Factors Associated with Length of Stay in Hospital Patients with and Without Dementia

Abstract

BACKGROUND:

Hospital care of older adults, especially of those with dementia, is associated with a high risk of complications and increased mortality. Adverse events are often triggered by hospital-related factors, hence the time spent in hospitals should be limited. There is little knowledge of the specific factors influencing hospitalizations of older persons.

OBJECTIVES:

To assess the duration of length of stay (LOS) and risk factors of increased LOS, and, specifically, the role of delirium and neuropsychiatric symptoms (NPS) among a large sample of older adults with and without dementia in Germany.

METHODS:

A claims data based dynamic retrospective cohort study from 2004 to 2015 was conducted. People with dementia (PWD) were identified using ICD-10 codes and the application of diagnostic measures. A control group without diagnosis of dementia (CG) were matched in a 3: 1 ratio. Multivariate methods were used to investigate the factors associated with LOS.

RESULTS:

7,139 PWD and 21,417 controls were included. PWD had longer hospitalizations (first LOS: +4.3 days; second LOS: +0.2 days) than the CG. Diagnosis of delirium was associated with LOS, both for PWD (first LOS: +9.6 days; second LOS: +5.3 days) and CG (first LOS: +13.7 days; second LOS: +7.2 days).

CONCLUSION:

Major determinants of LOS were similar in PWD and the CG. The strongest association was found for the presence of delirium and NPS. Future research should focus on prevention and intervention strategies that may reduce the impact of delirium as well as NPS on the length of stay especially for PWD.

Keywords

Claims data dementia hospitalization neuropsychiatric symptoms observational study

INTRODUCTION

People with dementia (PWD) have a higher risk of hospitalization than people without dementia [1], and during hospitalization PWD suffer from higher complication and mortality rates than people without dementia [2, 3]. Hospital-related conditions, such as immobilization [4, 5] and malnutrition [6, 7], often contribute to adverse outcomes, hence it may be advisable to limit their length of stay (LOS). However, a recent systematic review showed longer hospital stays of up to 22 days for PWD compared to people without dementia in 52 out of 60 studies [8].

Nevertheless, there have been only few studies that analyzed the factors associated with longer periods of hospitalization in PWD and, in particular, there has been little research that focused on the role of delirium and neuropsychiatric symptoms (NPS) during hospitalization. The elucidation of the specific role of delirium and NPS for LOS may be of special relevance as NPS are present in around 75% of PWD during their hospitalization [9-11] and increase the risk of mortality [12]. Up to now only two studies have analyzed the influence of delirium on LOS and both studies have shown longer hospital stays for PWD with delirium [13, 14]. The effect of NPS on LOS remains largely unexplored to date.

The identification of factors associated with an increased LOS among PWD is of utmost importance, both to decrease the deleterious effects of the hospitalization for older patients and to avoid unnecessary costs for the health care system. However, to date there is a lack of large studies analyzing determinants of LOS among PWD as compared to people without dementia. Against this background, the aims of this study were 1) to identify the duration of hospital stays of PWD compared PwoD as well as 2) to assess determinants of LOS, and, specifically, the role of delirium and NPS among a large sample of older adults in Germany.

METHODS

Design

A retrospective dynamic cohort study using claims data from 2004 to 2015 was constructed. The database consists of a statutory health insurance (SHI) sample beginning in 1998 (18.75% random sample of all subjects insured by “AOK Hessen”), an ongoing joint project between the SHI fund “AOK Hessen”, the Hessian Association of Statutory Insurance Physicians, the Hessian Ministry of Social Affairs, and the PMV research group at the University of Cologne, Germany [15]. “AOK Hessen” is the largest SHI in the federal state of Hesse with 1.5 Mill insurees (2015), covering 25% of the total population. All partners approved the utilization of the database for research purposes. Patient informed consent was not required by law as the study was based on pseudonymous data. Cohort entry (earliest in 2006) and exit was possible in every year of the study period. Reasons for cohort exit were limited to death, end of insurance period, or end of study period. End of insurance period relates to participants switching to another SHI.

Study population

Eligible for cohort entry were all members of the SHI aged ≥55 years during the study period with information on age and sex, and a continuous insurance period of at least two years prior to cohort entry. A study participant was classified as a PWD, if the participant had two confirmed outpatient diagnoses of any dementia according to the 10^th Revision of the ICD, German modification (ICD-10-GM) in the same or consecutive 3-month periods or had an inpatient diagnosis of dementia. Dementia diagnoses were ascertained from ICD-10 Chapter V (F-codes) as well as Chapter VI (G30, G31.0, and G31.82). Furthermore, the application of one of the following diagnostic measures based on the physicians’ fee scale for outpatient procedures or the German Procedure Classification for inpatient procedures was required: Testing of cerebrospinal fluid, computed tomography scan (CT), magnetic resonance imaging of the head (MRI), positron-emission tomography of the brain (PET). Sensitivity analyses were conducted comparing the PWD according to our definition (before matching was applied) with participants having dementia diagnoses without the application of diagnostic measures and hence were excluded from the PWD group (Supplementary Tables 1 and 2).

The dementia free control group (CG) was drawn from the study population with 3: 1 matching on age at cohort entry (±3 years), sex, and year of cohort entry. A meta-analysis showed that the rate of missed dementia was around 53.7% in Europe [16]. Hence, a relevant number of people in the CG might have had dementia without a proper diagnosis. In order to reduce this number, participants with a diagnosis of dementia (not fulfilling the case definition), cognitive impairment, vascular encephalopathy, or severe depressive disorder two years preceding cohort entry or during their follow-up, were excluded from the CG.

