SK. SC;. Med. Vol. 35. No. 8. pp. 997-1001. Printed in Great Britain
0277-9536 92 55.00 + 0.00 Pergamon Press Ltd
1992
REGULAR PATTERN OF PREVENTIVE DENTAL SERVICES-A MEASURE OF ACCESS JOHN F. NEWMAN’ and HELEN C. GIFT?*
‘Institute of Health Administration, Georgia State University, 30 Pryor Street, Atlanta, GA 30303, U.S.A. and 2Disease Prevention and Health Promotion Branch, National Institute of Dental Research, Westwood Building, Room 534, 5333 Westbard Avenue, Bethesda, MD 20892, U.S.A. Abstract-Having a regular pattern of care should be an indicator of access to and periodic use of preventive and health maintenance services. The analyses reported in this study are intended to provide a better understanding of the factors related to having a regular pattern of preventive dental care. The data were collected in 1981 as part of a U.S. household survey, ‘A Study of Dental Health Related and Process Outcomes Associated with Prepaid Dental Care’, the most comprehensive cross-sectional data base available in dentistry. Descriptive analyses of a constructed variable, representing perceived and realized access and a preventive orientation, indicate that 53% of the population had a regulur pattern ofpreventivecare.Those with a regular pattern of care were more likely to be white, younger, have dental insurance, have no cost barriers, have more than 12 years of education, be dentate, have no perceived symptoms, and no fear of pain. Logistic regressions indicated that there was an increased probability of having a regular pattern of preventive care if individuals had no economic access problems, had positive attitudes, had higher income, reported few oral symptoms, and were non-Black. Overall, the descriptive models used suggested that individuals with resources in the form of finances and education, and a sense of self-efficacy as expressed in attitudes toward oral health, had the greatest probability of having a regular pattern of preventive care. Key words-dental
utilization, preventive services. regular care
BACKGROUND
that a useable resource is there if needed. A broader concept, having a regular pattern of preventive dental care, should be an indicator of access to and periodic use of preventive and health maintenance services. Also, the concept should reflect any value placed on using personal resources for preventive dental care. The assumption made here is that those who visit a health care professional regularly have a better chance of preventing disease or of receiving early treatment. Lack of regular care may result in increased severity of disease or in extensive treatment and a poor general level of health. It is important to consider differences in attitudes, knowledge, and social factors in relation to regular preventive dental care, and ultimately to health status.
In the literature, access to and utilization of professional health services have been conceptualized
and measured differently: (1) as a function of structural characteristics; (2) availability of resources; (3) an individual’s knowledge of the system; (4) an individual’s ability to use the system; and (5) an individual’s interest in using the system [l-l I]. Having a regular source of care has been measured by different, single-dimension, quantitative indicators [4-51: (I) realized access or actual behavior (interval since last visit, visit during the past 12 months, the proportion of preventive to restorative services), and (2) perception of ‘usual source of cure’, e.g. asking the individual: “Is there a particular doctor/dentist or clinic you usually go to when sick, or for advice about health?” [4-51. Implicit in the concepts, ‘regular care’, or ‘usual source of care’, is the assumption that an individual receives periodic and repetitive services appropriate to the prevention of diseases or maintenance of health [3, 51. Yet, the reported indicators of usual source of care or routine care do not adequately measure the concept. Neither the periodic behavior nor the preventive nature of the visit is reflected in most of the reported measures. For example, ‘visit’ only reflects acting on a propensity or perceived need. Similarly, ‘having a usual source of care’ is only a perception *To whom correspondence should be addressed.
METHODS AND VARIABLES
Methodology
The data for these analyses were collected in 1981 as part of a U.S. household survey, ‘A Study of Dental Health Related and Process Outcomes Associated with Prepaid Dental Care’, sponsored by Health Resources and Services Administration of the U.S. Public Health Service [12, 131. This is the most comprehensive, yet under-analyzed, cross-sectional data base in dentistry. It includes payment and oral health status information. Participants in the survey were selected using a multi-staged household probability sample of 48 states (excluding Alaska and Hawaii) and the District of Columbia. The final sample
997
998
JOHS
F.
