This study was a double-blind, cluster-randomized, controlled trial with 1 month and 6 months follow-ups. The study involved 10 public hospitals in Malaysia, which have participated in a survey of oral hygiene practice for stroke patients. These hospitals were selected because they provide rehabilitation services that are led by rehabilitation medicine specialists. Hospitals were first stratified by size into either large, medium, or small in terms of number of health care providers. From each stratified group, hospitals were block-randomized in groups of 4 (“ABBA”) by a computer-generated randomization method. In total, 5 hospitals were assigned to the test group (277 registered nurses) and 5 hospitals were assigned to the control group (270 registered nurses). The allocation sequence was concealed from the investigator coordinating the trial (who had contact with the centers). Through concealment, the assessor was “blind” as to what group participants had been assigned and participants were also blind as to what groups they were assigned to, as both received a form of Web-based CPD.

The study population was registered nurses caring for stroke patients at the hospitals, mainly from the rehabilitation and general medical wards. All the registered nurses from these identified wards were invited to take part in this trial. Information sheet related to the study and written informed consent were given to all the nurses before commencing the study. The forms were distributed to the nurses by the ward managers or chief nurses. Nurses who provided their written consent were those who participated in the trial from the study population. Participation was voluntary and no contact was made with the nurses to ensure confidentiality and reduce the potential for “social bias.”

This clinical trial was registered with the National Institutes of Health, Ministry of Health, Malaysia; NMRR-13-1540-18833(IIR). Before the commencement of the study, ethical approval was obtained from the Institute for Health Behavioral Research and Medical Research and Ethics Committee of the National Institutes of Health, Ministry of Malaysia. Permissions to conduct the study were also obtained from the directors of the respective hospitals. Recruitment and baseline assessments were from September 2014 to November 2014 at 10 hospitals across the country (both the Peninsular Malaysia and island of Borneo Malaysia). This study followed CONSORT guidelines.

The nurses self-completed a questionnaire on the practice of providing oral hygiene care to stroke patients, which contained 12 items specific to attitudes, SN, PBC, and GI to providing oral hygiene care related to TPB. These items were derived from the manual of “Constructing Questionnaires Based on the Theory of Planned Behavior” developed by the Centre of Health Services Research, University of New Castle, UK (2004) [24]. Items related to direct measure of the domains were chosen and modified to the oral health context. For example, in the GI domain, “I expect to measure the blood pressure of my patients with diabetes in each consultation” was modified to “I expect to perform oral care (including denture) for patients in every session.” Each domain had 3 items that were rated on a 5-point Likert scale (strongly disagree, disagree, not disagree or agree, agree, and strongly agree). Domain scores can range from 3 to 15, with higher scores reflecting more positive attitude, stronger subjective norms, greater perceived behavior control, and greater general intention to provide oral hygiene care. Sociodemographic and environmental characteristic (eg, attended oral care training, availability of oral health guidelines and oral hygiene kits, and having dental professional support in the ward) information were also obtained from the participants.

In addition, knowledge of oral health care was assessed using 5 items related to dental plaque, gum bleeding, consequences of dental plaque, how to prevent gingivitis, and how oral health affects general health [25]. Knowledge scores can range from 0 to 5, with higher scores indicative of greater oral health knowledge. Assessments were carried out preintervention and at 1 month and 6 months postintervention.

A Web-based CPD program was developed for the test and control groups. The test group program was specific to provision of oral hygiene care to stroke patients and covered details of oral health knowledge, attitudes, subjective norms, means of behavioral control, and intention (ie, based on TPB). The test group contents include, for example, information on good oral condition and the importance of having good oral health, the consequences of poor oral hygiene, and the importance of nurse’s roles and care of stroke patients. The development of the contents was guided by the definition of the TPB domains and scope of the study. The control group received an analogous Web-based CPD program related to “bundles of care” for stroke patients that included some details on oral hygiene care but not specific to TPB [26]. The CPD programs were developed by stroke physicians (rehabilitation medicine) and dentists and followed good practices of CAL for oral health [27]. Following the assignment to the groups, the participants were provided with details of the Web-based programs through a secure internet portal. Participants were reminded and encouraged to complete the Web-based CPD program every 6 weeks.

With the assumption that this practice is at 50% and that it will not change without education intervention, whereas there will be a 25% improvement in practices following CAL intervention (ie, 63% of nurses will practice oral care in rehabilitation). Then a proposed sample size of 247 in each group is required with sample power at 80%. Allowing for nonparticipation and a dropout rate of ~20%, thus it was prudent to attempt to recruit over 600 nurses (300 per group) in total to test the hypothesis.

The changes in knowledge, attitudes, SN, PBC, and GI were determined overtime and compared between the test and control groups using Friedman two-way analysis of variance (ANOVA) and Mann-Whitney U test analysis, respectively. Multiple linear regression analyses were performed to determine key factors associated with changes in GI to provide oral hygiene care at 1 month and 6 months.