Covariates

Comorbidities were assessed using hospital discharge diagnoses and confirmed outpatient diagnoses based on ICD-10-GM starting two years prior to cohort entry. The following coded diagnoses of delirium and NPS (ICD-10 code) were assessed during hospitalization: delirium (F05), eating disorder (R63.0, F50), sleeping disorder (G47, F51), aberrant motor behavior (R25), agitation (R45.1, R46.3), hallucination (R44), acute psychotic state (F23), delusional disorder (F22), apathy (R45.3), anger (R45.4), violence (R45.6, F91), hostility (R45.5), anxiety (F41), somnolence (R40.0). Since aberrant motor behavior has no specific ICD-10 code we used the mentioned ICD-10 code as proxy coding. Outpatient dispensation of selected medication was assessed using the Anatomical Therapeutic Chemical Classification and were classified in the same manner as the comorbidities. Information regarding over-the-counter and inpatient medication was not available. The level of care needed, which is part of the statutory nursing care and qualifies for specific nursing care reimbursements in Germany, is defined in three stages, with people in stage three being in need of the most care. The last billing for level of care prior to each hospitalization was used to determine the level of care needed.

Outcomes

The main outcome was the LOS of the first two hospitalizations with at least one overnight hospital stay and their determinants of the study participants after cohort entry. If a participant had overlapping periods of hospitalizations, the LOS was summed up. The approach using only the first and second hospitalization after cohort entry was chosen to analyze independently and in detail the differences in LOS between groups and their determinants.

Statistical methods

Differences of study characteristics between PWD and the CG were investigated with chi-square tests for categorical variables, t-tests, or, in case of violations of the normal distribution assumption, Wilcoxon-Mann-Whitney tests for continuous variables, and Kruskal-Wallis-tests for median values with a level of statistical significance of p <0.05.

The LOS and its determinants were analyzed by using multivariate linear regression models for comparability to other studies analyzing LOS. In addition, Poisson regression was used to display rate ratios of the mean inpatient days. For the determinants of LOS a stepwise procedure was used. Variables with a p <0.05 were allowed to stay in the model. To compare the adjusted differences between groups, four models with an increasing number of covariates were utilized. Comorbidities and medication were diagnosed or dispensed from two years prior cohort entry until cohort entry. All statistical analyses were done using SAS 9.4 (SAS Institute Inc., Cary, NC, USA).

RESULTS

Sample characteristics

Main characteristics of eligible study participants are shown in Table 1. Based on the aforementioned case definition and matching criteria 7,139 PWD and 21,417 controls were identified. Mean age of PWD and the CG was 78.2 and 77.2 years, respectively, and 57% of the study population was female. Among PWD, 7% did not have any hospitalization compared to 34% of the CG. The rate of hospitalizations per 100 person years was 116.4 among PWD compared to 51.4 among the CG. PWD had a higher burden of cardiovascular, neurologic, pulmonary, and bone/joint diseases as well as of depression (excluding severe depressive disorder), diabetes, and other comorbidities including chronic kidney disease, alcohol abuse, and dehydration.

Table 1

Sample characteristics for PWD and the control group (CG)

	PWD (N = 7,139)	CG (N = 21,417)	p^d
Demographic and hospitalization characteristics
Sex (female, %)	56.9	56.9	1.000
Age in years (Mean, SD)	78.2 (7.2)	77.2 (7.2)	<0.001
Follow-up time in years (Median, IQR)	2.2 (3.4)	3.0 (4.4)	<0.001
LOS first hospitalization in days (Mean, SD)	18.0 (18.0)	12.6 (16.0)	<0.001
LOS second hospitalization in days (Mean, SD)	13.9 (13.7)	12.6 (13.6)	<0.001
Number of Hospitalizations during follow-up (Median, IQR)	3 (4)	1 (3)	<0.001
Hospitalization Rate (per 100 Person Years)	116.4	51.4	-
People without Hospitalisation (n, %)	489 (6.9)	7,361 (34.4)	-
Reasons for leaving study (%)
Deceased	58.4	34.9	<0.001
End of study period	39.7	62.2
End of insurance period	1.9	2.9
Comorbidities at cohort entry (%)
Cardiovascular diseases	56.7	34.9	<0.001
Neurologic diseases (excl. dementia related diagnoses)	18.6	3.5	<0.001
Depression (excl. severe depressive disorders)	34.0	16.0	<0.001
Pulmonary diseases	20.1	17.9	<0.001
Diabetes	36.4	29.2	<0.001
Cancer	20.7	20.8	0.814
Bone/Joint diseases	57.7	53.0	<0.001
Other comorbidities^c	26.3	11.5	<0.001
Drug intake at cohort entry (%)
Insulin	14.3	8.3	<0.001
Other antidiabetic drugs	22.5	17.8	<0.001
Antithrombotic agents	48.0	32.0	<0.001
Anticonvulsants	15.2	5.7	<0.001
Glucocorticoids	14.0	13.4	0.243
Opioids	26.4	19.9	<0.001
NSAIDS	51.3	51.8	<0.432
Antipsychotics	36.2	4.2	<0.001
Antidepressants	34.5	13.3	<0.001
Anxiolytics	14.3	7.5	<0.001
Hypnotics &sedatives	11.9	4.7	<0.001
Statins	34.6	28.4	<0.001
Angiotensin antagonist	16.0	17.4	<0.001
ACE inhibitors	59.0	50.9	0.067

Comorbidities, and medication were diagnosed or prescribed prior to the first hospitalization. A) PWD: 6,650; CG: 14,056. B) PWD: 5,122; CG: 9,050. C) Chronic kidney disease, alcohol abuse, dehydration. D) P-values derived from t-tests or (in case of violations of the normal distribution assumption) Wilcoxon-Mann-Whitney tests for continuous variables, chi-square tests for categorical variables, and Kruskal-Wallis tests for median variables.

