Risk of Hospitalization in Patients with Alzheimer’s Disease and Lewy Body Dementia: Time to and Length of Stay

Abstract

Background:

Patients with dementia are at high risk of being hospitalized, but there is little knowledge whether this applies to all forms of dementia.

Objective:

To investigate if there are differences in hospitalization between patients with Alzheimer’s disease (AD) and Lewy body dementia (LBD), and further, to compare admission rate with the general age-matched population.

Methods:

Patients (age 75.7±7.4) recently diagnosed with mild form of AD (n = 110) or LBD (n = 91) were included from outpatient clinics. The participants were followed from time of diagnosis, for five years or until death. Study outcomes were time to first hospitalization after diagnosis, number of admissions, total number of hospital days, and length of stay. Age-standardized admission ratios were calculated. Time to first admission was analyzed using competing risks regression models, and differences in number of hospitalizations and hospital days were addressed using negative binomial regression models.

Results:

More than 77% of the patients were admitted, largely as unplanned hospitalizations. Patients with LBD had significantly shorter time until first hospitalization (median 1.28 years, 95% CI 0.93–1.67 versus AD: 2.32 years, 95% CI 1.74–3.31) and more days in hospital (median 13 days, IQR 4, 38), than patients with AD (7 days, IQR 0, 18).

Conclusion:

Our data indicates that patients with LBD have shorter time until first admission and higher admission rate than AD patients. This imposes a great burden on patients, their family, and the health care system. More knowledge about hospital admissions of people with dementia is needed. Future studies should investigate strategies to avoid potentially preventable admissions.

Keywords

Alzheimer’s disease dementia hospitalization length of stay Lewy body dementia patient admission

Introduction

Hospitalization may constitute a major burden on patients with dementia as they might experience it as stressful and noisy [1]. In addition, the ability to communicate may be impaired, which increases the risk of specific needs being undertreated. Dementia itself is considered a risk factor for hospitalization [2], and a study from the UK found that patients with dementia had over two times more unplanned hospitalization than age-matched controls without dementia [3]. Lewy body dementia (LBD) is an umbrella term consisting of dementia with Lewy bodies (DLB) and Parkinson’s disease dementia (PDD) [4]. DLB and PDD share many clinical and neuropathological features [5], and distinction between DLB and PDD is difficult as the diseases progresses [4]. LBD is believed to have a poorer prognosis than AD [6], and previous studies have found patients with LBD to have shorter survival time after diagnosis [7, 8], faster admission to nursing homes [9], and to have a higher caregiver burden [10], compared to patients with AD. In one of the few studies including a relatively large group of patients with DLB [11], DLB patients were more frequently hospitalized and stayed hospitalized longer than patients with AD. So far, most dementia studies have focused on dementia in general, or just included patients with AD [2]. Knowledge about hospitalization in other forms of dementia is scarce. Therefore, the aims of this study were to investigate if there are differences in hospitalization between patients with AD and LBD, and to compare the admission rates with that of the general age-matched population in the catchment area. Key outcomes were time to first hospitalization after diagnosis of dementia, number of admissions, total number of hospital days, and length of stay. Based on previous studies, we hypothesize that patients with dementia will have a higher admission rate than patients from the general population. Further, patients with LBD will have shorter time until first admission and more hospital days compared with AD.

MATERIALS AND METHODS

Study cohort

Data were obtained from the DemVest study, a longitudinal cohort study on dementia in Western Norway. Patients with mild dementia were included from five outpatient clinics in geriatric medicine and psychiatry in the counties of Hordaland and Rogaland. Mild dementia was defined as having a Mini-Mental State Examination (MMSE) [12] score ≥20 or a Clinical Dementia Rating scale (CDR) equal to one or lower [13]. In total, 266 patients have completed the baseline assessment. The inclusion period went over two periods. From March 2005 to March 2007, all consecutive referrals with first time diagnosis of dementia were asked to participate. From April 2007 until 2013, only consecutive referrals with DLB or PDD were included. This was done to increase sample sizes for these diagnoses. A thorough description of the inclusion has been published earlier [14]. In this present study, 201 patients are included. Other forms of dementia (n = 30) were excluded. Two patients had too severe dementia at baseline (MMSE < 20), six patients did not have dementia after five years, and three patients lacked information about baseline MMSE or dementia diagnosis. In addition, twenty-four patients did not want to participate further.

Dementia diagnosis and assessments

Dementia was diagnosed according to the Diagnostic and Statistical Manual for Mental Disorders, 4th edition (DSM-IV). The consensus criteria for dementia with AD (NINCDS-ADRDA), DLB, and PDD were used to set the final clinical diagnosis of dementia [15 –17]. Three specialists in geriatric medicine and geriatric psychiatry independently applied the diagnostic criteria. In cases of discrepancy, the final diagnosis was based on consensus. DLB and PDD share clinical and pathological features [18], and they were therefore combined and referred to collectively as LBD in this study. At each follow-up, the diagnosis was re-evaluated and this study used final consensus diagnoses from 2017. Here 56 pathological diagnoses were included, which matched the clinical diagnoses of AD and LBD in more than 80% of the cases [19].

