210
Journal of Public Health Dentistry
A Profile of Black and Hispanic Subgroups’ Access to Dental Care: Findings from the National Health Interview Survey Lu Ann Aday, PhD Ronald N. Forthofer,PhD School of Public Health The University of Texas Health Science Center PO Box 20 186 Houston, TX 77225
Abstract Thisstudy examined which black and Hispanic minority subgroups were least likely to obtain dental care and why, based on logistic regression analyses of the 1986 National Health Interview Survey. Blacks and Hispanics were less apt to have private dental insurance coverage, to be knowledgeable about the purpose of fluoride, to have been to a dentist in the past year, and, when they did go, were more apt to have gone in response to symptoms rather than for preventive reasons, compared to whites. Logistic regression anaiyses foradults 18years of age and older and for children and adolescents 2 to 17 years of age showed that the following individuals had the lowest probability of having been to a dentist in the past year: males, members of larger families, adults who were unemployed or in blue-collar jobs, those who lived in the South or nonmetropolitan areas, people who perceived their health to be fair or poor, and those with no private dental insurance. Mexican-Americans were least likely to have been to a dentist regardless of their income of education. In general, the findings confirmed the importance of dental insurance, as well as suggesting a need for more school-baseddental programs andpublic health clinic-baseddental health education and outreach efforts for targeting minority children and adults. Key Words:dental care, dental insurance, ethnic groups.
Black and Hispanic minorities have poorer oral health than whites. Data from the National Center for Health Statistics(NCHS)Health and Nutrition Examination Survey conducted during 1970-74 documented that 77.9 percent of blacks, compared to 62.3 percent of whites, had unmet dental care needs, including gingivitis, periodontal disease, tooth decay, and fixed bridge and/or partial denture repair (1).The 1985436 National Survey of Oral This research was supported by the National Center for Health Statis tics, Centers for Disease Control, Public Health Service, US Department of Health and Human Services, Hyattsville, MD. Send correspondence and reprint requests to Dr.Aday. Manuscript received: 9/6/90; returned toauthors for revision: 1/1/90; accepted for publication: 9/3 /91.
J Public Health Dent 1992;52(4):210-15
Health in US Employed Adults and Seniors, conducted by theNational Institute of Health, similarly showed that the percentage of persons with decayed teeth was 22.1 percent in blacks compared to 6.8 percent in whites (1). Data from the 1982-84 National Center for Health Statistics Hispanic Health and Nutrition Examination Survey (HHANES)documented that over half of the 17-year-old Mexican-Americans had five or more filled or decayed teeth. Dental caries in the 5- to 17-yeardd MexicanAmericans in this survey was predominantly a disease of the occlusal surfaces of molars, a condition preventable with the proper use of fissure sealants. Mild gingivitis was also prevalent in 77 percent of the Mexican-American children in the HHANES (2-4). Other studies of Hispanic immigrant children-for example, in San Francisco (5) and children of migrant farm workers in Colorado (6I-also show large proportions of children with many preventable and treatable oral health problems. Blacks and Hispanics also have much lower rates of utilization of dental health care services than do whites (7-15). Based on the 1986 National Center for Health Statistics-National Health Interview Survey (NCHSNHIS), age-adjusted proportions of racial /ethnic groups that had been to a dentist in the past year were 59.4 percent for whites, 42.1 percent for blacks, 38.6 percent for Mexican-Americans, and 53.0 for other Hispanics. The age-adjusted percent of Mexican-Americans who had never been to a dentist (12.6 percent) was high compared to whites (4.6 percent) and other minority groups, such as blacks (7.1 percent) and other Hispanics (5.9 percent). Whites who had been to a dentist also averaged more visits per person per year (2.1) compared to the ethnic minorities-blacks (1.4), Mexican-Americans (1.31, and other Hispanics (1.8). The proportion with private dental insurance coverage was also higher for whites (39.3 percent) compared to blacks (28.4 percent), Mexican-Americans (28.7 percent) or other Hispanics (30.1 percent) (15). Since many of the dental problems experienced by minorities are preventable or treatable, it is important to find ways of enhancing minorities‘ access to dental care. In addition, it is important to examine subgroups of
Vol. 52, No. 4,Summer 1992
Hispanicsseparately since multivanate analyses examining Hispanics’ utilization of health services have found that Mexican-Americans, besides tending to have lower rates of utilization than whites and blacks, also tend to have lower utilization than do other groups of Hispanics (9,16-18). A number of studies have examined possible predictors of dental care use (7,8,12-16,19-23). Grembowski et al. (24), in their public health model of the dental care process, suggest that a complex array of structural, historical, and cognitive factors predict whether or not dental care is sought. There is inadequate evidence, however, on the relative importance of these factors in understanding the lower utilization rates of poor and minority popula tions. The analyses reported here are based on the 1986 NCHS National Health Interview Survey and associated dental supplement, which provides a reasonable number of observations to address these issues for blacks and Hispanic subgroups. The factors most important in predicting which racial/ethnic subgroups are least likely to obtain dental care are identified. The results are intended to facilitate the design of interventions for enhancing minorities’ access to dental care. Methods The 1986 National Health Interview Survey is a nationally representative area probability sample of the noninstitutionalized US population. More detail about the sample design is available in the descriptive reports based on that survey (25). The survey is based on household interviews of all adults 17 years of age or older in person; information on all children was obtained from a knowledgeable adult in the household. Proxy responses were also allowed for adult household members who were unavailable at the time of the interview. There were 62,052 persons for whom information was obtained in the survey. However, the analyses here were limited to whites, blacks, and Hispanics two years of age or older; these exclusions reduced the sample to 58,435 persons. The distribution in the respective racial/ethnic subgroups is: whites (44,573), blacks (9,438), HispanicMexican-Americans (2,3301, and Hispanic-Other (2,094). Selected opinion and knowledge questions regarding the use of fluoride were asked only of the adult who was the main household informant (total n=14,180). Unlike many other surveys that have low response rates, and hence are subject to possible selection bias, a strength of this study is that the response rate for this nationally representative survey was 96 percent. The framework used was the Behavioral Model of Health Services Utilization developed by Andersen and his colleagues (7,8). In this model, a series of predisposing, enabling, and need factors are hypothesized to be predictive of individuals’ utilization of health services. Predisposing factors reflect the propensity of the individ-