Main discharge diagnosis

The main discharge diagnosis relating to the first and second hospitalization are reported in Table 2. Diseases of the cardiovascular system were the most common main discharge diagnosis in the first (PWD: 22.8%, CG: 27.1%) and second hospitalization (PWD: 20.4%, CG: 26.7%) in both groups. Other conditions (PWD: 13.6%, CG: 10.4%), which include infectious diseases, diseases of the blood/blood forming organs, skin diseases, symptoms, signs and abnormal clinical/laboratory findings, and external causes of morbidity/mortality, as well as injuries (PWD: 19.1%; CG: 13.9%) were the second most common main discharge diagnosis for the first and second hospitalization, respectively.

Table 2

Main discharge diagnosis for hospitalizations among PWD and the CG

Main discharge diagnosis (%)	ICD 10	First hospitalization			Second hospitalization
		All (N = 20,706)	PWD (N = 6,650)	CG (N = 14,056)	All (N = 14,172)	PWD (N = 5,122)	CG (N = 9,050)
Circulatory system	I00-I99	25.7	22.8	27.1	24.4	20.4	26.7
Injuries^A	S00-T98	10.2	11.9	9.4	15.8	19.1	13.9
Digestive system	K00-K93	8.8	5.0	10.7	9.5	7.9	10.4
Musculoskeletal system	M00-M99	7.6	3.2	9.6	6.6	3.2	8.5
Neoplasms	C00-D09	7.4	3.0	9.5	8.8	3.8	11.6
Nervous system	G00-G99	6.0	13.2	2.5	4.5	8.7	2.0
Respiratory system	J00-J99	5.7	4.9	6.1	7.1	8.3	6.5
Genitourinary system	N00-N99	4.2	4.0	4.4	5.1	5.9	4.7
Endocrine diseases^B	E00-E90	3.6	5.7	2.6	4.5	7.1	3.1
Diseases of the eyes &ears	H00-H95	2.4	0.9	3.2	2.4	1.1	3.1
Other mental/behavioral disorders (excl. Depression)	F04-F99	1.8	4.6	0.4	1.4	3.3	0.3
Dementia	F00-F03	1.6	5.0	0.0	1.0	2.8	0.0
Benign neoplasms	D10-D36	1.2	0.5	1.5	1.1	0.6	1.5
Depression	F32-F33	0.4	1.1	0.1	0.3	0.7	0.1
Other conditions^C	A00-B99, D50-D89,	11.4	13.6	10.4	5.6	6.2	5.2
	L00-L99, Q00-Q99,
Unknown	-	2.1	0.9	2.6	2.0	1.1	2.5

A) Injuries, poisoning, other consequences of external causes. B) Endocrine, nutritional and metabolic diseases. C) Includes: Infectious and parasitic diseases, Diseases of the blood/blood-forming organs, skin/subcutaneous tissue diseases; Symptoms, signs and abnormal clinical/laboratory findings; External causes of morbidity/mortality; Factors influencing health status/contact with health services; Codes for special purposes; Congenital malformations, deformations and chromosomal abnormalities.

Outcomes

The adjusted differences of LOS are presented in Table 3. The first hospitalization was 5.3 days longer for PWD when adjusted for age and sex. In the fully adjusted model, the LOS was 4.3 days higher for PWD compared to the CG. As for the second hospitalization, PWD were 1.3 days longer hospitalized when adjusting for age and sex. Additional adjustment for year of hospitalization, main discharge diagnosis, comorbidities, medication, and level of care resulted in a 0.2 day longer hospitalization for PWD compared to the CG.

Table 3

Adjusted estimates in outcome variables for PWD (reference group: CG)

	Model 1^A Estimate (95% CI)	Model 2^B Estimate (95% CI)	Model 3^C Estimate (95% CI)	Model 4^D Estimate (95% CI)
LOS first hospitalization
mean difference in days	5.3 (4.8; 5.8)	5.1 (4.6; 5.6)	5.2 (4.7; 5.8)	4.3 (3.8; 4.9)
rate ratio	1.4 (1.4; 1.4)	1.4 (1.4; 1.4)	1.4 (1.4; 1.4)	1.3 (1.3; 1.3)
LOS second hospitalization
mean difference in days	1.3 (0.8; 1.7)	1.0 (0.6; 1.5)	0.8 (0.3; 1.4)	0.2 (-0.4; 0.8)
rate ratio	1.1 (1.1; 1.1)	1.1 (1.1; 1.1)	1.1 (1.1; 1.1)	1.0 (1.0; 1.0)

LOS, length of stay. A) Adjusted for age, sex. B) Adjusted additionally for year of hospitalization, main discharge diagnosis. C) Adjusted additionally for comorbidities, medication present/prescribed before the hospitalization. D) Adjusted additionally for level of care needed before the hospitalization. Multivariate linear regression and Poisson regression used for LOS.

Risk factors for LOS among PWD and the CG

The association of NPS with LOS are shown in Figure 1. For the complete list of risk factors associated with the LOS please see Tables 4 and 5. Among PWD there was a pronounced trend showing that the older the patient, the shorter the LOS. In particular, PWD aged 75 to 84 years (first LOS: -2.4 days; second LOS: -1.4 days), 85 to 94 years (first LOS: -5.3 days; second LOS: -2.8 days) or older than 95 years (first LOS: -8.1 days; second LOS: -4.0 days) had a shorter stay than PWD aged 55 to 64. The CG did not show a similar trend.