Patients were followed annually, with structured assessments for demographic and clinical data. Baseline and the follow-up examinations included the following instruments: MMSE [12] and the global CDR [13] was applied to measure the cognitive impairment and to stage the degree of dementia, respectively. The global score of CDR is rated from zero (no dementia) to three (severe dementia). Neuropsychiatric Inventory (NPI) consist of 12 items and was used to assess the total burden of neuropsychiatric symptoms [20]. We used the frequency times severity score for each item which were summed up to give a total sum score of maximum 144. The total score from the Montgomery Aasberg Depression Rating Scale (MADRS) was used to grade depression [21]. Comorbidity at baseline was assessed retrospectively using the Cumulative Illness Rating Scale (CIRS) [22]. The total CIRS score (range 0–52) is the sum of all 13 items, which represents individual body systems. Each item is rated from zero (no impairment) to four (extremely severe impairment). Dementia is rated in the psychiatry item. To avoid adjusting for dementia severity during the social and clinical analyses, we excluded the psychiatry item when assessing the confounding effect of comorbidity on relation between AD/LBD and the outcomes.

Outcomes

Information about hospitalization in the study cohort during the study period were extracted from the electronic patient record system used in the Western Norway health region. This area consists of two university hospitals and some local hospitals and serves approximately one million people. Planned admissions with at least one overnight stay and all unplanned admission were extracted. To be able to estimate the total hospitalization load, admissions from both psychiatric and somatic units were included. Information included date of admission, date of discharge, and information about whether the admission was planned or unplanned. Length of hospital stay was defined as time between date of admission to hospital and date of discharge from hospital. The number of hospital days per person is the number of hospital days a person has in a given period.

Age-standardized admission ratios, defined as the ratio between observed and expected number of admissions, were estimated for planned and unplanned admissions for the first 12 months after dementia diagnosis. The general age-matched population in the catchment areas served as a comparison group. All admissions made by the general population in the study period were extracted from the same electronic patient record system.

Statistical analyses

Statistical analyses were performed using IBM SPSS Statistics version 24 unless otherwise specified. Baseline and demographic variables are presented as counts and percentages of categorical variables, means and standard deviations (SD) for symmetrically distributed continuous variables, and medians and interquartile ranges (IQR) for skewed continuous variables. Comparison between AD and LBD were performed using Pearson Chi-square tests, alternatively Fisher’s exact test, independent samples t-test, or Mann-Whitney U tests as appropriate. Admission rates per year for the general population were calculated using the number of admissions from the general population together with population level data in catchment area [23]. The admission rate per year for the general population was then multiplied with the number of patients in the study cohort eligible for admission that specific year to yield the expected number of admissions. This was done for all years in the study period (2005-2013) to retrieve the age-standardized admission ratios for the first 12 months after inclusion. Differences between AD and LBD were further explored in regression analyses: For the number of hospitalizations and the number of hospital days during one and five years from inclusion, negative binomial regression was used. Incidence rate ratios (IRRs) are presented with 95% confidence intervals (CI) and Wald tests of effects, with and without adjustments for demographic and clinical variables. Cumulative incidences of hospitalization and death during the first five years were estimated using the non-parametric Aalen-Johansen estimator and plotted for AD and LBD separately. Time to first hospitalization was analyzed with Fine-Gray models, allowing for competing risk of dying before first hospitalization [24]. These results are presented as sub-distribution hazard ratios (SDHR) with 95% CI and Wald tests of effects, and was performed with and without adjustment for demographic variables, somatic comorbidities, dementia-related variables, and medication use. Predicted cumulative incidence curves from the unadjusted model were compared with the Aalen-Johansen estimator for assessment of model fit [25]. Predicted cumulative incidences at specific time points and median times to first hospitalization were based on the same model and presented with 95% CI. The competing risks analyses were performed in R version 3.4 with functions finegray, survfit, and coxph from the survival package. As Supplementary results, we present analyses of hospitalization rates using negative binomial models with an offset given as the logarithm of time under observation; and cause-specific hazard ratios (CSHR) obtained by treating death as a non-informative censoring event [24].

Ethical issues

The regional ethics committee approved the study (REK ID 2010/633). The study procedure was explained in detail and written consent was obtained from all participants or next of kin.

RESULTS

Baseline characteristics

The cohort statistics of the 201 included patients are shown in Table 1. The AD and LBD patients were similar regarding to age, cognitive impairment and dementia stage, living condition, and smoking status, but patients with LBD were more often men, had higher comorbidity, more neuropsychiatric symptoms, including depression, and more often used antipsychotic drugs. They also reported longer duration of dementia symptoms before diagnosis than patients with AD.