211
ual to seek care. Those used in this study were race, age, sex, family size, education, employment/occupa tion, region, and place of residence. Enabling variables capture the resources available to the individual as well as those available in the community. The enabling variables used here refer to the individual and included poverty level and private dental insurance. Need refers either to individual or provider judgments of the presence and severity of conditions requiring treatment. A question concerning whether respondents perceived their health to be excellent, good, fair, or poor was used to adjust for their overall health status. A focus of the analysis was on understanding access for different subgroups of minorities-reflected in the interactionsof race with other characteristics (age,education, and poverty level). Other variables based on the Behavioral Model that would have been useful to include in the analyses but were not available in the 1986 NCHS-NHIS include knowledge and attitudes toward dental care, Medicaid coverage, the availability of dentists in the areas in which respondents resided, and direct indicators of the need for dental care. Logistic regression was chosen as the procedure for examining the linkage between the dichotomous dependent variable, whether or not the person had been to a dentist in the past year, and the predisposing, enabling, and need predictor variables. The procedure controlled for racial/ethnic group membership and its interactions with age, education, and poverty level (26). Models were developed separately for those aged 2 to 17 and for adults 18 years of age or older. The education, employment status, and occupation of the head of the family (the adult family member in whose name the home was owned or rented) were attached to the child’s record, and used as predictors of the child’s utilization. In the analyses of these data, collected using a complex sampling design, it was necessary to incorporate the following: (1 ) sample weights to adjust for different probabilities of selection, and (2) design effects, reflecting the extent to which the variances of estimates obtained from stratified and cluster sample-based designs differedfrom those of a simple random sample. The analysis was performed with PROC LOGISTIC in SAS, which allowed the sample weights to be incorporated. The SUPERCARPprogram along with PROC GLM in the SAS package were used to obtain estimates of the design effects for the predictor variables (27,281. In assessing the statistical significance of the terms in the l o p t i c regression models, a design effect of 2.0 was used-meaning that the variances were estimated to be twice that of a simple random sample. This results in a conservative procedure for identifying the statistically significantpredictors.
Results Table 1 provides specific information on Americans’ knowledge related to fluoride and reasons for seeing a
Journal of Public Health Dentistry
212
TABLE 1 Selected Characteristics of Whites, Blacks, and Hispanics, 2+ Years of Age, Regarding Dental Health and Health CareUnited States, 1986 Racial/Ethnic Groups (%) Hispanic Characteristics
White
Black
Mexican
Other
Adult household respondent's knowledge about purpose of fluoride in public water systems* 70 38 38 45 Prevent tooth decay 12 24 10 15 Some other reason 18 37 51 40 Not known Toothpaste with fluoride 97 99 98 Yes 97 Toothpaste for plaque 38 48 39 Yes 52 Mouthrinse with fluoride 11 7 9 Yes 10 Main reason for last visit in previous two years 41 44 51 56 Exam or cleaning self-initiated 24 12 7 9 Exam or cleaning dentist-initiated 23 34 31 24 Symptoms 11 10 10 11 Follow-up on condition 1 1 0 1 Other reason
Total
Total (%)
42 12 46
65
98
97
44
50
8
10
53 8 27
42 21 24
11 1
11 1
13
22
*This question was asked only of the main adult household informant (total n=14,180).
dentist. Seventy percent of the adult household informants in white families knew that the purpose of fluoride in the water supply was to prevent tooth decay, whereas only 38 percent of the black and 42 percent of the Hispanic informants knew this. In the 1986 NHIS, information was obtained on the brand names of any toothpastes or mouthrinses used by each household member. The brands were then compared with Food and Drug Administration (FDA) and American Dental Association (ADA) standards regarding their fluoride content and plaque reduction potential. The vast majority of residents of US households (97%) used toothpastes containing fluoride regardless of racial/ethnic group. About half of Americans used toothpastes that, according to FDA and/or ADA standards, also have ingredients to reduce the accumulation of plaque. Whites (52%)were more likely to use these types of toothpaste than were blacks (38%).About one in ten Americans used a mouthrinse with fluoride, and usage of such productsdid not vary greatly across racial/ethnic groups. Information was also obtained for all household members regarding the reason for their last visit to the dentist in the previous two years. Blacks and Mexican-heritage Hispanics were more likely than whites to have gone in response to symptoms. Table 2 presents the results of the logistic regression of the predictors of whether a dentist was seen for adults 18
years of age or older and children and adolescents 2-17 years of age. Reference-cellcoding was used in the analysis-that is, the effectsof the variables in the model were measured from a reference category formed by the combination of the levels of the variable excluded from the model. In Table 2 these are the categoriesappearing after "vs" in the listing of categories for each variable (the last category in combinations of categories for interaction terms). Adults. Among adults 18 years of age and older, persons with the following characteristics were least likely to have been to a dentist: males, members of larger families, those who were unemployed or in blue-collar jobs, persons residing in the southern states or nonmetropolitan areas, those with no private dental insurance, and people who felt that they had fair to poor health. Mexican-Americans were least likely to have been seen by a dentist-regardless of their education and income. Children and Adolescents. The characteristics in general of children and adolescents for whom the probability of seeing a dentist were lower were similar to those for adults: boys more so than girls; those in larger families, or ones in which the head of the family had no post-high school education, or was unemployed, or worked in a blue-collar job; children who lived in nonmetropolitan areas or the southern states; and those who were poor or had no private dental insurance. Mexican-Americanchildren were least likely to have seen a dentist.