Fig.1

Association of neuropsychiatric symptoms with LOS for hospitalizations as mean difference in days (reference group: No diagnosis of neuropsychiatric symptom). CG, Control group; PWD, people with dementia; LOS, length of hospital stay. In both groups adjusted for age, sex, year of hospitalization, main discharge diagnosis, level of care needed.

Table 4

Factors associated with LOS for the first hospitalization. Associations are expressed as mean difference (MD) in days and rate ratios (RR) among PWD and the CG

	PWD		CG
	MD (95% CI)	RR (95% CI)	MD (95% CI)	RR (95% CI)
Age group
55-64	Reference		Reference
65-74	-1.0 (-3.1; 1.2)	0.9 (0.9; 1.0)	0.7 (-0.6; 2.1)	1.1 (1.0; 1.1)
75-84	-2.4 (-4.4; -0.4)	0.9 (0.8; 0.9)	2.1 (0.8; 3.4)	1.2 (1.2; 1.2)
85-94	-5.3 (-7.6; -3.1)	0.7 (0.7; 0.8)	-0.6 (-2.1; 0.9)	1.0 (1.0; 1.0)
95+	-8.1 (-12.8; -3.3)	0.6 (0.6; 0.7)	-7.4 (-11.3; -3.5)	0.6 (0.5; 0.6)
Sex
Women	Reference		Reference
Men	0.3 (-0.5; 1.2)	1.0 (1.0; 1.0)	-0.1 (-0.6; 0.4)	1.0 (1.0; 1.0)
Year of hospitalization
2006-2007	Reference		Reference
2008-2009	0.1 (-1.2; 1.3)	1.0 (1.0; 1.0)	-0.2 (-1.0; 0.7)	1.0 (1.0; 1.0)
2010-2011	-2.1 (-3.4; -0.8)	0.9 (0.9; 0.9)	-2.1 (-3.0; -1.2)	0.9 (0.8; 0.9)
2012-2013	-3.5 (-4.8; -2.2)	0.8 (0.8; 0.8)	-3.2 (-4.1; -2.3)	0.8 (0.8; 0.8)
2014-2015	-4.0 (-5.3; -2.7)	0.8 (0.8; 0.8)	-4.0 (-4.9; -3.2)	0.7 (0.7; 0.7)
Hospital indication
Circulatory system	Reference		Reference
Respiratory system	-3.9 (-5.9; -1.9)	0.8 (0.8; 0.8)	-0.2 (-1.4; 0.9)	1.0 (1.0; 1.0)
Musculoskeletal system	0.6 (-1.9; 3.0)	1.0 (1.0; 1.1)	3.3 (2.3; 4.3)	1.3 (1.3; 1.3)
Endocrine system	-7.5 (-9.4; -5.5)	0.6 (0.6; 0.6)	-1.9 (-3.5; -0.2)	0.9 (0.8; 0.9)
Digestive system	-4.8 (-6.9; -2.8)	0.7 (0.7; 0.8)	-2.5 (-3.4; -1.6)	0.8 (0.8; 0.8)
Genitourinary system	-4.5 (-6.8; -2.3)	0.8 (0.7; 0.8)	-2.4 (-3.7; -1.1)	0.8 (0.8; 0.8)
Eyes &ears	-5.8 (-10.2; -1.5)	0.7 (0.6; 0.7)	-3.2 (-4.7; -1.7)	0.7 (0.7; 0.7)
Neoplasms	-1.0 (-3.5; 1.6)	0.9 (0.9; 1.0)	2.2 (1.2; 3.1)	1.2 (1.2; 1.2)
Benign neoplasms	-3.5 (-9.6; 2.5)	0.8 (0.7; 0.9)	-2.2 (-4.4; -0.1)	0.8 (0.8; 0.8)
Nervous system	-3.9 (-5.3; -2.4)	0.8 (0.8; 0.8)	-1.6 (-3.3; 0.1)	0.9 (0.9; 0.9)
Depression	18.7 (14.6; 22.7)	1.9 (1.9; 2.0)	5.6 (-5.1; 16.3)	1.5 (1.3; 1.8)
Dementia	-1.0 (-3.1; 1.0)	1.0 (0.9; 1.0)	-	-
Other mental/behavioral disorders	2.8 (0.6; 5.0)	1.1 (1.1; 1.1)	-1.2 (-5.2; 2.9)	1.0 (0.9; 1.0)
Injury, Poisoning, other consequences of external causes	0.1 (-1.4; 1.6)	1.0 (1.0; 1.0)	3.4 (2.4; 4.3)	1.3 (1.2; 1.3)
Other conditions^A	-2.0 (-3.4; -0.5)	0.9 (0.9; 0.9)	0.3 (-0.6; 1.3)	1.0 (1.0; 1.0)
Unknown	-6.1 (-10.4; -1.7)	0.6 (0.5; 0.6)	-2.5 (-4.1; -0.8)	0.8 (0.7; 0.8)
Comorbidities at baseline
Not present	Reference		Reference
Diabetes (Type I or Type II)	-0.3 (-1.1; 0.6)	1.0 (1.0; 1.0)	0.4 (-0.1; 1.0)	1.0 (1.0; 1.0)
Cardiovascular diseases	0.8 (-0.1; 1.7)	1.0 (1.0; 1.1)	-	-
Other comorbidities^B	-0.4 (-1.4; 0.5)	1.0 (1.0; 1.0)	-	-
Neurologic diseases	-2.4 (-3.5; -1.3)	0.9 (0.9; 0.9)	-	-
Pulmonary diseases	-	-	0.0 (-0.6; 0.7)	1.0 (1.0; 1.0)
Neuropsychiatric symptoms during hospitalization
Not present	Reference		Reference
Hallucinations, acute psychotic state, delusions	13.9 (10.2; 17.6)	1.6 (1.5; 1.6)	24.1 (17.4; 30.8)	2.4 (2.2; 2.5)
Agitation, anger, hostility^C	5.0 (3.0; 7.0)	1.3 (1.2; 1.3)	8.4 (6.7; 10.2)	1.6 (1.6; 1.7)
Delirium	9.6 (8.1; 11.1)	1.5 (1.5; 1.5)	13.7 (11.4; 16.0)	2.0 (1.9; 2.0)
Eating disorders, somnolence	6.0 (3.9; 8.2)	1.3 (1.3; 1.3)	6.2 (3.6; 8.8)	1.4 (1.4; 1.5)
Apathy, anxiety	3.9 (-0.7; 8.6)	1.2 (1.1; 1.2)	-	-
Level of care needed
No level of care	Reference		Reference
Level one care	2.1 (1.1; 3.1)	1.1 (1.1; 1.1)	4.1 (3.4; 4.9)	1.4 (1.3; 1.4)
Level two care	1.2 (0.1; 2.4)	1.1 (1.1; 1.1)	6.7 (5.7; 7.7)	1.6 (1.5; 1.6)
Level three care	1.2 (-0.3; 2.7)	1.1 (1.0; 1.1)	7.1 (5.2; 9.0)	1.6 (1.5; 1.6)

