Abstract
INTRODUCTION
Dementia is the generic term used for different forms of cognitive degeneration. According to the International Classification of Diseases (ICD), dementia is a syndrome resulting from chronic or progressive disease of the brain with impairment of several high cortical functions including memory, orientation, thinking, perception, calculating, learning, language, and judgment [1].
Alzheimer’s disease (AD) caused by neurodegenerative processes in the brain is the most common type of dementia in the elderly accounting for two-thirds of dementia cases [2]. Vascular dementia caused by circulatory disturbances accounts for 15% to 20% of dementia cases with the remaining cases being mixed forms of the degenerative-vascular type or other forms secondary to diseases like HIV, primary Parkinson syndrome, chronic intoxication, and others [1, 2].
Currently about 1.3 million people suffer from dementia of any type in Germany [3]. Prevalence of AD depends on age and rises from about 1% in the 65- to 69-year-old age category to more than 20% in the 90- to 94-year-old age category [2, 5]. Prevalence is higher among women than men [2–5]. Both prevalence and incidence increase with age [5, 6]. In Germany, it is estimated that between 120,000 and 160,000 new AD cases are diagnosed each year [2].
However, exact prevalence of AD remains unknown, whereas AD is highly likely the most common form of dementia [7]. AD has a discrete disease onset, progresses slowly, and is currently incurable. The course of the disease is characterized by increasing loss of cognitive function and reductions in activities of daily living (ADL). With AD progressing patients are reaching daily need of care. Current therapeutic options with medications and psycho-social interventions are limited to reducing disease symptoms, preserving cognitive and non-cognitive functions, and improving quality of life [8]. There are strong associations between development of AD and several diseases. Vascular diseases, elevated blood pressure, and diabetes are identified as contributing risk factors for developing AD [9].
Burden of medical comorbidities in AD is greater than in matched individuals without AD [10–12]. Furthermore, in patients affected by AD physical comorbidities make also important contributions to cognitive and functional decline in patients with AD [9, 12].
Generally, number of comorbidities was found to be significantly associated with lower cognition [13, 14]. In a 2-year follow-up, Solomon et al. found only a tendency that comorbidity, measured with a general cumulative comorbidity score, increased functional decline [14]. Atrial fibrillation, systolic hypertension, and angina were associated with more rapid decline on both the Clinical Dementia Rating-Sum of Boxes and Mini-Mental State Examination, while history of coronary artery bypass graft surgery, diabetes, and antihypertensive medications were associated with a slower rate of decline [15]. Effects of cardiovascular medications on rate of functional decline in AD was analyzed by Rosenberg et al.: Use of statins and beta blocker were associated with delay of functional decline, while diuretic use was associated with a faster rate of functional decline[16].
Comorbidities are both associated with cognitive decline, as well as with poorer self-care and decreased mobility and hence may adversely affect patients’ ADL and independence. Dijkstra et al. showed that cerebrovascular, cardiovascular, or musculoskeletal comorbidities, as well as malignancies, urinary disorders, or cachexia are significant predictors of care dependency for AD [16, 17]. AD patients also show higher rates of falls and an excess of hip fractures [18].
Identifying relevant factors deteriorating the disease and inducing need for nursing care would be of high relevance for healthcare planning both from an individual and societal point of view as well as for better recognizing severe cases and identifying early treatment options.
Using secondary data for analyzing the course of AD depends both on the criteria used to identify AD and the granularity of data available to approximate progression. In general, claims data substantially differ from assessments in clinical surveys [19], however, are considered a reasonable source to generate hypotheses [20].
Identifying AD disease progression in German Statutory Health Insurance (SHI) claims data is not directly feasible via coded information. However, data from clinical trial indicate that nursing home placement is closely related to dementia progression [21]. Furthermore, in a recent German study, severity of dementia and cost of SHI-financed long-term care were significantly different between mild and moderate dementia (960€ and 2,367€ resp.), where SHI-covered home health care costs increase from 522€ to 962€. With the robust face validity of need for nursing home as a deterioration indicator for severe dementia [21], an approximation of disease progression by analyzing need for nursing care seems justified. A scientific cooperation with the DAK-Gesundheit, a German SHI covering more than 6 million insurees, allowed an assessment of medical comorbidities, patterns of drug use, and clinical course in diagnosed AD based on DAK-Gesundheits’s claims related to deterioration.
Objective
Physical comorbidities are important contributors to cognitive and functional decline in AD and are potential factors for extending need for nursing care. The main objective of this study was to assess factors related to patient’s dependency and to identify potential early predictors of need for care a year prior to the AD diagnosis within a cohort study, modeling the time to increased need for nursing care.
