Association between socioeconomic contextual factor, dental care service availability, and prevalence of periodontitis in Brazil: a multilevel analysis

Associação entre o fator contextual socioeconômico, a disponibilidade de serviços odontológicos e a prevalência de periodontite no Brasil: uma análise multinível

Asociación entre el factor socioeconómico, la disponibilidad de servicios odontológicos y la prevalencia de periodontitis en Brasil: un análisis multinivel

Abstracts

This study aimed to examine the effect of dental care services on periodontitis cases in Brazilian municipalities. The sample comprised 3,426 individuals aged 35-44 years. Moderate to severe periodontitis with clinical attachment loss and probing depth was the dependent variable, both > 3mm. Its exploratory variables were grouped into four categories: (1) individual characteristics; (2) contextual development indicators; (3) health service and structural factors; and (4) dental care use. Data were collected using the SBBrasil 2010 Project, the Brazilian Institute of Geography and Statistics, the Brazilian Information System of Primary and Secondary Care, and the Program to Improve Access and Quality of Dental Specialization Centers (PMAQ-CEO). Multilevel logistic regression was used to assess associations of periodontitis with individual and context variables. Municipalities with > 1 CEO or > 1 of any centers were associated with periodontitis, with OR = 0.97 (95%CI: 0.55-1.71) and OR = 0.41 (95%CI: 0.17-0.97), respectively. Prevalence of periodontitis was more likely in older people, lower education levels, and individuals that sought dental visits for pain/extraction and periodontal treatment. Other dental care services availability were not associated with the prevalence of periodontitis.

Keywords:
Dental Care; Periodontitis; Dental Health Care

Este estudo teve como objetivo examinar o efeito dos serviços de assistência odontológica nos casos de periodontite em municípios brasileiros. A amostra foi composta por 3.426 indivíduos de 35-44 anos de idade. A variável dependente foi periodontite moderada a grave com perda de inserção clínica e profundidade de sondagem, ambas > 3mm. As variáveis exploratórias foram agrupadas em quatro categorias: (1) características individuais; (2) indicadores contextuais de desenvolvimento; (3) serviços de saúde e fatores estruturais; (4) uso da assistência odontológica. Os dados foram coletados por meio do Projeto SBBrasil 2010, do Instituto Brasileiro de Geografia e Estatística, do Sistema de Informação da Atenção Primária e Secundária e do Programa de Melhoria do Acesso e Qualidade da Atenção dos Centros de Especialidade Odontológica (PMAQ-CEO). A regressão logística multinível foi utilizada para avaliar associações de periodontite com variáveis individuais e de contexto. Municípios com mais de um CEO ou mais que um centro de qualquer tipo de assistência foram associados à periodontite com OR = 0,97 (IC95%: 0,55-1,71) e OR = 0,41 (IC95%: 0,17-0,97), respectivamente. A prevalência de periodontite foi mais provável em idosos, indivíduos com menor escolaridade e indivíduos que procuraram consultas odontológicas para dor/extração e tratamento periodontal. A disponibilidade de outros serviços odontológicos não se associou à prevalência de periodontite.

Palavras-chave:
Assistência Odontológica; Periodontite; Serviços de Saúde Bucal

Este estudio tuvo como objetivo evaluar la eficacia de los servicios de atención odontológica en los casos de periodontitis en municipios brasileños. La muestra estuvo compuesta por 3.426 individuos de entre 35 y 44 años de edad. La variable dependiente fue periodontitis de moderada a severa, con pérdida de inserción clínica y profundidad de sondaje, que eran > 3mm. Las variables exploratorias se agruparon en cuatro categorías: (1) características individuales; (2) indicadores contextuales de desarrollo; (3) servicios de salud y factores estructurales; (4) uso de atención odontológica. Los datos se recabaron de Proyecto SBBrasil 2010, del Instituto Brasileño de Geografía y Estadística, del Sistema de Información de la Atención Primaria y Secundaria y del Programa para Mejorar el Acceso y la Calidad de los Centros de Especialización Odontológicas (PMAQ-CEO). Se utilizó la regresión logística multinivel para evaluar las asociaciones de periodontitis con variables individuales y contextuales. Los municipios con más de un CEO u otro tipo de centro de asistencia se asociaron a periodontitis con OR = 0,97 (IC95%: 0,55-1,71) y OR = 0,41 (IC95%: 0,17-0,97), respectivamente. La prevalencia de la periodontitis fue más probable en los ancianos, en las personas con menos nivel de instrucción y en las personas que buscaron citas odontológicas por dolor/extracción y tratamiento periodontal. La disponibilidad de otros servicios odontológicos no se asoció con la prevalencia de periodontitis.