NEWMAN and HELEN C. GIFT
percent reported a dental visit “during the past two years”. Approximately 36% of the respondents reported the most recent dental visit in 1981, while one third (32%) had been to the dentist most recently in 1980. Three quarters had a dentist to whom they normally went; and two thirds reported that their most recent visit (not part of a series) was for a checkup or cleaning. Analyses reported elsewhere on these data illustrated the expected statistical associations of predisposing and enabling variables with interval since last dental visit, number of dental visits, and reporting a usual source of care [14. 151. For example, older adults, blacks, individuals with lower socioeconomic status (as measured by education and income), those without dental insurance, and those with poor attitudes toward oral health had less dental utilization than their comparison groups. Figure 1 illustrates the development of the construct to represent regular pattern of precentice care based on three of the measures cited above. ‘Realized’ contact with the dentist was measured by a reported visit within the past two years. This variable is linked directly with other construct variables on the interview, provides a suitable split in the sample for
consisted of 2428 households with 7078 individuals 3 years of age and older. The sample was stratified to create a balance of those with and without dental insurance. Personal interviews were conducted with members of the household and self-administered questionnaires were left behind to be completed. At the interview, efforts were made to schedule the entire family for an oral health screening examination. These were performed by trained and calibrated examination teams in the homes of respondents [l2, 131. Follow-ups with dental practices and dental insurance plans were completed to create additional data sets. Variables Variables available to assess routine or usual care in this data base are (I) “did you visit a dentist (dental specialist, hygienist, dental technician) during the past 2 years?“; (2) “when was the last time you received any type of dental care-month and year?“; (3) “do you have a regular dentist, that is, a dentist you normally go to for routine dental care?“; (4) purpose of most recent visit (not part of a series): checkup/cleaning; something was hurting or bothering you; wanted some kind of work done. Sixty-nine
Dental
past
visit
during
two years
Regular regular
Last visit
dentist/ location
I
part
of a series
I No
No
No (n=3870)
(n=l904)
(n=2337)
‘es (n=2085)
Last visit for checkup
Yes
Yes
Yes
(n=4763)
(n=4370)
(n=2721)
I
Regular
pattern
preventive Yes
of
care
: 53%
(n = 1910) No
: 47%
(n = 1740)
Fig. 1. Construction
of regular
pattern
of preventive
care
Regularpattern of preventive dental services
999
predismosinq Bqp: 3 or older RacelEthnicitv: white, black, other, Hispanic
Gender,
male,
female
perceived
svmwtoms: summary scale (O-6), presence of one or more symptoms (broken or chipped tooth that hurts when touched; gums that sometimes hurt, feel sore to touch or bleed for no apparent reason; any teeth that are sensitive to hot and cold liquids or foods; sores that sometimes develop on tongue, or on the inside of mouth or cheeks; any teeth that ache or throb sometimes for no apparent reason; and any teeth that hurt when eating or drinking very sweet things. (high value--more symptoms)
Attitude toward oral health/car e: summary scale (7-42) based on a series of attitude statements (5 point agree/disagree), scaled so a high value was a more positive attitude; e.g., some people are just born with good teeth and others are not--there's not much anyone can do about it. Enablinq Years of Education
(head of household)
I980 Income (0 and up) Dental Insurance Coveraae
(no, yes)
Economic Access: 4 point Guttman scale representing the extent to which cost of dental care is a barrier to use. (low value--no problem) Oral Hvaiene Behavior Behavior: Summary scale (O-10) representing the extent to which the individual reports oral hygiene behaviors including brushing, flossing, use of fluoride dentifrice, avoiding sweet snacks, use of wooden sticks. (higher value-- more oral hygiene behaviors) Measures
from Clinical File (examination)
# Permanent Teeth (O-32) / Tooth Surfaces Needino Restoration
(O-148)
Fig. 2. Variables in study.
