Clinical Implications of Oral Assessment in Elderly Patients with Acute Heart Failure: A Single-center, Retrospective Study

Yusuke Uemura (  yusuke0307@kosei.anjo.aichi.jp ) Anjo Kosei Hospital https://orcid.org/0000-0002-2262-1452 Rei Shibata Nagoya University Graduate School of Medicine Haruna Ishikawa Anjo Kosei Hospital Ayumi Nagahori Anjo Kosei Hospital Yuta Katsumi Anjo Kosei Hospital Kenji Takemoto Anjo Kosei Hospital Shinji Ishikawa Anjo Kosei Hospital Toyoaki Murohara Nagoya University Graduate School of Medicine Masato Watarai Anjo Kosei Hospital

in the assessment and management of HF patients. In the present study, we investigated the clinical implications of oral assessment in elderly patients with acute HF.

Study population
Patients admitted to Anjo Kosei Hospital for the treatment of HF between October 2018 and March 2019 were reviewed. All patients were diagnosed with HF using the Framingham criteria [19]. We enrolled 77 patients aged 65 years and older who underwent oral assessment during hospitalization.
A medical history was obtained to document past medical history, medications, and co-morbid disease.
Hypertension was de ned as systolic blood pressure (BP) ≥ 140 mmHg or diastolic BP ≥ 90 mmHg on repeated measurements, or receipt of antihypertensive treatment. Diabetes mellitus was de ned as having a blood hemoglobin A1c ≥ 6.5%, 2-hour value ≥ 200 mg/dL (≥ 11.1 mmol/L) on a 75 g oral glucose tolerance test, and/or taking glucose-modulating medication according to the diagnostic criteria of the Japan Diabetes Society [20].
Oral assessment Oral health was assessed by a certi ed dysphagia nurse using the revised oral assessment guide (ROAG) [21]. ROAG includes eight categories: voice, lips, mucous membranes, tongue, gums, teeth/dentures, saliva, and swallowing. Each category was described and rated from healthy (score 1) to severe (score 3).
The total score ranged from eight, representing a normal oral health, to twenty-four, which represents severe oral health problems. The ROAG score was obtained within 1 week of hospitalization when the patient's respiratory state was stabilized without the need for oxygen. In the present study, patients with ROAG scores of 8 were regarded to have a normal oral health and those with higher scores were regarded as having poor oral health.

Assessments for geriatric conditions
Physical functional status was evaluated using the Barthel Index (BI), handgrip, and 10-meter gait speed.
The BI was obtained by ward nurses at admission and at discharge, as previously reported [8]. Changes in the BI were calculated as the difference between the BI on admission and the BI on discharge.
Handgrip and 10-meter gait speed were evaluated by physical therapists before discharge.
Registered dieticians assessed nutritional status. Nutritional status was screened using the controlling nutritional status (CONUT) score and the geriatric nutritional risk index (GNRI) [22] [23]. Laboratory data at admission and body mass index (BMI) at the rst measurement within 72 hours of hospitalization were used for calculation of the scores. Dietary energy intake was assessed by the proportion of nutritional intake from food compared to the predicted calorie requirement. Nutritional intake was calculated based on the food intake for 3 days around the day of oral assessment. The predicted calorie requirement was de ned as the total energy expenditure estimated from the Harris-Benedict equation [24].

Biomarker analysis and echocardiography
Blood samples were obtained at the time of hospital admission. Complete blood counts were performed utilizing a Sysmex XE-5000 analyzer (Sysmex, Kobe, Japan). Plasma brain natriuretic peptide (BNP) was measured with the AIA-2000 enzymatic immunoassay analyzer (TOSOH, Tokyo, Japan). Other biomarkers were measured using a LABOSPECT 008 autoanalyzer (Hitachi Co., Tokyo, Japan). Estimated glomerular ltration rate (eGFR) was calculated by the Modi cation of Diet in Renal Disease formula [26].
Echocardiographic examination was performed by an experienced sonographer using Vivid E9 with XD clear (GE Healthcare, Tokyo, Japan). The images were recorded in a console and analyzed o ine. Left ventricular ejection fraction was calculated using the modi ed Simpson's rule.