Palabras-clave:
Atención Odontológica; Periodontitis; Servicios de Salud Dental

Introduction

The relationship between primary and secondary care is crucial since it connects adequate treatment to health needs. An ideal interface is defined as one that refers all appropriate cases in the population, with no barriers to receive specialized care after a referral ¹. Some countries provided this interface within a universal system with a complex network of services in which clinics and hospitals are staffed by professionals from various specialties, enabling their cooperation and a multidisciplinary approach to patient needs ^1,2,3. In Brazil, primary care services are the gateway to the public health system, and dental care is provided mainly under oral health teams in the Family Health Strategy (FHS). Secondary care has been expanded through Dental Specialization Centers (CEO), ensuring comprehensive referral treatment ^4,5. The CEO operates as specialized units of higher technological density providing services usually not offered at the primary level ⁶.

There is little evidence of the effects of primary dental care on oral health and perhaps none about the effects of secondary dental care. Some studies have shown that the presence of primary dental care services showed no significant impact on dental caries and tooth loss in young individuals ^7,8,9. However, one study suggested that a preventive-oriented system may confer fewer dental caries ¹⁰. Regarding periodontitis, a systematic review showed a poor association in the access to primary dental care at the municipal level ¹¹. Moreover, population-level periodontal services may not be effective in improving periodontal health as they focus only on those who seek help instead of the disease root causes, such as socioeconomic status, gender, or ethnicity ^12,13.

Access to dental services varies among Brazilian municipalities because of differences in population coverage of primary health care or the presence of specialized services. Some studies state that this variation in access may explain differences in the prevalence of periodontitis ^14,15,16. Dental care availability was associated with moderate and severe clinical attachment loss (CAL) levels ¹⁴, and specialized dental care was associated with higher periodontal procedure rates ¹⁷.

Studies analyzing oral care models have focused on process and structure ^18,19 or on implementing a program or health policies ²⁰. Few studies have investigated the effect of periodontal services on periodontal disease ^11,12. Nevertheless, there is no consensus on the influence of periodontal care services on periodontitis, and the role of confounding factors and mediators must be clarified ¹¹. To fill this knowledge gap, this study investigated the relationship between dental care services and prevalence of periodontitis among adults in Brazilian municipalities.

Methods

Study design, setting, and data

This cross-sectional study pooled data from various sources. The individual characteristics were obtained from the SBBrasil 2010: Brazilian National Oral Health Survey (SBBrasil 2010) ²¹. Contextual variables were collected from the health information systems of all 177 municipalities participating in the SBBrasil 2010. The original sample totaled 9,779 individuals aged 35 to 44 years. This study focused on the last consultation of users of public services, causing 6,353 to be excluded. This criterion was used because it is unlikely that public services would affect users of other types of healthcare systems.

Gender, age, income, education level, and last visit to the dentist and the reason for that visit were extracted from the SBBrasil 2010. Municipal-level variables, such as urban population, population size, Gini index, the income component of the Municipal Human Development Index (HDI-M), and Gross Domestic Product (GDP) were obtained from the 2010 census data, available from the Human Development Report²². Coverage of primary dental care and outpatient production data for periodontics procedures were collected from the 2010 Ambulatory Information System, Brazilian Unified National Health System (SIA/SUS). The type and characterization of the services related to secondary dental care, trained periodontist rate, and production were obtained from the Brazilian National Register of Health Establishments (CNES).