analysis, and is a reasonable estimate of recent contact with a dentist given the limitations of self-reported data. While historically it has been recommended that individuals visit a dentist l-2 times per year to prevent oral diseases, a visit in a two year interval, reported on an interview, should be sufficient to reflect a preventive and maintenance approach to oral health. ‘Dental visit’ was combined with reporting a regular source of care (perceived) and a preventive reason for the most recent visit (which was not
part of a series) to create the variable, regular pattern of preventive care. Fifty-three percent had a regular pattern of preventive care (had a visit in the past 2
years, reported a usual source of care, and visited the dentist most recently for a checkup or cleaning) (Fig. 1). The remainder of the respondents reported negatively on one or more of the three variables. For the purpose of analyses, comparisons were made between those who had a regular pattern of preventive care and those who did not (all others). Other variables were selected for analyses if they had been factors of significance in other studies of dental services utilization [l-l 1, 14-151. Emphasis *Each data file has a different base, depending on the overall was placed on variables representing the behavioral response rate for each survey component. As a conmodel of utilization (61. Variables were grouped to structed variable, the ‘n’ for the dependent variable is represent predisposing and enabling factors, oral determined by the responses to each component variable, which in turn affects the analytical sample size hygiene behaviors, and actual oral health status (see (3669) for these analyses. Bases for specific cross-tabulaFig. 2). Variables used in these analyses included tions and multivariate analyses are further reduced direct reports from the individual, calculations or depending on internal response rates. The data were constructs based on a series of questions, and analyzed both weighted and unweighted, with similar results for both analyses. measures abstracted from the clinical file.*
JOHN
1000
F. NEWMMAN and HELEN C. GIFT
RESULTS
Bivariate and multivariate analyses were conducted to determine the descriptive characteristics of individuals with a regular pattern of pretentice care *. In bivariate analyses, individuals with a regular pattern ofpreventive care were more likely than those without to be white, have dental insurance, have high income, have excellent attitudes and oral hygiene behaviors, have no perceived symptoms, and have no fear of pain. Those without a regular pattern of preventive care were more likely to have less than 12 years of education, to have great need for restorative treatment or be edentulous, to be older, and to have economic access barriers. Theoretically important and statistically significant variables were selected for the multivariate analyses from those illustrated in Fig. 2. The purpose of the multivariate analyses was to determine the total effect of a set of predisposing, enabling, behavioral and need variables on regular pattern ofprecentice care. Groups of variables, representing predisposing and enabling factors, were run separately and sequentially. Subsequently, these subsets of variables were combined and analyzed in a comprehensive model. Logistic stepwise regressions were used since the assumptions for a dichotomous outcome variable were more appropriate. The logistic regression analyses with only predisposing variables included demographic as well as attitude variables. Having a positive attitude toward oral health, fewer perceived symptoms, and being non-black were the only significant variables for predicting regular pattern of preventive care. In the analysis with only enabling variables, having fewer economic access problems, higher family income, and more education were the only significant variables. When predisposing, enabling and oral hygiene behavior variables were combined in a logistic regression model, having better economic access, more positive attitude, fewer symptoms, and being nonBlack were the most significant variables. Family income, age, oral hygiene behavior, and education were also significant. A final full model logistic analysis was run using predisposing, enabling, and oral hygiene behavior variables, as well as number of permanent teeth and restorative needs. Logistic regression indicated that there was an increased probability of having a regular pattern ofpreventive care if individuals had no economic access problems, had higher income, were nonBlack, had positive attitudes, reported few oral symptoms, had more teeth, and had fewer teeth needing restorations. The likelihood ratio test was significant (P ,< 0.001) for each of the four models. The likelihood ratio test comparing the final model with the third model
*Tables forming the basis of the reported available from the author.