Table 1

Descriptive characteristics of study cohort at baseline

Demographic and clinical characteristics	Total (n = 201)	AD (n = 110)	LBD (n = 91)	p
Female, n (%)	121 (60.2)	79 (71.8)	42 (46.2)	<0.001^*
Age at baseline in years, mean (SD)	75.7 (7.4)	75.5 (7.4)	75.8 (7.4)	0.79^†
Living with partner, n (%)⁽ⁿ^= 199)	114 (56.7)	58 (53.2)	56 (62.2)	0.20^*
Years of education⁽ⁿ^= 195)	9 (7, 11)	9 (7, 11)	9 (7, 11)	1.00
Smoking at baseline, n (%)	35 (17.4)	22 (20.4)	13 (15.3)	0.45^*
Duration of dementia symptoms in years⁽ⁿ^= 191)	2 (1, 3)	2 (1, 3)	3 (2, 4)	0.006
CIRS total score	5 (4, 7)	5 (4, 7)	6 (4, 9)	0.014
MMSE total score	24 (22, 26)	24 (22, 25)	24 (22, 26)	0.39
CDR global score ≤1, n (%)⁽ⁿ^{= 183,AD = 107,LBD = 76)}	172 (94.0)	104 (97.2)	68 (89.5)	0.07^*
NPI total score⁽ⁿ^= 198)	13 (4, 29)	9 (2, 25)	17 (7, 35)	0.003
MADRS total score⁽ⁿ^= 196)	6 (6, 11)	5 (2, 10)	8 (4, 13)	0.003
Medication use the first year after inclusion
Number of medications	4 (2, 6)	4 (2, 5)	4 (3, 6)	0.11
Polypharmacy (≥5 medications), n (%)	84 (41.8)	41 (37.3)	43 (47.3)	0.15^*
Use of anti-dementia, n (%)	151 (75.1)	88 (80)	63 (69.2)	0.08^*
Use of antidepressant, n (%)	96 (47.8)	53 (48.2)	43 (47.2)	0.90^*
Use of antipsychotics, n (%)	32 (15.9)	9 (8.2)	23 (25.3)	0.001^*

Statistics presented as median (IQR) and compared using Mann-Whitney U test unless otherwise noted. IQR, Interquartile range, given as (Q1, Q3); SD, standard deviation; CIRS, Cumulative illness rating scale, 0–52, where high scores indicate more illness; MMSE, Mini-Mental Status Examination, 0–30, where high scores indicate better cognitive function; CDR, Clinical Dementia Rating Scale (0–3), where high score indicates more severe dementia; NPI, Neuropsychiatric inventory, 0–144, where high scores indicate more neuropsychiatric symptoms; MADRS, Montgomery and Aasberg depression rating scale, 0–60, where high scores indicate more depressive symptoms. ^*p-value from Chi square test. ^†p-value from Independent samples t-test.

Hospitalizations after dementia diagnosis

A summary of the hospitalization data is presented in Table 2. Over five years, more than 77% of the patients were admitted, and in total, they had 418 admissions. The majority of admissions were unplanned; 73.0% and 81.3% within one and five years, respectively. Compared to AD, patients with LBD had more unplanned admissions and more unplanned hospital days, at one and five years. In addition, LBD patients had longer length of stay per unplanned admission after one year, with median 7 (IQR 2, 14) versus 2 (1, 6) days. No significant differences were detected in planned admissions between the two patient groups. In the regression analyses (Table 3), LBD patients had significantly more admissions, both planned and unplanned, and 84% more hospital days than AD patients (IRR 1.84, 95% CI 1.22, 2.77, p = 0.004), within five years after dementia diagnosis. After adjusting for demographic variables and somatic comorbidities, the difference in hospital days for LBD versus AD patients was reduced to 63% (IRR 1.63, 1.02, 2.61, p = 0.040). None of the other differences remained statistically significant. After adjusting for dementia-related variables, the difference in hospital days was further reduced and were no longer statistically significant (IRR 1.53, 0.92, 2.52, p = 0.099). When time under observation was accounted for, the unadjusted IRR were larger for all outcomes, and differences in all hospital admissions remained significant in the fully adjusted model (IRR 1.42, 1.02, 1.98, p = 0.0369). See Supplementary Table 1.