213
Vol. 52, No. 4, Summer 1992
TABLE 2 Logistic Regression Analysis of Predictors of Seeing a Dentist in the Year-United States, 1986,
TABLE 2 Logistic Regression Analysis of Predictors of Seeing a Dentist in the Year-United States, 1986 [Continued] Odds Ratios of Seeing a Dentist
Odds Ratios of Seeing a Dentist Selected Predictors
Predisposing Race Black Hispanic (Mexican) Hispanic (other) vs white Sex: Male vs female Family size (number) Education Less than high school High school graduate vs post-high school Employment/occupation Employed-blue collar Unemployed Not in labor force vs employed-white collar Region Northeast West central West vs south Location of residence Central city Not central city vs non-MSA Enabling Poverty level Above vs below Private dental insurance Covered vs not Need Perceived health Excellent Very good or good vs fair or poor Predisposing and enabling Race and age Black Hispanic (Mexican) Hispanic (other) vs white
Adults
-
0.737 0.964
Children
0.800 0.657 0.990 0.850 0.944
-
0.554 0.662
0.764 0.753 0.881
0.714 0.790 1.010
1.266 1.234 1.257
1.804 1.637 1.117
1.039 1.145
1.001t 1.156
-
1.268
1.711
1.690
1.379 1.260
1.230$ 1.102
0.924$ 1.057 0.999
-
Approximately 18 percent of adults and 15 percent of the children in the sample were not included in the logistic regression analysisprincipally due to missing values on income and insurance coverage.The characteristics of those not included were most like those who had not been to a dentist. Our expectation is that the odds ratios (or likelihood) of seeing a dentist for subgroups
Selected Predictors
Predisposing and enabling Race and education Black Less than high school High school graduate Post-high school Hispanic (Mexican) Less than high school High school graduate Post-high school Hispanic (other) Less than high school High school graduate Post-high school White Less than high school High school graduate Post-high school Race and poverty level Black Above poverty level Below poverty level Hispanic (Mexican) Above poverty level Below poverty level Hispanic (other) Above poverty level Below poverty level White Above poverty level Below poverty level
Adults
Children
.859 ,654 1.014
-
,287 .218 ,261
-
1.161 .874 1.206
-
,557 .643 1.ooo
-
,680 1.014
-
.326 .261
-
5.54 1.206
-
.842 1.ooo
-
-
-
-
-
*The main effects of race, age, education, and poverty level for adults wereincorporatedin theoddsratioscomputedfor interactionsbetween thesevariables. The interactions of these variables, reported for adults, were not significant for children. Age was, however, not includedin the model for children. All the effects reported were significant at P1.01, unless indicated as significant at P1.05(t)or not significant (t).
included in the logistic regression analysis that had high incomes and ratesof insurance coverage vs those that did not (people with a post-high school education in whitecollar jobs compared to those with less than a high school education who were unemployed, for example) would have increased, rather than diminished, had the cases with missing values been included.
Journal of Public Health Dentistry
214
Discussion Dental care is often viewed as elective, since poor dental health is generally not life-threatening. Caries or gum disease can, however, lead to the early and unnecessary loss of teeth, exacerbatesymptoms related to other physical conditions, reduce the quality of life associated with one’s physical appearance or the ability to chew food adequately, and, at its most severe, to edentulousness. A variety of programs have been suggested to enhance the access of minority children and adults to dental care: for example, through school-based preventive dentistry programs, through targeted interventions by public health departments or community health centers, through the establishment of resident training opportunities for dental and dental hygienist students in low-income minority areas, through the expansion of dental coverage and benefits available under Medicaid, as well as through the addition of dental coverage under employer-based health insurance plans (29-36). Not having dental coverage is an important deterrent to seeing a dentist. The provision of dental coverage to adult workers and their dependents would greatly enhance access to dental care. Coverage for routine exams and prophylactic cleaning, and for lower cost preventionoriented services could be emphasized in extending such coverage-thereby ultimately containing the overall costs of treating dental morbidities resulting from inadequate routine dental care. The effect of public coverage for dental care through Medicaid could not be examined directly in these analyses. However, according to the 1977 National Medical Care Expenditure Survey, almost half (46.7%) of Puerto Ricans had Medicaid coverage compared to 27.3 percent for Mexican-heritage Hispanics,28.7percent forblacks, and 5.8 percent for whites (18). The finding that poor non-Mexican-American Hispanics (mostly Puerto Ricans) had higher rates of use than their nonpoor counterparts points to the probable impact of such coverage on access. Hispanic men who are unemployed or not in the labor force were particularly likely to have inadequate dental care because of not having private insurance through work, or of not being eligible for Medicaid or Medicare unless disabled. Those groups with no public or private dental coverage particularly could benefit from the availability of subsidized resident training or public health clinics in their neighborhoods. Both adults and children in nonmetropolitan areas, as well as those residing in the southern United States, were less likely to have been to a dentist. This probably reflects the impact of the lower availability of dentists and the less generous Medicaid benefits available for dental care in these areas. There also appears to have been a considerable lack of awareness among adult informants in minority households in particular about the purposes of fluoridated public water supplies. However, the vast majority of
residents of US households use toothpastes containing fluoride. Using fluoride toothpaste may help to prevent tooth decay and reduce the need for related restorative care, which more often motivates this group to visit a dentist. The value of fluoridation, the use of fluoridated toothpastes and mouthrinses, proper toothbrushing and flossing techniques all should be emphasized in schoolbased dental programs and in public dental health clinics in which minority children and their parents are served. There is some evidence as well that Mexican-Americans and members of large families (which more often tend to characterize this Hispanic subgroup) are less likely to have seen a dentist. Public health nurses and outreach workers should pay particular attention to screening such children, and in facilitating their and their families’access to public health dental clinics or community-oriented dental residency training programs providing free or low-cost care to these populations. Acknowledgments The authors wish to express their appreciation to theNationalCenter for Health Statistics and NCHS s t a f f 4 e r r y Hendershot, Susan Jack, and Barbara Bloom in particular-for their support for the conduct of the analysesreported here. Wearealsograteful toBillSpears,Chih-Wen Chung. and Katherine V. Wilcox for their assistance in the construction andanalysesof thedata files,and to RonaldM. Andersen andMeei-Shia Chen, Center for Health Administration Studies, the University of Chicago, for their insightful and helpful comments on an earlier draft of the manuscript.
