Multimorbidity and associated outcomes among older adult inpatients with neurological disorders

ABSTRACT Background: Multimorbidity is common among adults and associated with socioeconomic deprivation, polypharmacy, poor quality of life, functional impairment, and mortality. Objectives: To identify the frequency of multimorbidity among older adults inpatients with neurological disorders (NDs), stratify clusters of chronic comorbidities associated with NDs in degrees, and verify whether multimorbidity was associated with demographic data, readmission, long length of hospital stay (LOS), and hospital mortality in this population. Methods: We enrolled patients aged ≥60 years successively admitted to a tertiary medical center with NDs between January 1, 2009, and December 31, 2010. Results: Overall, 1,154 NDs and 2,679 comorbidities were identified among 798 inpatients aged ≥60 years (mean: 75.76±9.12). Women comprised 435 (54.51%) of patients. Multimorbidity was detected in 92.61% (739) of patients, with a mean of 3.88±1.67 (median: 4.0), ranging from 2 to 10 chronic diseases. Patients with epilepsy, dementia, and movement disorders had the highest degrees of clusters of chronic morbidities (>50% of them with ≥5 chronic disorders), followed by those with cerebrovascular and neuromuscular disorders. Multimorbidity was associated with long LOS (p<0.001) and readmission (p=0.039), but not with hospital mortality (p=0.999). Conclusions: Multimorbidity was preponderant among older adults inpatients with NDs, and NDs had a high degree of associated chronic comorbidities. Multimorbidity, but not isolated NDs, was associated with readmission and long LOS. These results support ward-based, neurohospitalist-directed, interdisciplinary care for older adults inpatients with NDs to face multimorbidity.

Although with different meanings, the terms comorbidity and multimorbidity have been used interchangeably, thus clouding the interpretation of scientific publications 1 . However, the creation of the Medical Subject Heading (MeSH) "multimorbidity" in January 2018 indicates the important distinction that should be made in publications and systematic literature searches examining the concepts of multimorbidity and comorbidity 1,2 . While comorbidity refers to diseases (a combination of index diseases with other diseases), as in stroke (as an index disease) having chronic atrial fibrillation, hypertension, dementia, and diabetes as comorbidities, multimorbidity is the combination of two or more chronic diseases in the same individual, and so multimorbidity is related to patients, not diseases 1 . In addition, "no single condition [in multimorbidity] holds priority over any of the co-occurring conditions from the perspective of the patient, healthcare professional, or academic research team" 1 . It is worth clarifying the meaning of the related MeSH term risk factor, which is "an aspect of personal behavior or lifestyle, environmental exposure, inborn or inherited characteristic, which, on the basis of epidemiological evidence, is known to be associated with a health-related condition considered important to prevent" 3 . Population aging is a global phenomenon related to multimorbidity 4 . Additionally, in some lowor middle-income developing countries, aging is growing faster than in developed nations 5 . Multimorbidity is associated with low socioeconomic status 6 . Patient multimorbidity is also closely associated with functional impairment and the consequences of this interplay, i.e., psychological harm, poor quality of life, effects of inappropriate drug use and drug interactions, as well as high mortality 7 . Importantly, the chronic diseases that compose multimorbidity are usually grouped in clusters, forming several disease patterns 8 . Such beyond-chance disease associations may occur when one disease is directly responsible for others (i.e., complicating multimorbidity) or when several diseases share common or correlated risk factors (i.e., causal multimorbidity), which may be biological, socioeconomic, cultural, environmental, or behavioral in nature 9 . Despite all evidence, most epidemiological and clinical studies are designed for a single disease, rather than for multimorbidity 6 . Attention should be given to chronic neurological disorders (NDs) as components of patient multimorbidity, since NDs are common, complex, and severe, and neurological patients are frequently admitted to hospitals 10,11,12 . Persons aged 60 years and older already are more than 30 million individuals in Brazil 13 . Previous studies showed individual NDs were not predictors of hospital mortality, patient readmission, and long length of stay (LOS) in adults admitted to a tertiary medical center in Brazil; however, those studies revealed this population had an extraordinary prevalence of patient multimorbidity 14,15 . The present study aimed to identify the frequency of multimorbidity among older adults inpatients with NDs and degrees of clusters of chronic comorbidities associated with NDs, as well as investigate whether multimorbidity was associated with any particular demographic variable, long LOS, patient readmission, or hospital mortality in this population.

