Is Oral Health All About Tooth Brushing? A Longitudinal Study in Individuals With Brain Injury in A Neurorehabilitation Setting

Objective: To investigate oral health changes and its associated factors during hospitalization in individuals with acquired brain injury (ABI). Material and Methods: Sixty-one individuals were recruited to evaluate the acute changes in oral health by performing plaque, calculus, bleeding on probing (BOP) and bedside oral examination (BOE) at week 1 and week 5. Individuals’ brushing habits, eating diculties, onset of pneumonia etc. were retrieved from e-journal. Association between oral-health outcomes to systemic variables were investigated through multilevel regression models. Results: Dental plaque (P=0.01) and total BOE score (P<0.05) decreased over time but not the calculus (P=0.30), BOP (P=0.06) and increase in tooth brushing frequency (P=0.06). Reduction in plaque, and BOE over time were negatively associated with higher score of periodontitis at baseline (coef. -6.8; -1.0, respectively), which in turn were associated with an increased proportion of BOP (coef. ≈ 15.0). Increased proportion of calculus was associated with eating diculties (coef. 2.3) and onset of pneumonia (coef. 6.2). Conclusions: Nursing care has been fundamental in improving oral health but non-signicant improvement in calculus, BOP and brushing frequency indicates a need for development in existing oral care program through academic-clinical partnership keeping eating diculties and patients’ vegetative and cognitive state in consideration.


Introduction
Oral care is essential to maintain oral health and prevent complications such as periodontal diseases and tooth loss in patients with acquired brain injury (ABI) [1][2][3] . Poor oral hygiene among dependent hospitalized patients could lead to severe complications such as poor nutritional intake, increased length of hospital stay, and pneumonia 1,4,5 . In relation to oral health, stroke can cause hemiparesis and hemiplegia to the facial muscles and the muscles of the pharynx, tongue, palate, and mastication, resulting in impaired oral clearance 6,7 . Medications prescribed for patients after stroke may further impact oral health resulting in, for example, dry mouth, oral ulcers, and stomatitis 8 . ABI individuals with swallowing di culties have compromised oral clearance and leads to increased bacterial load 1 .
Swallowing impairment, along with poor oral health has a signi cant impact on an individual's nutritional intake 9 , increasing the risk of aspirational pneumonia 5,10 , which in turn has a negative impact on rehabilitation and other functional outcomes 5,11 .
Evidence suggests that stroke survivors with increased plaque and bacterial load experience a deterioration of the periodontal conditions 12,13 . Recently, a study showed that 40% of the ABI population had an abundant amount of dental plaque and increased bleeding on probing (BOP), a nding that may indicate an acute hospitalization effect 3 . In addition, 74% of the ABI individuals also had severe periodontitis, a condition, supported by their poor socio-behavioral and medical history, representing a chronic stage of an oral health disease 3 .
Post ABI, many patients are reliant on nursing staff to assist them with oral hygiene. Despite indications that healthcare staff is interested in improving this aspect of care, a recent survey among > 250 health professionals showed that oral care had not been their prime focus due to barriers such as, lack of time due to prioritizing other emergency tasks, unfocused oral care policies, lack of training and evidencebased continuing education 14 .
In the light of the current evidence on the importance of oral health among individuals with ABI, oral care management through oral care providers could play an important role in this area 9 . It is not clear whether the oral care provided by the healthcare professionals help improve or deteriorate the oral health in a neurorehabilitation setting 3,14,15 . This new knowledge may provide an overview to promote and manage oral health in individuals with ABI in a neurorehabilitation setting. Accordingly, the aim of this study was to investigate the changes in oral health status over time (5 weeks) and its associated factors during hospitalization in patients with ABI. We hypothesized that the current standard oral care provision requires further structural improvement and modi cations.

