The primary objective of this 6 month study is to compare the effectiveness of oral health promotion interventions on both clinical oral health, and oral opportunistic respiratory pathogens in institutionalized elders. Secondary objectives are to investigate changes in oral health-related quality of life, incidence of adverse side effects, pneumonia, as well as subject acceptability.

Elders residing in nursing homes in Hong Kong will be recruited into the clinical trial. Patients will be randomly allocated into one of the following groups: 0.2% chlorhexidine spray, 0.1% pH-balanced chlorine dioxide spray, or sterile water spray (placebo control), once daily. Dental plaque, gingival bleeding, oral opportunistic respiratory pathogens, oral health-related quality of life (OHRQoL), and pneumonia incidence will be assessed at baseline, 3 months and 6 months. Subject acceptability of the interventions will be assessed at the end of the clinical trial.

STUDY SITE AND PARTICIPANTS This study will be a randomized, double-blind, parallel-group clinical trial of 6 months duration. The study was approved by the Institutional Review Board of the University of Hong Kong (IRB reference number: UW 12-509) (Appendix II). A clinical trial certificate (No. 100269) was obtained from the Department of Health, Hong Kong.

The inclusion criteria for recruiting patients in this randomized clinical trial (RCT) are: (1) age 65 years and above; (2) not less than six natural teeth in the mouth; (3) not having an indwelling naso-gastric feeding tube.

SAMPLE SIZE CALCULATION Dental plaque will be the primary outcome variable in this study. A previous study conducted amongst institutionalized elders in Hong Kong documented a mean PI score of 2.0 (standard deviation [SD] =0.7) (Lo et al. 2005). Thus, based on 80% power and a statistical significance level set at 0.05, 61 participants in each group are required in order to detect a difference in PI scores of at least 0.4 between the three treatment groups. In order to account for an anticipated 25% dropout rate, the proposed sample size is 76 patients per group (228 in total).

SUBJECT RECRUITMENT An invitation letter will be sent to residential care homes informing that the dental hospital would like to provide oral health activities for elders at their care homes. Written informed consent will be obtained from all subjects who opt to participate before the commencement of the study.

STUDY PROCEDURE A one-hour oral health presentation covering oral health care and common oral diseases, as well as oral hygiene demonstrations and practical exercises encompassing proper techniques for assisting residents with toothbrushing, denture cleaning, and clearance of remaining food debris in the oral cavity following meals, will be provided by a registered dentist to the nursing home staff and residents. In addition, hand-over-hand demonstrations on how to use the oral spray, and the specific areas of the oral cavity to be sprayed will be performed on a tooth block model. Caregivers will apply a daily dose of 12 sprays (each delivering a volume of 0.13ml) to the buccal and lingual gingival third of tooth surfaces in each sextant.

Baseline assessment will be conducted within one week after the presentation and demonstration. Following baseline assessments, participants will be allocated randomly via block randomization to receive either 0.2% chlorhexidine (Corsodyl) spray (Group 1), 0.1% pH-balanced chlorine dioxide (CloSYS) spray (Group 2), or sterile water spray (placebo control) (Group 3), once daily. The randomization sequence will be generated by the research assistant. Chemotherapeutic agents will be assigned individually to participants, and allocated in unmarked (labeled with subject's identity only) propellant-free atomizer bottles by the research assistant. Both the principal examiner conducting the clinical assessments, and the nursing home staff will be blinded to treatment group allocation. All participants will receive a powered toothbrush and a standard sodium fluoride toothpaste. Sufficient bottles of spray will be prepared for a 6-month volume of usage and sprays will be administered (one hour following toothbrushing) by nursing home staff once daily. Scheduled reports of patient compliance will be provided by nursing home staff to the investigator. Nursing home staff compliance with the administration of oral sprays will be confirmed by the completion of logbooks, and bi-weekly weighing of spray bottles by the research assistant.

BASELINE DATA COLLECTION Clinical oral health assessment Primary outcome measures (plaque levels) will be assessed using the Silness and Löe Plaque Index (PI) (Silness and Löe 1964) which has been validated and widely used in clinical trials assessing changes in oral hygiene status. PI scores will be recorded on all permanent teeth and at six sites per tooth. Gingival health was assessed by using the Gingival Bleeding Index (GBI) (Ainamo and Bay 1975).

In order to evaluate the side effects of the chemotherapeutic agents (i.e. chlorhexidine and chlorine dioxide), extrinsic tooth staining and supragingival calculus accumulation will be assessed by using the Modified Lobene extrinsic tooth stain index (Macpherson et al. 2000) and Volpe-Manhold calculus index (CI) (Volpe et al. 1965), respectively.

Microbiological assessment Supragingival plaque will be collected with a sterile Gracey curette from the six index teeth (upper right first molar, lower left first molar, upper left first premolar, lower right first premolar, upper left incisor and lower right incisor) (McInnes and Cutting 2010). Samples will be stored in 2 ml of sterile phosphate buffered saline (PBS) and placed in an icebox, and subsequently transported back to the laboratory within six hours for processing and identification of aerobic and facultatively anaerobic Gram-negative bacilli (AGNB), Staphylococcus aureus, Streptococcus pneumoniae, Haemophilus influenzae, and Moraxella catarrhalis.

