Thank you for having me here. It is my pleasure to share my research with all of you.
Oral health is a serious concern in OAD because various reasons. First of all, it is associated with systemic health. It can cause pain and discomfort and therefore affect patient’s wellbeing. Poor oral health can also cause malnutrition as we all know. Research also shows that poor oral health is associated with uncontrolled diabetes. It may cause recurrent respiratory infection such as aspiration pneumonia, which is one of the leading cause of death in functionally dependent elderly. Some evidences also show that poor oral health can increase the risk of heart attack. Therefore, maintaining good oral health is important for elderly persons, especially those with dementia, because many of them lose capacities to communicate with caregivers for their discomfort and needs and therefore have to suffer excessive and treatable pain and infection. However, oral health is poor in patients with dementia.
Due to cognitive impairment and functional dependence, demented patients may gradually lose their ability to appropriately manage their oral hygiene. At the early stage of the disease, patients may frequently forget brushing teeth, along with disease progress, demented patients will gradually lose their oral hygiene skills, such as brushing teeth, cleaning dentures and use of dental floss. As a result, oral hygiene is poor in these patients.
Because of poor oral hygiene, functional dependence and anticholinergic side effects of medications, risk of dental caries is increased in this population. Research shows that demented patients have high prevalence of dental caries.
As a result of increased risk of dental caries and periodontal disease, risk of tooth loss may also increase in this population Compared to those without dementia, more demented elders lose all of their natural teeth and become edentulism. Some research showed that up to 65% of the demented patients lost all of their natural teeth.
As mentioned, Along with disease progress, patients may gradually lose their cognitive and physical function. Their capacities to appropriately manage dental appliances are also decreased. Therefore, use of denture will decline in this population. Also, many demented patients may have denture-related soft tissue problems such as denture stomatitis and denture epulis, which is an overgrowth of fibrous connective tissue caused by over-extended denture.
Soft tissue lesions such as yeast infection, cancer, trauma, are also commonly seen in elderly patients with dementia
Currently, little is known about tooth survival in OAD. During the past decades, only a couple longitudinal studies have been conducted and found OAD tend to have a higher risk of tooth loss compared to non-demented. However, due to short follow-up time, questionable study designs and methods to address potential confounders, the findings of these studies might be questionable and are inconclusive. However, Given the increased risk of oral disease, impaired capacity in maintaining oral hygiene, cognitive deterioration, among OAD, clinicians believe risk of tooth loss among OAD may be higher.
Along with the increased elderly population, dementia is becoming more common. In response to this increased clinical need to improve oral health in patients with dementia, this study was conducted with the hypothesis that tooth survival does not differ in patients with and without dementia. The goals of this study were to examine the association between dementia and tooth loss, and to detail tooth loss pattern of OAD under the current model of care. Knowing this information is important for clinicians. It can not only help us develop appropriate preventive care plan for those with increased risk to lose teeth, it can also help us optimize prosthodontic treatment plan for those with high risk of tooth loss and therefore optimize the treatment outcome.
This study was a retrospective cohort study. There are two important features worth to emphasize here. This study was conducted in a community-based geriatric dental clinic in Minnesota. Here, an important feature to appreciate is that this study was not to explore the natural history of tooth loss in patients with dementia, but was intended to explore the patterns of dentition survival under current model of care. Under current model of care, tooth loss is not solely due to oral diseases, but a function of the collective judgments of both patients or their representatives and dental professionals based on their perceptions, values, and available resources. This outcome could closely reflect cumulative real-world clinical results that have occurred not only due to clinical pathology, but also the treatment planning judgments of the dental professionals involved in care, and the preferences and values of patients or their responsible parties, thereby helping dental professionals better treatment plan for elderly patients with dementia.
To be able to enroll into the study, patients must meet the following criteria: Medical history available in dental records was used to identify cases with dementia. Patients met the selection criteria were assigned into two groups. Our preliminary study indicated that demented and not-demented patients were different according to characteristics associated with tooth loss, therefore the propensity score matching (PSM) method was used to remove the impacts of these differences on the study outcomes
This figure presents the mechanism used to determine the enrollment period for a subject. ** If a subject was disease-free at arrival and no treatment other than comprehensive oral exam, xray and adult prophylaxis was needed, the follow-up started at the date when comprehensive oral exam was completed. If a subject presented with untreated dental caries or other oral health needs, dental care was provided to this subject. This subject's follow-up started on the date when the subject finished the last treatment of the initial treatment plan. At this point, the subject was considered free of oral disease and the oral function was considered relatively stable. *** Considering the fact that many elderly patients used dental services episodically and did not receive dental care regularly at the study clinic, the clinic used 3 years as a cutoff line to determine patient status (active vs. inactive). If a patient did not present to the clinic for three years since the patient’s last visit, this patient would be considered inactive and the patient’s dental record would be removed from the clinic and placed in storage. We adapted this rule to determine the end-point of follow-up for subjects. *If the last visit of a subject was more than three years before the end date of the study (12/31/2006), this subject was considered lost to follow up. *The date of the last visit was considered as the endpoint for the subject. * *If the last visit of a subject was less than or equal to three years before the end date of the study (12/31/2006), this subject was considered as remaining in the study at the end of observation. *The endpoint was 12/31/2006, which was the end of the study period.