analyses are
indicated a significant improvement in z’ as a result of the addition of the two oral health status variables (P < 0.001). Odds ratios and 95% confidence intervals were calculated for significant variables. These represent the odds of having a regular pattern of preventice care with a unit change in an independent variable. By way of illustration, in the enabling model, individuals with few (1) economic access problems were 0.64 as likely to have a regular pattern ofprecentire care as those with no (0) economic access problems. Those with one child were 1.1 times more likely to have a regular pattern of preventive care than those with no children. Looking at the total model, those with the most economic access problems (4), were 0.69 as likely to have a regular pattern of preventive care as those with some (3) economic access problems. As the number of teeth increased by one, individuals were 1.06 times more likely to have a regular pattern of preventive dental care. As the number of teeth increased by IO, individuals were 1.88 times more likely to have a regular pattern of preventive dental care. Similarly, an increase in 10 points on the attitude scale increased the odds of having a regular pattern of preventive care by 1.60. SUMMARY
AND CONCLUSIONS
Analyses suggest that regular pattern of preventive care, a constructed variable representing perceived and realized access to care and a preventive orientation, is a valuable measure. In bivariate analyses, the construct was correlated with the same predisposing, enabling, and oral status variables as more traditional measures, such as usual source of care or dental visits. In multivariate analyses, the results suggested that regular pattern of precentice care is a sensitive measure of access to care. Overall, the results suggested that individuals with resources in the form of finances and education, and a sense of self-efficacy, as expressed in attitudes toward oral health, had the greatest probability of having a regular pattern of preventice care. The relations among behavioral and social variables and the construct indicated that more extensive development of a model would be appropriate. Also, analyses of population subgroups might improve understanding of regular pattern of preventive care. Considerably more research is needed on those groups of individuals with unequal access to oral health care. Overall, oral health status in the U.S. is improving, yet minorities and lower socioeconomic status groups often suffer unnecessarily from oral diseases which are basically preventable. Given that these same groups suffer from many other social and medical burdens, relief from unnecessary oral diseases would be a considerable benefit. Yet, improved understanding of the groups which appear to have these social inequities is not as easy as it appears. Large scale social surveys, such as the one which was
Regular
pattern of preventive dental services
the basis of the analyses here, are few and costly. Often the data from such surveys are limited in scope and depth. Beyond this, there are often insufficient numbers in the sample of the very individuals about whom we need to gain more information. This affects interpretation of any analyses of multiple social indicators and oral health variables. Large national surveys, such as the one used in these analyses, designed to report national statistics, may not be the solution for providing the detailed understanding of groups with multiple social inequities. Smaller scale, targeted local-area or community-based research may be the best solution for understanding social inequities in oral health. REFERENCES
1. Aday L. A., Fleming G. V. and Andersen R. M. Access to Medical Care in the U.S.: Equity for Whom? Sage, Beverly Hills, CA, 1980. R. A Behavioral Model of Families’ Use of 2. Andersen Health Services, Research Series No. 25. Center for Health Administration Studies, University of Chicago, 1968. 3. Andersen R. M., Giachello A. L. and Aday L. A. Access of Hispanics to health care and cuts in services: a state-of-the-art overview. P&1. Hlth Reports 101, 238-252, 1986. 4. Andersen R., Lion J. and Anderson 0. W. Two Decades of Health Services. Ballinger, Cambridge, MA, 1976. Aday L. A., Chiu G. Y. 5. Andersen R. M., McCutcheon and Bell R. Exploring dimensions of access to medical care. Hlth Services Res. 18, 49-74, 1983. R. M. and Newman J. F. Societal and 6. Andersen individual determinants of medical care utilization in
the United States. The Milbank Memorial Fund Q. 51, 1973. Evashwick C., Rowe G., Diehr P. and Branch L. Factors explaining the use of health care services by the elderly. Hlth Services Res. 19, 358-382, 1984. Gift H. C. Utilization of professional dental services. In Social Sciences and Dentistry (Edited by Cohen L. K. and Bryant P. S.). pp. 202-226. Quintessence, London, 1984. Kasper J. A. and Barrish G. Usual sources of medical care and their characteristics. Data Preview 12. National Medical Care E.rpenditures Study, DHHS Publ. No (PHS) 82-3324. U.S. Gov. Printing Office, Washington, DC, 1982. Newman J. F. and Anderson 0. W. Patterns ofDental Service Utilization in the United States: A Nationwide Social Survey, Research Series No. 30. Center for Health Administration Studies, University of Chicago, 1972. Ware J. E., Davies A. R., Allen H. M., Manning W. G. and Holty S. Explaining Dental Utilization Behavior. Final Report of the Rand Corporation to DHHS PHS OASH NCHSR & HCFA, 1987. Bonito A., Iannachione V., Jones A. and Stuart C. A Study of Dental Health Related Process Outcomes Associated with Prepaid Dental Care. Final Report, Part I, DHEW Contract No. HRA 236-76 0093. Research Triangle Institute, Research Triangle Park, NC, 1984. Brown L. J., Oliver R. C. and L6e H. Periodontal Diseases in the U.S. in 1981. J. Periodontol. 60,363-370, 1989. Kleinman D. V. and Gift H. C. Dental health attitudes and behaviors of the United States public. Paper presented at the Annual Meeting of the American Public Health Association, 1987. Gift H. C. Measuring and understanding social inequities in oral health. Centerfor Health Adminisfration Studies Syposium, University of Chicago, 1990.
,, 95-124,
8.
9,
10.