Outcome and data measurement

The outcome variable was moderate to severe periodontitis measured at the individual level and defined as a combination of CAL and probing depth (PD) using The World Health Organization’s (WHO) probe ^14,23,24. Six index teeth, one in each sextant, were examined, and the highest score was used to represent individual disease history in each sextant. Moderate to severe disease was defined as having at least two sites with > 3mm of CAL, and at least one site with > 3mm of PD, not necessarily at the same site.

Exposure variables selection

The exploratory variables were grouped into four categories: (1) individual characteristics: gender (men, women), age (years), education level (elementary school, high school, higher education), household income per capita ²⁵ (< USD 50, USD 50-100, USD 150-300, > USD 300); (2) socioeconomic contextual development indicators: urban population (< 75%, > 75%), Gini index (< 0.50, 0.50-0.60, > 0.60), GDP per capita (< BRL 13,700, BRL 13,700-20,000, > BRL 20,000) HDI-M (< 0.7, ≥ 0.7), and population size (< 20,000, 20,000-100,000, > 100,000); (3) health services and structural factors: secondary dental care centers (none, ≥ 1 CEOs, ≥ 1 of any referral centers), primary dental care coverage (< 25%, 25-75%, > 75%) rate of primary dental care per 100,000 inhabitants (up to 10, 10-20, > 20), and presence of trained periodontist (none, at least one periodontist); (4) dental care use: root and coronal scaling rates per 100,000 inhabitants (lowest national tertile, second national tertile, highest national tertile), reason for last dental visit (check-up/prevention, pain/tooth extraction, treatment), and time since last dental visit (less than 1 year, 1-2 years, 3 or more years).

Statistical analysis

Descriptive data were transformed using sampling weights into the representative prevalence and means. Sampling weights were calculated based on the sampling fractions of each sample stage and were calibrated to correct for non-response ²⁶.

Multilevel logistic regression with a random intercept was used to estimate the adjusted effect of the main variables on periodontitis. Individuals were nested in their respective municipalities, in the second-level units. The logit link function was used, and its exponentiated β values are odds ratio (OR). The variance partitioning coefficient (VPC) was calculated using Method D proposed by Goldstein et al. ²⁷, in which the first level variance is fixed at 3.29 (p2/3) for dichotomous variables, assuming a threshold model. The percentage of variance explained was calculated, including the variance of the fixed linear predictor model, according to Equation 14.21 of Snijders & Bosker ²⁸.

A total of five additive models were fitted based on four categories of variables from our theoretical model (Figure 1), using a hierarchical analysis framework ²⁹. The first model (M1) included individual-level confounding factors as variables; only those with p < 0.10 were retained in the next model. The second model (M2) included significant M1 factors and contextual indicators of socioeconomic development as variables; only non-M1 variables with p < 0.10 were retained in the next model. The third model (M3) included significant M2 indicators and structural healthcare factors as variables. The fourth model (M4) additionally included M3 variables and a dental service use factor; only variables with p < 0.10 were retained in the final model. The final model (M5) included the significant variables from each previous model. Each model fit was assessed using the Bayesian information criterion (BIC) and the percentage of variance was explained. All analyses were performed using Stata program (https://www.stata.com).

Figure 1
Theoretical model of explanatory variables.

Results

The study sample comprised 3,574 individuals, of which 101 were edentulous, and 47 had insufficient dental sextants for CAL examination. The final analytical sample comprised 3,426 individuals, of which 34.5% were men and 65.5% were women. The response rate was 38.9% at the household level, mainly because many houses were closed or uninhabited. Among participating households, 2.2% of individuals refused to be examined. The prevalence of moderate to severe periodontitis in Brazilian adults was 14.5%. Individuals with elementary school and lower household income per capita (< USD 50) had a higher prevalence of moderate to severe periodontitis (15.6% and 16.3%, respectively). Regarding service use, individuals who underwent check-up/prevention exams had a lower prevalence of periodontitis (7%). However, those who accessed the service in < 1 year had a higher prevalence of periodontitis (17.2%). Regarding secondary dental care, among 177 municipalities, 92 have CEOs.