The study data came from two sources. Dental utilization data and insurance information was abstracted from the office management system used in the study clinic. Other data such as oral assessment, medical history and so on was abstracted from dental records. Based on literature and clinical relevance, 27 variables were selected and used to pretict tooth loss in OAD.
How to appropriately measure and control the impacts of comorbidity and medications on study outcome is a great challenge in clinical dental research. Previous studies tried to address this issue, but their methods were questionable. For instance, Chalmers used a binary variable, with or without 3 or more medical conditions to measure the impact of comorbidity on dental caries in OAD. This method was questionable because a patient with 3 medical conditions such as osteoporosis, hyperlipidemia and anemia is not necessary to have a higher risk to develop caries than a patient who has only one medical condition such as Hemiplegia which is the complete paralysis of one side of body as a result of stroke. Another big challenge in clinical dental research is to quantitatively measure drug-induced anticholinergic burden for polypharmacy patients. As we know many medications have anticholinergic potency which can cause dry mouth and therefore increase risk of dental caries and tooth loss. To address this issue, ADS was used in our study. This scale has been validated and showed a strong association with serum anticholinergic activity, which is the gold standard to measure the Anticholinergic burdens of medications. Based on this scale, Drugs are rated in an ordinal fashion from 0 to 3, 0 indicating no anticholinergic potency and 3 indicating marked anticholinergic burden. To measure the total anticholinergic burden for a patient taking multiple medications,
In addition to dementia, several other factors, such as age, residential status, anticholinergic burden of medications and physical mobility, may also associate with tooth survival in elderly patients. These factors may also associate with dementia and therefore they could serve as confounders and mask the impact of dementia on tooth survival. To address this issue, propensity score matching was used.
A logistic regression was developed to calculate the propensity score for study subjects. The dependent variable Y described whether the individual was demented (Yes/No); the predictor variables X were all of the variables associated with both dementia and tooth loss, and those whose associations with dementia and tooth loss were to be determined; ß = coefficients of independent variables; p = number of independent variables. Each subject was given an estimated propensity score based on this model, which was the subject’s probability of being demented given his/her age, gender, and other confounders. All the subjects (demented and not-demented) were then sorted in a increasing order by propensity score and divided into eight roughly equal-sized strata. Demented and not-demented patients were then compared using a stratified analysis with the strata based on the propensity scores.
Three regression models were developed to explore the patterns of tooth loss in older adults with dementia. Cox
Compared with non-demented subjects, demented subjects were older. More demented subjects had dental insurance and most of these patients were covered by state Medicaid program. great majority of the demented subjects were from nursing homes, but 2/3 of the non-demented subjects were community-living elders.
On average, demented subjects presented with 4 decayed teeth or retained root upon arrival, higher than the mean number in non-demented group. Compared to non-demented group, more subjects in demented group presented with poor oral hygiene. Nearly 1/3 of the subjects with dementia had high amount of dental plaque, calculus or gingival bleeding at first arrival.
Also, demented subjects presented with more chronic conditions. The mean number of medical conditions upon arrival was 10 in demented group, much higher than that in the non-demented group. Burden of comorbidity was similar in both groups because dementia itself actually account for 1 out of 1.8 Charlson comorbidity score in the demented group. So without counting the weight of dementia, the average burden of comorbidity in these two groups was similar. On average, demented subjects took more medications at arrival. Drug-induced anticholinergic burden was also higher in demented group. Another noteworthy finding is that considerable proportion of the subjects in both group took medications with anticholinergic potency upon arrival. Therefore, dry mouth may be a concern in the study subjects.
In the demented group, more than half of the subjects were moderate to severely impaired in cognition and most of them need help in transfer into the dental chair which indicated moderate to severe impairment in physical mobility. As a result of cognitive and motor disorder, more than two third of the demented subjects lost their capacity to maintain oral hygiene at first arrival.