II.
12.
13.
14.
IS.
1001
0277-9536 92 55.00 + 0.00 Pergamon Press Ltd
1992
REGULAR PATTERN OF PREVENTIVE DENTAL SERVICES-A MEASURE OF ACCESS JOHN F. NEWMAN’ and HELEN C. GIFT?*
‘Institute of Health Administration, Georgia State University, 30 Pryor Street, Atlanta, GA 30303, U.S.A. and 2Disease Prevention and Health Promotion Branch, National Institute of Dental Research, Westwood Building, Room 534, 5333 Westbard Avenue, Bethesda, MD 20892, U.S.A. Abstract-Having a regular pattern of care should be an indicator of access to and periodic use of preventive and health maintenance services. The analyses reported in this study are intended to provide a better understanding of the factors related to having a regular pattern of preventive dental care. The data were collected in 1981 as part of a U.S. household survey, ‘A Study of Dental Health Related and Process Outcomes Associated with Prepaid Dental Care’, the most comprehensive cross-sectional data base available in dentistry. Descriptive analyses of a constructed variable, representing perceived and realized access and a preventive orientation, indicate that 53% of the population had a regulur pattern ofpreventivecare.Those with a regular pattern of care were more likely to be white, younger, have dental insurance, have no cost barriers, have more than 12 years of education, be dentate, have no perceived symptoms, and no fear of pain. Logistic regressions indicated that there was an increased probability of having a regular pattern of preventive care if individuals had no economic access problems, had positive attitudes, had higher income, reported few oral symptoms, and were non-Black. Overall, the descriptive models used suggested that individuals with resources in the form of finances and education, and a sense of self-efficacy as expressed in attitudes toward oral health, had the greatest probability of having a regular pattern of preventive care. Key words-dental
utilization, preventive services. regular care
BACKGROUND
that a useable resource is there if needed. A broader concept, having a regular pattern of preventive dental care, should be an indicator of access to and periodic use of preventive and health maintenance services. Also, the concept should reflect any value placed on using personal resources for preventive dental care. The assumption made here is that those who visit a health care professional regularly have a better chance of preventing disease or of receiving early treatment. Lack of regular care may result in increased severity of disease or in extensive treatment and a poor general level of health. It is important to consider differences in attitudes, knowledge, and social factors in relation to regular preventive dental care, and ultimately to health status.
In the literature, access to and utilization of professional health services have been conceptualized
and measured differently: (1) as a function of structural characteristics; (2) availability of resources; (3) an individual’s knowledge of the system; (4) an individual’s ability to use the system; and (5) an individual’s interest in using the system [l-l I]. Having a regular source of care has been measured by different, single-dimension, quantitative indicators [4-51: (I) realized access or actual behavior (interval since last visit, visit during the past 12 months, the proportion of preventive to restorative services), and (2) perception of ‘usual source of cure’, e.g. asking the individual: “Is there a particular doctor/dentist or clinic you usually go to when sick, or for advice about health?” [4-51. Implicit in the concepts, ‘regular care’, or ‘usual source of care’, is the assumption that an individual receives periodic and repetitive services appropriate to the prevention of diseases or maintenance of health [3, 51. Yet, the reported indicators of usual source of care or routine care do not adequately measure the concept. Neither the periodic behavior nor the preventive nature of the visit is reflected in most of the reported measures. For example, ‘visit’ only reflects acting on a propensity or perceived need. Similarly, ‘having a usual source of care’ is only a perception *To whom correspondence should be addressed.
METHODS AND VARIABLES
Methodology
The data for these analyses were collected in 1981 as part of a U.S. household survey, ‘A Study of Dental Health Related and Process Outcomes Associated with Prepaid Dental Care’, sponsored by Health Resources and Services Administration of the U.S. Public Health Service [12, 131. This is the most comprehensive, yet under-analyzed, cross-sectional data base in dentistry. It includes payment and oral health status information. Participants in the survey were selected using a multi-staged household probability sample of 48 states (excluding Alaska and Hawaii) and the District of Columbia. The final sample
997
998
JOHS
F.