Statistical analysis
All analyses were performed using PASW Statistics 21 software (SPSS Inc., Chicago, IL, USA). Continuous variables were presented as the mean ± standard deviation or median (interquartile range). Categorical variables were presented as the count and/or percentage. The student's t-test or Mann-Whitney U-test was used for group comparisons. Univariate correlations between changes in the BI during hospitalization and other variables were investigated using the Pearson's rank correlation test, and then a multiple linear regression analysis was performed. Variables with P < 0.05 in the univariate analyses were incorporated into the multivariable model. In all analyses, P < 0.05 was considered statistically signi cant.

Baseline characteristics
Baseline characteristics of the patients are shown in Table 1. The mean age of patients was 80.0 ± 9.1 years, and 58.4% of patients were men. The mean ROAG score was 9.9 ± 2.2. Poor oral health (ROAG score ≥ 9) was identi ed in 66.2% of the enrolled patients. Details of ROAG evaluations are shown in Table S1.
Patients were divided into two groups based on the ROAG score: a normal oral health group (ROAG score = 8, n = 26) and a poor oral health group (ROAG score ≥ 9, n = 51) ( Table 1). Patients with poor oral health were older and had lower albumin levels. C-reactive protein levels were higher in patients with poor oral health than in those with normal oral health. There were no signi cant differences in gender, BMI, history of HF, etiology of coronary arterial disease (CAD) and stroke, prevalence of hypertension, diabetes and atrial brillation, renal function, hemoglobin, sodium, BNP level, left ventricular ejection fraction, and the use of medications on admission between two groups (Table 1). Association between oral health and geriatric assessments on admission We examined the association between oral health and common geriatric assessments on admission.
Barthel Index, handgrip strength, GNRI, dietary energy intake, and MMSE were signi cantly lower in the poor oral health group than in the normal oral health group. The COUNT score was signi cantly higher in patients with poor oral health than those with normal oral health ( Table 2). There were no signi cant differences between the groups in gait speed for the 10-meter walk. Thus, patients with poor oral health showed a higher prevalence of decreased physical function, undernutrition, and cognitive impairment. Association between ROAG score and activities of daily living (ADL) preservation Efforts to preserve patients' ability to perform ADL during hospitalization are important in the management of acute heart failure. The BI has been reported as a common tool to evaluate the ability to perform ADL. Here, we investigated the changes in the BI during hospitalization. The actual BI values on admission and discharge, and the BI changes in each group are shown in Table 3. Reduction in the BI during hospitalization was signi cant in the enrolled patients in both the normal oral health group and poor oral health group (P < 0.01). Of note, the BI decreased more in patients with poor oral health than those with normal oral health (P < 0.01). Finally, to determine factors associated with ADL preservation during hospitalization in elderly patients with acute HF, we compared the clinical parameters on admission that were associated with the change in the BI during hospitalization using univariate and multivariate regression analyses. Changes in the BI during hospitalization were signi cantly correlated with age, hemoglobin levels, ROAG score, GNRI, and dietary energy intake on admission in the univariate regression analysis. Of those, ROAG score on admission was the only independent predictor of changes in the BI during hospitalization in the multivariate regression analyses (Table 4).