A) Diseases: Infectious/parasitic, blood/blood-forming organs, skin/subcutaneous tissue, symptoms, abnormal clinical/laboratory findings, external causes of morbidity/mortality; Factors influencing health status/contact with health services; Codes for special purposes; Congenital malformations, deformations and chromosomal abnormalities. B) Chronic kidney disease, alcohol abuse, dehydration. C) +Physical violence, sleep disturbances, aberrant motor behavior.

Table 5

Factors associated with LOS for the second hospitalization. Associations are expressed as mean difference (MD) in days and rate ratios (RR) among PWD and the CG

	PWD		CG
	MD (95% CI)	RR(95% CI)	MD (95% CI)	RR(95% CI)
Age group
55-64	Reference		Reference
65-74	-1.3 (-3.2; 0.5)	0.9 (0.9; 0.9)	1.8 (0.4; 3.2)	1.2 (1.2; 1.2)
75-84	-1.4 (-3.2; 0.4)	0.9 (0.9; 0.9)	2.1 (0.8; 3.5)	1.2 (1.2; 1.3)
85-94	-2.8 (-4.8; -0.8)	0.8 (0.8; 0.8)	1.2 (-0.4; 2.8)	1.2 (1.1; 1.2)
95+	-4.0 (-8.6; 0.6)	0.7 (0.7; 0.8)	-0.8 (-6.0; 4.3)	1.0 (0.9; 1.1)
Sex
Women	Reference		Reference
Men	-0.1 (-0.8; 0.7)	1.0 (1.0; 1.0)	-0.5 (-1.0; 0.1)	1.0 (1.0; 1.0)
Year of hospitalization
2006-2007	Reference		Reference
2008-2009	-0.9 (-2.1; 0.4)	0.9 (0.9; 1.0)	0.5 (-0.6; 1.5)	1.0 (1.0; 1.1)
2010-2011	-1.2 (-2.4; 0.1)	0.9 (0.9; 0.9)	-0.8 (-1.9; 0.2)	0.9 (0.9; 1.0)
2012-2013	-1.2 (-2.4; 0.0)	0.9 (0.9; 0.9)	-2.5 (-3.5; -1.5)	0.8 (0.8; 0.8)
2014-2015	-2.2 (-3.4; -1.0)	0.9 (0.8; 0.9)	-2.9 (-3.9; -1.9)	0.8 (0.8; 0.8)
Hospital indication
Circulatory system	Reference		Reference
Respiratory system	-2.0 (-3.6; -0.5)	0.9 (0.8; 0.9)	-1.0 (-2.2; 0.2)	0.9 (0.9; 1.0)
Musculoskeletal system	2.7 (0.5; 4.9)	1.2 (1.1; 1.2)	3.3 (2.2; 4.4)	1.3 (1.3; 1.3)
Endocrine system	-4.4 (-6.0; 2-.9)	0.7 (0.7; 0.7)	-2.7 (-4.3; -1.0)	0.8 (0.8; 0.8)
Digestive system	-3.8 (-5.3; -2.3)	0.7 (0.7; 0.8)	-2.9 (-3.9; -1.9)	0.8 (0.7; 0.8)
Genitourinary system	-2.6 (-4.3; -0.9)	0.8 (0.8; 0.8)	-3.3 (-4.6; -1.9)	0.7 (0.7; 0.8)
Eyes &ears	-5.3 (-8.9; -1.7)	0.6 (0.6; 0.7)	-3.6 (-5.2; -2.0)	0.7 (0.7; 0.7)
Neoplasms	-0.1 (-2.2; 1.9)	1.0 (1.0; 1.0)	1.3 (0.4; 2.3)	1.1 (1.1; 1.1)
Benign neoplasms	-2.0 (-7.0; 2.9)	0.9 (0.8; 1.0)	-2.8 (-5.1; -0.5)	0.8 (0.7; 0.8)
Nervous system	0.8 (-0.7; 2.2)	1.1 (1.0; 1.1)	-0.2 (-2.1; 1.8)	1.0 (1.0; 1.0)
Depression	21.1 (16.7; 25.4)	2.4 (2.3; 2.6)	11.1 (-0.4; 22.7)	1.7 (1.4; 1.9)
Dementia	5.7 (3.4; 8.0)	1.4 (1.3; 1.4)	-	-
Other mental/behavioral disorders	6.5 (4.3; 8.7)	1.4 (1.3; 1.5)	6.2 (0.9; 11.5)	1.4 (1.3; 1.5)
Injury, Poisoning, other consequences of external causes	0.3 (-0.9; 1.4)	1.0 (1.0; 1.0)	2.1 (1.2; 3.0)	1.2 (1.1; 1.2)
Other conditions^A	-1.9 (-3.6; -0.3)	0.9 (0.8; 0.9)	0.5 (-0.8; 1.7)	1.0 (1.0; 1.1)
Unknown	-7.5 (-11.2; -3.9)	0.4 (0.4; 0.5)	-2.8 (-4.6; -1.0)	0.8 (0.7; 0.8)
Comorbidities at baseline
Not present	Reference		Reference
Diabetes (Type I or Type II)	0.3 (-0.5; 1.0)	1.0 (1.0; 1.0)	-	-
Other comorbidities^B	0.6 (-0.2; 1.5)	1.0 (1.0; 1.1)	-	-
Pulmonary diseases	0.8 (-0.1; 1.7)	1.1 (1.0; 1.1)	-	-
Cancer	-	-	-0.3 (-0.9; 0.3)	1.0 (1.0; 1.0)
Neuropsychiatric symptoms during hospitalization
Not present	Reference		Reference
Delirium	5.3 (3.8; 6.7)	1.4 (1.3; 1.4)	7.2 (4.7; 9.7)	1.5 (1.4; 1.6)
Hallucinations, acute psychotic state, delusions	9.3 (4.6; 14.0)	1.5 (1.4; 1.7)	-	-
Agitation, anger, hostility^C	-	-	7.3 (5.5; 9.0)	1.6 (1.5; 1.6)
Eating disorders, somnolence	-	-	10.8 (8.3; 13.3)	1.7 (1.6; 1.8)
Level of care needed
No level of care	Reference		Reference
Level one care	1.4 (0.3; 2.5)	1.1 (1.1; 1.1)	3.1 (2.4; 3.9)	1.3 (1.2; 1.3)
Level two care	1.2 (0.2; 2.3)	1.1 (1.1; 1.1)	3.9 (2.9; 4.8)	1.3 (1.3; 1.3)
Level three care	-0.5 (-1.8; 0.8)	1.0 (1.0; 1.0)	2.1 (0.2; 4.0)	1.2 (1.1; 1.2)