Results
Of the 90 patients included at baseline, 61 provided data for the 5-week follow-up. The mean age was 55.1 years (± 14.0), and 64% of the patients were male. More information about the sociodemographic data of the participants can be found elsewhere 3 .
Paired analysis revealed that the proportion of sites with visible dental plaque (P = 0.01) signi cantly decreased over the time but BOP (P = 0.06), calculus (P = 0.30) and the frequency of tooth brushing (P = 0.06) did not achieve statistically signi cant changes after 5 weeks of hospitalization. In addition, the total BOE score (P < 0.001) signi cantly improved over time and also most of the BOE domains like, swallow, saliva, mucosa, teeth and odor (P < 0.05) ( Table 2).
Mixed-effects regression models indicated that individuals with 'moderate' periodontitis at baseline (coef. -6.8) and those hospitalized at the regional ward (coef. -15.6) had decreased proportion of sites with dental plaque. In addition, the number of extracted teeth (coef. -1.0), the proportion of calculus (-0.5), and the time (-8.8) were also associated with a reduction in the proportion of plaque (Table 3A).
Patients with higher scores of both 'moderate' (coef. 14.3) and 'severe' (coef. 15.6) periodontitis at baseline had an increase in the proportion of sites with BOP over the study period (Table 3B), whereas those who improved their 'cognitive' domain (coef. -6.6) had a decrease in the proportion of BOP. The number of extracted tooth at baseline (coef. 0.5) and increased proportion of plaque over the study period (coef. 0.4) were also associated with an increased proportion of BOP after the 5-week follow up. As displayed in Table 3C, those who developed pneumonia during hospitalization (coef. 6.2) and those with eating di culties over the study period (coef. 2.3) had an increase in the proportion of sites with dental calculus.
Finally, mixed-effects regression models indicated that the individuals with higher scores of 'moderate' periodontitis at baseline (coef. -1.0), those hospitalized at the regional ward (coef. -1.7), and those who improved their 'motor' skills during the study period (coef. -0.6) had a reduction in their total BOE score, whereas those with dysphagia at baseline (coef. 0.5) and the old individuals (coef. 0.04) had an increased total BOE score after ve weeks (Table 3D).

Discussion
The main nding of the study was that the oral health parameters such as visible plaque and BOE scores signi cantly improved during 5 week stay at neurorehabilitation setting but the proportion of sites with BOP and frequency of tooth brushing over time (5-week stay) did not reach statistical signi cance. These ndings demonstrate that although there was an improvement in the oral health status in hospitalized individuals, it was not substantiated, indicating a need for further development in oral care program.
A signi cant reduction in the amount of dental plaque was observed over time. Plaque is a bio lm comprised of diverse community of microorganisms which is formed regularly on the tooth surface and can be maintained by proper tooth brushing and ossing 1,16,17 . Individuals with less severe ABI are usually admitted to the regional ward at HNRC instead of highly-specialized ward due to their better motor and cognitive functions, which makes them more co-operative then severely affected individuals with ABI 18 . It was also evident from the mixed regression analysis that individuals admitted to the 'regional ward', showed an strong association in reduction of plaque compare to highly-specialized ward 19 . Interestingly, the proportion of plaque was also reduced in individuals with 'moderate' periodontitis and with an increased proportion of calculus and BOP, indicating that there are also other factors such as host-immune response 5 , which were not taken into account in this study, that might have in uenced BOP.
Gingival bleeding (BOP) is an acute reversible in ammatory condition that occurs as a response to plaque accumulation on the periodontal tissues 20 . In general, good oral hygiene practices are su cient to control and reduce gingival bleeding 17,21 , which was also shown in the current study with strong association between plaque and BOP (Table 3C). However, despite the signi cant plaque reduction, proportion of sites with gingival in ammation (BOP) did not reduce in the same individuals over the study period. It is also important to discuss that the SD values of BOP were probably quite higher as few individuals had a very low BOP, while some (especially those with periodontitis) had high BOP. Such a nding suggests that factors other than plaque might play a role in the onset and progression of gingival in ammation. It has been shown that a more exacerbated and rapid immune response, acute hospitalization and cognitive and systematic complications are linked to a higher neutrophilic activity, which is able to mount an immediate response when exposed to plaque 22,23 . Interestingly, our ndings demonstrated that BOP decreased over time in individuals who showed an improvement in their 'cognitive' function, indicating reduction in confusion and agitation leading to increased cooperation with oral care, which very well correlates with previous research 5 . It has been shown that BOP is closely associated with 'severe' periodontitis, which in addition to an already existing cognitive impairment, may contribute to other chronic conditions that share a common biological background to ABI 5,24 . Assuming that such an exacerbated immune response is not restricted to the oral cavity, this may interfere with other in ammatory processes, especially in a hospital setting and in the presence of other comorbidities, explaining partially our ndings 5 . On the other hand, the proportion of sites with BOP increased over time among patients with both 'moderate' and 'severe' periodontitis, despite the increase in tooth brushing frequency over the same period. This nding indicates that the oral health status in these patients were poor and tooth brushing alone may not be enough to tackle oral health problems. Despite the efforts made by nurses to maintain oral hygiene, there was still deterioration of the in ammatory periodontal condition 1 . This suggests the need for involvement of dental personnel in hospitals for providing adequate oral care to ABI patients 3 .
Calculus, de ned as hard deposit around the gingiva as a result of long-term plaque accumulation, showed no signi cant improvement over time. It is important to highlight that calculus does not indicate disease, but it makes oral hygiene more di cult to maintain as well as works as a plaque-retaining factor 25 . Even though the removal of calculus is not possible without professional dental assistance, it is possible to maintain proper oral hygiene by preventing calculus formation. Such a nding supports the idea that chronic oral changes require professional help from dental personnel as well as changes in socio-behavioral factors for the improvement of oral health 20 . Our ndings also revealed that individuals with eating di culty and those who developed pneumonia during hospitalization had an increase in the proportion of sites with dental calculus. One may speculate whether the combination of dental calculus and eating di culties may in uence the onset of pneumonia. A recent study on patients with ABI has shown a robust association between periodontitis and debilitating conditions like dysphagia, dependency on a feeding tube, which is a major concern, as they lead to pneumonia 5 . Although our study does not allow us to disentangle the causal relationship between these conditions, our overall ndings suggest the need for increased focus on oral care especially for ABI individuals with conditions like eating di culties and severe cognitive disturbances.
Interestingly, BOE scores decreased in individuals with higher scores of 'moderate' periodontitis. As discussed, 'moderate' periodontitis originates essentially from neglected oral hygiene, so does most the BOE domains 3,5,21,26 . Thus, the combined effect of plaque reduction and increased frequency of oral hygiene can explain this association. It should be noted that, although BOE is a simple and easy to use tool in hospital settings, especially in intensive care units, its usefulness is questioned in ABI patients and therefore, the BOE results may be carefully interpreted 2,3 . This is because the instrument seems not to re ect the real clinical conditions of patients with ABI, thus, affecting the treatment plan. Further, it has been shown that 'ageing' patients have more compromised function than young individuals, making them more vulnerable to dysphagia and unable to perform and maintain good oral hygiene procedures 3,5 .
A recent survey conducted among 157 oral caregivers at HNRC showed that the majority of oral caregivers were aware of the existing 'Danish National Clinical Guidelines for Oral Care' 19 . However, a signi cant number of oral care providers did not follow the guidelines systematically, expressing it as an ineffective, time-consuming, and di cult to follow 14 . Professionals were aware that patients with eating di culties have challenges and different requirements 14 and on top cognitive and motor de cits adds an extra challenge to maintain the oral hygiene 5 . In addition, there is always a professional dilemma to maintain oral hygiene standard whilst respecting patient's autonomy once they refuse for the oral hygiene care, even if it is required. Therefore, all these factors should be considered while formulating and designing oral care training and guidelines to improve oral care in a neurorehabilitation setting.