References 1. Sinkford JC. Status of dental health in black and white Americans. JNatlMed Assoc 1988;8@1127-31. 2. Ismail AI, Burt BA, Brunelle JA. Prevalence of dental caries and periodontal disease in Mexican-American children aged 5 to 17 years: results from southwestern I-II-IANES,1982-83.Am J Public Health 1987;m967-70. 3. Ismail AI, Burt BA, Brunelle]A. Prevalenceof total tooth loss, dental caries, andperiodontaldiseasein Mexican-Americanadults:results from the southwestern HHANES. J Dent Res 1987;65:1183-8. 4. Ismail AI, Szpunar SM. The prevalence of total tooth loss, dental caries, and periodontal disease among Mexican Americans, Cuban Americans, and Puerto Ricans: findings from 1-WANES1982-84. Am J Public Health 199O;So(Suppl):66-70. 5. PollickHF,RiceAJ, EchenbergD. Dental healthofrecentimmigrant children in the newcomer schools, San Francisco. Am J Public Health 1987;m731-2. 6. Call RL, Entwistle B, Swanson T. Dental caries in permanent teeth in children of migrant farm workers. Am j P u b l i c Health 1 9 8 7 n1002-3. 7. Aday LA, Fleming GV, Andersen R. Access to medical care in the U S who has it, who doesn’t. Chicago: Pluribus Press, Inc. 1984. 8. Andersen R, Aday LA, Lyttle CS,Cornelius LJ, Chen MS. Ambulatory care and insurance coverage in an era of constraint. Chicago: Pluribus Press, 1987. (CHAS Research Series no 35.) 9. Markides KS, Levin JS, Ray LA. Determinants of physician utilization among Mexican-Americans: a three-generations study. Med Care 1985;23:236-46. 10. Newacheck PW. Access to ambulatory care for poor persons. Health Services Res 1988;23:3:401-19. 11. Orr ST,Charney E, Straus J. Use of health services by black children according to payment mechanism. Med Care 1988;26939-47. 12. National Center for Health Statistics. Health indicators for Hispanic, black, and white Americans. Washington, DC USGovenunent Printing Office, 1984;DHHSpubno(PHS)&l576(Vitaland health statistics; series 10; no 148). 13. Hayward RA,Meetz HK, Shapiro MF, Freeman HE. Utilization of dental seMces: 1986 patterns and trends. J Public Health Dent
Vol. 52, No. 4,S u m m e r 1992 1989i49147-52. 14. Ismail AI,Szpunar SM. Oral health status of Mexican-Americans with low and high acculturation status: findings from southwestern HHANES, 1982-84.J Public Health Dent 1990;502431. 15. National Center for Health Statistics. Use of dental services and dental health, United States, 1986. Washington, DC: US Government Printing Office, 1988; DHHS pub no (PHS)88-1593(Vital and health statistics; series 10; no 165). 16. Garcia JA, Juarez RZ. Utilization of dental health services by Chicanos and Anglos. J Health Soc Behav 1978;19:42836. 17. Guendelman S, Schwalbe J. Medical care utilization by Hispanic children: how does it differ from black and white peers? Med Care 1986;24925-4O. 18. Schur CL, Bemstein B, Berk ML. The importance of distinguishing Hispanic subpopulations in the use of medical care. Med Care 1987;25:627-41. 19. Jack SS. Use of dental services: United States, 1983. Washington, D C USGovemment Printing Office, 1986; DHHSpub no (PHSI861250 (Advance data; no 122). 20. lsmail AI, Burt BA, Hendershot GE, Jack S, Corbin SB. Findings from the dental care supplement of the national health interview survey, 1983.J Am Dent Assoc 1987;114;617-21. 21. Kovar MG, Jack S, Bloom B. Dental care and dental health: NHIS. Am J Public Health 1988;78:1496-7. 22. Hunt N, Silverman HA. Use of dental services in 1980. Health Care Finanang Rev 1987;9:31-42. 23. Rossiter LF. Dental services: use, expenditures, and sources of payment. Washington, DC: US Government Printing Office, 1982; DHHS pub no (PHS)82-3319 (National health care expenditures study; Data preview 8). 24. Grernbowsla D, Andersen RM, Chen MS. A public health model of the dental care process. Med Care Rev 1989;46:439-96.
215 25. National Center for Health Statistics. Current estimates from the national health interview survey, Unitedstates, 1986. Washington, Dc:US Government Printing Office, 1987;DHHS pub no ( P H S W 1592 (Vital and health statistics; series 10; no 164). 26. Cleary PD, Angel R. The analysis of relationship involving dichotomous dependent variables. J Health Soc Behav 1984;25:334-48. 27. Hidiroglou MA, Fuller WA, Hidunan RD. SUPER CARP. Ames, IA:Survey Section, Statistical Laboratory, Iowa State University, 1980. 28. SAS Institute. SAS language guide for personal computers, version 6 edition. Cary, NC: SAS Institute, 1985. 29. Fleiss JL. Review of the national preventive dentistry demonstration program. Am J Public Health 1986;76434-47. 30. Klein SP, Bohannan HM, Bell RM, et al. The cost and effectiveness of school-based preventive dental care. Am J Public Health 1985;75:382-91. 31. KleinSP, Bohannan HM, Bell RM, et al. Conjectureversusempirical data: a response to concerns raised about the national preventive dentistry demonstration program. Am J Public Health 1986;76:44852. 32. RebichT.Wool-basedpreventivedentalcare: adifferent view. Am J Public Health 1985;75:392-4. 33. Rebich T. Different strokes, regarding the NPDDP. Am J Public Health 1986;76453. 34. Corbin SB, Kleinman DV, Lane JM. New opportunities for enhancing oral health: moving toward the 1990 objectives for the nation. Public Health Rep 1985;100:515-24. 35. GrembowskiD,ConradD,Milgrom P. Utilizationofdentalservices in the Unitedstates and an insured population. Am J Public Health 1985;7587-9. 36. Grembowski D, Conrad D, Milgrom P. Dental care demand among children with dental insurance. Health Serv Res 1987;21:755-75.