MATERIALS AND METHODS
Study setting
This analysis is based on routine data from DAK-Gesundheit, which cover anonymized datasets from 2005 to 2008 of 2.1 Mio insurees aged 50 years and above comprising approximately 20,000 insurants with specific AD diagnosis. The evaluation is based on a routine dataset from a large German SHI. The anonymized data included information on ICD diagnosis, hospitalizations, drug prescriptions, outpatient contacts, and sickness benefits provided by the health insurance for nursing care related toAD.
Following a cohort study design, all subjects with incident dementia and no level of ambulatory nursing care were identified and followed-up. Dementia was defined as at least having one ambulatory or in-hospital diagnosis within the study period. Ambulatory diagnoses of German statutory health insurance data are related to diagnoses given in a specific quarter of a year whereas hospital diagnoses are related to an individual hospital stay. Diagnoses are coded using the International Classification of Disease (ICD-10 GM), 10th revision for Germany [1]. Table 1 gives an overview of all ICD-10 GM codes considered during the selection process. The subgroup of patients with AD was defined as having at least one diagnosis of AD without any diagnosis indicating other specific forms of dementia.
Insurees were eligible for entering the cohort after at least one year of available person time. Potential quarter of cohort entry was the first quarter with at least one diagnosis of AD after becoming eligible. To be classified as an incident case of AD, insurants were required to have had no diagnosis of dementia and no prescriptions of antidementia drugs in the year prior to the quarter of potential cohort entry. In addition, only insurants with no level of ambulatory nursing care at the start of cohort-entry were included. All insurants entered the cohort at the 45th day of the quarter in which the first AD diagnosis was made. All cohort members were followed up until either a first level of ambulatory nursing care was assigned, the end of study period (November 30, 2008) or SHI coverage reached, or death occurred.
Exposure assessment considered the year (i.e., the 4 quarters) preceding the quarter of cohort entry and covered gender, age, number of hospitalizations, number of different drugs prescribed, number of physician contacts, and a total of 240 diagnostic subclasses defined by International Classification of Diseases 10th revision (ICD 10; 3-digit) and 96 drug categories based on Anatomical Therapeutic Chemical Classification System (ATC).
In Germany, the level of care needed (“Pflegestufe”) represents the extent of daily nursing care needed by the patient and is classified into three severity levels based on the care time— for Level 1 requiring at least 1.5 hours per day. Corresponding to the assessed level of care, SHI either issues a payment to the caregiver or the caregiver receives benefits for ambulatory care administered by a nurse.
Statistical analysis
Time from cohort entry until disease deterioration (i.e., assignment of a level of ambulatory care) was modeled using COX-proportional hazard regression. In these models, the diagnostic categories and drug classes were entered as categorical variables in a stepwise selection model. Variables remained in the model, if they were significantly associated with deterioration (p < 0.05) and were otherwise removed from the model. All other variables (age, gender, number of previous hospitalizations, etc.) were forced into the model. All calculations were performed with SAS 9.2 (SAS Institute, Cary, North Carolina).
RESULTS
The characteristics of the study population are displayed in Table 2. 10,005 patients entered the cohort fulfilling the study criteria. Of them 31.4% were males and 68.6% females. Mean age at point of start was 75.1 years. Mean follow-up time was 506.9 days. For 2,696 patients (26.9%) the event, i.e., the assignment of the first level of ambulatory nursing care, could be identified within the study period.
Rates of need for nursing care increased substantially with increasing age and were similar in males and females by 6% with every year age (Fig. 1). In the COX-proportional hazard model, number of hospitalization and number of different drugs used was a significant indication of aggravation, whereas more than 5 outpatient visits per year indicates a predictor of reduced need for care (compared to 0 to 4 visits). Gender did not indicate significant influence on progression (Table 3).
The model identified comorbidities with increasing (up-arrow) and decreasing (down-arrow) influence on the risk of decrease progression, as shown in Table 4. Conditions, which increase significantly the risk of need for nursing care, were malignant neoplasms of illdefined, secondary, and unspecified sites, malnutrition, renal failure, chromosomal abnormalities, illdefined and unknown causes of mortality, injuries to abdomen and lower back, injuries involving multiple body regions, and burns confined to eye and internal organs.
Behavioral syndromes associated with physical factors, disorders of nerves, glaucoma, acute upper respiratory infections, other soft tissue disorders, diseases of male genital organs, symptoms of digestive system, encounters due to examination and investigation, and potential health hazards related to communicable diseases were associated with a decrease in risk (Table 4).