Table 1 shows the distribution of moderate to severe prevalence of periodontitis based on the covariables at the city level among public healthcare users. Moderate to severe periodontitis was more prevalent in municipalities with > 75% primary dental care coverage (21.8%) or no CEO or other centers (18.2%). In addition, municipalities with > 20 primary dental care centers (per 100,000 inhabitants) had a higher prevalence of periodontitis (19%). The absence of a trained periodontist was higher for the prevalence of periodontitis (16.9%). Regarding treatment type, adults in cities in the upper tertile of root and coronal scaling rates per 100,000 inhabitants had 18.5% and 18.2% prevalence of periodontitis, respectively.

Table 1
Weighted prevalence in individuals with two or more sites with > 3mm of clinical attachment loss and > 3mm pocket depth according to city-level covariates among public healthcare users in Brazil.

Characteristics	Total sample		Moderate-severe periodontitis		p-value
	%	n	Weighted prevalence (%)	Cases (n)
Total	100.0	3,426	14.5	435
CEO
None	33.2	513	18.2	64	0.14
≥ 1 centers	57.6	2,709	13.8	360
≥ 1 of any referral centers	9.1	201	5.5	11
Primary dental care coverage (%)
< 25	52.3	1,283	12.3	144	0.04
25-75	23.4	1,526	12.0	211
> 75	24.3	617	21.8	80
Rate of primary care centers [per 100,000 inhabitants]
Up to 10	38.7	1,219	11.8	153	0.15
10-20	22.0	1,432	11.2	181
> 20	39.3	775	19.0	101
Presence of a trained periodontist
None	40.9	616	16.9	65	0.33
At least one periodontist	59.1	2,810	12.8	370
Root scaling rate [per 100,000 inhabitants] (national tercile)
Lowest	28.2	960	10.4	125	0.16
Second	30.7	944	13.1	101
Highest	41.2	1,486	18.5	207
Coronal scaling rate [per 100,000 inhabitants] (national tercile)
Lowest	27.4	966	11.2	130	0.22
Second	30.1	918	12.5	94
Highest	42.5	1,506	18.2	209
GDP per capita (BRL)
< 13,700	32.9	1,284	12.5	176	0.69
13,700-20,000	20.5	1,152	14.4	123
> 20,000	46.6	990	16.0	136
M-HDI (income component)
< 0.70	21.0	523	14.8	62	0.92
≥ 0.70	79.0	2,903	14.4	373
Population size (inhabitants)
< 20,000	19.3	311	21.3	37	0.21
20,000-100,000	26.3	419	12.9	46
> 100,000	54.5	2,696	12.9	352
Urban population (%)
< 75	11.0	367	10.3	38	0.23
> 75	89.0	3,059	15.0	397
Gini index
< 0.50	35.8	364	20.8	49	0.01
0.50-0.60	37.0	1,541	10.5	207
> 0.60	27.2	1,521	11.7	179

• CEO: Dental Especializations Centers; GDP: Gross Domestic Product; HDI-M: Municipal Human Development Index.

Table 2 shows the crude association between individuals and contextual predictors for moderate to severe periodontitis. In the adjusted analysis presented in Table 3, M1 (individual socioeconomic factors) and M2 (contextual socioeconomic development indicators) were not shown because they were the control, but the p-values were presented at the Table’s note. Age and education level were significant in M1 and remained associated in M2. Therefore, the M3 (contextual health service factors) showed that municipalities that have > 1 CEO (only) or > 1 center of any type were associated with periodontitis, with OR = 0.97 (95%CI: 0.55-1.71) and OR = 0.41 (95%CI: 0.17-0.97), respectively, when adjusted for age, education level (M1 + M2), primary dental care coverage, primary dental care rate, and trained periodontist level. Age, education level, > 1 CEO, and > 1 center of any type remained associated in M4 (dental care users). Individual and contextual service factors associated with the presence of a CEO or any center remained associated with periodontitis in the final model (M5). The risk of periodontitis was greater in older adults, individuals who sought dental visits for pain/extraction periodontal treatment, and municipalities with > 1 CEO. In contrast, individuals with higher education levels and/or living in cities with > 1 center of any type had a lower risk of periodontitis. The models indicated variation in the prevalence of periodontitis among cities based on the variance partition coefficient, which ranged from 15% to 17%.