Journal of Public Health Dentistry
A Profile of Black and Hispanic Subgroups’ Access to Dental Care: Findings from the National Health Interview Survey Lu Ann Aday, PhD Ronald N. Forthofer,PhD School of Public Health The University of Texas Health Science Center PO Box 20 186 Houston, TX 77225
Abstract Thisstudy examined which black and Hispanic minority subgroups were least likely to obtain dental care and why, based on logistic regression analyses of the 1986 National Health Interview Survey. Blacks and Hispanics were less apt to have private dental insurance coverage, to be knowledgeable about the purpose of fluoride, to have been to a dentist in the past year, and, when they did go, were more apt to have gone in response to symptoms rather than for preventive reasons, compared to whites. Logistic regression anaiyses foradults 18years of age and older and for children and adolescents 2 to 17 years of age showed that the following individuals had the lowest probability of having been to a dentist in the past year: males, members of larger families, adults who were unemployed or in blue-collar jobs, those who lived in the South or nonmetropolitan areas, people who perceived their health to be fair or poor, and those with no private dental insurance. Mexican-Americans were least likely to have been to a dentist regardless of their income of education. In general, the findings confirmed the importance of dental insurance, as well as suggesting a need for more school-baseddental programs andpublic health clinic-baseddental health education and outreach efforts for targeting minority children and adults. Key Words:dental care, dental insurance, ethnic groups.
Black and Hispanic minorities have poorer oral health than whites. Data from the National Center for Health Statistics(NCHS)Health and Nutrition Examination Survey conducted during 1970-74 documented that 77.9 percent of blacks, compared to 62.3 percent of whites, had unmet dental care needs, including gingivitis, periodontal disease, tooth decay, and fixed bridge and/or partial denture repair (1).The 1985436 National Survey of Oral This research was supported by the National Center for Health Statis tics, Centers for Disease Control, Public Health Service, US Department of Health and Human Services, Hyattsville, MD. Send correspondence and reprint requests to Dr.Aday. Manuscript received: 9/6/90; returned toauthors for revision: 1/1/90; accepted for publication: 9/3 /91.
J Public Health Dent 1992;52(4):210-15
Health in US Employed Adults and Seniors, conducted by theNational Institute of Health, similarly showed that the percentage of persons with decayed teeth was 22.1 percent in blacks compared to 6.8 percent in whites (1). Data from the 1982-84 National Center for Health Statistics Hispanic Health and Nutrition Examination Survey (HHANES)documented that over half of the 17-year-old Mexican-Americans had five or more filled or decayed teeth. Dental caries in the 5- to 17-yeardd MexicanAmericans in this survey was predominantly a disease of the occlusal surfaces of molars, a condition preventable with the proper use of fissure sealants. Mild gingivitis was also prevalent in 77 percent of the Mexican-American children in the HHANES (2-4). Other studies of Hispanic immigrant children-for example, in San Francisco (5) and children of migrant farm workers in Colorado (6I-also show large proportions of children with many preventable and treatable oral health problems. Blacks and Hispanics also have much lower rates of utilization of dental health care services than do whites (7-15). Based on the 1986 National Center for Health Statistics-National Health Interview Survey (NCHSNHIS), age-adjusted proportions of racial /ethnic groups that had been to a dentist in the past year were 59.4 percent for whites, 42.1 percent for blacks, 38.6 percent for Mexican-Americans, and 53.0 for other Hispanics. The age-adjusted percent of Mexican-Americans who had never been to a dentist (12.6 percent) was high compared to whites (4.6 percent) and other minority groups, such as blacks (7.1 percent) and other Hispanics (5.9 percent). Whites who had been to a dentist also averaged more visits per person per year (2.1) compared to the ethnic minorities-blacks (1.4), Mexican-Americans (1.31, and other Hispanics (1.8). The proportion with private dental insurance coverage was also higher for whites (39.3 percent) compared to blacks (28.4 percent), Mexican-Americans (28.7 percent) or other Hispanics (30.1 percent) (15). Since many of the dental problems experienced by minorities are preventable or treatable, it is important to find ways of enhancing minorities‘ access to dental care. In addition, it is important to examine subgroups of
Vol. 52, No. 4,Summer 1992
Hispanicsseparately since multivanate analyses examining Hispanics’ utilization of health services have found that Mexican-Americans, besides tending to have lower rates of utilization than whites and blacks, also tend to have lower utilization than do other groups of Hispanics (9,16-18). A number of studies have examined possible predictors of dental care use (7,8,12-16,19-23). Grembowski et al. (24), in their public health model of the dental care process, suggest that a complex array of structural, historical, and cognitive factors predict whether or not dental care is sought. There is inadequate evidence, however, on the relative importance of these factors in understanding the lower utilization rates of poor and minority popula tions. The analyses reported here are based on the 1986 NCHS National Health Interview Survey and associated dental supplement, which provides a reasonable number of observations to address these issues for blacks and Hispanic subgroups. The factors most important in predicting which racial/ethnic subgroups are least likely to obtain dental care are identified. The results are intended to facilitate the design of interventions for enhancing minorities’ access to dental care. Methods The 1986 National Health Interview Survey is a nationally representative area probability sample of the noninstitutionalized US population. More detail about the sample design is available in the descriptive reports based on that survey (25). The survey is based on household interviews of all adults 17 years of age or older in person; information on all children was obtained from a knowledgeable adult in the household. Proxy responses were also allowed for adult household members who were unavailable at the time of the interview. There were 62,052 persons for whom information was obtained in the survey. However, the analyses here were limited to whites, blacks, and Hispanics two years of age or older; these exclusions reduced the sample to 58,435 persons. The distribution in the respective racial/ethnic subgroups is: whites (44,573), blacks (9,438), HispanicMexican-Americans (2,3301, and Hispanic-Other (2,094). Selected opinion and knowledge questions regarding the use of fluoride were asked only of the adult who was the main household informant (total n=14,180). Unlike many other surveys that have low response rates, and hence are subject to possible selection bias, a strength of this study is that the response rate for this nationally representative survey was 96 percent. The framework used was the Behavioral Model of Health Services Utilization developed by Andersen and his colleagues (7,8). In this model, a series of predisposing, enabling, and need factors are hypothesized to be predictive of individuals’ utilization of health services. Predisposing factors reflect the propensity of the individ-