NEWMAN and HELEN C. GIFT
percent reported a dental visit “during the past two years”. Approximately 36% of the respondents reported the most recent dental visit in 1981, while one third (32%) had been to the dentist most recently in 1980. Three quarters had a dentist to whom they normally went; and two thirds reported that their most recent visit (not part of a series) was for a checkup or cleaning. Analyses reported elsewhere on these data illustrated the expected statistical associations of predisposing and enabling variables with interval since last dental visit, number of dental visits, and reporting a usual source of care [14. 151. For example, older adults, blacks, individuals with lower socioeconomic status (as measured by education and income), those without dental insurance, and those with poor attitudes toward oral health had less dental utilization than their comparison groups. Figure 1 illustrates the development of the construct to represent regular pattern of precentice care based on three of the measures cited above. ‘Realized’ contact with the dentist was measured by a reported visit within the past two years. This variable is linked directly with other construct variables on the interview, provides a suitable split in the sample for
consisted of 2428 households with 7078 individuals 3 years of age and older. The sample was stratified to create a balance of those with and without dental insurance. Personal interviews were conducted with members of the household and self-administered questionnaires were left behind to be completed. At the interview, efforts were made to schedule the entire family for an oral health screening examination. These were performed by trained and calibrated examination teams in the homes of respondents [l2, 131. Follow-ups with dental practices and dental insurance plans were completed to create additional data sets. Variables Variables available to assess routine or usual care in this data base are (I) “did you visit a dentist (dental specialist, hygienist, dental technician) during the past 2 years?“; (2) “when was the last time you received any type of dental care-month and year?“; (3) “do you have a regular dentist, that is, a dentist you normally go to for routine dental care?“; (4) purpose of most recent visit (not part of a series): checkup/cleaning; something was hurting or bothering you; wanted some kind of work done. Sixty-nine
Dental
past
visit
during
two years
Regular regular
Last visit
dentist/ location
I
part
of a series
I No
No
No (n=3870)
(n=l904)
(n=2337)
‘es (n=2085)
Last visit for checkup
Yes
Yes
Yes
(n=4763)
(n=4370)
(n=2721)
I
Regular
pattern
preventive Yes
of
care
: 53%
(n = 1910) No
: 47%
(n = 1740)
Fig. 1. Construction
of regular
pattern
of preventive
care
Regularpattern of preventive dental services
999
predismosinq Bqp: 3 or older RacelEthnicitv: white, black, other, Hispanic
Gender,
male,
female
perceived
svmwtoms: summary scale (O-6), presence of one or more symptoms (broken or chipped tooth that hurts when touched; gums that sometimes hurt, feel sore to touch or bleed for no apparent reason; any teeth that are sensitive to hot and cold liquids or foods; sores that sometimes develop on tongue, or on the inside of mouth or cheeks; any teeth that ache or throb sometimes for no apparent reason; and any teeth that hurt when eating or drinking very sweet things. (high value--more symptoms)
Attitude toward oral health/car e: summary scale (7-42) based on a series of attitude statements (5 point agree/disagree), scaled so a high value was a more positive attitude; e.g., some people are just born with good teeth and others are not--there's not much anyone can do about it. Enablinq Years of Education
(head of household)
I980 Income (0 and up) Dental Insurance Coveraae
(no, yes)
Economic Access: 4 point Guttman scale representing the extent to which cost of dental care is a barrier to use. (low value--no problem) Oral Hvaiene Behavior Behavior: Summary scale (O-10) representing the extent to which the individual reports oral hygiene behaviors including brushing, flossing, use of fluoride dentifrice, avoiding sweet snacks, use of wooden sticks. (higher value-- more oral hygiene behaviors) Measures
from Clinical File (examination)
# Permanent Teeth (O-32) / Tooth Surfaces Needino Restoration
(O-148)
Fig. 2. Variables in study.
