Community Dent Oral Epidemiol 2012; 40: 5361 All rights reserved

2011 John Wiley & Sons A/S

The inuence of dental attendance on change in oral healthrelated quality of life

Crocombe LA, Brennan DS, Slade GD. The inuence of dental attendance on change in oral healthrelated quality of life. Community Dent Oral Epidemiol 2012; 40: 5361. 2011 John Wiley & Sons A S Abstract Background: Few longitudinal studies have investigated the association between dental attendance and oral healthrelated quality of life (OHRQoL). These studies were limited to older adults, or to study participants with an oral disadvantage and did not assess if dental attendance had a different effect on OHRQoL for different people. Objective: This project was designed to test whether routine dental attendance improved the OHRQoL of survey participants and whether any patient factors inuenced the effect of dental attendance on change in OHRQoL. Methods: Collection instruments of a service use log book and a 12 month follow-up mail self-complete questionnaire were added to the Tasmanian component of the National Survey of Adult Oral Health 2004 06. The dependent variable was change in OHIP-14 severity and the independent variable was dental attendance. Many putative confounders effect modiers were analysed in bivariate, stratied and threemodel multivariate analyses. These included indicators of treatment need, sociodemographic characteristics, socioeconomic status, pattern of dental attendance and access to dental care. Results: None of the putative confounders were associated with both dental attendance and the change in mean OHIP-14 severity. The only statistically signicant interaction for change in OHIP-14 severity was observed for dental attendance by residential location (P < 0.01). In multivariate analysis, there was a statistically signicant association of dental attendance with change in mean OHIP-14 severity. It also showed that the difference in association of attendance between Hobart, the capital city of Tasmania, and other places was statistically signicant based on the interaction between residential location and attendance (P < 0.05). Conclusion: The effect of dental attendance on OHRQoL was inuenced by a patients residential location.

Leonard A. Crocombe1, D. S. Brennan1 and G. D. Slade2

Australian Research Centre for Population Oral Health, School of Dentistry, The University of Adelaide, Adelaide, SA, Australia, 2Department of Dental Ecology, UNC School of Dentistry, University of North Carolina, Chapel Hill, NC, USA
1

Key words: dental services research; epidemiology; health services research; oral health; quality of life Dr Leonard Crocombe, Australian Research Centre for Population Oral Health, School of Dentistry, The University of Adelaide, Adelaide, SA 5005, Australia Tel.: +61 0419 597 756 Fax: +61 3 6226 7704 e-mail: leonard.crocombe@adelaide.edu.au Submitted 9 September 2010; accepted 18 July 2011

Although dental clinical indicators are important in clinical practice and may be used to provide an indication of treatment need (1), problems arise when these indices are used as measures of oral health (2). Clinical indicators reect the endpoint of the disease and do not indicate the impact of the disease process on function or the persons wellbeing. Oral healthrelated quality of life (OHRQoL) measures give the patients perspective on the impact of dental treatment. Few longitudinal studies have investigated the association between dental attendance and OHRdoi: 10.1111/j.1600-0528.2011.00634.x

QoL. Fiske et al. (3) studied 100 elderly British people and found that 74% beneted from treatment. Petersen and Nortov (4) studied 187 Denmark pensioners aged 6770 years who were given dental care, and after 3 years, the percentage of participants who reported poor denture function had declined and, at follow-up, less felt embarrassed by teeth or preferred food that was easy to chew. Locker (5) conducted a study over a 3-year period in an older adult Canadian population and found that people whose oral health status had improved had made signicantly more dental

53

Crocombe et al.