Table 2
Crude association between individuals and contextual variables, and moderate to severe periodontitis among individuals aged 35-44 years using multilevel regression analysis.

Characteristics	Unadjusted coefficients
	OR	95%CI
Gender
Men	1.00
Women	0.99	0.79-1.23
Age	1.09	1.05-1.13
Education level
Elementary school	1.00
High school	0.73	0.56-0.93
College/University	0.45	0.32-0.64
Household income per capita (USD)
< 50	1.00
50-100	0.93	0.63-1.77
150-300	0.77	0.53-1.11
> 300	0.78	0.48-1.25
Gini index
< 0.50	1.00
0.50-0.60	0.81	0.48-1.38
> 0.60	0.78	0.43-1.41
HDI-M (income component)
< 0.70	1.00
≥ 0.70	1.09	0.70-1.71
Urban population (%)
< 75	1.00
> 75	1.24	0.73-2.12
GDP per capita (BRL)
< 13,700	1.00
13,700-20,000	0.87	0.52-1.44
> 20,000	1.16	0.71-1.89
Population size (inhabitants)
< 20,000	1.00
20,000-100,000	0.89	0.42-1.87
> 100,000	1.09	0.55-2.17
CEO
None	1.00
≥ 1 center	1.04	0.66-1.64
≥ 1 of any referral centers	0.44	0.20-1.01
Primary dental care coverage (%)
< 25	1.00
25-75	1.14	0.71-1.82
> 75	1.33	0.77-2.32
Rate of primary dental care centers [per 100,000 inhabitants]
Up to 10	1.00
10-20	1.19	0.71-1.99
> 20	1.22	0.70-2.12
Presence of trained periodontist
None	1.00
At least one periodontist	1.23	0.79-1.92
Root scaling rate [per 100,000 inhabitants] (national tertile)
Lowest	1.00
Second	0.92	0.53-1.60
Highest	1.57	0.97-2.55
Coronal scaling rate [per 100,000 inhabitants] (national tertile)
Lowest	1.00
Second	0.84	0.48-1.49
Highest	1.46	0.92-2.38
Reason for last dental visit
Check-up/Prevention	1.00
Pain/Tooth extraction	2.04	1.46-2.86
Treatment	1.43	0.99-2.08
Time since last dental visit (years)
Less than 1	1.00
1-2	1.06	0.86-1.30
3 or more	1.22	0.95-1.57
Fit indices
Second level variance	0.666
VPC	16.6%
R2	0.0%
Akaike information criterion	2,510.67
Bayesian information criterion	2,522.92

• 95%CI: 95% confidence interval; CEO: Dental Specialization Centers; GDP: Gross Domestic Product; OR: odds ratio; VPC: variance partitioning coefficient.

Table 3
Adjusted association between individuals and contextual variables, and moderate to severe periodontitis among individuals aged 35-44 years using multilevel regression analysis.

Characteristics	M3 *		M4 **		M5 ***
	OR	95%CI	OR	95%CI	OR	95%CI
CEO
None	1.00		1.00		1.00
≥ 1	0.97	0.55-1.71	1.03	0.62-1.71	1.07	0.57-1.73
≥ 1 of any referral centers	0.41	0.17-0.97	0.49	0.19-1.23	0.45	0.20-1.03
Primary dental care coverage (%)
≤ 25	1.00
25-75	1.04	0.59-1.54
> 75	1.40	0.69-2.80
Rate of primary care centers [per 100,000 inhabitants] (%)
Up to 10	1.00
10-20	1.21	0.67-2.19
> 20	1.22	0.59-2.55
Presence of trained periodontist
None	1.00
At least one periodontist	1.19	0.67-2.10
Root scaling rate [per 100,000 inhabitants] (national tercile)
Lowest			1.00
Second			1.00	0.49-2.05
Highest			1.44	0.41-5.07
Coronal scaling rate [per 100.000 inhabitants] (national tertile)
Lowest			1.00
Second			0.80	0.39-1.63
Highest			0.99	0.31-3.16
Reason for last dental visit
Check-up/Prevention			1.00
Pain/Tooth extraction			1.81	1.27-2.58
Treatment			1.39	0.94-2.04
Time since last dental visit (years)
Less than 1			1.00		1.00
1-2			0.98	0.81-1.20	1.81	1.28-2.56
3 or more			1.03	0.79-1.36	1.40	0.95-2.06
Fit indices
Second level variance	0.644		0.669		0.707
VPC	16.4%		16.9%		17.7%
R2	4.8%		6.2%		5.3%
Akaike information criterion	2,464.3		2,437.6		2,446.0
Bayesian information criterion	2,537.7		2,529.1		2,501.0

