1 INTRODUCTION

According to the 2018 Japanese government's Vital Statistics,1 pneumonia and aspiration pneumonia are the third leading causes of death. The aspiration of oral bacteria with saliva or food is the predominant cause of pneumonia in older adult patients.2, 3 Yoneyama et al4, 5 showed the effectiveness of professional and mechanical oral cleaning by dentists or dental hygienists in the prevention of pneumonia when undertaken in addition to self-oral cleaning by residents or by caregivers in special nursing homes in Japan. Subsequently, oral health management (OHM) by oral health professionals was included as a service in the long-term care (LTC) insurance system of Japan, whereby dental hygienists provide dentist-supervised oral health care for residents of LTC facilities. OHM means that the nurse selects the resident, and the dental hygienist provides oral hygiene (hereafter professional OHM) at least twice a month and instructs the nursing staff for those who have given consent. This service does not include dental visits for dental treatment.

Since the study by Yoneyama et al4, 5 was conducted approximately 20 years ago, the necessity for oral healthcare (OHC) provision is more crucial than before because of the increased number of older residents in LTC facilities, high demand for nursing care,6 greater number of older adult patients with dementia,7 and most notably, more teeth present in older individuals8, 9 than was observed 20 years ago. However, there have been no large-scale reports in Japan since the report by Yoneyama et al4, 5 in 2002. Therefore, there is limited available information on the current status of professional OHM and its effects on the incidence reduction in pneumonia in residents at LTC facilities.

Accordingly, we conducted a 1-year longitudinal multicentre study in Japanese LTC facilities to test the hypothesis that professional OHM by dental hygienists is associated with a lower incidence of pneumonia among residents of Japanese LTC insurance facilities, in which the number of older adults with many teeth and severe health care needs has increased.

2 MATERIALS AND METHODS

Thirty members of the Special Committee of the Japanese Society of Gerodontology conducted a workshop to discuss the concept of the prospective cohort study and define the evaluation criteria for the contents of the survey. The facilities to be surveyed were LTC insurance facilities throughout Japan, which were visited by the Special Committee members of the Japanese Society of Gerodontology. The directors and staff of the LTC facilities were briefed about the study, and permission for study participation was obtained from the directors of 37 facilities situated throughout Japan. A baseline survey was conducted between October 2018 and February 2019. One year after the baseline survey, permission to conduct a follow-up survey was obtained from the directors of 22 of the 37 facilities. Individuals receiving parenteral feeding and those who did not require professional OHM at the time of the baseline survey were excluded from this study.

This study was conducted in accordance with the ethical principles of the Declaration of Helsinki and was approved by the Ethics Committee of the Japanese Society of Gerodontology (approval numbers 2018-1 and 2019-3). Written informed consent for participation in the follow-up survey was obtained from participants or their proxies.

2.1 Baseline surveys 2.1.1 Questionnaire administration

Data on age, sex, body mass index (BMI), medical history (pneumonia, including aspiration pneumonia, stroke, diabetes mellitus and respiratory disease (chronic obstructive pulmonary disease, asthma and other)), willingness to engage in oral hygiene behaviours and the provision of professional OHM by dental hygienists were obtained with the help of the facility staff using a structured questionnaire. Professional OHM by dental hygienists was indicated if the facility staff could not maintain optimal oral health of the residents. A facility nurse determined the need for dental hygienist-delivered professional OHM for residents based on the oral hygiene management manual published by the Japanese Society of Gerodontology. Ultimately, professional OHM was provided by a dental hygienist if the individual or family consented to the procedure. Activities of Daily Living were assessed by a facility nurse using the Barthel Index (BI).10 Cognitive impairment was assessed by a trained specialist using the Clinical Dementia Rating (CDR) scale.11

2.1.2 Assessment of oral conditions

The number of dentists and dental hygienists who visited each facility to conduct assessments varied according to the size of the facility. In total, 50 dentists and dental hygienist conducted assessments, ranging from three to ten per facility. They were trained in the evaluation criteria and had no prior knowledge of the participants’ existing oral health status. During the survey, at least one researcher with survey experience always participated, to confirm that the standards were consistent. The dentists and dental hygienists who conducted the actual survey were not informed about the availability of oral health care for the residents. Only mouth mirrors and headlights were used to assess the oral conditions. The number of teeth present was defined as the number of erupted teeth in the oral cavity, and those who had no teeth present were defined as edentulous. Residents who used a denture during meals were defined as prosthesis users.