visits and received signicantly more dental services than those who oral health status had deteriorated or did not change. Fisher et al. (6) using data from the longitudinal Florida Dental Care Study found that dental services were effective in resolving oral disadvantage, where an oral disadvantage was dened as avoiding laughing or smiling because of unattractive teeth or gums, avoiding talking to someone because of unattractive teeth or gums or bad breath, or being embarrassed by the appearance of teeth or gums. Gagliardi et al. (7) surveyed South Australian community dwelling adults aged 75 years or older and found that dental care improved the study participants subjective oral health. However, the above studies were limited to older adults (35, 7), or to study participants with an oral disadvantage (3, 6, 7), and did not assess whether dental attendance had a different effect on OHRQoL for different people. Knowing whether the effect of dental attendance varies between different groups of people would assist politicians, policy makers and administrators when allocating limited valuable resources to improve population OHRQoL. There is a need for research investigating the association between dental attendance, patient factors and OHRQoL that is prospective to demonstrate temporal sequence and that is based on a population sample for representativeness. This project was designed to test whether dental attendance improved the OHRQoL of survey participants and whether any patient factors inuenced the effect of dental attendance on change in OHRQoL.

complete questionnaire of interviewed people who were not examined, a service-use log book issued to study participants after the examination, a follow-up mail self-complete questionnaire after 12 months; and if the logbooks were misplaced, a dental treatment audit. Change in oral health was obtained from the follow-up mail self-complete questionnaire, dental attendance from the service-use log book or the dental treatment audit, study participant characteristics from the CATI.

Dependent variable
The Oral Health Impact Prole (OHIP) measures peoples perception of the social impact of oral disorders on their well-being (9). The short-form OHIP (OHIP-14) has 14 questions, two of which answer each of seven content areas of functional limitation, physical pain, psychological discomfort, physical disability, psychological disability, social disability and handicap, and clinical data from the oral epidemiological examination. OHIP-14 severity was calculated for baseline and 12-month follow-up OHIP-14 severity was dened as the sum of the ordinal responses, where never was coded as 0, hardly ever as 1, occasionally as 2, fairly often as 3 and very often as 4 (9). To calculate change in OHIP-14 severity, the baseline score was subtracted from the follow-up score so that a change score less than zero indicated that the survey participant had improved OHRQoL. A survey participant could theoretically have a change in OHIP-14 severity score ranging from )56 to +56.

Independent variable

Methods
Sampling and data collection
The baseline data were obtained from the Tasmanian component of the National Survey of Adult Oral Health 2004 06 (NSAOH, 8), a cross-sectional study of a clustered stratied random sample of dentate Australians aged 15 years or more. The collection procedures for the NSAOH comprised a computer-assisted telephone interview (CATI) by trained interviewers, an oral epidemiological examination conducted by calibrated dentist examiners and a mailed self-complete questionnaire. The collection instruments that were added to the Tasmanian component of the NSAOH for this study were a backup baseline mail self-

Dental treatment was collected at the level of individual service items, using the three-digit coding scheme of the Australian Dental Associations Schedule of Dental Services and Glossary, Edition 9. Survey participants for whom at least one service item number was recorded were classied as having made a dental visit.

Putative confounders and effect modiers

Study participants were asked eight questions on treatment need if they had a need for extractions, llings, a scale and clean, dentures, a dental check-up, gum treatment, a dental crown or bridge or any other treatment. For this study, participants who answered that they had any selfperceived treatment need, except a check-up or a scale and clean, were considered to have a dental

54

Dental attendance and quality of life

treatment need. Survey participants were dened as having an oral clinical disease if they had at least one decayed tooth or if they had at least one tooth with a periodontal pocket depth of 4 mm or more. The survey participants sociodemographic characteristics analysed were age (1544, 4559 and 60+ years), sex (male, female), country of birth (Australia or overseas) or whether the study participant resided in Hobart (Tasmanias state capital) or not. Socioeconomic status was measured by education (Degree Teacher Nursing, Trade Diploma Certicate or no Postsecondary school education), level of annual household income (AU$30,000, AU$30,000AU$59,999 or AU$60,000+), occupation (manager professional paraprofessional, trades clerical or blue-collar worker labourer) and employment status (employed or unemployed). Occupation was based on the Australian Standard Classication of Occupations (10). People in part-time employment were classied as employed. The survey participant pattern of attendance at baseline was measured by the survey participants perception of their regularity of attendance (at least once every year or less often than once a year) and usual reason for attendance (check-up or a dental problem). Access to dental care was measured by whether the study participants were eligible for public dental care, whether they had much difculty in paying a AU$100 dental bill or whether they had avoided or delayed dental treatment because of cost.

analyses were repeated, and compared mean difference in OHRQoL and their 95% condence intervals among strata to assess potential effect modication. Interaction p values were calculated. In the multivariate analysis, a three-part model was used. A crude model, labelled Model 1, estimated the association of dental attendance with change in OHIP-14 severity. The multivariate modelling was then extended by adding the rstorder terms to the crude model (Model 2). In Model 3, interaction effects between the potential confounders/effect modiers and dental attendance on the change in OHIP-14 severity were included.