Hospitalization in more recent years were associated with shorter hospital stays among PWD and the CG. The first hospitalization was shorter in 2010/2011 (PWD: -2.1 days; CG: -2.1 days), 2012/2013 (PWD: -3.5 days; CG: -3.2 days), and 2014/2015 (PWD: -4.0 days; CG: -4.0 days) compared to 2006/2007 in both groups. The second hospitalization showed a similar pattern.

Compared to a main discharge diagnosis of diseases of the cardiovascular system, depression was significantly associated with longer hospital stays in the first and second hospitalization among PWD (first LOS: +18.7 days; second LOS: +21.1 days). Other mental/behavioral disorders than depression or dementia were associated with longer hospitalizations in PWD (first LOS: +2.8 days; second LOS: +6.5 days) and the CG (second LOS: +6.2 days).

Delirium was strongly associated with LOS both among PWD (first LOS: +9.6 days; second LOS: +5.3 days) and the CG (first LOS: +13.7 days; second LOS: +7.2 days). Hallucinations, an acute psychotic state or delusional disorders during hospitalization prolonged the hospitalizations in PWD (First LOS: +13.9 days; second LOS: +9.3 days) and the CG (First LOS: +24.1 days).

Sensitivity analyses

Sensitivity analyses comparing PWD with excluded participants that had a coded dementia diagnosis without fulfilling the aforementioned diagnostic criteria are shown in Supplementary Tables 1 and 2. The first hospitalization was 2.9 days longer for PWD compared to excluded participants. There was no significant difference for the second hospitalization in the fully adjusted model.

DISCUSSION

This comprehensive study with a large sample size analyzed risk factors for length of hospital stay among older patients with and without dementia. An important novelty of the study is the strong focus on the diagnosis of delirium and NPS during hospitalization across hospitalizations. It was shown that LOS is strongly associated with delirium and NPS both among PWD and the CG. Hospitalization year, main discharge diagnosis, and the presence of delirium or NPS were risk factors of similar relevance for longer hospital stays in PWD and in the CG as well.

The findings of the present study with longer and more frequent hospitalizations among PWD compared to the CG are in line with the available scientific literature [1, 8]. However, the current study describes for the first time a reduced difference in LOS between the first and second hospitalization among PWD as compared to the CG. This observation may be interpreted as a consequence of an increasing assimilation between PWD and people without dementia over time. Increasing age was associated with a shorter hospital stay among PWD. This was also reported by a previous study[17] and might contribute in explaining the increasing assimilation of PWD and people without dementia.

Hospitalizations were shorter in more recent years in both PWD and the CG, which is in agreement with a general reduction in LOS over time in Germany which is likely the consequence of a variety of health policy changes [18].

