Predictors for oral and general health outcomes and quality of life among older people attending general practice clinics in South Australia.

BACKGROUND
As the proportion and number of older people in Australia continue to grow, innovative means to tackle primary care and prevention are necessary to combat the individual, social and economic challenges of non-communicable diseases.


OBJECTIVE
To assess risk factors (or predictors) for oral and general health outcomes and quality of life of older people (75+ years.) attending general practice (GP) clinics in South Australia.


METHODS
Data were collected from older people attending 48 GP clinics in metropolitan South Australia. Age, sex, education, living arrangement, material standards, chronic conditions and nutrition were assessed as risk factors. Global self-rated oral and general health and quality of life (OHIP Severity and EQ-5D Utility) were included as outcome measures.


RESULTS
A total of 459 participants completed the study; response rate was 78%. In the adjusted models, high satisfaction with material standards and good nutritional health were positively associated with all four oral and general health measures. Sex (β = -0.08), age (β = -0.09) and number of chronic conditions (β = -0.12) were negatively associated with EQ-5D, while living arrangement (β = 0.07) was positively associated. Further, having four or more chronic conditions (RR:1.47) was significantly associated with self-rated general health.


CONCLUSION
Satisfaction with material standards and nutritional risk were consistent predictors for oral and general health outcomes and quality of life of older people visiting GP clinics.


Introduction
Population ageing is a consistent demographic trend and as in most developed countries both the proportion and number of older people is expected to grow in Australia. 1,2In 2016, about 15% of all Australians (3.7 million) were aged 65 years and over; 2 Over the next three decades older people will account for 20.3% (or 8.1 million) of Australia's overall population. 3By 2046, people aged 75-84 years and 85+ years will account for 35% (2.6 million) and 19% (1.4 million) respectively of all older people (aged 65+yrs) in Australia. 4Based on the 2015 Survey of Disability, Ageing and Carers, one in two of older Australians had some form of disability. 5Many however, manage to live independently; nearly 95% of all older Australians live in households, either at their homes or self care retirement villages. 5Ageing increases the risk for compromised oral and general health conditions and associated disability, leading to an additional burden on health care resources, and challenges towards service provision and health workforce availability.
Oral health plays an essential role in healthy ageing.Tooth decay and periodontal disease continue to have a signi cant impact on older people's oral health, general health and quality of life. 6In Australia, nearly 45% of older people reported having tooth decay as the leading cause of tooth extraction 7 ; over half (54.4%) had periodontal disease, compared with 2.7% of people aged 15-24 years. 8Older people carry a higher risk of periodontal disease and are more likely to experience advanced stages of periodontitis, leading to consequences such as root caries, impaired eating and reduced socialisation. 9Poor oral health also affects nutritional intake 10,11 , contributes to physical frailty 12 , and is linked with several other chronic conditions such as diabetes 13 , cardiovascular disease 14 , kidney disease 15 and respiratory infections. 16ltimorbidity or co-occurrence of chronic conditions is also common among older people.The Australian Institute of Health and Welfare (AIHW) reports 87% of people aged 65 years and older had at least one chronic condition in 2014-15. 17Among Australians aged 65 years and over 60% had two or more chronic conditions 18 and 29% had three or more chronic conditions. 19Arthritis, chronic back pain, cardiovascular diseases and psychological problems were among the commonly co-occurring chronic conditions among older people. 20,21ronic conditions are a signi cant burden to the health care system in terms of growing hospitalisations. 193][24] With over 10,000 annual potentially preventable hospitalisations in older people due to oral and dental conditions, strengthening early intervention and primary care in older people is a vital issue. 25,26.
In Australia, public dental care for older people is means tested and require an appropriate health care card (such as pensioner card, veterans card or seniors card) to avail a large majority of dental services.Only about one in four of older people (25.5%) aged 65 years and over are eligible for public dental care 8 , and just about one in ten (12%) older people had received public dental care in the last 12 months. 27With a majority of dentists in Australia (85% in 2016) providing care in private dental clinics, costs are covered by private health insurance or through out of pocket payments. 28Costs have been commonly reported as a barrier to accessing dental care, and about 32% of older people who had no insurance avoided seeing a dentist. 2,8To date, there is an increasing emphasis for innovative primary care pathways for improving oral health for older people and the use of multiple points of care for prevention, screening and referral that are easily accessible by older people. 29,30e growing importance of the older age group and need for innovative solutions that focus on prevention and primary care have led the World Health Organization to produce a 'Call for Public Health Action' on the oral health of older people. 31The approach calls for a better understanding of common risk factors for oral and general health, and encourages the use of integrated care solutions to tackle the gap in service provision for older people. 29,31An integrated approach towards screening, diagnosis and treatment of oral health conditions could help reduce waiting times and improve the e ciency of an already burdened public dental system. 32The use of a broader medical, nursing and allied health team could help in the early identi cation of risk factors, and thereby promote healthy ageing.Medical practice or General Practitioner (GP) visit is the most commonly availed primary care service in Australia and covered by universal health insurance (or Medicare).In 2016-17, there were over 38 million Medicare claims for unreferred GP visits by older people, which is about 30% of all Medicare claims. 33e integration of oral health assessment and screening at GP clinics presents as a viable solution, but its potential use and care pathways require a better understanding.
The objective of the study was to assess risk factors (or predictors) for oral and general health outcomes and quality of life among older people visiting general practice (GP) medical clinics in South Australia.It was anticipated that by gaining a better understanding of predictors, one might be able to contribute towards health service planning and integrated care models for the management of oral disease at a primary care setting.