211
ual to seek care. Those used in this study were race, age, sex, family size, education, employment/occupa tion, region, and place of residence. Enabling variables capture the resources available to the individual as well as those available in the community. The enabling variables used here refer to the individual and included poverty level and private dental insurance. Need refers either to individual or provider judgments of the presence and severity of conditions requiring treatment. A question concerning whether respondents perceived their health to be excellent, good, fair, or poor was used to adjust for their overall health status. A focus of the analysis was on understanding access for different subgroups of minorities-reflected in the interactionsof race with other characteristics (age,education, and poverty level). Other variables based on the Behavioral Model that would have been useful to include in the analyses but were not available in the 1986 NCHS-NHIS include knowledge and attitudes toward dental care, Medicaid coverage, the availability of dentists in the areas in which respondents resided, and direct indicators of the need for dental care. Logistic regression was chosen as the procedure for examining the linkage between the dichotomous dependent variable, whether or not the person had been to a dentist in the past year, and the predisposing, enabling, and need predictor variables. The procedure controlled for racial/ethnic group membership and its interactions with age, education, and poverty level (26). Models were developed separately for those aged 2 to 17 and for adults 18 years of age or older. The education, employment status, and occupation of the head of the family (the adult family member in whose name the home was owned or rented) were attached to the child’s record, and used as predictors of the child’s utilization. In the analyses of these data, collected using a complex sampling design, it was necessary to incorporate the following: (1 ) sample weights to adjust for different probabilities of selection, and (2) design effects, reflecting the extent to which the variances of estimates obtained from stratified and cluster sample-based designs differedfrom those of a simple random sample. The analysis was performed with PROC LOGISTIC in SAS, which allowed the sample weights to be incorporated. The SUPERCARPprogram along with PROC GLM in the SAS package were used to obtain estimates of the design effects for the predictor variables (27,281. In assessing the statistical significance of the terms in the l o p t i c regression models, a design effect of 2.0 was used-meaning that the variances were estimated to be twice that of a simple random sample. This results in a conservative procedure for identifying the statistically significantpredictors.
Results Table 1 provides specific information on Americans’ knowledge related to fluoride and reasons for seeing a
Journal of Public Health Dentistry
212
TABLE 1 Selected Characteristics of Whites, Blacks, and Hispanics, 2+ Years of Age, Regarding Dental Health and Health CareUnited States, 1986 Racial/Ethnic Groups (%) Hispanic Characteristics
White
Black
Mexican
Other
Adult household respondent's knowledge about purpose of fluoride in public water systems* 70 38 38 45 Prevent tooth decay 12 24 10 15 Some other reason 18 37 51 40 Not known Toothpaste with fluoride 97 99 98 Yes 97 Toothpaste for plaque 38 48 39 Yes 52 Mouthrinse with fluoride 11 7 9 Yes 10 Main reason for last visit in previous two years 41 44 51 56 Exam or cleaning self-initiated 24 12 7 9 Exam or cleaning dentist-initiated 23 34 31 24 Symptoms 11 10 10 11 Follow-up on condition 1 1 0 1 Other reason
Total
Total (%)
42 12 46
65
98
97
44
50
8
10
53 8 27
42 21 24
11 1
11 1
13
22
*This question was asked only of the main adult household informant (total n=14,180).
dentist. Seventy percent of the adult household informants in white families knew that the purpose of fluoride in the water supply was to prevent tooth decay, whereas only 38 percent of the black and 42 percent of the Hispanic informants knew this. In the 1986 NHIS, information was obtained on the brand names of any toothpastes or mouthrinses used by each household member. The brands were then compared with Food and Drug Administration (FDA) and American Dental Association (ADA) standards regarding their fluoride content and plaque reduction potential. The vast majority of residents of US households (97%) used toothpastes containing fluoride regardless of racial/ethnic group. About half of Americans used toothpastes that, according to FDA and/or ADA standards, also have ingredients to reduce the accumulation of plaque. Whites (52%)were more likely to use these types of toothpaste than were blacks (38%).About one in ten Americans used a mouthrinse with fluoride, and usage of such productsdid not vary greatly across racial/ethnic groups. Information was also obtained for all household members regarding the reason for their last visit to the dentist in the previous two years. Blacks and Mexican-heritage Hispanics were more likely than whites to have gone in response to symptoms. Table 2 presents the results of the logistic regression of the predictors of whether a dentist was seen for adults 18
years of age or older and children and adolescents 2-17 years of age. Reference-cellcoding was used in the analysis-that is, the effectsof the variables in the model were measured from a reference category formed by the combination of the levels of the variable excluded from the model. In Table 2 these are the categoriesappearing after "vs" in the listing of categories for each variable (the last category in combinations of categories for interaction terms). Adults. Among adults 18 years of age and older, persons with the following characteristics were least likely to have been to a dentist: males, members of larger families, those who were unemployed or in blue-collar jobs, persons residing in the southern states or nonmetropolitan areas, those with no private dental insurance, and people who felt that they had fair to poor health. Mexican-Americans were least likely to have been seen by a dentist-regardless of their education and income. Children and Adolescents. The characteristics in general of children and adolescents for whom the probability of seeing a dentist were lower were similar to those for adults: boys more so than girls; those in larger families, or ones in which the head of the family had no post-high school education, or was unemployed, or worked in a blue-collar job; children who lived in nonmetropolitan areas or the southern states; and those who were poor or had no private dental insurance. Mexican-Americanchildren were least likely to have seen a dentist.