The Oral Health Assessment Tool (OHAT) is a convenient oral screening tool for nurses and nursing staff to assess oral problems in people with disabilities and those in need of care.12 This study used the Japanese version of the OHAT (OHAT-J), which has been evaluated for reliability and validity.13 The OHAT-J was administered by a dentist. In addition, the Tongue Coating Index (TCI), which assesses oral health status by the degree of attachment of the tongue coat based on a visual inspection,14 was used. Oral function was assessed using oral diadochokinesis, modified water swallow test and repetitive saliva swallowing test.

2.2 Follow-up questionnaire

Information on pertinent events during the observation period was obtained from the staff of each facility. These events included the incidence of pneumonia, hospitalisations (and causes), discharge from the facility and death.

2.3 Statistical analysis

Descriptive characteristics of the study participants (Table 1) were compared according to their need for professional OHM. Based on the necessity for professional OHM as determined by the nurse, and consent provision by the participants or their families, participants requiring professional OHM were stratified into two groups (those not requiring oral hygiene management and those requiring oral hygiene management). Professional OHM was provided by a dental hygienist. Intergroup comparisons of the participant characteristics were conducted. Student's t test was used to assess differences in normally distributed continuous data, while the Mann-Whitney U test was used for non-normally distributed continuous data. Cross tabulations and chi-squared tests were used for categorical variables.

TABLE 1. Characteristics of study participants according to the requirement for oral health management Variables Overall (N = 504) No requirements for oral hygiene management(N = 155) With requirements for oral hygiene management(N = 349) P-value Mean ± SD/N (%) Median, [Q1, Q3] Mean ± SD/N (%) Median, [Q1, Q3] Mean ± SD/N (%) Median, [Q1, Q3] Age, years 87.4 ± 7.8 88.0 [83.0, 93.0] 86.4 ± 7.6 87.0 [82.0, 92.0] 87.8 ± 7.9 88.0 [83.0, 93.5] .059 Sex (female) 441 (81.5) 124 (80.0) 287 (82.2) .537 Barthel Index 31.7 ± 26.0 30.0 [5.0, 50.0] 39.1 ± 27.7 40.0 [10.0, 60.0] 28.4 ± 24.6 25.0 [5.0, 45.0] <.001 Body mass index 20.5 ± 3.6 20.1 [17.8, 22.7] 21.1 ± 4.1 20.9 [17.7, 23.6] 20.2 ± 3.4 19.9 [17.9, 22.3] .009 Clinical dementia rating 0, 0.5 49 (9.9) 20 (12.9) 29 (8.5) .190 1 89 (17.9) 33 (21.3) 56 (16.4) 2 131 (26.4) 37 (23.9) 94 (27.6) 3 227 (45.8) 65 (41.9) 162 (47.5) Onset of pneumonia/1 year, N (%) 61 (12.1) 22 (14.2) 39 (11.2) .375 Oral conditions Willingness for oral hygiene, N (%) 267 (53.2) 96 (62.3) 171 (49.0) .007 Oral Health Assessment Tool-J Score 2.9 ± 2.2 2.0 [1.0, 4.0] 3.1 ± 2.0 3.0 [1.0, 4.0] 2.8 ± 2.3 2.0 [1.0, 4.0] .04 Tongue Coating Index 20.9 ± 22.6 11.1 [0.0, 33.3] 21.1 ± 22.6 16.7 [0.0, 33.3] 20.8 ± 22.6 11.1 [0.0, 33.3] .8 Number of teeth present 9.9 ± 9.4 7.5 [0.0, 18.0] 11.0 ± 9.5 9.0 [0.0, 20.0] 9.4 ± 8.9 7.0 [0.0, 17.0] .136 Edentulous, N (%) 171 (34.2) 45 (29.4) 126 (36.3) .152 Prosthesis use, N (%) 265 (52.6) 84 (54.2) 181 (51.9) .699 Medical history Stroke, N (%) 190 (37.8) 55 (35.5) 135 (38.7) .561 Diabetes mellitus, N (%) 82 (16.3) 23 (14.8) 59 (16.9) .694 Respiratory diseasea, N (%) 75 (14.9) 19 (12.4) 56 (16.0) .342 Note P < .05 was considered statistically significant. Q1, first quartile; Q3, third quartile; SD, standard deviation. a Respiratory disease: chronic obstructive pulmonary disease, asthma, other (without pneumonia)