There was no statistical difference in tooth loss between demented and non-demented subjects. On average, 27% of the subjects lost at least one tooth during the follow-up.
This figure shows the tooth survival curve in patients with and without dementia. As you can see, there is no much difference between these two groups in tooth survival. In detail,
Further analysis indicted there was no statistical difference between demented and non-demented groups in rate of tooth loss events. Each year, there were about 15
On average, demented patient lost
In the next few minutes, we will spend some time to discuss the significant findings of this study and their clinical indications. These findings indicated that risk of oral disease, such as dental caries and perio disease, is high in patients with dementia. The increased risk may result from dry mouth caused by the side effects of medication, it may also result from the impaired capacity to manage their oral hygiene. To address these risks, adequate preventive care need to be in place. In addition to use of fluoride, development of personalized oral hygiene training program corresponding to the cognitive function of demented patients is also necessary to maintain oral health for these patients. Given the fact that 70% of the subjects unable to efficiently manage oral hygiene, providing training and education to caregiver to help them better provide oral hygiene and identify sign and symptoms of toothache and infection is also important
Our analysis showed no difference in the patterns of tooth loss between demented and non-demented patients. 27% of the subjects These results indicated risk of tooth loss and rate of tooth loss was high in a group of subjects, regardless they were demented or not. Therefore, from clinical stand point, it is necessary to better target and distinguish these high risk patients from those low-risk elderly patients. In addition, individualized treatment plan need to be developed to appropriately address the oral health needs of these patients, not only aggressive preventive care plan need to be implement to prevent the rapid loss of oral function. The rapid tooth loss also needs to be anticipated and considered during the denture treatment planning process so that the optimal treatment outcome can be achieved.
The analysis indicated that dementia was not associated with tooth survival. However, post-hoc analysis showed that statistic power was adequate. Several factors may contribute to this negative association. First of all, As discussed, drug-related anticholinergic burden was high in the subjects. Consequently, xerostomia might be a big concern for these patients. On top of the increased risk of tooth loss resulting from functional dependency, decreased salivary flow could not only remarkably elevate the risk of caries, but also accelerate the progress of caries and periodontal disease and further increase risk of tooth loss. Although the potential difference in drug-related anticholinergic burden between demented and non-demented groups was well controlled by the propensity scores, the impact of anticholinergic adverse effect of medications on tooth survival might have been elevated to a significant level and become a dominant contributing factor for tooth loss, especially in those with functional dependency. Under this circumstance, dementia’s impact on tooth survival was undetectable. Another possible explanation for this negative association could be associated with the current model of care. Under current model of care, tooth loss is not solely due to clinical pathology, but also the treatment planning judgments of the dental professionals involved in care, and the preferences and values of patients or their responsible parties. Research showed that dentist’s decision play an important role in tooth loss among older adults. For instance, Johnson’s study showed that prosthetic considerations accounted for 46% of the total extraction in their study. Non-dental factors such as no resource for proposed care, patient’s preference also accounted for 13-17% of the total extraction and therefore is an important contribution factor for tooth loss in older adults. These findings show that dental practice pattern, dental provider’s treatment philosophy, and other non-dental factors can all contribute to tooth loss. Under this circumstance, tooth survival in demented patients may be artificially reduced and therefore, the impact of dementia was masked.
As a retrospective study based on existing data, this study has several limitations. First of all, because standard cognitive assessment tool such MMSE were not widely used in dental practice, including the study clinic. We were unable to precisely measure association between severity of cognitive impairment and risk of tooth loss. Another issue associated with the lack of use of standard instrument to assess cognitive impairment is the reliability of data regarding cognitive assessment for demented patients. Our further analysis indicated that without help from the standard instruments, nearly all the demented subjects were identified as cognitively impaired. 71% of the subjects from nursing homes were considered having cognitive impairment, and 47 percent of them were considered moderately or severely impaired. These findings consist with the results of previous studies and indicate the cognitive assessments available in the dental records were fairly reliable The 2 nd potential limitation is that we couldn’t identify the exact cause of tooth loss for the subjects since this information was recorded in paper chart so it would request a lot of effort and time to go through all the progress notes to identify the cause of tooth loss. Under this circumstance, the regression used in this study might overestimate the rate of tooth loss for the subjects of this study when subjects had multiple teeth extracted for prosthetic considerations. In this situation, all of the teeth extracted during the same cycle of prosthetic treatment were inter-correlated, and only one tooth loss event should be counted for this subject. However, because the causes of tooth loss were unknown, a tooth loss event might be counted more than once, if these extractions were completed in several visits. Due to this inherent limitation of a retrospective design, the rate of tooth loss in these subjects could be overestimated in this situation when the Poisson regression model was applied. Finally, this study was completed based on the data available in one clinic. The patient population and the practice model in this clinic may be different from those in other states. Therefore, generalizability may be an issue.