Results
Participation
Of the 1745 eligible (in-scope) household phone numbers in Tasmania, 40.3% (n = 703) were classied as nonrespondents. Four hundred and ninety people completed either the baseline mail or the back-up questionnaires, which represented over half (56.1%) of those people in scope who completed the telephone interview. Nearly threequarters of those who completed either of the baseline self-complete questionnaires completed the 12-month follow-up questionnaire (73.5%). The vast majority of people who completed the follow-up questionnaire provided treatment details (n = 320: 88.9%). At every stage of the survey, the proportion of subjects born in Australia was similar to that in the Tasmanian component of the 2006 Census (Table 1), though Aboriginal or Torres Strait Islanders were under-represented at the followup questionnaires. The proportion of subjects who spoke a language other than English at home was under-represented at baseline, but representative of the Tasmanian population at follow-up. Managers administrators were representative of the census at the follow-up questionnaires stage, but paraprofessionals were under-represented at both stages. The proportions of professionals, tradespeople, clerical workers and blue-collar operators or labourers were similar at all stages of the survey and were representatives of the Tasmanian population.

Analysis
Data were weighted by sex, age and residential location to generate all descriptive statistics, thereby producing population estimates for the target population of dentate Tasmanian adults. Categorical variables were summarized as percentages and corresponding 95% intervals, while continuous variables were summarized as means and 95% condence intervals. An analysis of the differences between responders and nonresponders to the follow-up questionnaire was performed by comparing some demographic variables at follow-up to those at baseline and to that in the 2006 Tasmanian Census. Bivariate analysis was undertaken to estimate crude effects of dental care on quality of life and to nd potential confounders. Stratied analyses were undertaken with the potential confounders/ effect modiers. Within each stratum the bivariate

Distributions
Over half the survey participants were 45 years of age or older (55.1%), there was an even split of the sexes (male 50.8%, female 49.2%), and under half

55

Crocombe et al. Table 1. Characteristics of subjects at baseline and follow-up stages of the study compared to the Census Baseline questionnaires Follow-up questionnaire 2006 Census Tasmaniab n = 362 n = 490 % % (95% CI) % (95% CI) 84.1 (79.2, 88.9) 0.3 (0.0, 0.8) 5.3 (1.5, 9.1) 11.7 22.6 5.7 11.9 26.0 22.1 (5.3, 18.2) (13.0, 32.3) (2.0, 9.4) (4.8, 19.2) (17.9, 34.0) (13.7, 30.3) 84.9 3.5 8.0 17.9 16.9 11.8 12.8 30.8 19.7

Characteristics of subjectsa

Born in Australia 87.4 (83.6, 91.2) Aboriginal or Torres Strait Islander 4.8 (0.00, 10.2) Language other than English spoken at home 4.0 (1.4, 6.6) Occupation Manager administration 9.3 (4.4, 14.2) Professional 18.9 (11.6, 26.1) Paraprofessional 6.4 (3.0, 9.8) Tradesperson 13.0 (6.4, 19.7) Clerical 26.9 (18.7, 35.1) Blue-collar operator or labourer 25.5 (16.1, 34.9)
a b

Weighted by sex, age and residential location for the appropriate stage to Australian Bureau of Statistics Census in 2006. The 2006 Census was held on 8th August 2006 and found 476 481 persons usually resident in Tasmania.