To investigate the factors associated with pneumonia incidence among study participants, the characteristics of the study groups were compared based on whether they had had pneumonia during the follow-up period. A multivariable Poisson regression analysis with robust standard errors was used to estimate the relative risk (RR) of having pneumonia during that period. The main exposure variable was professional OHM provision. Models were adjusted for variables reported to be associated with pneumonia, including age, sex, BI, BMI, CDR, stroke, diabetes and respiratory disease. To examine the possibility of multicollinearity, correlations between each factor were checked. As we collected data from multiple institutions, we conducted a likelihood ratio test comparing the multilevel model with a single-level model containing the same predictors. As logistic regression might have overestimated the risk of association in a high incidence of outcomes,15 a Poisson regression model with robust standard errors was used.16 Statistical analyses used Stata version 16.1 (StataCorp, College Station) and SPSS Statistics version 26 (IBM), with the significance level set at P < .05.

3 RESULTS

Of 889 residents in 37 facilities surveyed at baseline, 525 from 22 facilities (59.1%) participated in this follow-up study (Figure 1). Twenty-one residents who required parenteral feeding and 155 who did not require OHM were excluded from the analysis dataset. The final analysis dataset included data from 349 individuals (82.2% women, mean age: 87.8 ± 7.9 years) who required professional OHM. Eighteen (7.5%) participants of the 238 (68.2%) who were provided professional OHM developed pneumonia, and of the 111 participants who were not provided with professional OHM, 21 (18.9%) developed pneumonia. In addition, of the 155 participants who did not require professional OHM, 22 (14.2%) developed pneumonia.

Flow chart of study participants. A baseline survey of 889 residents of long-term care insurance facilities was conducted. For follow-up, 525 (59.0%) residents gave consent to participate. Of these, 21 (2.3%) participants receiving parenteral feeding at the time of the baseline survey were excluded. The remaining 504 (56.7%) participants were divided into groups based on their need for oral health management (OHM) by a dental hygienist. Of the 349 (69.2%) residents who needed OHM by a dental hygienist, 238 (68.2%) were provided the service, and 111(31.8%) were not provided the service. Among the 238 residents who were provided OHM by a dental hygienist, 18 (7.5%) residents had pneumonia, and 220 (92.5%) residents did not during that year. Among the 111 residents who did not receive OHM by a dental hygienist, 21 (18.9%) residents had pneumonia, and 90 (81.1%) residents did not during that year

Table 1 lists the baseline characteristics of the study participants. The group that required professional OHM had a significantly lower BI score, BMI, proportion of participants willing to engage in oral hygiene care and OHAT-J scores than the participants in the group that did not require professional OHM.

Table 2 shows that 111 (31.8%) of the 349 residents were considered by the nurses to require OHM, but were not provided with professional OHM because they could not give consent. The participants who received professional OHM had significantly lower rates of pneumonia and a lower TCI over the 1-year follow-up period than those who did not receive professional OHM.