In conclusion, our study confirmed that
Pattern of Tooth loss in Older Adults with Dementia Under Current Model of Care Xi Chen, DDS, PhD Assistant Professor Department of Dental Ecology 08/10/10 Xi Chen, UNC School of Dentistry
08/10/10 Xi Chen, UNC School of Dentistry Non-demented Group (N=372) Demented Group (N=119) P value Length of enrollment 39.2 37.5 0.4598 Age at enrollment 73.8 81.5 <.0001 Gender Male 29.6 25.2 0.3592 Female 70.4 74.8 Dental insurance No 33.1 15.1 0.0002 Yes 66.9 84.9 Residential status Community 65.6 10.1 <.0001 Assisted living 9.4 4.2 Nursing home 25.0 85.7
08/10/10 Xi Chen, UNC School of Dentistry Non-demented Group (N=372) Demented Group (N=119) P value Number of remaining teeth 19.6 18.1 0.0610 Number of decayed/broken teeth 3.1 4.2 0.0056 Number of teeth with restoration 11.4 10.4 0.1439 Percent of decayed/broken teeth among the remaining teeth 18.5 27.4 0.0006 Percent of filled teeth among the remaining teeth 57.5 56.2 0.6070 Calculus / Plaque / Gingival bleeding (%) None 1.2 0.9 <.0001 Small to moderate 85.5 67.9 High 13.3 31.3 Use of prosthesis at arrival (%) No 65.6 67.2 0.7431 Yes 34.4 32.8
* ADS – Anticholinergic Drug Scale 08/10/10 Xi Chen, UNC School of Dentistry Non-demented Group (N=372) Demented Group (N=119) P value Number of medical conditions 5.9 9.5 <.0001 Burden of comorbidity (Charlson comorbidity index) 1.0 1.8 <.0001 Number of medications 6.2 7.9 0.0003 Sum of ADS * of current medications 1.8 2.3 0.0433 Maximum of ADS * of current medications (%) 0 39.7 18.1 0.0002 1 37.0 56.0 2 7.6 9.5 3 15.8 16.4
08/10/10 Xi Chen, UNC School of Dentistry Non-demented Group (N=372) Demented Group (N=119) P value Cognitive impairment (%) None 82.9 2.5 <0.0001 Questionable 4.1 0.9 Slight 8.7 43.2 Moderate to severe 4.4 53.4 Physical mobility (%) Walk independently 66.5 17.1 <0.0001 Need walker 19.2 30.8 Need help in transfer 14.3 51.3 Bedridden 0 0.9 Capacity to perform oral hygiene (%) Self sufficient 84.0 21.0 <0.0001 Need help 16.0 74.0 Won’t cooperate 0 5.0
Characteristics of tooth loss between demented group and non-demented group
08/10/10 Xi Chen, UNC School of Dentistry Demented Group Non-demented Group P value Percent of subjects with tooth loss events 28.6 26.9 0.7187 Mean number of teeth lost among the subjects with tooth loss events 2.7 2.4 0.4737
Results Tooth survival P = 0.50; Hazard Ratio = 0.92 for demented vs. non-demented subjects with 95% confidence interval (0.59, 1.63) 08/10/10 Xi Chen, UNC School of Dentistry Time Percent with tooth loss event Non-demented Demented 12 m 11.3 10.8 24 m 21.1 23.8 36 m 26.4 33.2 48 m 31.0 37.3 60 m 38.4 37.3
Ratio of tooth loss events for demented and non-demented subjects = 0.93, with 95% confidence interval (0.62, 1.39) 08/10/10 Xi Chen, UNC School of Dentistry Rate of tooth loss per 100 patient-year (SE) 95% confidence interval P Value Demented group 14.9 (2.04) (11.4, 19.5) 0.9943 Non-demented group 14.9 (1.36) (12.4, 17.8)
Ratio of rate of teeth lost per patient per 5 years for demented and non-demented subjects = 1.05, with confidence interval (0.55, 1.98) 08/10/10 Xi Chen, UNC School of Dentistry Number of teeth lost per patient per 5 years (SE) 95% confidence interval P Value Demented group 1.21 (0.25) ( 0.80, 1.82) 0.4764 Non-demented group 1.01 (0.15) ( 0.76, 1.34)