analysis, and is a reasonable estimate of recent contact with a dentist given the limitations of self-reported data. While historically it has been recommended that individuals visit a dentist l-2 times per year to prevent oral diseases, a visit in a two year interval, reported on an interview, should be sufficient to reflect a preventive and maintenance approach to oral health. ‘Dental visit’ was combined with reporting a regular source of care (perceived) and a preventive reason for the most recent visit (which was not
part of a series) to create the variable, regular pattern of preventive care. Fifty-three percent had a regular pattern of preventive care (had a visit in the past 2
years, reported a usual source of care, and visited the dentist most recently for a checkup or cleaning) (Fig. 1). The remainder of the respondents reported negatively on one or more of the three variables. For the purpose of analyses, comparisons were made between those who had a regular pattern of preventive care and those who did not (all others). Other variables were selected for analyses if they had been factors of significance in other studies of dental services utilization [l-l 1, 14-151. Emphasis *Each data file has a different base, depending on the overall was placed on variables representing the behavioral response rate for each survey component. As a conmodel of utilization (61. Variables were grouped to structed variable, the ‘n’ for the dependent variable is represent predisposing and enabling factors, oral determined by the responses to each component variable, which in turn affects the analytical sample size hygiene behaviors, and actual oral health status (see (3669) for these analyses. Bases for specific cross-tabulaFig. 2). Variables used in these analyses included tions and multivariate analyses are further reduced direct reports from the individual, calculations or depending on internal response rates. The data were constructs based on a series of questions, and analyzed both weighted and unweighted, with similar results for both analyses. measures abstracted from the clinical file.*
JOHN
1000
F. NEWMMAN and HELEN C. GIFT
RESULTS
Bivariate and multivariate analyses were conducted to determine the descriptive characteristics of individuals with a regular pattern of pretentice care *. In bivariate analyses, individuals with a regular pattern ofpreventive care were more likely than those without to be white, have dental insurance, have high income, have excellent attitudes and oral hygiene behaviors, have no perceived symptoms, and have no fear of pain. Those without a regular pattern of preventive care were more likely to have less than 12 years of education, to have great need for restorative treatment or be edentulous, to be older, and to have economic access barriers. Theoretically important and statistically significant variables were selected for the multivariate analyses from those illustrated in Fig. 2. The purpose of the multivariate analyses was to determine the total effect of a set of predisposing, enabling, behavioral and need variables on regular pattern ofprecentice care. Groups of variables, representing predisposing and enabling factors, were run separately and sequentially. Subsequently, these subsets of variables were combined and analyzed in a comprehensive model. Logistic stepwise regressions were used since the assumptions for a dichotomous outcome variable were more appropriate. The logistic regression analyses with only predisposing variables included demographic as well as attitude variables. Having a positive attitude toward oral health, fewer perceived symptoms, and being non-black were the only significant variables for predicting regular pattern of preventive care. In the analysis with only enabling variables, having fewer economic access problems, higher family income, and more education were the only significant variables. When predisposing, enabling and oral hygiene behavior variables were combined in a logistic regression model, having better economic access, more positive attitude, fewer symptoms, and being nonBlack were the most significant variables. Family income, age, oral hygiene behavior, and education were also significant. A final full model logistic analysis was run using predisposing, enabling, and oral hygiene behavior variables, as well as number of permanent teeth and restorative needs. Logistic regression indicated that there was an increased probability of having a regular pattern ofpreventive care if individuals had no economic access problems, had higher income, were nonBlack, had positive attitudes, reported few oral symptoms, had more teeth, and had fewer teeth needing restorations. The likelihood ratio test was significant (P ,< 0.001) for each of the four models. The likelihood ratio test comparing the final model with the third model
*Tables forming the basis of the reported available from the author.