Several studies have reported longer hospitalizations for PWD compared to people without dementia across different indications for hospital admission, including falls,[19] fractures,[20] infectious,[21] and cardiovascular diseases [22]. Depression or other mental/behavioral problems as reasons for hospitalization, which are generally associated with longer hospitalizations,[18] had a great impact on LOS among PWD and the CG. As we excluded participants with severe depressive disorders from the CG we cannot assess with certainty whether depression as a reason for hospitalization would explain significant differences in LOS between PWD and the CG. A cohort study among AD patients found that depression as a comorbidity did not significantly prolong hospitalization,[23] indicating that only cases of depression severe enough to lead to hospitalization might have an impact on LOS. However, future studies should further explore this issue.

The presence of delirium and NPS during hospitalization was strongly associated with longer hospital stays in both groups. Since NPS have been reported in up to 75% of PWD these findings may have contributed to a large degree to the difference in LOS between PWD and the CG [10, 11]. They also highlight the need for early diagnosis and treatment of delirium and NPS in hospitals, where they may often stay undiagnosed [24]. Furthermore, their presence might impede the treatment of the leading pathology and the treatment of delirium or NPS itself may also prolong hospitalization. The presence of delirium or NPS might also result in a transfer to a psychiatric ward which in turn prolongs the hospital stay as well and the latter may explain the strong association of delirium and NPS with LOS among the CG. Therefore, our findings also illustrate the urgent need to prevent and successfully treat delirium and NPS which may decrease their impact on LOS. Additionally, it is known that the prevalence of delirium is underestimated using the ICD-10 system [25]. It is very likely that the same applies for the NPS. This assumption is supported by the low prevalence of NPS during hospitalization in our cohort (Supplementary Table 3). An underreporting of delirium and NPS might have resulted in a reduction of the association of delirium and NPS with LOS.

Strengths and weaknesses

Our study has several limitations. First of all, we analyzed SHI data from a single German region. However, it should be noted that German studies from other geographical areas also reported a longer stay for PWD compared to people without dementia [26, 27]. The current analysis relied on ICD-10-GM codes used for billing purposes in the SHI-System. These data may ensure a high degree of validity, but they have not been externally validated. As a further limitation of our study, information on possible confounders, including nutritional status, stage of dementia, education, and clinical parameters were not available in the administrative data. In particular, the nutritional status might have explained at least partly the differences in LOS between PWD and people without dementia, since malnutrition is associated with an increased LOS [28, 29] and dementia itself is an independent predictor of malnutrition [28, 30]. Among PWD, the type of dementia might be an important predictor as well as one study showed longer LOS for PWD with Lewy body disease compared to people with Alzheimer’s disease [31]. We tried to ensure a valid diagnosis of dementia by including diagnostic criteria in our case definition. Based on the latter it is likely that we included primarily moderate to severe cases of dementia. This assumption is supported by our sensitivity analyses, which showed a longer first stay for PWD compared to excluded participants with coded dementia diagnoses but without the application of our diagnostic criteria. Finally, screening for NPS is often not a routine in general hospitals [24] and this might have led to an underestimation of the impact of NPS on LOS in our analyses.

The strengths of our study were the large sample size, the differentiation between multiple hospitalizations, and the inclusion of patients independently of their living situation, health status, or nationality. Furthermore, by using SHI data recall or interviewer bias were avoided.

In conclusion, the identification of the causal factors that lead to a longer hospitalization in PWD is of high relevance if unnecessarily long hospitalizations are to be avoided in the future. In the current study, PWD showed longer hospitalizations than the CG. The determinants of the LOS were partly identical in both groups as longer hospitalization were strongly associated with the presence of delirium and NPS. Our results underline the necessity to 1) conduct studies to fully understand the impact of delirium and NPS on LOS which are based on primary data collection with reliable assessments for delirium and NPS, and 2) develop interventional studies focusing on approaches for prevention and treatment of delirium and NPS in older hospital patients, especially in those with PWD.

Footnotes

ACKNOWLEDGMENTS

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. TM is a PhD candidate funded by the Robert Bosch Foundation Stuttgart within the Graduate Program People with Dementia in Acute Care Hospitals. The Robert Bosch Foundation had no role in any part of the study. The authors would like to thank AOK Hesse and the KV Hesse for providing the data for the AOK Hesse/KV Hesse sample.

Authors’ disclosures available online ().

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

References

Phelan

, Borson

, Grothaus

, Balch

, Larson

(2012) Association of incident dementia with hospitalizations. JAMA 307, 165–172.

Holmes

, House

(2000) Psychiatric illness predict poor outcome after surgery for hip fracture: a prospective cohort study. Psychol Med 30, 921–929.

Watkin

, Blanchard

, Tookman

, Sampson

(2012) Prospective cohort study of adverse events in older people admitted to the acute general hospital: risk factors and the impact of dementia. Int J Geriatr Psychiatry 27, 76–82.

Brown

, Friedkin

, Inouye

(2004) Prevalence and outcomes of low mobility in hospitalized older adults. J Am Geriatr Soc 52, 1263–1270.

Zisberg

, Shadmi

, Sinoff

, Gur-Yaish

, Srulovici

, Admi

(2011) Low mobility during hospitalization and functional decline in older adults. J Am Geriatr Soc 59, 266–273.

Avelino-Silva

, Marfel

, Curiati

JAE

, Amaral

JRG

, Campora

, Jacob-Filho

(2014) Comprehensive geriatric assessment predicts mortality and adverse outcomes in hospitalized older adults. BMC Geriat 14, 129.

Corkins

, Guenter

, DiMaria-Ghalili

, Jensen

, Malone

, Miller

, Patel

, Plogsted

, Resnick

, American Society for Parenteral and Enteral Nutrition (2014) Malnutrition diagnoses in hospitalized patients: United States, 2010. JPEN J Parenter Enteral Nutr 38, 186–195.