Table 2

Hospitalization in patients with Alzheimer’s disease and Lewy body dementia

	One year after diagnosis				Five years after diagnosis
	Outcome	Total (n = 201)	AD (n = 110)	LBD (n = 91)	p	Total (n = 201)	AD (n = 110)	LBD (n = 91)	p
Patients admitted, n (%)	73 (36.3)	31 (28.2)	42 (46.2)	0.021	156 (77.6)	76 (69.1)	80 (87.9)	<0.001
Hospitalizations, n (%)
Total	126	59	67		418	190	228
Planned	34 (27.0)	17 (28.8)	17 (25.4)		78 (18.7)	33 (17.4)	45 (19.7)
Unplanned	92 (73.0)	42 (71.2)	50 (74.6)		340 (81.3)	157 (82.6)	183 (80.3)
Number of admissions per person, median (IQR)
Total	0 (0, 1)	0 (0, 1)	0 (0, 1)	0.018	2 (1, 3)	1 (0, 3)	2 (1, 3)	0.001
Planned	0 (0, 0)	0 (0, 0)	0 (0, 0)	0.48	0 (0, 1)	0 (0, 0)	0 (0, 1)	0.15
Unplanned	0 (0, 1)	0 (0, 0)	0 (0, 1)	0.013	1 (0, 2)	1 (0, 2)	2 (1, 3)	0.004
Number of hospital days per person, median (IQR)
Total	0 (0, 5)	0 (0, 2)	0 (0, 8)	0.003	9 (0, 14)	7 (0, 18)	13 (4, 38)	<0.001
Planned	0 (0, 0)	0 (0, 0)	0 (0, 0)	0.46	0 (0, 3)	0 (0, 0)	0 (0, 6)	0.21
Unplanned	0 (0, 2)	0 (0, 0)	0 (0, 7)	0.002	4 (0, 14)	2 (0, 11)	7 (2, 26)	0.001
Length of stay in days per admission, median (IQR)
Total	5 (2, 14)	4 (2, 10)	7 (2, 21)	0.15	4 (2, 10)	4 (1, 10)	4 (2, 12)	0.29
Planned	14 (2, 24)	14 (2, 22)	6 (2, 32)	0.52	14 (2, 23)	14 (5, 22)	7 (2, 24)	0.23
Unplanned	4 (2, 10)	2 (1, 6)	7 (2, 14)	0.006	3 (1, 8)	3 (1, 7)	3 (2, 9)	0.06

Statistics presented as median (IQR) and compared using Mann-Whitney U test. IQR, interquartile range, given as (Q1, Q3); AD, Alzheimer’s disease; LBD, Lewy body dementia.

Table 3

Negative binomial model comparing counts of admissions and hospital days in patients with Lewy body dementia with those of Alzheimer’s disease

	Unadjusted (n = 201)		Adjusted for demographics^* (n = 193)		Adjusted for demographics and somatic comorbidity^† (n = 188)		Adjusted for demographics, somatic comorbidity and dementia-related variables (n = 185)
Outcome	IRR (95% CI)	p	IRR (95% CI)	p	IRR (95% CI)	p	IRR (95% CI)	p
Within one year after dementia diagnosis
All hospital admissions	1.37 (0.86, 2.20)	0.19	1.08 (0.64, 1.81)	0.79	1.14 (0.68, 1.92)	0.61	1.13 (0.65, 1.95)	0.66
Planned admissions	1.21 (0.56, 2.63)	0.63	1.17 (0.50, 2.75)	0.72	1.29 (0.54, 3.11)	0.57	1.09 (0.44, 2.75)	0.85
Unplanned admissions	1.44 (0.85, 2.43)	0.17	1.07 (0.67, 1.72)	0.78	1.04 (0.65, 1.66)	0.87	1.05 (0.64, 1.71)	0.85
Days in hospital	2.32 (1.01, 5.34)	0.047	1.93 (0.77, 4.85)	0.16	2.32 (0.87, 6.18)	0.09	2.21 (0.72, 6.78)	0.17
Within five years after dementia diagnosis
All hospital admissions	1.45 (1.11, 1.90)	0.007	1.27 (0.96, 1.68)	0.10	1.25 (0.94, 1.66)	0.12	1.22 (0.91,.1.63)	0.19
Planned admissions	1.65 (1.04, 2.62)	0.035	1.63 (0.91, 2.90)	0.10	1.72 (0.96, 3.11)	0.07	1.46 (0.78, 2.71)	0.24
Unplanned admissions	1.41 (1.05, 1.89)	0.022	1.21 (0.89, 1.62)	0.22	1.16 (0.86, 1.56)	0.33	1.17 (0.86, 1.58)	0.33
Days in hospital	1.84 (1.22, 2.77)	0.004	1.59 (1.01, 2.52)	0.048	1.63 (1.02, 2.61)	0.040	1.53 (0.92, 2.52)	0.099

IRR, incidence rate ratio; CI, confidence interval; CIRS, Cumulative Illness Rating Scale; NPI, Neuropsychiatric inventory; MMSE, Mini-Mental Status Examination; MADRS, Montgomery Aasberg Depression Rating Scale. ^*Demographics: Age, sex, education, social status. ^†Somatic comorbidities: CIRS minus psychiatric dimension, smoking. ^‡Dementia related variables: CIRS psychiatric dimension, NPI, MMSE, and MADRS.