the respondents lived in Hobart (43.8%). More than three-quarters (84.1%) of the people were born in Australia. The highest categories of education (degree, teacher, nurse) and income ($60 000+) contained the lowest proportion of survey participants, but less than a quarter of respondents reported a blue-collar occupation. Just over half the respondents were employed, visited a dentist every 12 months or less and usually visited a dentist for a check-up. Over three-quarters of respondents had no or only a little difculty in paying a $100 dental bill, over 40% were eligible for public dental care and just over a third had avoided or delayed dental treatment because of cost. Over half (54.8%) the survey participants visited a dentist over the 1 year of this study, the vast majority of whom (94.9%) saw a private sector dentist. The most common types of dental care received were diagnostic, preventive and restorative services. The mean baseline OHIP-14 severity was 6.96 (95%CI=5.848.08) and mean change in OHIP-14 severity was )0.53 ()1.14, 0.81). The
200

frequency distribution of change in OHIP-14 severity over the year of this study was approximately bell shaped although highly leptokurtic (sharp peak) (Fig. 1). Just under half the survey participants felt they had a treatment need other than a check-up or a scale and clean (42.4%, 35.149.7). Over a third of the examined study participants (37.4%) had one or more dental diseases.

Bivariate analysis: dental attendance by putative confounders

Oral disease (P = 0.93), sex (P = 0.23), age (P = 0.45), country of birth (P = 0.47), occupation (P = 0.40) and employment status (P = 0.11) were not associated with dental attendance. The percentage of people who visited a dentist was higher for subjects in the degree teacher nursing category compared to both people in the trade diploma certicate category and people with no postsecondary school education (P = 0.02). Dental visiting was less frequent at lower compared to higher levels of household income (P = 0.02). People, who at baseline, reported a more frequent history of dental attendance (P < 0.01), and those who usually visited a dentist for a check-up rather than a problem (P < 0.01) were more likely to have visited a dentist over the 12 months of this study.

150

n = 358 Mean = 0.53 SD = 95%CI = 1.140.81

Frequency

100

50

Bivariate analysis: change in OHIP-14 by putative confounders

There was no statistically signicant association of dental attendance (P = 0.14) with change in mean OHIP-14 severity. Neither treatment need (P = 0.84), oral disease (P = 0.10), nor any of the survey participant sociodemographic factors (sex P = 0.87, age P = 0.56, country of birth P = 0.16,

0 30 25 20 15 10 5

Change in OHIP severity

5 10 15 20 25 30 35

Fig. 1. Change in OHIP severity frequency statement.

56

Dental attendance and quality of life

residential location P = 0.42), socioeconomic factors (education P = 0.66, household income P = 0.98, occupation P = 0.94, employment status P = 0.35), pattern of attendance measures (regularity P = 0.13, usual reason P = 0.27), or access to dental care measures (eligibility for public dental care P = 0.55, difculty paying a AU$100 dental bill P = 0.31, avoided or delayed dental care because of cost P = 0.50) were statistically associated with change in mean OHIP-14 severity (Table 2). None of the putative confounders were associated with both dental attendance and the change in mean OHIP-14 severity and were dened as not being confounders.

Table 2. Relationship between putative confounders effect modiers and change in mean OHIP-14 severity Change in mean OHIP-14 severity n Mean )0.73 )0.69 )0.59 )0.49 )1.26 )0.27 )0.48 )0.58 )0.16 )1.51 0.12 )0.32 )1.65 )0.81 )0.31 )1.11 )0.28 )0.49 )0.87 )0.38 )0.64 )0.79 )0.87 )0.48 )0.75 )0.23 )0.92 )0.09 )0.81 )0.18 )0.22 )0.69 )0.60 )0.30 )0.23 )0.69 95% CIs )1.61, )0.15 )1.44, )0.05 )0.64, 0.46 )1.06, 0.08 )2.54, 0.02 )0.83, 0.28 )1.48, 0.51 )1.31, 0.15 )0.79, 0.47 )2.63, )0.39 )1.26, 1.51 )0.92, 0.27 )2.91, )0.40 )1.89, 0.27 )1.11, 0.48 )2.18, )0.03 )1.44, 0.87 )1.45, 0.48 )2.29, 0.55 )1.61, 0.84 )1.35, 0.07 )1.77, 0.20 )2.05, 0.31 )2.63, 1.68 )1.51, 0.01 )1.26, 0.80 )1.63, )0.21 )1.14, 0.95 )1.51, )0.12 )1.19, 0.82 )1.50, 1.06 )1.46, 0.08 )1.31, 0.12 )2.21, 1.61 )1.46, 0.99 )1.31, )0.06 0.50 P 0.14 0.84 0.10