Möllers

, Stocker

, Wei

, Perna

, Brenner

(2019) Length of hospital stay and dementia: A systematic review of observational studies. Int J Geriatr Psychiatry 34, 8–21.

Chung

, Cummings

(2000) Neurobehavioral and neuropsychiatric symptoms in Alzheimer’s disease: characteristics and treatment. Neurol Clin 18, 829–846.

10.

Hessler

, Schaufele

, Hendlmeier

, Junge

, Leonhardt

, Weber

, Bickel

(2017) 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.

11.

Sampson

, White

, Leurent

, Scott

, Lord

, Round

, Jones

(2014) Behavioural and psychiatric symptoms in people with dementia admitted to the acute hospital: prospective cohort study. Br J Psychiatry 205, 189–196.

12.

Peters

, Schwartz

, Han

, Rabins

, Steinberg

, Tschanz

, Lyketsos

(2015) Neuropsychiatric symptoms as predictors of progression to severe Alzheimer’s dementia and death: the Cache County Dementia Progression Study. Am J Psychiatry 172, 460–465.

13.

Fick

, Steis

, Waller

, Inouye

(2013) Delirium superimposed on dementia is associated with prolonged length of stay and poor outcomes in hospitalized older adults. J Hosp Med 8, 500–505.

14.

Reynish

, Hapca

, De Souza

, Cvoro

, Donnan

, Guthrie

(2017) Epidemiology and outcomes of people with dementia, delirium, and unspecified cognitive impairment in the general hospital: prospective cohort study of 10,014 admissions. BMC Med 15, 140.

15.

Ihle

, Koster

, Herholz

, Rambow-Bertram

, Schardt

, Schubert

(2005) [Sample survey of persons insured in statutory health insurance institutions in Hessen–concept and realisation of person-related data base]. Gesundheitswesen 67, 638–645.

16.

Lang

, Clifford

, Wei

, Zhang

, Leung

, Augustine

, Danat

, Zhou

, Copeland

, Anstey

, Chen

(2017) Prevalence and determinants of undetected dementia in the community: a systematic literature review and a meta-analysis. BMJ Open 7, e011146.

17.

Draper

, Karmel

, Gibson

, Peut

, Anderson

(2011) The Hospital Dementia Services Project: age differences in hospital stays for older people with and without dementia. Int Psychogeriatr 23, 1649–1658.

18.

Eurostat (2016) Health in the European Union - facts and figures. Hospital discharges and length of stay statistics. http://ec.europa.eu/eurostat/sstatistics-explained/index.php/Hospital_discharges_and_length_of_stay_statistics#Average_length_of_hospital_stay_for_in-patients, Last updated November 28, 2018, Accessed December 4, 2018.

19.

Harvey

, Mitchell

, Brodaty

, Draper

, Close

(2016) Differing trends in fall-related fracture and non-fracture injuries in older people with and without dementia. Arch Gerontol Geriatr 67, 61–67.

20.

Mitchell

, Harvey

, Brodaty

, Draper

, Close

(2016) Hip fracture and the influence of dementia on health outcomes and access to hospital-based rehabilitation for older individuals. Disabil Rehabil 38, 2286–2295.

21.

Briggs

, Coary

, Collins

, Coughlan

, O’Neill

, Kennelly

(2016) Acute hospital care: how much activity is attributable to caring for patients with dementia? QJM 109, 41–44.

22.

Carter

, Reynolds

, Carter

, Potluri

, Uppal

, Chandran

, Potluri

(2016) The impact of psychiatric comorbidities on the length of hospital stay in patients with heart failure. Int J Cardiol 207, 292–296.

23.

Beydoun

, Beydoun

, Gamaldo

, Rostant

, Dore

, Zonderman

, Eid

(2015) Nationwide inpatient prevalence, predictors, and outcomes of Alzheimer’s disease among older adults in the United States, 2002-2012. J Alzheimers Dis 48, 361–375.

24.

Timmons

, O’Shea

, O’Neill

, Gallagher

, de Siun

, McArdle

, Gibbons

, Kennelly

(2016) Acute hospital dementia care: results from a national audit. BMC Geriatr 16, 113.

25.

Laurila

, Pitkala

, Strandberg

, Tilvis

(2003) The impact of different diagnostic criteria on prevalence rates for delirium. Dement Geriatr Cogn Disord 16, 156–162.

26.

Bruggenjurgen

, Andersohn

, Ezzat

, Lacey

, Willich

(2015) Medical management, costs, and consequences of Alzheimer’s disease in Germany: an analysis of health claims data. J Med Econ 18, 466–473.

27.

Motzek

, Junge

, Marquardt

(2016) [Impact of dementia on length of stay and costs in acute care hospitals]. Z Gerontol Geriatr 50, 59–66.

28.

O’Shea

, Trawley

, Manning

, Barrett

, Browne

, Timmons

(2017) Malnutrition in hospitalised older adults: A multicentre observational study of prevalence, associations and outcomes. J Nutr Health Aging 21, 830–836.

29.

Rasheed

, Woods

(2013) Malnutrition and associated clinical outcomes in hospitalized patients aged 60 and older: an observational study in Rural Wales. J Nutr Gerontol Geriatr 32, 71–80.

30.

Favaro-Moreira

, Krausch-Hofmann

, Matthys

, Vereecken

, Vanhauwaert

, Declercq

, Bekkering

, Duyck

(2016) Risk factors for malnutrition in older adults: a systematic review of the literature based on longitudinal data. Adv Nutr 7, 507–522.

31.

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.