Methods
The overall design of the study involved a follow-up of non-institutionalised[*] older people (75+ years) attending publicly funded annual health assessments in GP medical clinics in South Australia.More detailed information on the study design and methods are published elsewhere 30,[34][35][36] .The methods presented in this paper are limited to the analysis of risk factors.

Sample and data collection
Participants for the study were drawn from a convenience sample of 48 GP clinics, mainly in metropolitan South Australia.All older people aged 75 years and above visiting these GP clinics for annual health assessments were asked to participate in a questionnaire study on oral and general health.General practice staff (nurse or doctor) evaluated the cognitive ability of all eligible candidates.Some 590 participants consented to participate in the study.A 12 page self-report questionnaire was mailed to each participant.Two follow-up reminders were sent across a three-week interval to improve the response rate.All data collection occurred between December 2013 and September 2014.

Data items
The self-report questionnaire collected a variety of information including: demographic details (such as age, sex and educational level); socio-economic data (such as nature of living arrangements and satisfaction with material standards); chronic medical conditions; nutritional risk information; along with oral and general health measures.Regarding living arrangement, participants were questioned "who do you live with?" with four possible responses (spouse, other family, other persons or live alone).Satisfaction with material standards was worded as "overall, how satis ed are you with the material standards of your life?" and responses were collected on a Likert scale (0 =totally dissatis ed to 10=totally satis ed).Participants self-assessed their experience of chronic medical conditions by selecting from a list of 22 commonly occurring conditions [ †].Multiple responses were allowed.Nutritional risk was assessed using the 10-item "Nutrition Screening Initiative" checklist 37 .A dichotomous (yes or no) response was used for these items.
Oral health outcomes were measured using a global self-rated item, and a 14-item Oral Health Impact Pro le (OHIP).The global self-rated oral health (SROH) question (How would you rate the overall health of your teeth, dentures, and gums?) essentially asked the participant to rate their oral health in a ve-point Likert scale (ranging from 1=poor to 5= excellent) 38 .Oral health-related quality of life was measured using a 14-item version of the Oral Health Impact Pro le (OHIP) 39 .The OHIP-14 measured seven dimensions that included functional limitation, physical pain, psychological discomfort, physical disability, psychological disability, social disability and handicap 39,40 .Item level responses were recorded in a 5 point Likert scale: 0=never; 1=hardly ever; 2= occasionally; 3=fairly often and 4=very often.
General health outcomes were measured using a global self-rated item and a quality of life measure.The selfrated general health (SRGH) question was worded as, "In general, how would you say your health is?"; and was measured on a ve-point Likert scale (ranging from 1=poor to 5= excellent).Quality of life was measured using the EuroQol ve-dimension (EQ-5D) instrument that assessed health problems on a 3-point scale 40,41 .The ve dimensions included mobility, self-care, usual activities, pain/discomfort and anxiety/depression 41 .Participants were asked to rate their level of di culty (or) problems in each of these dimensions.For example, the dimension pain/discomfort had three response categories: "I have no pain or discomfort"; "I have moderate pain or discomfort"; "I have extreme pain or discomfort."