Putative confounders effect modier

Stratied analysis
The only statistically signicant interaction for change in OHIP-14 severity was observed for dental attendance by residential location (P < 0.01). For people who resided in Hobart, there was a greater reduction in mean OHIP-14 severity scores for people did not attend a dentist ()2.02) than those that did ()0.38, P = 0.01). People from areas outside Hobart who visited a dentist had a greater decrease in OHIP-14 severity than dental attenders who resided in Hobart, while the opposite association was true for those who did not visit a dentist (Table 3). There was a statistically signicant lesser reduction in mean OHIP-14 severity for those who visited a dentist than those who did not visit a dentist in the strata of people with a household income of AU$60,000 or more (Visit = )0.82, Not visit = )1.28, P = 0.01) and of people with none, hardly any or a little difculty paying a AU$100 dental bill (Visit = ).73, Not visit = )1.25, P = 0.03).

Multivariate analysis
In Model 1 (the crude model), there was no meaningful relationship between dental attendance and change in mean OHIP-14 severity (Table 4). The parameter estimate ()0.04) was very close to zero and was statistically nonsignicant (P = 0.94). After adjustment for rst-order terms in Model 2, dental attendance was associated with an increase in mean OHIP-14 severity (parameter estimate = 0.60), although the association was not statistically signicant (P = 0.23). In contrast to Model 2, in Model 3, there was a statistically signicant association of dental attendance with change in mean OHIP-14 severity. It also showed that the difference in association of attendance between Hobart and other places was

Dental attendance Yes 175 No 143 Treatment need Yes 150 No need 208 Oral disease Yes 89 No disease 269 Sociodemographic factor Sex Male 132 Female 226 Age 15 to <45 years 114 45 to <60 years 127 60+ years 117 Country of birth Australia 301 Other 57 Residential location Hobart 156 Other 202 Socioeconomic factor Level of education Deg. Teach Nurse 77 Trade Dip. Cert. 122 No Postsec. Edu. 139 Household income <$30 000 141 $30 000<$60 000 103 $60 000+ 93 Occupation Manage Prof Para. 80 Trades Clerical 76 Blue Col. Lab. 32 Employment status Employed 188 Not employed 165 Pattern of attendance Regularity 12 months 183 >12 months 175 Usual reason Check-up 194 Problem 163 Access to dental care Healthcare card Yes 151 No 207 Diff. pay $100 dental bill None, hardly any, 280 a little A lot 78 Avoided because of cost Yes 126 No 232

0.87 0.56

0.16 0.42

0.66

0.98

0.94

0.35

0.13 0.27

0.55 0.31

57

Crocombe et al. Table 3. Mean change in OHIP-14 severity and dental attendance stratied by residential location Change in OHIP-14 severity Mean )0.38 )2.02 )1.13 )0.06 95% CI )1.77, )3.21, )2.23, )0.86, 1.02 )0.84 )0.03 0.75

Residential location Hobart Outside Hobart

Dental attendance Yes No Yes No

statistically signicant based on the interaction between residential location and attendance (P < 0.05). The interaction terms between income and dental attendance (AU$30,000AU$59,999: P = 0.41, AU$60,000+: P = 0.58, ref: less than AU$30,000), and between difculty in paying AU$100 bill and dental attendance (P = 0.42) were statistically insignicant.