• 95%CI: 95% confidence interval; CEO: Dental Specialization Centers; OR: odds ratio; VPC: variance partitioning coefficient.
• Note: p-values obtained from chunk test in logistic regression, equivalent to chi-square test. Model 1 (not shown in table): variables in the model: gender (p = 0.95), age (p < 0.01), education level (p < 0.01), income (p = 0.61); Model 2 (not shown in table): variables in the model: age (p < 0.01), education level (p < 0.01), urban population (p = 0.60), Gini coefficient (p = 0.44), GDP (p = 0.57), Human Development Index (p = 0.81), city size (p = 0.33).
• * Model 3: variables in the model: age (p < 0.01), education level (p < 0.01), CEO (p = 0.10), primary dental care coverage (p = 0.54), primary dental care center rate (p = 0.88), presence of periodontist (p = 0.55);
• ** Model 4: variables in the model: age (p < 0.01), education level (p < 0.01), CEO (p = 0.25), root scaling (p = 0.77), coronal scaling (p = 0.79), time since last dental visit (p = 0.94), and reason for last dental visit (p < 0.01);
• *** Model 5: variables in the model: age (p < 0.01), education level (p < 0.01), CEO (p = 0.08), and reason for last dental visit (p < 0.01).

As a sensitivity analysis, we included the number of remaining teeth as a control variable, since the likelihood of the outcome changes according to the number of teeth, but the results remained the same.

Discussion

This study had two primary findings. Firstly, we found conflicting findings on the association of secondary dental care services with the prevalence of moderate to severe periodontitis. Secondly, the primary dental care findings were robust, and all three indicators showed higher prevalence in municipalities with higher coverage. These results disagree with a systematic review suggesting that higher coverage might be associated with a lower prevalence of periodontitis ¹¹. Although the relationship between secondary dental care and periodontitis had not been studied previously, the hypothesis that dental care service might be associated with a lower prevalence of periodontitis was not confirmed. Overall, our findings suggest that neither primary nor secondary dental services influence periodontitis at the population level.

This study had limitations and strengths. A limitation common to all observational studies is the impossibility to establish a temporal relationship, thus allowing reverse causality to occur. Dental care access might be associated with a lower prevalence of periodontitis because of treatment and prevention, but it may also be associated with higher levels since those seeking care get access. Moreover, this study could not control for some confounding variables, such as smoking and diabetes. Its strengths included data from representative general population samples using only public health service users. The study used valid and official information from health information systems. Finally, it was not possible to estimate the extension and severity of periodontitis due to methodological limitations.

Findings based on M3 showed that the chance of having periodontitis was lower when municipalities had > 1 referral center of any type (CEO or otherwise) as they may have some incentive to reach high-risk populations ³⁰. The effect of CEO occurred due to periodontal procedures, such as root and coronal scaling, being used to control the disease. However, neither variable was associated in model M4. Any explanation is speculative, but CEOs might have a different referral system, reaching individuals with major diseases who use primary healthcare or due to some residual confounding. Finally, the absence of an ideal interface between primary and secondary dental care for referral cases may increase demand-led procedures ³¹. This study results may explain such differences in the prevalence of periodontitis among the municipalities.