Data analysis
The self-ratings of oral health (outcome variable) were dichotomised into two groups: 'poor/fair' as poor oral health, and 'good/very good/excellent' as good oral health.The identical approach was adopted for self-rated general health.The OHIP scores (outcome variable) were derived by summing responses to all 14 questions.
Potential range for scores is 0 to 56 higher scores indicated more adverse impact (or severity).OHIP severity as opposed to OHIP prevalence (which is the percentage of people responding with fairy often or very often) has been used, as prevalence score evaluate only frequently occurring impacts -while the severity score by using all response categories overcomes limitations inherent in restricting summary scores to an arbitrary threshold. 42EQ-5D responses were converted to health utility values, where each set of responses on the standard instrument was matched to a health state value (0 = death and 1.0 = perfect health) using an algorithm derived from modelling values using health state preferences from the Australian population; higher scores indicated better health 40 .
All the other independent variables were either already represented as binary variables (sex, age group, education level) or dichotomised into two groups.Living with spouse, family or others was coded as 'living with others'; the rest coded as living alone.A response ≥ 8 to satisfaction for material standards was considered as highly satis ed; the rest coded as less satis ed.The number of chronic conditions for each participant was counted; based on the sample distribution a new variable measuring the extent of these conditions was derived (less than four conditions and 4+ conditions).Nutritional risk items were scored based on recommended values provided by the Nutritional Screening Initiative checklist 37 .These scores were later dichotomised into two groups: 'low to moderate risk' (classi ed here as good nutritional health) and 'high risk'.Sample characteristics for the full sample were mainly analysed through counts and proportions.All four outcome variables were rst cross-tabulated with chronic risk factors, and described through row counts and proportions/means for poor health.The strength of these bivariate associations was determined through a chisquare or t-test, with a level of signi cance set at p ≤ 0.05.In the multivariate analysis we rst used log binomial regression for self-rated oral and general health to estimate prevalence ratios.However, due to non convergence issues, a Cox regression with constant time was used to produce hazard ratio equivalent to relative risk 43 .Linear regression was used for quality of life measures.All data were analysed using IBM SPSS Version 24.

Ethical considerations
Ethical approval for the study was obtained from the University of Adelaide (Ethical approval number: H-2013-057), and the study was conducted in accordance with the Declaration of Helsinki.Written informed consent was obtained from all participants at the time of their recruitment in the general practice clinics (by a nurse or a general practitioner).
[*] Non-institutionalised older people included only those who were living at home, and not those hospitalised, or living in nursing homes or community care facilities.

Results
A total of 459 participants returned the completed questionnaires, providing an overall response rate of 78 per cent.

Sample characteristics
A larger proportion of the respondents were female (54%), and nearly a quarter of all participants were 85 years or older (24%).About 40 % of the respondents were living alone.Over half of all respondents (55%) reported 4 or more chronic conditions, and more than one-third of the participants (37.2%) were classi ed as having high nutritional risk (see Table 1).

Bivariate analysis
The bivariate analysis of the explanatory variables with the oral and general health outcomes is presented in Table 1.Satisfaction with material standards and nutritional risk were signi cantly associated with all four outcome variables.Sex, age and education level were associated only with quality of life.Females and older adults aged 85+ years had lower health utility scores, while participants with tertiary education or above had higher health utility scores.A larger proportion of participants 'living with others' reported poor self-rated general health.Having 4 or more chronic conditions was signi cantly associated with self-rated general health and quality of life.

Regression analysis
Table 2 presents the adjusted regression analysis of the explanatory variables with the four oral and general health outcomes.Self-rated oral health was signi cantly associated with nutritional risk.Participants classi ed as having high nutritional risk were more likely to report poor self-rated oral health (RR: 1.50), compared to those classi ed as having good nutritional health.Oral health-related quality of life was signi cantly associated with satisfaction with material standards and nutritional risk.The OHIP severity scores were negatively associated with satisfaction with material standards i.e. the higher the satisfaction the lower the OHIP severity.In contrast, older people having high nutritional risk were more likely to have higher OHIP severity scores (B: 8.55) compared to older people having good nutritional health.
In regard to general health measures, self-rated general health was signi cantly associated with satisfaction with material standards, chronic conditions and nutritional risk.Participants highly satis ed with material standards were less likely to report poor general health, compared to those less satis ed with material standards (RR: 0.70).
Further, older adults having 4 or more chronic conditions, and those classi ed as having high nutritional risk were more likely to report poorer general health, when compared to those having less than four chronic conditions (RR:1.47) and good nutritional risk (RR: 1.72) respectively.Quality of life was signi cantly associated with sex, age, living arrangements, satisfaction with material standards, chronic conditions and nutritional risk.Health utility scores were negatively associated with sex (females, B:-0.08), and age (85+yrs, B:-0.03).Older adults living alone were more likely to have higher health utility scores compared to those living with others.Health utility also had a negative association with the number of chronic conditions and nutritional risk, and a positive association with satisfaction with material standards.