Hospitalization compared to the general population

Age-standardized admission ratio the first year after diagnosis showed differences between patients with dementia and the general population. Patients with LBD in the age group 60–74 years had almost four times more admissions (unplanned: 3.94, 95% CI 2.48–5.75; planned: 3.66, 1.98–6.22) than the general population in the same age-group. Patients with AD had about 30% more admissions than the general population. This applied to both planned and unplanned admissions, see Fig. 1.

Fig.1

Age-standardized admission ratio for people with Alzheimer’s disease and Lewy body dementia compared to the general population in the catchment area, 12 months after inclusion in study (Somatic and psychiatric hospitalizations).

Time to first hospitalization

Figure 2 shows cumulative hospitalization incidences during five years after dementia diagnosis. Patients with LBD had significantly shorter median time until first hospitalization (1.28 years, 95% CI 0.93, 1.67) compared to patients with AD (2.32 years, 1.74, 3.31). Predicted cumulative incidences of hospitalizations for one to five years from inclusion are given for both patient groups in Table 4, where it can be seen that within five years, 87% (95% CI 80–92%) of LBD patients will be expected to have been hospitalized at least once, as compared with 69% (61–78%) of AD patients. Correspondingly, unadjusted analysis showed that patients with LBD had increased risk of being hospitalized at any time after inclusion (SDHR 1.72, 95% CI 1.25, 2.35, p < 0.001), see Table 5. The difference in risk was reduced after adjustment for demographics and somatic comorbidities (SDHR 1.42, 1.01, 2.00, p = 0.046) and after further adjustment for dementia related variables (SDHR 1.37, 0.96, 1.95, p = 0.09). When treating death as a censoring event, i.e., when considering those patients that were still alive at each time point, the LBD diagnosis was associated with 86% higher hazard of being hospitalized as compared with AD (CSHR 1.86, 95% CI 1.35, 2.56, p < 0.001), see Supplementary Table 2. After adjusting for demographics, somatic comorbidities, and dementia-related variables, the CSHR of hospitalization was still higher for patients with LBD, now borderline non-significant (CSHR 1.42, 95% CI 0.99, 2.03, p = 0.054). Adjustments for medication variables did not substantially affect the estimated SDHR or the CSHR.

Fig.2

Observed cumulative incidences of hospitalization and death during five years after dementia diagnosis. AD, Alzheimer’s disease; LBD, Lewy body dementia.

Table 4

Predicted cumulative incidences of hospitalization

	AD	LBD
Time	CIF (95% CI)	CIF (95% CI)
1 year	0.30 (0.23, 0.37)	0.45 (0.37, 0.55)
2 years	0.47 (0.39, 0.56)	0.66 (0.57, 0.75)
3 years	0.58 (0.49, 0.67)	0.77 (0.69, 0.85)
4 years	0.65 (0.56, 0.74)	0.83 (0.76, 0.90)
5 years	0.69 (0.61, 0.78)	0.87 (0.80, 0.92)

AD, Alzheimer’s disease; LBD, Lewy body dementia; CIF, cumulative incidence function; CI, confidence interval.

Table 5

Competing risks – assessing risk of first hospitalization in patients with Lewy body dementia with those of Alzheimer’s disease

Adjustment variables	n (no. events)	SDHR (95% CI)	p	Concordance index (SE)	p for interaction
Unadjusted	201 (156)	1.72 (1.25, 2.35)	<0.001	0.563 (0.021)
Adjusted for demographics^*	193 (149)	1.48 (1.06, 2.06)	0.021	0.591 (0.025)
Adjusted for demographics and somatic comorbidities^†	188 (144)	1.42 (1.01, 2.00)	0.046	0.614 (0.026)
Adjusted for demographics, somatic comorbidities	185 (142)	1.37 (0.96, 1.95)	0.086	0.620 (0.026)
and dementia related variables^‡
Adjusted for the above plus, respectively:
polypharmacy	185 (142)	1.36 (0.95, 1.95)	0.089	0.619 (0.026)
no. of medicines	185 (142)	1.36 (0.95, 1.94)	0.092	0.618 (0.026)
use of antidepressants	185 (142)	1.37 (0.96, 1.96)	0.082	0.619 (0.027)	0.57
use of anti-dementia	185 (142)	1.37 (0.96, 1.97)	0.082	0.620 (0.027)	0.23
use of antipsychotics	185 (142)	1.38 (0.96, 1.98)	0.081	0.621 (0.026)	0.59

SDHR, sub distributional hazard ratio; CI, confidence interval; CIRS, Cumulative Illness Rating Scale; NPI, Neuropsychiatric Inventory; MMSE Mini-Mental Status Examination; MADRS, Montgomery Aasberg Depression Rating Scale. ^*Demographics: Age, sex, education, social status. ^†Somatic comorbidities: CIRS minus psychiatric dimension, smoking. ^‡Dementia related variables: CIRS psychiatric dimension, NPI, MMSE, and MADRS.