Study population, site, and period
The patients enrolled were those aged 60 years or older and successively admitted to Hospital São Rafael (HSR) between January 1, 2009, and December 31, 2010. HSR is a general tertiary teaching hospital in Salvador, Brazil. When the database for this study was created, HSR used to admit users of several private health insurance services and also of Brazil's public health system (Sistema Único de Saúde -SUS).
Data were collected from electronic medical records, making it easier to find patients with NDs in discharge lists and registries of all neurological procedures. We enrolled older adults inpatients with NDs using the following methods: searches for NDs listed in discharge summaries conducted by our local Information Technology Department (IT-HSR); and IT-HSR identification of registered neurological procedures. After the IT-HSR selection, we carefully examined all written patient health records. We selected older adults inpatients with NDs based on the following criteria: patients with acute NDs admitted to be treated by neurologists; patients admitted with acute clinical disorders owing to an underlying chronic ND that needed care or follow-on therapy by a neurologist; and medical or surgical patients who suffered neurological complications during their hospital stay and required consultation with a neurologist.
Exclusion criteria were as follows: patients who were admitted by or consulted a neurologist but had no NDs; patients whose medical records lacked important data; patients with acute trauma, subarachnoid hemorrhage, central nervous system (CNS) tumor, or other neurosurgical diseases who were directly referred to a neurosurgeon; patients who underwent neurological consultation solely for the diagnosis of brain death due to head trauma, cardiac arrest, or neurosurgical or oncological conditions; and patients transferred to another hospital without an established diagnosis.

Demographics
We considered the influences of age, gender, marital status, and socioeconomic status on patient multimorbidity. Differences in patient multimorbidity with regard to socioeconomic status were estimated comparing SUS and private health insurance users 16,17 .

Neurological disorders
Diagnostic criteria for NDs were based on the Tenth Revision of the International Statistical Classification of Diseases and Related Health Problems (ICD-10) 18 . Neurocognitive disorders were also included if cases of delirium or dementia were diagnosed based on the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV) criteria 19 .

Comorbidities
We used the following criteria to compile important comorbidities that could influence the studied outcomes:  19 .

Patient multimorbidity
Patients were dichotomized according to the presence (two or more chronic diseases) or the absence of multimorbidity (zero or one chronic disease).
Although recorded as index ND, headache, syncope, delirium, CNS infections, and brain injury were not counted to determine multimorbidity. Table 1 shows the chronic disorders included to establish patient multimorbidity.

Degrees of clusters of chronic comorbidities related to index neurological disorders
In order to estimate the degrees of clusters of chronic comorbidities, which accounted for patient multimorbidity associated with each index ND, we summed all chronic diseases of individual patients. This sum included the index ND (if a chronic disorder) and other chronic NDs, plus chronic clinical comorbidities; we then divided the patients into five categories as follows: 0=no chronic diseases (in this situation, the index ND was an acute disorder and patients had no chronic comorbidities); 1=one chronic disease (the index ND could be either an acute ND with one chronic comorbidity or a chronic disease without a chronic comorbidity); 2=two chronic diseases; 3=three chronic diseases; 4=four chronic diseases; and 5=five or more chronic diseases.

Associations of patient multimorbidity
Univariate analysis was performed to detect associations of patient multimorbidity with patient demographic data, long LOS, readmissions, and hospital mortality.

Readmission and long length of stay
Hospital readmissions during the 2-year study period were recorded. The median LOS for the entire population was 9 days, hence hospitalizations of 9 days or longer were conventionally considered long LOS.

Hospital mortality
Hospital mortality was defined as patient death during hospitalization.

Statistics
Quantitative variables with normal distribution were reported as means and standard deviations, whereas variables with non-normal distribution were expressed as medians and interquartile ranges. Normal variables were identified by graphic analysis and the Shapiro-Wilk test. Categorical variables were reported as frequencies and percentages.
We used Student's t-test to perform bivariate comparisons between groups for numerical variables with normal distribution.

Ethics
This study was approved by the HSR Research Ethics Committee (no. 8/11) on August 25, 2011.

