Study design and setting

This study protocol was developed in accordance with the Consolidated Standards of Reporting Trials (CONSORT) guidelines.17 The study design is a multicentre cluster randomised controlled clinical trial, with care homes as units of randomisation (figure 1). The study will be performed in different care home units in the Netherlands. After approaching the care homes, the care homes will be allocated randomly to either the intervention group or the control group, assigned by lot. The sample size of the participants in both groups will be monitored. The enrolment of all care home units will take place at different moments within 1 year. However, in one care home unit, all participants will start at the same time.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Presents a flow chart of the multicentre cluster randomised controlled clinical trial.

Power calculation

On the basis of a power calculation, 500 physically disabled care home residents with dysphagia will be followed during 1 year. The care homes are the units of randomisation and will be allocated randomly to the intervention or control group (figure 1). Factors taken into account in the sample size calculations are the distribution of the outcome variable in the population (σ) and the presumed effect of the intervention (μ¹−μ²), the requested power (1−β), the α and the value of the intraclass correlation (design effect). The primary formula used to calculate the sample size is:

Calculating the power, the incidence of pneumonia in care home residents is, in agreement with not internationally published incidence data for the year 2011, largely assessed as 25%,18 the presumed reduction of the incidence is 40%, the requested power is 0.8, and the α is 0.05. If the true relative risk of failure of residents of the intervention group relative to residents of the control group is 0.6, 250 residents are needed in the intervention group and 250 residents in the control group in order to be able to reject the hypothesis that the risk difference equals 0% with probability (power) 0.8.

Recruitment of care homes and participants

The care homes will be selected from a database. Care home units with residents for short-term care or with cognitively impaired residents (mainly suffering from dementia) will be excluded from this study. After recruiting the care homes, the care homes will randomly be allocated to either the intervention group or the control group, assigned by lot. The care homes will receive information on the study in detail by post and email. After this mailing, all care homes will be called. If the care home agrees to participate, the principal investigator (VH) will organise an information meeting. Written informed consent by the board of the care home will be received before the start of the study. After this informed consent, recruitment of residents will take place. Only physically disabled residents with dysphagia are eligible to participate.

In the Netherlands, care homes contain somatic and psychogeriatric wards of primarily physically disabled residents and/or primarily cognitively impaired residents, respectively. The most common medical diagnoses in primarily physically disabled residents are cerebrovascular accidents (26%), other neurological diseases (7%), post-hip surgery status and other locomotor disorders (30%), malignancies (8%) and other miscellaneous diagnoses (29%).19 More than 85% of residents of psychogeriatric wards suffer from dementia. Most residents show a wide range of disabilities and handicaps, affecting autonomous activities of daily living, personal care, mobility and mental functions. The nursing staff, speech therapist and the responsible elderly care physician20 indicate if the resident on a somatic ward is cognitively and physically able to participate and is able to conduct the given instructions by the nursing staff.

Participants are recruited by a physician and a speech therapist of the care home. The recruited residents or their legal representatives will receive information, both by word of mouth and written, about the study. An information letter and the informed consent letter will be presented by the principal investigator or the nursing staff. Before giving their written informed consent, participants and/or their legal representatives will be given the opportunity to ask additional questions about the study.

Inclusion

Residents will be included when they meet the following criteria: aged 65 years or older, physically disabled and diagnosed with dysphagia. Dysphagia is defined as: “Difficulty with any stage of swallowing (oral, pharyngeal, oesophageal) and dysfunction in any stage or process of oral intake; includes any difficulty in the passage of food, liquid, or medicine during any stage of swallowing that impairs the client's ability to swallow independently or safely”.21 In this study, dysphagia will be determined as a dysfunction in the process of oral intake using the Functional Oral Intake Scale (FOIS).22 ,23 The FOIS reflects the functional oral intake of patients with dysphagia and has adequate reliability, validity and sensitivity to change in functional oral intake in patients with stroke.23 ,24 It is a seven-point scale (level 1–7), developed to systematically rate the functional level of oral intake of food and liquid. Residents with level 1 (nothing by mouth) till level 6 (total oral diet with multiple consistencies without special preparation but with specific food limitations) will be included. Level 7 shows a total oral diet with no restrictions and therefore no dysphagia. Dysphagia will be diagnosed by a speech therapist.

Residents who are cognitively impaired (mainly suffering from dementia), in a coma or vegetative state, terminally ill, dependent on mechanical ventilation, in day care, in short-term care, or already using an oral hygiene care solution, will be excluded. A physician has to withdraw a resident from the study when the resident's conditions have altered according to one of the exclusion criteria. However, in case pneumonia is the reason for the condition alteration, the pneumonia will be registered as a case with respect to the outcome variable. Withdrawn residents will not be replaced because otherwise, in order to determine the incidence of pneumonia, significant bias would be introduced by replacement.