Discussion
The statistically signicant interaction term indicated that the relationship between dental attendance. Dental attendance was associated with a greater improvement in OHRQoL if the survey participant was located outside the State capital city of Hobart. Previous reports have shown benets from regular dental attendance (11) the change in OHIP-14 severity depended on the residential location, and adverse effects from problem-oriented visit patterns (1214). This study goes further

and shows that the effect of dental attendance on OHRQoL is inuenced by residential location. A reason may be the different type of dental care likely to be given in rural compared to metropolitan residents. The oral health status of Australian adults residing outside capital cities is poorer than those of people living in capital cities (15). People residing outside capital cities are more likely to have complete tooth loss, an inadequate dentition (<21 teeth), missing teeth, a higher dental caries experience and untreated coronal dental caries than capital city residents (13). It has also been suggested that rural residents have a different concept of health (16, 17), commonly describing health as an absence of disease, and spend their money on disease management rather than on primary care and health promotion (17). This attitude has been shown to be reected in dental visiting behaviour. People living outside capital areas were more likely than those living in capital cities to have a problem-orientated pattern of dental attendance, less likely to make an annual dental visit and less likely to have a particular dentist that they usually visit (15). These factors suggest that dental care provided to rural residents is more likely for a problem, and once that problem is treated, the patients OHRQoL improves. On the other hand, people living in capital city areas who made a dental visit within the previous 12 months were less likely to have had one or more teeth extracted and more likely to have received a professional clean and polish than people from noncapital city areas (15). OHIP-14 has shown to be susceptible to oor effects. If people do not have

Table 4. Regression analysis for change in mean OHIP-14 severity Model 1 Crude model: visit Parameter Intercept Location (Hobart = 0, Other = 1) Interaction term: location visit Income (AU$30 000AU$59 999, ref: AU$29 999) Income (AU$60 000+, ref: AU$29 999) Diff. in paying AU$100 bill (No = 0, Yes = 1) Visit a dentist (No = 0, Yes = 1) Estimate )0.69 )0.04 SE 0.36 0.53 P value 0.07 0.94 Model 2 Crude model plus rst-order terms Estimate )1.03 )1.06 0.63 1.66 0.60 SE 0.69 0.77 0.71 0.96 0.49 P value 0.15 0.18 0.38 0.10 0.23 Model 3 Model 2 plus interaction terms Estimate )1.97 1.35 )1.93 )0.95 0.55 1.65 1.73 SE 0.98 0.80 0.92 0.76 0.72 0.95 0.66 P value 0.05 0.10 <0.05 0.22 0.45 0.10 0.01

R2 < 0.01, Model P = 0.94

R2 = 0.02, Model P = 0.24

R2 = 0.03, Model P = 0.08

58

Dental attendance and quality of life

any social impacts from their oral health, the resulting treatment cannot improve that situation. Shortcomings of the study should be discussed. The randomized control trial (RCT) is the gold standard of study design. However, an RCT was not feasible for this study for ethical reasons, and inferences had to be made from on an observational epidemiological study. Being a prospective study, it was able to demonstrate temporal sequence and, being based on a population sample, allowed the results to be representative of the Tasmanian population. The 1-year follow-up period, allowed the measurement of longer-term changes in OHRQoL and reduced the inuence of posttreatment cognitive dissonance (18). An extensive array of likely confounders and effect modiers was investigated, including stratied analyses to search for possible heterogeneity of effects among population subgroups. No follow-up clinical examination was undertaken, and so it is possible that new or untreated oral disease is a factor. New dental disease was not expected to have developed over the 1-year observation period of this study to a stage where it would have a deleterious effect on OHRQoL. Untreated oral disease, particularly in the case of people who usually attend a dentist for a problem and not a check-up, could have inuenced change in OHRQoL. However, the inuence of both untreated and new oral disease on change in OHRQoL would have occurred to both those people who did and did not attend a dentist. A further complication is response shift, where people are primed at follow-up to a previously asked question. In this study, response shift would have occurred if asking about, say, a persons social disability at baseline will inuence their response at follow-up. However, the likelihood that the survey participants will be inuenced by one or two questions of an eight-page self-complete questionnaire asked some 12 months earlier and when they had not been contacted in the time of this study was unlikely. It was not possible to assess a clinical as opposed to a statistical meaningful change in OHRQoL because a clinical signicant change in OHIP-14 severity has never been dened. However, using another dependent variable such as the oral health transition statement could suffer from posttreatment cognitive dissonance. The change in OHIP-14 severity was moderate, never going beyond the range of )3 to +3 when it could theoretically have ranged from )56 to +56. There are a few reasons why this occurred:

The benecial effects of dental care in one OHIP14 dimension may be negated by its harmful side effects in another dimension. For example, a dental extraction provided to relieve pain (the OHIP-14 painful aching dimension) may also have an unwanted side effect such as leaving a visible gap (the OHIP-14 self-conscious dimension). Locker et al. (19) suggested that the responsiveness of OHIP-14 is only moderate according to Cohens benchmarks and using the Global Oral Health Transition Statement as the gold standard. It is acknowledged that OHIP-14 severity is subject to residual confounding because of regression to the mean (20). This is where high baseline scores inuence overall change after treatment. For the analyses for change in OHIP14 severity, a more precise adjustment for regression to the mean was made through the use of residuals as the dependent variable. The estimate for change in OHIP-14 severity was )0.35 (P = 0.20), where this decrease was qualied as residual change in OHIP-14 severity. The estimate for change in OHIP-14 severity was both relatively small and statistically insignicant. The result for dental attendance was not remarkably different from that found in Model 2 in Table 4 in the multivariate analysis (0.60, P = 0.23). Unlike this study where there was not a statistically signicant association between dental attendance and OHRQoL, the ve longitudinal studies that have previously investigated the inuence of routine dental care on OHRQoL found an improvement in OHRQoL for those receiving dental care (37). However, a comparison of similar cohorts, where possible, showed that this current study conrmed the results of previous studies. In the Fiske et al. (3) study, one hundred elderly British people requesting dental care were interviewed and treated. In the current study, the closest group to that used in the Fiske study was the group of people with a treatment need who visited a dentist. When this group was investigated, the subjects had a reduction in OHIP-14 severity ()1.44), indicating an improvement in OHRQoL. In the Petersen and Nortov (4) study, 187 Danish pensioners aged 6770 years were given care that included comprehensive curative and preventive care as well as oral health education. After 3 years, the percentage of participants who reported poor

59

Crocombe et al.

function of dentures declined and less of the elderly felt embarrassed by teeth or preferred food that was easy to chew. The closest group in the current study to that in the Petersen and Nortov (4) study was the same as in the Fiske (3) study, with the same results. Lockers study (5) and this current study were not directly comparable because Lockers study had frequency of dental visits and volume of dental services as the independent variables, while this current study used dental attendance. The Fisher et al. (6) study found that dental services were effective in resolving oral disadvantage. The closest group in this current study with the Fisher et al. cohort was the group of subjects with a poor baseline OHRQoL who visited a dentist. It is acknowledged that difculties may arise in comparing this cohort with the Fisher et al. cohort because baseline and change in OHIP-14 severity were correlated (21), a problem termed mathematical coupling (22). However, for the poor baseline OHRQoL cohort in the current study, there was an improvement in mean OHIP-14 severity ()2.43). Gagliardi et al. (7) conducted a prospective single group intervention study of adults aged 75+ years receiving care through the South Australian Dental Service. The survey participants had a clinically determined treatment need, and the closest comparison group in this current study was those people who visited a dentist who had an oral disease. In this cohort, there was an improvement in mean OHIP-14 severity ()1.10). The difference in the effect of dental attendance on OHRQoL for Tasmanians who reside within and outside capital cities is important information for politicians, policy makers and administrators when allocating limited valuable resources to improve population OHRQoL. It suggests that resource allocation to improve access to dental care for people who reside outside the capital cities is more likely to improve population OHRQoL than a similar allocation within capital city areas. Further, understanding non-service factors for those who dont visit and the differential between capital city and non-capital city needs more investigation.