DISCUSSION

This is one of few studies comparing hospitalization in patients with AD and LBD. Three out of four patients had at least one hospitalization during the study period and the majority of the admissions were unplanned admissions. Patients with LBD had significantly shorter time until first hospitalization, longer length of stay per unplanned admission, and more unplanned hospital days during follow-up than patients with AD. Patients in the age-group 60–74 years with LBD had almost four times as many hospital admissions during the first year after diagnosis than the age-matched population in the same catchment area. Differences in hospitalization was partly explained by gender differences and more somatic comorbidity among the LBD patients.

Time to first admission

We found the median time to first admission after diagnosis to be 1.28 years for patients with LBD and 2.32 years for patients with AD. A longitudinal patient registry study from North America found median time to first admission to be 2.2 years for patients with AD [26], which is in between our findings for AD and LBD patients, though the American study only included unplanned admissions. Mueller et al.’s study [11] included both patients with AD and DLB. The direction of their findings corresponds to our findings with DLB (0.89 years±1.15) having shorter time until first admission after diagnosis compared to AD patients (1.67 years±1.62) [11].

Length of hospitalization (per admission and per person)

Results regarding length of hospitalization can be reported in many different ways. First, we reported number of days per admission, showing that patients with LBD had significantly longer length of stay during unplanned admissions than patients with AD, during the first year. Secondly, we compared median number of hospital days per person. Here we found that patients with LBD had significantly more hospital days, both during the first year (p = 0.003) and after five years (p < 0.001) compared to patients with AD. Three other studies have been found to assess the differences in inpatient stay between patients with AD and LBD [11 , 28]. Mueller et al. [11] included 194 DLB patients which were matched in according to age, gender, and MMSE. They reported their findings as mean number of hospital days per person-year, and found that patients with DLB (10.8 hospital days per person-year) had almost four hospital days more than patients with AD (6.9 hospital days per person-year). Murman et al. [27] found that patients with DLB had 10.3±17.4 hospital days per year, which was significantly more days than patient with AD (1.6±4.9 hospital days per year). A limitation with this study was the low number of patients with DLB (n = 15). Bostrom et al. [28] included 34 patients with DLB who were matched by age, gender, and MMSE to 34 patients with AD. They found that patients with DLB used more resources in total, but that patients with AD had more hospital days.

Although, these studies have reported their findings differently, our findings support the previous findings from Mueller et al. [11] and Murman et al. [27] suggesting that patients with LBD have a poorer prognosis in relation to hospitalization when compared with AD patients.

Age-standardized admission ratio compared to the general population

This study found that AD and LBD patients had a higher planned and unplanned admission ratio compared to an age-matched population. To our knowledge, only Mueller et al. [11] has reported similar findings. They found the unplanned admission ratio for AD and LBD to be 0.91 and 1.22, respectively [11]. The unplanned admission ratio for LBD patients was in the same direction while the unplanned admission ratio for patients with AD (0.91) was in opposite direction, compared to our findings. This difference might be explained by patients included in our study was removed from the general catchment population. Still, there is a potential risk that the general population could include patients with dementia as dementia is frequently undiagnosed [29]. Further, when we differentiated on age, our findings indicated that the greatest difference regarding hospitalization between patients with dementia and the general population was in patients between 60–74 years with LBD. A possible explanation for this might be that this age-group is usually with good health and therefore use the health care system less often [30]. In relation to this, a study reported that patient with LBD were frequently discharged to a higher care level than before admission [31], which may also apply for patients with AD. Increased knowledge about admissions is therefore of importance.

Adjustment for other risk factors related to hospitalization

Comorbidity, higher age, and being male have previously been found to be associated with increased risk of hospitalization in patients with dementia [3, 26]. The included LBD group had more comorbidities and more males than the AD group, and we saw that these differences did explain some of the observed differences in admission rate (Table 3) and risk of first hospitalization (Table 5). In accordance with a Scottish study [32], increased neuropsychiatric symptoms among LBD patients also explained some of the differences (Table 5). Use of antipsychotics is not recommended in patients with LBD due to increased risk of severe neuroleptic sensitivity [16] and polypharmacy is found to increase the risk of unplanned hospital admissions in patients with dementia [33]. However, we were not able to confirm any effects of use of antipsychotics or other psychotropic drugs in our study.