TABLE 2. Intragroup comparison by the requirement of oral health management according to the provision of oral health management Variables With requirements for oral hygiene management (N = 349) P-value Not provided (N = 111) Provided (N = 238) Mean ± SD/N (%) Median, [Q1, Q3] Mean ± SD/N (%) Median [Q1, Q3] Age, years 88.7 ± 7.5 89.0 [85.0, 94.0] 87.4 ± 8.0 88.0 [82.0, 93.0] .135 Sex (female), N (%) 92 (82.9) 195 (81.9) .881 Barthel Index 29.4 ± 28.1 20.0 [5.0, 50.0] 28.0 ± 22.9 25.0 [5.0, 45.0] .927 Body mass index 20.4 ± 3.1 20.2 [18.1, 22.4] 20.2 ± 3.5 19.8 [17.8, 22.3] .628 Clinical dementia rating 2.2 ± 0.8 2.0 [2.0, 3.0] 2.1 ± 0.9 2.0 [1.0, 3.0] .537 0, 0.5 4 (3.9) 25 (10.5) 1 20 (19.4) 36 (15.1) .207 2 29 (28.2) 65 (27.3) 3 50 (48.5) 112 (47.1) Onset of pneumonia/1 year, N (%) 21 (18.9) 18 (7.6) .003 Oral conditions Willingness for oral hygiene, N (%) 59 (53.2) 112 (47.1) .303 Oral Health Assessment Tool-J Score 2.8 ± 2.2 2.0 [1.0, 4.0] 2.8 ± 2.4 2.0 [1.0, 4.0] .547 Tongue Coating Index 27 ± 25.4 22.2 [0.0, 44.4] 17.9 ± 20.6 11.1 [0.0, 27.8] .002 Number of teeth present 7.6 ± 9.0 3.0 [0.0, 14.5] 7.9 ± 8.9 5.0 [0.0, 14.5] .791 Edentulous, N (%) 39 (35.5) 87 (36.7) .905 Prosthesis use, N (%) 57 (51.4) 124 (52.1) .909 Medical history Stroke, N (%) 39 (35.1) 96 (40.3) .409 Diabetes mellitus, N (%) 14 (12.6) 45 (18.9) .168 Respiratory diseasea, N (%) 17 (15.3) 39 (16.4) .876 Note Q1, first quartile; Q3, third quartile; SD, standard deviation. a Respiratory disease: chronic obstructive pulmonary disease, asthma, other (without pneumonia).

Table 3 shows that among the 349 participants who required professional OHM, 39 (11.2%) had pneumonia during the follow-up period. Moreover, the participants who had pneumonia had a significantly lower BI, were less likely to receive professional OHM and had a higher TCI than the residents who did not have pneumonia.

TABLE 3. Intragroup comparison by the requirement for oral health management according to the occurrence of pneumonia Variable With requirements for oral hygiene management (N = 349) Non-pneumonia (N = 310) Pneumonia (N = 39) P-value Mean ± SD/N (%) Median, [Q1, Q3] Mean ± SD/N (%) Median [Q1, Q3] Age, years 87.8 ± 8.1 88.0 [83.0, 93.3] 88.1 ± 6.5 87.0 [84.0, 94.0] .801 Sex (female), N (%) 258 (83.2) 29 (74.4) .184 Barthel Index 29.3 ± 24.6 50.0 [25.0, 45.0] 21.5 ± 23.9 10.0 [0.0, 40.0] 0.035 Body mass index 20.3 ± 3.4 19.9 [18.0, 22.3] 19.7 ± 2.9 19.5 [17.2, 21.6] .346 Clinical dementia rating 2.2 ± 0.9 2.0 [1.0, 3.0] 2.3 ± 0.9 3.0 [1.8, 3.0] .304 0, 0.5 26 (8.6) 3 (7.9) .513 1 50 (16.5) 6 (15.8) 2 87 (28.7) 7 (18.4) 3 140 (46.2) 22 (57.9) Oral conditions Oral hygiene management, N (%) 220 (71.0) 18 (46.2) .003 Willingness for oral hygiene, N (%) 149 (48.1) 22 (56.4) .671 Oral Health Assessment Tool-J Score 2.8 ± 2.3 2.0 [1.0, 4.0] 2.7 ± 2.3 2.0 [1.0, 4.0] .547 Tongue Coating Index 20.1±22.5 11.1 [0.0, 33.3] 26.4±22.3 22.2 [5.6, 50.0] .002 Number of teeth present 8.7 ± 9.1 6.0 [0.0, 17.0] 9.8 ± 9.3 7.0 [0.0, 17.5] .297 Edentulous, N (%) 117 (38.0) 9 (23.1) .078 Prosthesis use, N (%) 165 (53.2) 16 (41.0) .671 Medical history Stroke, N (%) 122 (39.4) 13 (33.3) .492 Diabetes mellitus, N (%) 56 (18.1) 3 (7.7) .117 Respiratory diseasea, N (%) 51 (16.5) 5 (12.8) .651 Note P < .05 was considered statistically significant. Q1, first quartile; Q3, third quartile; SD, standard deviation. a Respiratory disease: chronic obstructive pulmonary disease, asthma, other (without pneumonia)