213
Vol. 52, No. 4, Summer 1992
TABLE 2 Logistic Regression Analysis of Predictors of Seeing a Dentist in the Year-United States, 1986,
TABLE 2 Logistic Regression Analysis of Predictors of Seeing a Dentist in the Year-United States, 1986 [Continued] Odds Ratios of Seeing a Dentist
Odds Ratios of Seeing a Dentist Selected Predictors
Predisposing Race Black Hispanic (Mexican) Hispanic (other) vs white Sex: Male vs female Family size (number) Education Less than high school High school graduate vs post-high school Employment/occupation Employed-blue collar Unemployed Not in labor force vs employed-white collar Region Northeast West central West vs south Location of residence Central city Not central city vs non-MSA Enabling Poverty level Above vs below Private dental insurance Covered vs not Need Perceived health Excellent Very good or good vs fair or poor Predisposing and enabling Race and age Black Hispanic (Mexican) Hispanic (other) vs white
Adults
-
0.737 0.964
Children
0.800 0.657 0.990 0.850 0.944
-
0.554 0.662
0.764 0.753 0.881
0.714 0.790 1.010
1.266 1.234 1.257
1.804 1.637 1.117
1.039 1.145
1.001t 1.156
-
1.268
1.711
1.690
1.379 1.260
1.230$ 1.102
0.924$ 1.057 0.999
-
Approximately 18 percent of adults and 15 percent of the children in the sample were not included in the logistic regression analysisprincipally due to missing values on income and insurance coverage.The characteristics of those not included were most like those who had not been to a dentist. Our expectation is that the odds ratios (or likelihood) of seeing a dentist for subgroups
Selected Predictors
Predisposing and enabling Race and education Black Less than high school High school graduate Post-high school Hispanic (Mexican) Less than high school High school graduate Post-high school Hispanic (other) Less than high school High school graduate Post-high school White Less than high school High school graduate Post-high school Race and poverty level Black Above poverty level Below poverty level Hispanic (Mexican) Above poverty level Below poverty level Hispanic (other) Above poverty level Below poverty level White Above poverty level Below poverty level
Adults
Children
.859 ,654 1.014
-
,287 .218 ,261
-
1.161 .874 1.206
-
,557 .643 1.ooo
-
,680 1.014
-
.326 .261
-
5.54 1.206
-
.842 1.ooo
-
-
-
-
-
*The main effects of race, age, education, and poverty level for adults wereincorporatedin theoddsratioscomputedfor interactionsbetween thesevariables. The interactions of these variables, reported for adults, were not significant for children. Age was, however, not includedin the model for children. All the effects reported were significant at P1.01, unless indicated as significant at P1.05(t)or not significant (t).
included in the logistic regression analysis that had high incomes and ratesof insurance coverage vs those that did not (people with a post-high school education in whitecollar jobs compared to those with less than a high school education who were unemployed, for example) would have increased, rather than diminished, had the cases with missing values been included.
Journal of Public Health Dentistry
214
Discussion Dental care is often viewed as elective, since poor dental health is generally not life-threatening. Caries or gum disease can, however, lead to the early and unnecessary loss of teeth, exacerbatesymptoms related to other physical conditions, reduce the quality of life associated with one’s physical appearance or the ability to chew food adequately, and, at its most severe, to edentulousness. A variety of programs have been suggested to enhance the access of minority children and adults to dental care: for example, through school-based preventive dentistry programs, through targeted interventions by public health departments or community health centers, through the establishment of resident training opportunities for dental and dental hygienist students in low-income minority areas, through the expansion of dental coverage and benefits available under Medicaid, as well as through the addition of dental coverage under employer-based health insurance plans (29-36). Not having dental coverage is an important deterrent to seeing a dentist. The provision of dental coverage to adult workers and their dependents would greatly enhance access to dental care. Coverage for routine exams and prophylactic cleaning, and for lower cost preventionoriented services could be emphasized in extending such coverage-thereby ultimately containing the overall costs of treating dental morbidities resulting from inadequate routine dental care. The effect of public coverage for dental care through Medicaid could not be examined directly in these analyses. However, according to the 1977 National Medical Care Expenditure Survey, almost half (46.7%) of Puerto Ricans had Medicaid coverage compared to 27.3 percent for Mexican-heritage Hispanics,28.7percent forblacks, and 5.8 percent for whites (18). The finding that poor non-Mexican-American Hispanics (mostly Puerto Ricans) had higher rates of use than their nonpoor counterparts points to the probable impact of such coverage on access. Hispanic men who are unemployed or not in the labor force were particularly likely to have inadequate dental care because of not having private insurance through work, or of not being eligible for Medicaid or Medicare unless disabled. Those groups with no public or private dental coverage particularly could benefit from the availability of subsidized resident training or public health clinics in their neighborhoods. Both adults and children in nonmetropolitan areas, as well as those residing in the southern United States, were less likely to have been to a dentist. This probably reflects the impact of the lower availability of dentists and the less generous Medicaid benefits available for dental care in these areas. There also appears to have been a considerable lack of awareness among adult informants in minority households in particular about the purposes of fluoridated public water supplies. However, the vast majority of
residents of US households use toothpastes containing fluoride. Using fluoride toothpaste may help to prevent tooth decay and reduce the need for related restorative care, which more often motivates this group to visit a dentist. The value of fluoridation, the use of fluoridated toothpastes and mouthrinses, proper toothbrushing and flossing techniques all should be emphasized in schoolbased dental programs and in public dental health clinics in which minority children and their parents are served. There is some evidence as well that Mexican-Americans and members of large families (which more often tend to characterize this Hispanic subgroup) are less likely to have seen a dentist. Public health nurses and outreach workers should pay particular attention to screening such children, and in facilitating their and their families’access to public health dental clinics or community-oriented dental residency training programs providing free or low-cost care to these populations. Acknowledgments The authors wish to express their appreciation to theNationalCenter for Health Statistics and NCHS s t a f f 4 e r r y Hendershot, Susan Jack, and Barbara Bloom in particular-for their support for the conduct of the analysesreported here. Wearealsograteful toBillSpears,Chih-Wen Chung. and Katherine V. Wilcox for their assistance in the construction andanalysesof thedata files,and to RonaldM. Andersen andMeei-Shia Chen, Center for Health Administration Studies, the University of Chicago, for their insightful and helpful comments on an earlier draft of the manuscript.