Demographics
We selected 798 adults with NDs who were admitted to HSR. The mean age of patients was 75.76±9.12 years Neurological disorders and comorbidities: frequency and association with patient multimorbidity In our population, 555 (70%) individuals had primary NDs diagnosed by neurologists and identified in discharge summaries. The remaining 243 (30%) patients were admitted with clinical complications of underlying chronic NDs or developed neurological complications during their hospital stay. Overall, 312 (39%) patients were affected by more than 1 ND, totaling 668 additional NDs classified as neurological comorbidities in this population. Therefore, we found 1,154 NDs among 798 older adults inpatients and a wide range in the frequency of NDs (0.3 to 50.8%), with cerebrovascular disease being the most common (50.8%), as shown in Figure 1. We identified 2,679 comorbidities and an extensive range in the frequency of comorbidities (5.6 to 84.5%), with hypertension (84.5%) and diabetes (57.5%) being the most common ( Figure 2). The mean number of chronic morbidities was 3.875±1.666 per patient, with a median of 4.0. Total patient multimorbidity amounted to 739 (92.6%) chronic morbidities, ranging from 2 to 10 (identified in one patient). Figure 3 shows the degrees of clusters of chronic comorbidities related to the most prevalent index NDs, whose frequencies were over 5% among the 798 patients. In this sample, more than 50% of patients admitted with dementia, epilepsy, and movement disorders (mainly Parkinson's disease -PD) had 5 or more chronic diseases (including these index NDs). The intermediate group for the number of chronic morbidities included inpatients with cerebrovascular and neuromuscular diseases (mostly diabetic polyneuropathy), with 45% of them having 5 or more chronic diseases. As expected, only up to 30% of patients admitted with syncope or headache had 5 or more chronic diseases.

Length of stay
The median hospital LOS for these patients was 9 days, with an interquartile range of 1 to 20 days; 409 (51%) patients thus had an LOS of ≥9 days (95%CI 48-55%).

Hospital mortality
The hospital mortality rate for older adults inpatients with NDs in our sample was 18% (95%CI 15-21%).

Association of patient multimorbidity with demographic data, patient readmission, long length of stay, and hospital mortality
Patient multimorbidity was strongly associated with long LOS (p<0.001) and patient readmission to the hospital in the study period (p=0.039). However, we found no association between patient multimorbidity and demographic data or hospital mortality (Table 2).