Discussion
This study assessed predictors of oral and general health outcomes and quality of life of older people (75+years) visiting GP clinics in South Australia.Nutritional status and satisfaction with material standards offered better predictive capacity across the observed oral and general health measures.Multimorbidity (4+ chronic conditions) was signi cantly related with the both the general health measures; living arrangement with general health related quality of life.
In this study, older people classi ed as high nutritional risk had poor oral and general health outcomes and quality of life.Nutrition undoubtedly plays a major role in healthy ageing. 44Poor nutritional intake is associated with frailty, 44,45 which can adversely affect quality of life.Prior studies have identi ed that the prevalence of malnutrition is increasing among older people. 46Malnutrition is also associated with a reduced functional status and cognitive function, immune dysfunction, higher hospitalisation rate and mortality. 46,47The Nutrition Screening Initiative Checklist was employed on non-institutionalised older people attending a general practice setting in South Australia.Similar to the original study based in New England, United States (where the Checklist originated) 37 , the current study suggests that nutritional risk assessment is capable of enhancing older people's understanding of the determinants of nutritional well-being and also encouraging early assessment and intervention strategies at a primary care level.The relative ease of using the nutritional screening questions provides an opportunity for public health or aged care personnel or older people themselves to use the questionnaire to understand their nutritional risk status and thereby encourage visiting necessary medical (general practitioner) or allied health personnel (such as a nutritionist).It should be noted that nutritional risk assessment is only a screening tool 37 , and is not meant to replace regular assessment by quali ed medical or allied health personnel.
Poor nutritional intake both affects and is affected by oral conditions. 48Compromised oral health status due to tooth loss, periodontal disease or edentulousness affects the ability to maintain good nutritional status among older people. 49,50This two-way relationship between oral health and nutrition stresses the importance of including nutritional risk screening as a part of regular prevention and health promotion activities at primary care facilities including both general medical and dental practice clinics.Nevertheless, the Nutritional Screening Checklist mainly measures nutritional inadequacy 37 ; nutritional problems related to dietary excess and its consequences (such as obesity) may not be well represented.In a population-based study by Brennan and Singh  (2012) among older people in South Australia, it was found that compliance to dietary guidelines promotes good nutrition through healthy eating. 51Therefore, there is evidence to suggest that dietary guidelines along with a nutritional screening can be used as a health promotion tool under the guidance of appropriately trained health personnel.
A subjective measure for socioeconomic status (satisfaction with material standards) was used in the study, rather than an income-based measure.Prior research has suggested that subjective measurement of socioeconomic status offers the capacity to integrate experiences of stress, social position, perception of inequality and highlight the cumulative effect of changing status over lifetime. 52As income measures lose utility among older age groups, a subjective measure is argued as being more useful as well as capturing multiple domains in one measure. 35,53Similar to prior research 52 , this study provides further evidence that older people with lower social status are prone to poorer oral and general health outcomes and quality of life.The predictive nature of this measure supports its usefulness as a standard measure in determining oral and general health risk both in a primary care clinical setting as well as population-based studies.
Multimorbidity, as examined in the study, showed poor general health outcomes and quality of life among older people.This nding is similar with both national and international literature identifying the growing relevance of multiple chronic conditions among older people. 19,20,23Chronic conditions cause a signi cant burden to the health system 18 ; a large number of hospitalisations due to chronic conditions are preventable stressing the importance to early detection and employing preventative strategies at primary care level. 19,54The selection of 4+ as a cut-off point was determined based on the underlying distribution of values; no attempt was made to thematically group the chronic conditions that might be more relevant to oral health.Future research is recommended that selectively identi es associations between groups of chronic conditions and oral and general health outcomes and quality of life among older people.This will improve evidence on health promotion and health education strategies for older people at both medical and dental facilities.Our ndings suggest that living alone was positively associated with general health related quality of life.The most parsimonious explanation for this association is self-selection, i.e. those 75+ with poor QOL are able to remain living in the community only due to having spousal carers (90% of those 'living with others' in our study live with their spouses). 36is study applied a few measures such as self rated oral and general health, OHIP-14, EQ-5D, nutrition checklist, material standards, chronic conditions along with other demographic characteristics.Single-item self rated global measures for oral and general health are a proven way of assessing self-perception of oral and general health. 55Due to the ease in deploying such measures and cost effectiveness, it provides a simple solution for individual assessment at small practice settings. 56OHIP-14 is widely used globally to measure self-reported functional limitation, discomfort and disability attributed to oral conditions.Assessing and understanding consequences or social impacts of oral disease can be used in understanding oral health related behaviours, improving care provision and in advocacy. 57Health related quality of life was represented by quality weights (health utility) that re ect an individual's preference for different health outcomes. 41,58EQ-5D is a popular multi attribute utility instrument used for indirectly estimating health state utility values.The question emerges as to how these various measures can be deployed and used in routine clinical settings such as general practice (GP) or dental or allied health practitioner clinics.In this study, a questionnaire was sent to consenting older people visiting GP clinics; returned questionnaires were then analysed separately by a research team.Currently, the use of these assessment tools as part of routine care has some limitations.Improving teamwork and collaboration among various health personnel groups, as well a re-examining scope of practice, interprofessional education and training philosophies and above all governance and policy support is essential for integrating oral and general health. 59Electronic health records are one such means that provides vertical integration across all systems or care, and has the potential to be used in such assessments for older people oral and general health. 59uture studies will need to examine the feasibility of incorporating such measures in electronic health record systems used in clinical settings and in providing prognostic tools at the point of care but more importantly for timely use of such information in management and policy decisions.
Nearly 95% of all older people aged 65years or over are community dwelling adults, and live in their own homes or retirement villages. 5Integration of oral health care into existing support systems available for older people such as Home Care Packages 60 as well as services that older people more commonly use (such as visiting GP clinics) provide a means for improving early intervention and primary care. 61Provision of appropriate information as well as referral pathways to a trained dental professional are necessary aspects in such multidisciplinary involvement. 32,59The growing relevance of the older age group for health systems in Australia calls for improved collaboration and teamwork that spans medical, dental and allied health personnel. 62,63ntegrated care provision through medical, dental and allied health personnel that caters to the needs of older people is vital to promote healthy ageing.Further research is required on how various health professional teams can work together to promote a preventative philosophy, as suggested by the WHO Call for Public Health Action, and at a primary care level to serve older people.