Intervention and control groups

The intervention consists of applying a 0.05% chlorhexidine-containing solution (Perio-aid Maintenance 0.05%, Dentaid BeNeLux BV) twice daily immediately after the usual oral hygiene care. The application method depends on the severity of the dysphagia. The speech therapist indicates the application method by considering the FOIS level with respect to the toleration of thin liquids by the resident. Residents who tolerate thin liquids have to rinse with the 0.05% chlorhexidine-containing solution for 30 s twice daily immediately after the usual oral hygiene care. Residents with severe dysphagia, who cannot tolerate thin liquids, have to clean their teeth, gums, tongue, palate, and buccal mucosa with a gauze containing a 0.05% chlorhexidine-containing solution twice daily, immediately after the usual oral hygiene care. The control group receives the usual oral hygiene care without addition of an oral rinse. In both groups, the usual oral hygiene care is not standardised. When the usual oral hygiene care would be standardised, this would be an intervention as well and no effect of the additional chlorhexidine-containing solution could be established.

Monitoring

In the intervention group, the application of the chlorhexidine-containing solution will be guided and monitored after each period of 2 weeks by a group of visiting study assistants who have been instructed by the principal investigator about the use of a strict guiding and monitoring protocol. Every study assistant will visit the same care home ward(s) each time. In each care home, two trained internal study supervisors will monitor the usual daily oral hygiene care, including the application of the chlorhexidine-containing solution. The internal study supervisors will be nurses, dental hygienists or dentists. After every 2 weeks, the study assistants will monitor whether the chlorhexidine-containing solution has been used. The study assistants will monitor the potential side effects of the chlorhexidine-containing solution by oral assessments and by evaluations with the internal study supervisors. After each visit, the study assistants will report their findings using a standardised Clinical Report Form (CRF). The principal investigator will visit the intervention groups to evaluate the study with the internal study supervisors after 6 and 12 months after the start of the study. During those visits, the principal investigator will also do oral assessments with the participants in the intervention group for potential side effects of the chlorhexidine-containing solution. On top of that, the control groups will be monitored with the same monitoring protocol after each period of 4 weeks. In the control groups, the principal investigator will evaluate the study with the internal study supervisors at 6 months and at the end of the study. No follow-up monitoring of the participated care home residents will take place after the end of the study.

Data collection

Research data will be gathered at baseline by the principal investigator (the first author of this article) and the assistant examiners (dental hygienists and bachelor dental hygiene students). The assistant examiners are educated by the principal investigator to perform a short standardised oral examination and to gather data from the residents’ medical records. This short standardised oral examination will determine the number of teeth and implants present and the presence of removable dentures. The following additional data are gathered from the patient's medical record: age, gender, diagnosed diseases and actually used medication. An overview of the data which will be collected is presented in table 1. Subsequently, the level of care dependency will be determined by a care provider using the Care Dependency Scale.25

In case symptoms of pneumonia occur in the participating residents during this study, they will be physically examined by a physician. If pneumonia is diagnosed, the physician will inform the principal investigator in order to register the pneumonia as a case with respect to the outcome variable. The diagnosis and related clinical symptoms of pneumonia will be registered using a CRF. When a participant has more than one episode of pneumonia during the trial, all episodes will be monitored and reported. Any other alterations in medical conditions of the participant, such as hospitalisation, death or discontinuation of the study, will be reported by the physician using different CRFs. All CRFs will be collected by the assistant examiners every 2 weeks.

Outcome measures

The primary outcome variable is the incidence of pneumonia diagnosed by a physician, using a set of strictly described criteria. Since uniform criteria for diagnosing pneumonia in care home residents are lacking, a set of criteria will be assessed by a Delphi panel of experts before the start of the study. To determine the incidence of pneumonia, every episode of pneumonia with the included residents within the study period of 1 year will be registered.

Secondary outcomes are correlations between age, gender, diagnosed diseases, dysphagia severity, care dependency, actually used medication, number of teeth and implants present, and presence of removable dentures and the incidence of pneumonia in physically disabled care home residents with dysphagia who did or did not apply a 0.05% chlorhexidine-containing solution to their usual oral hygiene care.

Statistical analysis

The categorical and continuous variables will initially be subjected to an exploratory data analysis, applying a variety of mostly graphical methods and methods for testing the necessary assumptions. The effect of the intervention on the incidence of pneumonia will be determined using a Cox regression analysis. The intervention will be the independent variable, whereas the dependent variable is the occurrence of pneumonia. In addition to this main analysis, the influence of the other research variables on the incidence of pneumonia will be determined using an elaborate Cox regression analysis (age, gender, diseases diagnosed, dysphagia severity, care dependency, actually used medication, number of teeth and implants present and presence of removable dentures). These variables can be used in case any imbalance emerges in the clustering of the care homes. In this survival model, a frailty model which aims to account for heterogeneity caused by unmeasured covariates will be included to take into account the clustering of data within care homes. A 95% CI will be used with all estimated values. Recording the moment and the reason of withdrawal of participating residents facilitates the use of the data gathered from all participating residents for statistical analysis, even if they were withdrawn prematurely.