Strengths and limitations

Strengths of this study is the long follow-up time, the continuous evaluation of diagnosis, and the high diagnostic accuracy, which have been confirmed by autopsy [19]. An additional strength is the applied time-to-event analyses. They were performed allowing for competing risk of death, thus the results reflect the cumulative incidences of hospitalizations observed in the community better than an analysis treating death as independent censoring events [24]. Correspondingly, we did not account time under observation in the analysis of hospitalization models, as this was not relevant for the predicted load of hospitalizations from the patient groups over time. However, alternative analyses are presented in the Supplementary Material. This is a referral cohort, which might be a limitation. To limit this bias, the general practitioners were contacted in advance and encouraged to refer all patients with suspected dementia. Still, there might be a possibility that only patients with more complex symptoms were referred to the outpatient clinic. To provide an overview of the hospitalization load from patients with dementia, both somatic and psychiatric admissions were included. However, psychiatric admissions may be longer than somatic admissions and may skew the outcome. Additional sub-analyses were performed indicating that the higher number of unplanned hospital days for patients with LBD was related to unplanned psychiatric admissions. This study used time of diagnosis as starting point for reporting time to first hospitalization. This can be a limitation, especially for patients with DLB. Caregivers have reported that it may take over a year for patients with DLB to be properly diagnosed, and many are often given other diagnoses first [34]. However, this is considered a better starting point than symptom onset, as this may be limited by caregiver or patients recall of retrospective memory and sensitivity towards recognizing symptoms [35]. Finally, although being one of the largest LBD samples followed longitudinally in a single study, a larger sample size would have permitted analysis of DLB and PDD separately, and further stratification between age and gender would probably been possible. However, additional analyses showed significant difference in time to first admission when DLB patients were compared to AD, and with no significant differences between DLB and PDD.

Conclusion

Our data indicates that patients with dementia have more frequent hospitalization than the general population, and that LBD patients have more hospitalizations and days in hospital compared to AD patients. This imposes a great burden on patient and their family and the health care system. However, more studies are necessary to evaluate this fully. Information about hospitalization is important for healthcare providers to be able to plan care services and to provide adequate and relevant information to patients and families as the dementia progresses. Future studies should investigate frequent causes of possibly preventable admissions in patients with dementia and LBD, and develop strategies to target these causes in order to reduce the frequency and duration of potentially preventable hospitalizations.

Footnotes

ACKNOWLEDGMENTS

The authors wish to thank the DemWest Study Group and all the patients who participated in this study. The Resource Council of Norway (Grant 213375) supported this work.

Dag Aarsland is partly funded by the National Institute for Health Research (NIHR) Biomedical Research Centre at South London and Maudsley NHS Foundation Trust and King’s College London. The views expressed are those of the author and not necessarily those of the NIHR or the Department of Health and Social Care.

Authors’ disclosures available online ().

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

References

Cowdell

(2010) Care of older people with dementia in an acute hospital setting. Nurs Stand 24, 42–48.

Shepherd

, Livingston

, Chan

, Sommerlad

(2019) Hospitalisation rates and predictors in people with dementia: A systematic review and meta-analysis. BMC Med 17, 130.

Sommerlad

, Perera

, Mueller

, Singh-Manoux

, Lewis

, Stewart

, Livingston

(2019) Hospitalisation of people with dementia: Evidence from English electronic health records from 2008 to 2016. Eur J Epidemiol 34, 567–577.

Walker

, Possin

, Boeve

, Aarsland

(2015) Lewy body dementias. Lancet 386, 1683–1697.

Sezgin

, Bilgic

, Tinaz

, Emre

(2019) Parkinson’s disease dementia and Lewy body disease. Semin Neurol 39, 274–282.

Mueller

, Ballard

, Corbett

, Aarsland

(2017) The prognosis of dementia with Lewy bodies. Lancet Neurol 16, 390–398.

Oesterhus

, Soennesyn

, Rongve

, Ballard

, Aarsland

, Vossius

(2014) Long-term mortality in a cohort of home-dwelling elderly with mild Alzheimer’s disease and Lewy body dementia. Dement Geriatr Cogn Disord 38, 161–169.

Price

, Farooq

, Yuan

, Menon

, Cardinal

, O’Brien

(2017) Mortality in dementia with Lewy bodies compared with Alzheimer’s dementia: A retrospective naturalistic cohort study. BMJ Open 7, e017504.

Rongve

, Vossius

, Nore

, Testad

, Aarsland

(2014) Time until nursing home admission in people with mild dementia: Comparison of dementia with Lewy bodies and Alzheimer’s dementia. Int J Geriatr Psychiatry 29, 392–398.

10.

Svendsboe

, Terum

, Testad

, Aarsland

, Ulstein

, Corbett

, Rongve

(2016) Caregiver burden in family carers of people with dementia with Lewy bodies and Alzheimer’s disease. Int J Geriatr Psychiatry 31, 1075–1083.

11.

Mueller

, Perera

, Rajkumar

, Bhattarai

, Price

, O’Brien

, Ballard

, Stewart

, Aarsland

(2018) Hospitalization in people with dementia with Lewy bodies: Frequency, duration, and cost implications. Alzheimers Dement (Amst) 10, 143–152.

12.

Folstein

, Folstein

, McHugh

(1975) “Mini-Mental State” A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 12, 189–198.

13.

Morris

(1997) Clinical dementia rating: A reliable and valid diagnostic and staging measure for dementia of the Alzheimer type. Int Psychogeriatr 9, 173–176.

14.