References 1. Sinkford JC. Status of dental health in black and white Americans. JNatlMed Assoc 1988;8@1127-31. 2. Ismail AI, Burt BA, Brunelle JA. Prevalence of dental caries and periodontal disease in Mexican-American children aged 5 to 17 years: results from southwestern I-II-IANES,1982-83.Am J Public Health 1987;m967-70. 3. Ismail AI, Burt BA, Brunelle]A. Prevalenceof total tooth loss, dental caries, andperiodontaldiseasein Mexican-Americanadults:results from the southwestern HHANES. J Dent Res 1987;65:1183-8. 4. Ismail AI, Szpunar SM. The prevalence of total tooth loss, dental caries, and periodontal disease among Mexican Americans, Cuban Americans, and Puerto Ricans: findings from 1-WANES1982-84. Am J Public Health 199O;So(Suppl):66-70. 5. PollickHF,RiceAJ, EchenbergD. Dental healthofrecentimmigrant children in the newcomer schools, San Francisco. Am J Public Health 1987;m731-2. 6. Call RL, Entwistle B, Swanson T. Dental caries in permanent teeth in children of migrant farm workers. Am j P u b l i c Health 1 9 8 7 n1002-3. 7. Aday LA, Fleming GV, Andersen R. Access to medical care in the U S who has it, who doesn’t. Chicago: Pluribus Press, Inc. 1984. 8. Andersen R, Aday LA, Lyttle CS,Cornelius LJ, Chen MS. Ambulatory care and insurance coverage in an era of constraint. Chicago: Pluribus Press, 1987. (CHAS Research Series no 35.) 9. Markides KS, Levin JS, Ray LA. Determinants of physician utilization among Mexican-Americans: a three-generations study. Med Care 1985;23:236-46. 10. Newacheck PW. Access to ambulatory care for poor persons. Health Services Res 1988;23:3:401-19. 11. Orr ST,Charney E, Straus J. Use of health services by black children according to payment mechanism. Med Care 1988;26939-47. 12. National Center for Health Statistics. Health indicators for Hispanic, black, and white Americans. Washington, DC USGovenunent Printing Office, 1984;DHHSpubno(PHS)&l576(Vitaland health statistics; series 10; no 148). 13. Hayward RA,Meetz HK, Shapiro MF, Freeman HE. Utilization of dental seMces: 1986 patterns and trends. J Public Health Dent
Vol. 52, No. 4,S u m m e r 1992 1989i49147-52. 14. Ismail AI,Szpunar SM. Oral health status of Mexican-Americans with low and high acculturation status: findings from southwestern HHANES, 1982-84.J Public Health Dent 1990;502431. 15. National Center for Health Statistics. Use of dental services and dental health, United States, 1986. Washington, DC: US Government Printing Office, 1988; DHHS pub no (PHS)88-1593(Vital and health statistics; series 10; no 165). 16. Garcia JA, Juarez RZ. Utilization of dental health services by Chicanos and Anglos. J Health Soc Behav 1978;19:42836. 17. Guendelman S, Schwalbe J. Medical care utilization by Hispanic children: how does it differ from black and white peers? Med Care 1986;24925-4O. 18. Schur CL, Bemstein B, Berk ML. The importance of distinguishing Hispanic subpopulations in the use of medical care. Med Care 1987;25:627-41. 19. Jack SS. Use of dental services: United States, 1983. Washington, D C USGovemment Printing Office, 1986; DHHSpub no (PHSI861250 (Advance data; no 122). 20. lsmail AI, Burt BA, Hendershot GE, Jack S, Corbin SB. Findings from the dental care supplement of the national health interview survey, 1983.J Am Dent Assoc 1987;114;617-21. 21. Kovar MG, Jack S, Bloom B. Dental care and dental health: NHIS. Am J Public Health 1988;78:1496-7. 22. Hunt N, Silverman HA. Use of dental services in 1980. Health Care Finanang Rev 1987;9:31-42. 23. Rossiter LF. Dental services: use, expenditures, and sources of payment. Washington, DC: US Government Printing Office, 1982; DHHS pub no (PHS)82-3319 (National health care expenditures study; Data preview 8). 24. Grernbowsla D, Andersen RM, Chen MS. A public health model of the dental care process. Med Care Rev 1989;46:439-96.
215 25. National Center for Health Statistics. Current estimates from the national health interview survey, Unitedstates, 1986. Washington, Dc:US Government Printing Office, 1987;DHHS pub no ( P H S W 1592 (Vital and health statistics; series 10; no 164). 26. Cleary PD, Angel R. The analysis of relationship involving dichotomous dependent variables. J Health Soc Behav 1984;25:334-48. 27. Hidiroglou MA, Fuller WA, Hidunan RD. SUPER CARP. Ames, IA:Survey Section, Statistical Laboratory, Iowa State University, 1980. 28. SAS Institute. SAS language guide for personal computers, version 6 edition. Cary, NC: SAS Institute, 1985. 29. Fleiss JL. Review of the national preventive dentistry demonstration program. Am J Public Health 1986;76434-47. 30. Klein SP, Bohannan HM, Bell RM, et al. The cost and effectiveness of school-based preventive dental care. Am J Public Health 1985;75:382-91. 31. KleinSP, Bohannan HM, Bell RM, et al. Conjectureversusempirical data: a response to concerns raised about the national preventive dentistry demonstration program. Am J Public Health 1986;76:44852. 32. RebichT.Wool-basedpreventivedentalcare: adifferent view. Am J Public Health 1985;75:392-4. 33. Rebich T. Different strokes, regarding the NPDDP. Am J Public Health 1986;76453. 34. Corbin SB, Kleinman DV, Lane JM. New opportunities for enhancing oral health: moving toward the 1990 objectives for the nation. Public Health Rep 1985;100:515-24. 35. GrembowskiD,ConradD,Milgrom P. Utilizationofdentalservices in the Unitedstates and an insured population. Am J Public Health 1985;7587-9. 36. Grembowski D, Conrad D, Milgrom P. Dental care demand among children with dental insurance. Health Serv Res 1987;21:755-75.