Methodological Considerations
The current study sample originates from a single hospital setting, and therefore, our ndings may have limited external validity. However, it is worth mentioning that this hospital is a reference center for the treatment of ABI patients and receives patients from most regions of Denmark. In addition, limited sample size and 30% lost to follow-up might have reduced the analytical power, as can be noted by borderline P-values. Future studies with large samples originating from several centers are needed. Another limitation of the study was short follow-up time, given the chronicity of the most common oral diseases, i.e., dental caries and periodontitis. However, treatment of these conditions demands the involvement of dental personnel with appropriate armamentarium, which was not within the scope of the study. As our purpose was to observe the effect of hospitalization on oral health, we decided to evaluate conditions such as the proportion of dental plaque and of BOP, as those parameters can rapidly change.
We also need to be aware that few patients were excluded due to extreme fatigue, agitation, motorcognitive de cits etc., leaving us with no opportunity for clinical examination, which might be a bias in representing the entire oral health status.

Conclusions
A signi cant reduction in dental plaque and total BOE score was observed over time. However, nonsigni cant improvement in gingivitis, the proportion of calculus, and brushing frequency indicate the need for a further development of oral care program in individuals with ABI keeping motor-cognitive de cits and eating di culties in consideration. This study also enforces the need for the involvement of dentist in educating and supervising non-dental professionals at an early stage to provide a better and an integrated oral care program in ABI individuals in hospital settings.

Participants and recruitment
All individuals with ABI admitted between February-June 2019 at neurorehabilitation center were recruited in a longitudinal observational study. In total, 132 individuals with ABI were screened and examined within the rst week (baseline) from the admission day and later at week 5, re-examined to assess the acute changes in oral health. Out of 132 individuals, 90 were eligible for the week 1 assessment, and 61 individuals were eligible for week 5 assessment after ful lling all the above-mentioned eligibility criteria (For a detailed eligibility criteria, see Appendix 1).
Procedures  Table 1 for details.

Data Analyses
Data on the proportion of plaque, calculus, BOP, BOE scores, and frequency of tooth brushing were submitted to descriptive analyses. In addition, paired analyses (t-test for normally distributed variables and Wilcoxon signed-rand test for non-normally distributed variables) were also conducted. Using multilevel mixed-effects regression models, we were able to investigate the association between changes in oral health outcomes with both time-varying, collected at both baseline and week 5, and non-varying (elicited at baseline only) variables. To select potential confounders, we used the "backward" stepwise procedure, in which all variables were entered in the model, and then subsequently removed. Only variables with a P-value < 0.20 were maintained in the model and those with a P-value < 0.05 were considered statistically signi cant. The data analysis was carried out using the software Stata 14.2 (StataCorp., College Station, TX, USA).

Declarations
Tables