Limitations
The study used a convenience sample of non-institutionalised older people attending participating GP clinics in metropolitan South Australia.The study sample was relatively younger and included more males when compared to national estimates. 30It is suggested that readers observe caution in generalising ndings, as the study sample is less likely to be fully representative of all older people aged 75+ years, especially those living in rural areas or based in other settings such as residential care facilities or hospitals.The use of dichotomised predictors re ected the aim of identifying potential risk predictors and comparing the patterns of their associations across different measures of oral and general health.The nature of the sample was limited to community dwelling older persons centred around metropolitan Adelaide.So, while broadly representative of that population, it is likely to be more homogenous than other samples over wider geographic regions and larger age ranges.The crosssectional nature of the study also limits us in making any causal inferences.The study used existing primary care infrastructure and medical/nursing personnel to screen or determine the eligibility of participants, which was later followed up by a self-reported questionnaire.The response rate for the study was high.The data for the study were based on self-reported measures.No attempt was made to clinically examine/verify oral or general health conditions, which was beyond the scope of this study.[66]

Conclusion
This study shed light on risk factors for oral and general health in older people aged 75+ years and living in South Australia.Satisfaction with material standards and nutritional status seemed to offer consistent prediction towards understanding both oral and general health.Primary care teams involving general practitioners, nurses and allied health practitioners are in a good position to assess risk factors for older people, and work alongside the dental team.The involvement of multidisciplinary teams in primary care and prevention can contribute towards an integrated care approach in aged care and strengthen the goal of bringing oral health more closely aligned towards general health.Note: Chi-Square Test was used for SROH (Self Rated Oral Health) and SRGH (Self Rated General Health); t-test was used for OHIP (Oral Health Impact Pro le) Severity and EQ-5D (EuroQol 5 Dimension) Health Utility scores; The Full Sample presents column percent, while the outcome variables present row percent.

Table 2 :
Adjusted analysis of study variables and oral and general health outcomes *p<0.05; **p<0.01;PR: Prevalence Ratio; B: Regression Coe cient; Ref: reference group