Table 4 shows the outcome of the Poisson regression analysis used to assess the association between pneumonia incidence and professional OHM. The likelihood ratio test results comparing the multilevel model against a single-level model that contained the same predictors were not significant (P = .18); therefore, we used the single-level model. Provision of professional OHM resulted in significantly lower RR for pneumonia incidence (0.39, 95% CI: 0.22-0.69). In addition, the number of teeth present was associated with pneumonia incidence (RR: 1.04, 95% CI: 1.01-1.08). There was no correlation of r ≥ 0.6 between the factors.

TABLE 4. Effect of oral hygiene management on incident pneumonia in Poisson regression analysis with robust standard errors RR 95% CI P-value Age 1.012 0.972 - 1.053 .569 Sex (0: male, 1: female) 0.674 0.358 - 1.270 .223 Barthel Index 0.988 0.973 - 1.003 .120 Body mass index 0.955 0.861- 1.059 .383 Clinical dementia rating 0, 0.5 Reference 1 0.704 0.206-2.401 .575 2 0.516 0.137-1.942 .328 3 0.604 0.183-1.988 .406 Diabetes mellitus (0: no, 1: yes) 0.451 0.148-1.370 .160 Number of teeth present 1.045 1.012-1.079 .007 Oral hygiene management (0: no, 1: yes) 0.374 0.210-0.665 .001 Abbreviations: CI, confidence interval; RR, relative risk. 4 DISCUSSION

This study investigated the general and oral conditions, and provision of professional OHM, among residents of LTC facilities in Japan to examine the association between professional OHM and the incidence of pneumonia, and investigate the effectiveness of professional OHM provided by dental hygienists in preventing pneumonia. The incidence of pneumonia was significantly lower in residents who had received professional OHM than in those who had not.

Based on the effectiveness and current status of oral care,17-20 the Japanese LTC insurance system accepted dental professional OHM as a nursing care service in 2009. Subsequently, the content of this nursing care service was revised in 2018 to provide direct professional OHM at least twice a month to residents of LTC facilities and provide support for daily oral care, such as specific technical advice and guidance to the nursing staff. Thus, under the professional OHM service outlined by the Japanese LTC insurance system, dental hygienists need to perform direct professional OHM and promote the provision of professional OHM by LTC facility staff. However, this is the first study to evaluate the implementation and efficacy of these guidelines.

An interventional study by Yoneyama et al4, 5 verified the efficacy of oral care in preventing pneumonia. Since then, nurses and caregivers have widely provided oral care to patients in LTC facilities. As it is ethically challenging to conduct new interventional studies on oral care practices in such settings, this research was conducted as an observational study.

In a similar study conducted 20 years ago, the participants had fewer teeth present than the participants in our study, and they were younger and performed more ADLs.3, 4,