DISCUSSION
We investigated a population of adults with NDs who were admitted to a tertiary care hospital in order to evaluate patient multimorbidity and its burden. Our method of data retrieval was unusual because health systems are largely designed for individual disorders, and this study found a high level of patient multimorbidity. We studied a population of very old patients, as their mean age was 75 years, with a median of 76 years, and 25% of them were aged 82 years or older. In addition, these patients were suffering from very severe NDs, since HSR is the largest tertiary medical center in Salvador and one of the main facilities to treat complex NDs in our region. The studied population included complex disorders in that almost all patients were classified as having multimorbidity (92.6%). Notably, 85 (10.83%) inpatients were SUS users and hence we were able to compare outcomes between two socioeconomically distinct populations. The mean number of chronic disorders among the older adults inpatients admitted with NDs was 3.875±1.666 per patient (median: 4). This figure is similar to that found in a population of adults in which "multimorbidity as specific combinations of chronic conditions showed an effect on mortality, which would be higher than the risk attributable to individual morbidities" 23 .   Nevertheless, we recorded patient death only during hospitalization, and patients with multimorbidity had the same hospital mortality rate (17.9%) as the total study population. This short period for recording mortality might have underestimated its frequency. We also registered degrees of chronic comorbidities that accounted for patient multimorbidity, which demonstrated most patients had five or more chronic diseases. Most older adults inpatients with epilepsy, dementia, and movement disorders as index NDs had five or more chronic diseases. In fact, patient multimorbidity is common in adults admitted with epilepsy, who frequently suffer from chronic disorders such as cerebrovascular disease, dementia, CNS tumors, and uremic encephalopathy due to renal failure, and also have risk factors contributing to drug toxicity or drug withdrawal syndrome and chronic alcohol-induced disorders 24 . Dementia is often associated with patient multimorbidity, and multimorbidity strongly influences the clinical progression of dementia 25 . Dementia is considered one of the most common disorders among the so-called geriatric syndromes, which also encompass polypharmacy, urinary incontinence, malnutrition, depression, sarcopenia, and frailty 26 .
Mental and cognitive multimorbidity is independently associated with functional dependence, even more strongly than physical multimorbidity 27 . Among movement disorders, PD is largely associated with patient multimorbidity. This disease itself is very complex, with motor and non-motor symptoms, such as sleep disturbances, gastrointestinal disorders, and cardiovascular disease 28 . Moreover, patients with PD usually have other chronic disorders (anxiety, arthritis, cancer, emphysema, diabetes, heart disease, hypertension, depression, stroke, and thyroid disorders), which require special care and lead to higher healthcare costs 29 . Neuromuscular and cerebrovascular diseases were among the NDs with an intermediate to high degree of patient multimorbidity, with 40 to 50% of these patients having 5 or more chronic diseases. Peripheral neuropathy is very common in older adults inpatients, but it is often an underdiagnosed condition, particularly diabetic neuropathy, usually accompanied by several comorbidities, such as cardiovascular disease, dementia, pulmonary chronic disorders, renal failure, peripheral vascular disorders, cerebrovascular disorders, and peptic ulcer disease 30 . Peripheral neuropathy is also linked to cancer and its treatment, as well as to intensive care unit-acquired weakness, both with acute or subacute onset but usually with chronic progression and associated with patient multimorbidity 31,32 . Cerebrovascular disorder was one of the most common NDs associated with an intermediate to high degree of patient multimorbidity. Cerebrovascular disorder is usually associated with hypertension, diabetes, other cardiovascular disorders, dyslipidemia, obesity, peripheral vascular disorders, migraine, depression, and dementia 33 . Among the patients admitted with delirium or who developed delirium during hospitalization, 30% had 5 or more chronic diseases. Delirium is an acute condition associated with patient multimorbidity and polypharmacy, commonly found in the treatment of chronic disorders, particularly in frail older adults patients 34 . Patients admitted with headache and syncope had the lowest frequency of chronic comorbidities, and five of the patients admitted with syncope had no chronic comorbidities. Chronic headache, particularly migraine, is associated with chronic disorders, such as psychiatric, cerebrovascular, and endocrine diseases; these patients are older and have a higher frequency of headache, medication overuse, and lower socioeconomic status 35 . Syncope had the weakest association with patient multimorbidity; however, this condition, particularly in adults, is an occasional manifestation of several disorders, such as cardiac diseases (heart failure, coronary disease, atrial fibrillation), diabetes, kidney disease, peripheral nerve disorders, chronic alcoholism, hypertension (postural hypotension due to antihypertensive drugs), and some neurodegenerative diseases, e.g., multiple system atrophy, PD, Lewy body dementia, and cerebrovascular disorders 36,37 . Patient-centered outcomes, such as readmission, LOS, and hospital mortality, were analyzed in this study with respect to multimorbidity, rather than a single ND diagnosis. Notwithstanding the high rates of patient readmission, hospital mortality, and long LOS in this population, no isolated ND diagnosis was recognized as a predictor of readmission or hospital mortality 14,15 . However, we demonstrated patient multimorbidity was associated with hospital readmission and long LOS, reinforcing the hypothesis that the effects of multimorbidity on clinical outcomes result from the synergistic interplay of chronic disorders, lifestyle factors, aging, and consequent mental and physical impairment, rather than the sum of their expected individual effects 38 . Finally, older adults inpatients with NDs who have limited physical activity are vulnerable to functional decline and reduced capacity for recovery; they are also usually submitted to procedures that could lead to additional functional decline 39 . Indeed, in hospitals, individual medical disorders have been receiving more attention than the logistic and organizational issues for ward-based care of adults, even though ward-based patient care might provide better support and more efficiency when treating older adults inpatients with multimorbidity 40 . The assistance of neurohospitalists in hospitalist-directed interdisciplinary teams is therefore recommended, as it seems to foster better outcomes, including short LOS, prevent harms associated with hospitalization, and reduce healthcare costs when treating older adults patients 40,41,42,43 .

Limitations of this study
This was a retrospective study in which we only recorded readmissions to the same hospital and registered patient death only during their hospital stay (hospital mortality rate). Furthermore, our database was created before the restructuring and improvement in the management of neurological care at HSR, which recently incorporated a neurological intensive care unit.

Strengths of this study
Our sample size, the methods and quality of patient data retrieval, and the inclusion of the index ND and other NDs when assessing multimorbidity could be considered robust features of this study.
Multimorbidity was very common among older adults inpatients admitted with NDs, which were associated with a high mean number of chronic comorbidities. Index NDs had a high degree of associated chronic comorbidities, and multimorbidity, rather than single NDs, was associated with long LOS and patient readmission. Our results support the management of older adults inpatients with NDs in ward-based, neurohospitalist-directed, interdisciplinary care settings in order to tackle multimorbidity.