Aarsland

, Rongve

, Nore

, Skogseth

, Skulstad

, Ehrt

, Hoprekstad

, Ballard

(2008) Frequency and case identification of dementia with Lewy bodies using the revised consensus criteria. Dement Geriatr Cogn Disord 26, 445–452.

15.

McKhann

, Drachman

, Folstein

, Katzman

, Price

, Stadlan

(1984) Clinical diagnosis of Alzheimer’s disease: Report of the NINCDS-ADRDA Work Group under the auspices of Department of Health and Human Services Task Force on Alzheimer’s Disease. Neurology 34, 939–944.

16.

McKeith

, Dickson

, Lowe

, Emre

, O’brien

, Feldman

, Cummings

, Duda

, Lippa

, Perry

(2005) Diagnosis and management of dementia with Lewy bodies. Third report of the DLB consortium. Neurology 65, 1863–1872.

17.

Emre

, Aarsland

, Brown

, Burn

, Duyckaerts

, Mizuno

, Broe

, Cummings

, Dickson

, Gauthier

(2007) Clinical diagnostic criteria for dementia associated with Parkinson’s disease. Mov Disord 22, 1689–1707.

18.

Friedman

(2018) Dementia with Lewy bodies and Parkinson disease dementia: It is the same disease! Parkinsonism Relat Disord 46 Suppl 1, S6–s9.

19.

Skogseth

, Hortobagyi

, Soennesyn

, Chwiszczuk

, Ffytche

, Rongve

, Ballard

, Aarsland

(2017) Accuracy of clinical diagnosis of dementia with Lewy bodies versus neuropathology. J Alzheimers Dis 59, 1139–1152.

20.

Cummings

, Mega

, Gray

, Rosenberg-Thompson

, Carusi

, Gornbein

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

21.

Montgomery

, Asberg

(1979) A new depression scale designed to be sensitive to change. Br J Psychiatry 134, 382–389.

22.

Linn

, Linn

, Gurel

(1968) Cumulative Illness Rating Scale. J Am Geriatr Soc 16, 622–626.

23.

Statistics Norway, Table 07459: Population, by sex and one-year age groups (M) 1986–2019, https://www.ssb.no/en/statbank/table/07459, Accessed March 21, 2019.

24.

Austin

, Lee

, Fine

(2016) Introduction to the analysis of survival data in the presence of competing risks. Circulation 133, 601–609.

25.

Therneau

, Crowson

, Atkinson

, Multi-state models and competing risks. Technical Report, Mayo Clinic, Rochester, Minnesota, USA https://cran.r-project.org/web/packages/survival/vignettes/compete.pdf

26.

Rudolph

, Zanin

, Jones

, Marcantonio

, Fong

, Yang

, Yap

, Inouye

(2010) Hospitalization in community-dwelling persons with Alzheimer’s disease: Frequency and causes. J Am Geriatr Soc 58, 1542–1548.

27.

Murman

, Kuo

, Powell

, Colenda

(2003) The impact of parkinsonism on costs of care in patients with AD and dementia with Lewy bodies. Neurology 61, 944–949.

28.

Bostrom

, Jonsson

, Minthon

, Londos

(2007) Patients with Lewy body dementia use more resources than those with Alzheimer’s disease. Int J Geriatr Psychiatry 22, 713–719.

29.

Prince

, Bryce

, Ferri

, World Alzheimer Report 2011: The benefits of early diagnosis and intervention https://www.alz.co.uk/research/world-report-2011 Accessed July 5, 2019.

30.

Statistics Norway, 5 prosent av pasientene sto for en tredel av liggedagene på sykehus https://www.ssb.no/helse/artikler-og-publikasjoner/5-prosent-av-pasientene-sto-for-en-tredel-av-liggedagene-pa-sykehusAccessed October 7, 2019.

31.

Spears

, Besharat

, Monari

, Martinez-Ramirez

, Almeida

, Armstrong

(2019) Causes and outcomes of hospitalization in Lewy body dementia: A retrospective cohort study. Parkinsonism Relat Disord 64, 106–111.

32.

Russ

, Parra

, Lim

, Law

, Connelly

, Starr

(2015) Prediction of general hospital admission in people with dementia: Cohort study. Br J Psychiatry 206, 153–159.

33.

Mueller

, Molokhia

, Perera

, Veronese

, Stubbs

, Shetty

, Codling

, Huntley

, Stewart

(2018) Polypharmacy in people with dementia: Associations with adverse health outcomes. Exp Gerontol 106, 240–245.

34.

Galvin

, Duda

, Kaufer

, Lippa

, Taylor

, Zarit

(2010) Lewy body dementia: The caregiver experience of clinical care. Parkinsonism Relat Disord 16, 388–392.

35.

Brodaty

, Seeher

, Gibson

(2012) Dementia time to death: A systematic literature review on survival time and years of life lost in people with dementia. Int Psychogeriatr 24, 1034–1045.