzheimers Dis 50, 1217-1225.

51.

Sampson

, Gould

, Lee

, Blanchard

(2006) Differences in care received by patients with and without dementia who died during acute hospital admission: A retrospective case note study. Age Ageing 35, 187-189.

52.

Miu

DKY

, Chan

(2014) Under-detection of pain in elderly nursing home residents with moderate to severe dementia. J Clin Gerontol Geriatr 5, 23-27.

53.

Greiner

, Qualls

, Iwata

, White

, Molony

, Sullivan

, Burke

, Schulman

, Setoguchi

(2014) Predicting nursing home placement among home- and community-based services program participants. Am J Manag Care 20, e535-e536.

54.

Luppa

, Luck

, Weyerer

, Konig

H-H

, Brahler

, Riedel-Heller

(2010) Prediction of institutionalization in the elderly. Asystematic review. Age Ageing 39, 31-38.

55.

Connolly

, Gaehl

, Martin

, Morris

, Purandare

(2011) Underdiagnosis of dementia in primary care: Variations in the observed prevalence and comparisons to the expected prevalence. Aging Ment Health 15, 978-984.

56.

Papaioannou

E-SC

, Räihä

, Kivelä

S-L

(2012) Self-neglect of the elderly. An overview. Eur J Gen Pract 18, 187-190.

57.

Bail

, Berry

, Grealish

, Draper

, Karmel

, Gibson

, Peut

(2013) Potentially preventable complications of urinary tract infections, pressure areas, pneumonia, and delirium in hospitalised dementia patients: Retrospective cohort study. BMJ Open 3, e002770.

58.

Brand

, Sundararajan

(2010) A 10-year cohort study of the burden and risk of in-hospital falls and fractures using routinely collected hospital data. Qual Saf Health Care 19, e51.

59.

Dramé

, Lang

P-O

, Jolly

, Narbey

, Mahmoudi

, Lanièce

, Somme

, Gauvain

J-B

, Heitz

, Voisin

, deWazières

, Gonthier

, Ankri

, Saint-Jean

, Jeandel

, Couturier

, Blanchard

, Novella

J-L

(2012) Nursing home admission in elderly subjects with dementia: Predictive factors and future challenges. J Am MedDir Assoc 13, 83.e17-83.e20.

60.

Life expectancy Data by WHO region http://apps.who.int/gho/data/view.main.SDG2016LEXREGv?lang=en. Accessed on November 17, 2017 .

61.

Andersen

, Lolk

, Martinussen

, Kragh-Sorensen

(2010) Very mild to severe dementia and mortality: A 14-year follow-up - The Odense study. Dement Geriatr Cogn Disord 29, 61-67.

62.

Rao

, Suliman

, Vuik

, Aylin

, Darzi

(2016) Outcomes of dementia: Systematic review and meta-analysis of hospital administrative database studies. Arch Gerontol Geriatr 66, 198-204.

63.

Houghton

, Murphy

, Brooker

, Casey

(2016) Healthcare staffs’ experiences and perceptions of caring for people with dementia in the acute setting: Qualitative evidence synthesis. Int J Nurs Stud 61, 104-116.

64.

Kulmala

, Nykanen

, Manty

, Hartikainen

(2014) Association between frailty and dementia: A population-based study. Gerontology 60, 16-21.

65.

Zekry

, Hauw

J-J

, Gold

(2002) Mixed dementia: Epidemiology, diagnosis, and treatment. J Am Geriatr Soc 50, 1431-1438.

66.

WHO Media centre fact sheet dementia http://www.who.int/mediacentre/factsheets/fs362/en. Accessed on January 25, 2018.