Discussion
The major ndings of this study were as follows: 1) poor oral health, assessed using the ROAG score, is relatively common in elderly patients with acute HF, 2) patients with poor oral health had signi cantly more geriatric conditions than those with normal oral health, and 3) the ROAG score was independently correlated with changes in the BI during hospitalization in elderly patients with acute HF.
Oral health problems have been reported as common health conditions in older adults and hospitalized patients [17] [18] [27] [28]. Several points are known about the relationship between oral health problems and heart failure. Chronic in ammation caused by periodontal diseases is a risk factor for cardiac and cerebrovascular diseases, which are the most major comorbidities of HF [29] [30]. General fatigue, dyspnea, delirium, and sleep disturbance accompanied by decompensated HF might contribute to reducing adherence to oral hygiene. In addition, dry mouth is often a consequence of polypharmacy, particularly as a side effect of cardiovascular agents (angiotensin-converting enzyme inhibitors, betablockers, and diuretics) [31] [32]. However, oral health has been underrecognized in the assessment and management of patients with HF. Our data suggests the importance of oral assessment in the multidisciplinary management of elderly HF patients because poor oral health is highly prevalent and signi cantly associated with decline in physical function during hospitalization.
We have previously demonstrated that hospitalization for HF was signi cantly correlated with decreased BI as an assessment of ADL, and a decreased BI during hospitalization was associated with worse clinical outcomes [8]. It has also been reported that a decline in ADL due to acute HF is an independent risk factor of hospitalization for HF and mortality [33]. Therefore, it is important to identify predictors of ADL decline during hospitalization in patients with HF. It has been demonstrated in elderly patients and patients with HF that age and the nutritional index are associated with ADL decline during hospitalization [33] [34]. Our present study involving elderly patients with HF has shown that the ROAG score serves as a good predictor of changes in ADL, and its predictive ability is comparable to age, nutritional indices, and various other parameters. Therefore, oral assessment of patients with HF using the ROAG score to screen for poor oral health may allow us to provide comprehensive care and rehabilitation to patients with poor oral health at an earlier stage. This may help prevent ADL decline during hospitalization for HF, leading to a better prognosis.
In the present study, nutritional status, assessed by CONUT and GNRI, was worse in the poor oral health group than in the normal oral health group. Furthermore, patients with poor oral health had lower dietary energy intake. Previous studies have demonstrated that poor oral health is associated with periodontal disease, dental caries, hyposalivation, and tooth loss or edentulousness, which pose risks of chewing di culties, decreased masticatory function, and dysphagia [35] [36]. These oral problems may induce a preference for soft and easily chewable food and a need for changes in food texture to prevent aspiration and choking, leading to poor nutritional intake and undernutrition, and nally to sarcopenia, frailty, and decreased physical function [17] [37] [38].
In the acute setting of decompensated HF, nurses have a pivotal role in oral assessment and care, because of the small number of dentists and oral hygienists. Several oral health assessment tools have been developed for non-dental health care professionals [21] [39] [40]. Among these indices, the ROAG is not only a simple and comprehensive assessment but it also has favorable validity and reliability [41].
Oral assessment in patients with acute HF might have several in uences on multidisciplinary disease management. First, patients who require consultation with in-hospital or regional dental health care professionals are screened. Second, nurses receive feedback on oral care from their patients. Third, oral care is one of the fundamental self-care activities after discharge, and thus, the information that an oral assessment provides is useful for educating patients on dental health compliance. Forth, several management strategies could be considered for patients with poor oral health. Nutritionists and dietitians could help to increase the dietary intake of the patients by changing meal content and food texture. Doctors and pharmacists could choose orally disintegrating tablets for the ease of taking medicines.
Although the e cacy of multidisciplinary interventions for patients with poor oral health has not been fully elucidated, we believe that oral assessment could provide useful information for the multidisciplinary management of elderly patients with HF.
The present study had several limitations. First, this was a single-center, retrospective study, and the sample size was relatively small. Future prospective studies are necessary with larger patient populations. Second oral health was assessed by only one certi ed dysphagia nurse. Hence, the results could not be generalized to the routine assessment by ward nurses.

Conclusions
Poor oral health, as assessed by the ROAG, is highly prevalent and oral assessment using the ROAG predicts a decline in physical function during hospitalization in elderly patients with acute HF. Thus, oral assessment during hospitalization could provide useful information for the management of elderly HF patients.
Abbreviations HF: heart failure; BP: blood pressure; ROAG: revised oral assessment guide; BI: Barthel Index; CONUT: controlling nutritional status; GNRI: geriatric nutritional risk index; BMI: body mass index; MMSE: minimental state examination; BNP: brain natriuretic peptide; eGFR: estimated glomerular ltration rate; CAD: coronary arterial disease; ADL: activities of daily living Declarations Ethics approval and consent to participate The study was approved by the ethics committee of Anjo Kosei Hospital (Approval No. R19-032).
Because of its retrospective nature, informed consent was deemed unnecessary according to the national regulation issued by the Japanese Ministry of Health, Labour and Welfare. However, the present study was carried out by the opt-out method of our hospital website.