analyses are
indicated a significant improvement in z’ as a result of the addition of the two oral health status variables (P < 0.001). Odds ratios and 95% confidence intervals were calculated for significant variables. These represent the odds of having a regular pattern of preventice care with a unit change in an independent variable. By way of illustration, in the enabling model, individuals with few (1) economic access problems were 0.64 as likely to have a regular pattern ofprecentire care as those with no (0) economic access problems. Those with one child were 1.1 times more likely to have a regular pattern of preventive care than those with no children. Looking at the total model, those with the most economic access problems (4), were 0.69 as likely to have a regular pattern of preventive care as those with some (3) economic access problems. As the number of teeth increased by one, individuals were 1.06 times more likely to have a regular pattern of preventive dental care. As the number of teeth increased by IO, individuals were 1.88 times more likely to have a regular pattern of preventive dental care. Similarly, an increase in 10 points on the attitude scale increased the odds of having a regular pattern of preventive care by 1.60. SUMMARY
AND CONCLUSIONS
Analyses suggest that regular pattern of preventive care, a constructed variable representing perceived and realized access to care and a preventive orientation, is a valuable measure. In bivariate analyses, the construct was correlated with the same predisposing, enabling, and oral status variables as more traditional measures, such as usual source of care or dental visits. In multivariate analyses, the results suggested that regular pattern of precentice care is a sensitive measure of access to care. Overall, the results suggested that individuals with resources in the form of finances and education, and a sense of self-efficacy, as expressed in attitudes toward oral health, had the greatest probability of having a regular pattern of preventice care. The relations among behavioral and social variables and the construct indicated that more extensive development of a model would be appropriate. Also, analyses of population subgroups might improve understanding of regular pattern of preventive care. Considerably more research is needed on those groups of individuals with unequal access to oral health care. Overall, oral health status in the U.S. is improving, yet minorities and lower socioeconomic status groups often suffer unnecessarily from oral diseases which are basically preventable. Given that these same groups suffer from many other social and medical burdens, relief from unnecessary oral diseases would be a considerable benefit. Yet, improved understanding of the groups which appear to have these social inequities is not as easy as it appears. Large scale social surveys, such as the one which was
Regular
pattern of preventive dental services
the basis of the analyses here, are few and costly. Often the data from such surveys are limited in scope and depth. Beyond this, there are often insufficient numbers in the sample of the very individuals about whom we need to gain more information. This affects interpretation of any analyses of multiple social indicators and oral health variables. Large national surveys, such as the one used in these analyses, designed to report national statistics, may not be the solution for providing the detailed understanding of groups with multiple social inequities. Smaller scale, targeted local-area or community-based research may be the best solution for understanding social inequities in oral health. REFERENCES
1. Aday L. A., Fleming G. V. and Andersen R. M. Access to Medical Care in the U.S.: Equity for Whom? Sage, Beverly Hills, CA, 1980. R. A Behavioral Model of Families’ Use of 2. Andersen Health Services, Research Series No. 25. Center for Health Administration Studies, University of Chicago, 1968. 3. Andersen R. M., Giachello A. L. and Aday L. A. Access of Hispanics to health care and cuts in services: a state-of-the-art overview. P&1. Hlth Reports 101, 238-252, 1986. 4. Andersen R., Lion J. and Anderson 0. W. Two Decades of Health Services. Ballinger, Cambridge, MA, 1976. Aday L. A., Chiu G. Y. 5. Andersen R. M., McCutcheon and Bell R. Exploring dimensions of access to medical care. Hlth Services Res. 18, 49-74, 1983. R. M. and Newman J. F. Societal and 6. Andersen individual determinants of medical care utilization in
the United States. The Milbank Memorial Fund Q. 51, 1973. Evashwick C., Rowe G., Diehr P. and Branch L. Factors explaining the use of health care services by the elderly. Hlth Services Res. 19, 358-382, 1984. Gift H. C. Utilization of professional dental services. In Social Sciences and Dentistry (Edited by Cohen L. K. and Bryant P. S.). pp. 202-226. Quintessence, London, 1984. Kasper J. A. and Barrish G. Usual sources of medical care and their characteristics. Data Preview 12. National Medical Care E.rpenditures Study, DHHS Publ. No (PHS) 82-3324. U.S. Gov. Printing Office, Washington, DC, 1982. Newman J. F. and Anderson 0. W. Patterns ofDental Service Utilization in the United States: A Nationwide Social Survey, Research Series No. 30. Center for Health Administration Studies, University of Chicago, 1972. Ware J. E., Davies A. R., Allen H. M., Manning W. G. and Holty S. Explaining Dental Utilization Behavior. Final Report of the Rand Corporation to DHHS PHS OASH NCHSR & HCFA, 1987. Bonito A., Iannachione V., Jones A. and Stuart C. A Study of Dental Health Related Process Outcomes Associated with Prepaid Dental Care. Final Report, Part I, DHEW Contract No. HRA 236-76 0093. Research Triangle Institute, Research Triangle Park, NC, 1984. Brown L. J., Oliver R. C. and L6e H. Periodontal Diseases in the U.S. in 1981. J. Periodontol. 60,363-370, 1989. Kleinman D. V. and Gift H. C. Dental health attitudes and behaviors of the United States public. Paper presented at the Annual Meeting of the American Public Health Association, 1987. Gift H. C. Measuring and understanding social inequities in oral health. Centerfor Health Adminisfration Studies Syposium, University of Chicago, 1990.
,, 95-124,
8.
9,
10.
II.
12.
13.
14.
IS.
1001