Oral health-related quality of life assessment The Chinese version of the NHANES-OHIP will be used to assess oral health-related quality of life.

Functional disability assessment The Barthel Index (BI) is a valid and reliable measure of disability, which scores patients in 10 activities of daily living (Mahoney and Barthel 1965). Scored items include mobility (stair climbing, transfer, ambulation), and self-care (bowel and bladder care, toilet use, bathing, dressing, grooming, feeding). The BI scores range from 0-100, with higher scores representing better physical function, and will be assessed by a trained research assistant in this study.

The Mini-Mental State Examination (MMSE) is a simplified, practical and effective test in evaluating psychiatric patients (Folstein et al. 1975). It includes 11 questions and concentrates only on the cognitive aspect of mental disability, with scores ranging from 0-30. The results are interpreted as: 0-10=severe cognitive impairment; 10-20=moderate cognitive impairment; 20-25=mild cognitive impairment; 25-30=no cognitive impairment. The MMSE test will be conducted by a trained research assistant at each nursing home.

Hand function disability will be assessed using the Jebsen and Taylor Hand Function Test, which assesses manual dexterity (Jebsen 1969). Seven items are designed for the patients to perform, including 1) writing a short sentence; 2) simulated page turning; 3) picking up small objects and placing into a container; 4) stacking checkers; 5) simulated eating; 6) moving large empty cans; and 7) moving weighted large cans. The time of performance will be recorded for both dominant and non-dominant hands. The Jebsen and Taylor Hand Function Test will be performed by a trained research assistant at the nursing home.

Swallowing disability will be assessed using the Functional Oral Intake Scale (FOIS) (Crary et al. 2005). FOIS assessments will be recorded by senior nursing home staff at each institution.

The Oral Food Debris Index (OFDI) (Migliario and Rimondini 2011) will also be scored and documented.

FOLLOW-UP DATA COLLECTION At the 3-month review, clinical assessments and the NHANES-OHIP questionnaire will be conducted, and microbiological samples collected.

At the 6-month review, subjects will be queried regarding their satisfaction with the oral health promotion interventions provided, as well as the condition of their mouth. Responses were coded as: 1=totally satisfied, 2=very satisfied, 3=reasonably satisfied, 4=not very satisfied, 5=not at all satisfied. Clinical assessments, questionnaires (NHANES-OHIP) and microbiological sampling will be repeated. Medical records will be audited to assess the incidence of pneumonia during the study period.

STATISTICAL ANALYSIS All the data collected from the three time points will be analyzed using the statistics software package PASW 23.0 for Windows (SPSS Inc., Chicago, USA).

Baseline differences between groups in categorical demographic variables, oral mucosa status, CPI, LOA and prosthetic status, as well as the pathogen prevalence will be compared using Chi-square tests. As for continuous variables, one-way ANOVA for independents samples and paired t tests for related samples will be undertaken if they follow normal distribution; if they do not follow normal distribution, Kruskal-Wallis one-way ANOVA for independent samples and Wilcoxon Signed Rank test for related samples will be performed. Within-group comparisons (baseline to 3-month) of oral mucosa status, CPI, LOA and prosthetic status, adverse effect, and pathogen prevalence will be done via the McNemar test. Within-group comparisons (baseline to 3-month) of median DMFT, calculus index, stain index and GBI, as well as viable counts (cfu/ml) will be evaluated using the Wilcoxon Signed Rank test, while mean PI scores and NHANES-OHIP scores will be made by Paired-Sample T test. Comparison of change scores in PI between groups will be performed using One-Way ANOVA and Bonferroni pairwise comparisons. Comparison of change scores in GBI, calculus index, stain index and viable counts (cfu/ml) between groups will be determined by Kruskal-Wallis one-way ANOVA, and the Mann-Whitney U test for individual pairwise comparisons. Pathogen prevalence (persistence, acquisition, transience, loss, and absence) between groups will be made using the chi-square test.

Comparisons of mean PI and NHANES-OHIP scores across baseline, 3-month, and 6-month assessments will be completed by two-way ANOVA test, while pairwise comparisons will be made via the Bonferroni test. Median GBI, calculus index, stain index and viable counts (cfu/ml) at the three assessment points will be compared with the Friedman two-way ANOVA test, while pairwise comparisons will be done with the Wilcoxon Signed Rank test. Comparisons of pathogen prevalence at the three assessment points will be accomplished via Cochran Q test, and individual pairwise will be made with the McNemar test.

Univariate analyses of potential independent factors associated with PI, GBI, NHANES-OHIP, and viable counts of AGNB, S. aureus, S. pneumoniae, H. influenzae and M. catarrhalis at review assessments will be done with multiple linear regression analyses. Variables with p<0.05 will be entered into the multiple linear regression model (Forward Wald). Potential independent factors associated with the prevalence of AGNB, S. aureus, S. pneumoniae, H. influenzae and M. catarrhalis at review assessments will be subjected to multiple logistic regression analyses.

Inclusion Criteria:

• not less than six natural teeth
• not having an indwelling nasogastric tube

Exclusion Criteria:

-