As mentioned before, there is a need to improve the practice of oral hygiene care in the acute hospital setting, and some would argue particularly for stroke patients [28-31]. CAL has been used widely to deliver information and educate patients and caregivers in the medical field but few were reported to be related to oral health [32,33]. Thus, this study aimed to evaluate the effectiveness of a Web-based (CPD) program to change the GI of health carers to perform daily mouth cleaning for stroke patients using TPB and to identify the key factors associated with changes in GI among health care workers to provide oral hygiene care to stroke patients. Increasingly, there are calls to consider theory-based health promotion strategies in implementing programs and to provide evidence of whether theory does, in fact, dictates practice [34,35].

This study provided support for the application of TPB among stroke carers to perform oral care daily. Over time, in the test group, there was evidence of significant improvements in general intentions to perform oral hygiene care, SN, and knowledge. Even by 1 month, there was a significant improvement in GI, SN, and knowledge, and between 1 month and 6 months this persisted. Whereas in the control group there was a significant change in GI and attitudes, but it was transient; a significant difference was between 1 month and baseline but not between 6 months and baseline. At 1 month, there were significant differences in the change of attitude scores and in the change of SN scores between the test and control groups. By 6 months, significant differences in the change in GI, attitudes, and SN to provide oral hygiene care were evident in all TPB domains except in PBC. In addition, a significant difference in the change of knowledge scores over the 6-month period was also evident between the test and control groups. This provides evidence of the effectiveness of the Web-based intervention and suggests that its effectiveness increases over time. Nonetheless, PBC did not significantly change. It is plausible that this may take longer to change or needs to be supported by environment changes [36].

Regression analyses identified that the key factors associated with GI (to provide oral hygiene care) at 1 month were attitudes (at 1 month) and SN (at 1 month); and that general intention at 6 months was associated with SN (at 6 months) and PBC (at 6 months). Furthermore, the change in GI (between baseline and 1 month) was associated with the changes that occurred in attitudes and SN; and the change in GI (between baseline and 6 months) was associated with the changes that occurred in SN and PBC. The fact that several of TPB domain scores were associated with GI at the respective time periods and that the changes in several of the TPB scores were associated with changes in GI at the respective periods provide evidence of the interrelationship between domains of TPB and GI as the theory hypothesizes. The findings of this trial support other findings with respect to predictors of GI at specific time periods [37] and changes in GI [13]. Thus in line with TPB, carers who had positive attitudes, perceived positive social pressure, and were in control of their action would have high intention of performing oral care daily to patients with stroke [13]. Of note, some domains did not change nor were some domains associated with GI or change in GI, thus the influence of the domains varies across the study [37]. This is not an unusual feature in studies as not all domain changes are associated with GI over time [38,39]. This perhaps can be attributed to the context-specific settings of the study, the type of intention, and indeed the general intention practice [40,41].

Studies have shown that a significant improvement in GI, SN, and knowledge among the participants increased their confidence in promoting oral health and educating patients [19]. A noteworthy finding of this study was that both groups were given the CAL program, but only the intervention group had detailed information on oral health related to stroke patients. Thus this contributed to a significant increase in specific knowledge and GI among the intervention group compared with the control group [18].

The study benefits from its relatively large sample size, the diverse geographical areas within a country (major centers for stroke rehabilitation), and being a double-blind randomized controlled clinical trial. The response rate was close to 70% and loss to follow-up was largely attributed to nurses having changed working environments because of being assigned to other wards or hospitals. There was no significant difference in response rate between the test and control groups.

In this trial, participants were block-randomized by hospital (stratified by size) rather than by participants to avoid the potential of bias of participants discussing the Web-based CPD program within centers. The use of Web-based CPD in both the test and control group interventions allowed for a double-blind trial. In the test group, the program was built on TPB and included specific information to promote increases in knowledge, attitudes, SN, PBC, and GI to provide oral hygiene care. Although the control group program did contain some information on oral hygiene care, it was limited and not based on the TPB.

Going forward, it would be worth investing if changes in the practice of provision of oral hygiene care do occur and that the practice is maintained over time. Furthermore, studies can be done to examine the pathways that influence changes in behavior by considering specific beliefs of TPB attributes and the impact of Web-based CPD. The impact of CPD depends on various factors such as individual carer characteristics (eg, personal compliance or interest to update themselves with new information) [42,43], and environmental factors (eg, lack of Internet access or software compatibility) [22,44]. It would also be useful to monitor engagement in the Web-based learning (which at the time keeping anonymity) to determine how this effects outcomes or indeed other methods of learning.

A Web-based CPD program based on TPB was effective in increasing GI to perform oral hygiene care for stroke survivors in the acute hospital setting. In addition, the program was effective in changing attitudes and SN. Furthermore, the study found that changing SNs and PBC are key factors associated with changes in GI to provide oral hygiene care. These findings support and have implications for the use of theory-based health education CPD programs and oral health promotion programs.