Among prescribed drugs, significant increased risks were observed for drugs used in diabetes, diuretics, e.g., used in heart failure, preparations for wound treatment, antiseptics, and analgesics. Vasoprotectives and systemic antibiotics were associated with a decreased risk of needing nursing care(Table 5).
DISCUSSION
Physical comorbidities are relevant contributors to cognitive and functional decline in AD and the need for nursing care. The focus of this analysis has been the identification of predictors of disease progression defined as first assignment of long-term care. In this analysis of a German SHI database, several medical conditions were identified, which were associated with increasing or decreasing risk in patients with AD for reaching benefits for long-term nursing care, either paid to the caregiver or received for ambulatory care administered by a nurse.
The findings indicate that hospitalization is a significant predictor for deterioration of AD. This is supported by other studies that showed the association between higher medical comorbidity and lower cognition [13, 22].
The medical conditions, which were identified as significant factors for increasing the risk of need for nursing care covered a variety of diseases, which can be interpreted as directly linked to cognition, like malnutrition or diseases, which affect the functions of daily life and need of nursing care, like injuries, or pain by nerve disorders or soft tissue disorders.
Some findings comply with previous scientific results. In our study, injuries were associated with an elevated risk for nursing care. It is known that AD patients show higher rates of falls and an excess in hip fractures [18]. However, this analysis did not find “injuries to the hip and thigh” as significant contributors to the COX-proportional hazard regression models.
Also higher rates of glaucoma were noted in [23, 24]. In addition, higher prevalence of glaucoma in patients with AD visual field defects increase risk dependency on caregivers. In our model, glaucoma decreases the risk for need of care by 15%, which may be related to under-diagnosis of glaucoma in patients with poor health status.
Among conditions that increased the risk of need for care were “malign neoplasm if illdefined, secondary and unspecified sites”. Other malign conditions were not significant in the stepwise model. This finding may be explained by information bias, i.e., that these diagnoses were more often coded in patients with worse general health conditions and prognosis.
Prescribed drugs for diabetes medication, diuretics, preparations for wound treatment, antiseptics, and analgesics increased the risk for need of care, while vasoprotectives and systemic antibiotics decreased the risk. The main decreasing function of prescribed drugs in this statistical model could be interpreted as an indicator for health care utilization, similar to the decreasing association of office-based visits and significant contributors of ICD groups “persons encountering health services” and “persons with potential health hazards relate to communicable diseases”. Availability and utilization of these health care services could be a relevant factor for delay of need for care.
It is important to note, that while we have demonstrated associations between risk for need of care, and medical conditions or prescribed drugs, we have not demonstrated a causal relationship. Poorer cognition and function or increased other syndromes could lead to physical decline through self-neglect, improper medication, or monitoring of chronic conditions. Also general health could be related to cognitive health by placing additional stressors on the brain with potentially impacting in both directions. In addition, regarding the analysis of patients’ prescriptions and assigned disease codes, the disease prognosis and treatment predication could influence the selection of medication or physicians’ judgment of relevance of comorbidity in clinical practice. Conditions which were considered as minor relevant in patients with better prognosis would appear as protective factors in our analysis.
A major limitation of this analysis is the underlying secondary dataset. In general, AD-related analyses of claims data are affected by a low level of AD available. Taylor et al. analyzed both a clinically obtained AD database and US Medicare claims data [20]. Compared to the linked clinical study patients, Medicare identified 85% of dementia cases and 64% of AD patients. Despite false positive and negative cases Medicare claims data authors concluded analyses for prevalence estimates being reasonable whereas cost estimates might be overstated. Ziegler et al. analyzed age- and gender-specific prevalence and incidence rates of dementia in German SHI data estimating about 1.07 million moderately or severely demented people in Germany in 2007 [25]. They assumed that the identified low level of AD of only 23% in particular reflected missing diagnostic opportunities and awareness for AD in 2002.
Due to the low level of AD diagnoses being available in claims databases, burden and cost related to the AD subgroup are overestimated and are a substantial issue in cost of disease studies [20]. However, for identifying or predicting progression this subgroup is particularly of high relevance to decision maker in order to be able to most appropriately address the needs of these patients and their carers for treatment and care planning.
This analysis showed a variety of medical conditions associated with increased or decreased risk. Some could be explained by known scientific findings from previous studies, while some are not directly allegeable. AD patients with comorbidities that increase the risk of deterioration should be monitored with special attention, as they may be under an increased risk of care dependency. However, better understanding of clinical factors involved in the risk of care dependency is needed before definitive conclusions and recommendations can be made.