Acknowledgements
Organizations that supported the National Survey of Adult Oral Health 20042006 were the National Health and Medical Research Council (Nos. 299060, 349514, 349537), the Australian Government Department of Health and Aging, Population Health Division, the Australian Institute of Health and Welfare, Colgate Oral Care, the Australian Dental Association and the US Centers for Disease Control and Prevention. Funding sources that supported this study were an Australian Dental Association Research Foundation Grant (27 2006) and an Australian Postgraduate Award.

References
1. Cunningham SJ, Hunt NP. Quality of life and its importance in orthodontics. J Orthod 2001;28:1528. 2. Sheiham A, Croog SH. The psychosocial impact of dental diseases on individuals and communities. J Behav Med 1981;4:25772. 3. Fiske J, Gelbier S, Watson RM. The benet of dental care to an elderly population assessed using a sociodental measure or oral handicap. Br Dent J 1990;168:1536. 4. Petersen PE, Nortov B. The effect of a three-year trial of a community dental care program for aged pensioners in Denmark [in Danish]. Ugeskr Laeger 1995;157:27126. 5. Locker D. Does dental care improve the oral health of older adults? Community Dent Health 2001;18:715. 6. Fisher MA, Gilbert GH, Shelton BJ. Effectiveness of dental services in facilitating recovery from oral disadvantage. Qual Life Res 2005;14:197206. 7. Gagliardi D, Slade GD, Sanders AE. Impact of dental care on patients treatment goals and quality of life. Aust Dent J 2008;53:2633. 8. Slade GD, Spencer AJ, Roberts-Thompson K. Australias dental generations: the National Survey of Adult Oral Health 200406. Canberra: Australian Institute of Health and Welfare; 2007. 9. Slade GD. Derivation and validation of a short-form oral health impact prole. Community Dent Oral Epidemiol 1997;25:28490. 10. ABS. Australian standard classication of occupations (ASCO), 2nd edn. Canberra: Australian Bureau of Statistics; 1997. 11. Thomson WM, Williams SM, Broadbent JM, Poulton R, Locker D. Long-term dental visiting patterns and adult oral health. J Dent Res 2010;89:307. R, Garcia R, Kazis L. 12. Kressin NR, Spiro A III, Bosse Assessing oral health-related quality of life: ndings from the normative aging study. Med Care 1996;34: 41627. 13. Roberts-Thomson K, Do L. Oral health status. In: Slade GD, Spencer AJ, Roberts-Thompson K, editors. Australias dental generations: The national survey of adult oral health 200406. Canberra: Australias Institute of Health and Welfare, 2007; 81142. 14. Spencer AJ, Harford J. Dental care. In: Slade GD, Spencer AJ, Roberts-Thompson K, editors. Australias dental generations: The national survey of adult oral health 200406. Canberra: Australian Institute of Health and Welfare, 2007; 14372.

Conclusions
While on a population basis dental attendance was not signicantly associated with OHRQoL, the effect of dental attendance on OHRQoL was inuenced by a patients residential location.

60

Dental attendance and quality of life 15. Australian Institute of Health and Welfare. Geographic variation in oral health and use of dental services in the Australian population 200406. AIHW Dental Statistics and Research Unit Research Report No. 41. AIHW Cat. no. DEN 188. Adelaide. 16. Humphreys JS, Mathews-Cowey S, Weinand HC. Factors in accessibility of general practice in rural Australia. Med J Aust 1997;166:57780. 17. Coster G, Gribben B. Primary care models for delivering population-based health outcome. Wellington, New Zealand: National Health Committee; 1999. 18. Festinger L. A theory of cognitive dissonance. Stanford, SA: Stanford University Press; 1957. 19. Locker D, Jokovic A, Clarke M. Assessing the responsiveness of measures of oral health-related quality of life. Community Dent Oral Epidemiol 2004;32:108. 20. Slade GD. Assessing change in quality of life using the Oral Health Impact Prole. Community Dent Oral Epidemiol 1998;26:5261. 21. Altman DG. Statistics in medical journals. Stat Med 1982;1:5971. 22. Tu Y-K, Gilthorpe MS. Revisiting the relation between change and initial value: a review and evaluation. Stat Med 2007;26:44357.

61

This document is a scanned copy of a printed document. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material.