Another remarkable point to be stressed is the lack of association between primary care and periodontitis. This study did not confirm the results found in previous studies showing associations between primary care coverage and periodontitis ^14,32,33. Methodological differences may explain this, since this study included only public service users within the last three years, unlike previous studies that included private service users. Another aspect to consider is the heterogeneity in outcome measurement among published studies. Prevalence estimates are influenced by the methodology used, including periodontitis definitions and registration protocols ³⁴. Furthermore, there is the possibility that some determinants of both dental care and periodontitis may confound the current findings, such as regional networks (one city center serving several others) and differences in work processes among oral health teams ²⁶.

This study did not confirm the association between dental care and periodontitis. Nevertheless, the literature shows that the influence of health services is weak or absent regarding dental caries ^{10,35,36,37,38} and calculus and gingival bleeding ⁷. In addition, health services may have no epidemiological impact since they can only reach a small proportion of those in need ³⁹. Regarding periodontitis, some cases might go undiagnosed or be asymptomatic, leading to some data that is qualitatively submerged ⁴⁰ and possibly untreated, explaining the “iceberg phenomenon” ³⁹. The importance of socio-economic factors on dental care in the development of periodontal disease mirrors a pattern described for dental caries ⁸.

In this study, a few factors remained independently associated with periodontitis: the reason for the last dental visit (check-up/prevention, pain/tooth extraction, and treatment), age, and education level. The contextual effects of any secondary oral care services were moderately associated with a lower prevalence of periodontitis, suggesting that municipalities need specialized care models in addition to CEOs and the investigation of other health determinants. Studies on periodontitis and dental care services remain rare.

Acknowledgments

R. K. Celeste has a PQ2 fellowship from the Brazilian National Research Council (CNPq; n. 311592/2019-8).

References

• ¹
  Morris AJ, Burke FJT. Primary and secondary dental care: how ideal is the interface? Br Dent J 2001; 191:666-70.
• ²
  Marmot M, Friel S, Bell R, Houweling TA, Taylor S. Closing the gap in a generation: health equity through action on the social determinants of health. Lancet 2008; 372:1661-9.
• ³
  Kandelman D, Arpin S, Baez RJ, Baehni PC, Petersen PE. Oral health care systems in developing and developed countries. Periodontol 2000 2012; 60:98-109.
• ⁴
  Pucca GA, Gabriel M, De Araujo MED, De Almeida FCS. Ten years of a national oral health policy in Brazil: innovation, boldness, and numerous challenges. J Dent Res 2015; 94:1333-7.
• ⁵
  Goes PSA, Figueiredo N, Neves JC, Silveira FMM, Costa JFR, Pucca Júnior GA, et al. Avaliação da atenção secundária em saúde bucal: uma investigação nos centros de especialidades do Brasil. Cad Saúde Pública 2012; 28 Suppl:S81-9.
• ⁶
  Mendes EV. As redes de atenção à saúde. Ciênc Saúde Colet 2010; 15:2297-305.
• ⁷
  Ely HC, Abegg C, Celeste RK, Pattussi MP. Impact of oral health teams of the Family Health Strategy on the oral health of adolescents in the south of Brazil. Ciênc Saúde Colet 2016; 21:1607-16.
• ⁸
  Nadanovsky P, Sheiham A. Relative contribution of dental services to the changes in caries levels of 12-year-old children in 18 industrialised countries in the 1970s and early 1980s. Community Dent Oral Epidemiol 1995; 23:331-9.
• ⁹
  Davenport C, Elley K, Salas C, Taylor-Weetman CL, Fry-Smith A, Bryan S, et al. The clinical effectiveness and cost-effectiveness of routine dental checks: a systematic review and economic evaluation. Health Technol Assess 2003; 7:iii-v, 1-127.
• ¹⁰
  Celeste RK, Nadanovsky P, Leon AP. Association between preventive care provide in public dental services and caries prevalence. Rev Saúde Pública 2007; 41:830-8.
• ¹¹
  Filgueiras LV, Konflanz W, Haas AN, Celeste RK. Assessment of the contextual effects on the prevalence of periodontitis: a systematic review. Braz Oral Res 2023; 36:e0125.
• ¹²
  Knight ET, Murray Thomson W. A public health perspective on personalised periodontics. Periodontol 2000 2018; 78:195-200.
• ¹³
  Thomson WM, Sheiham A, Spencer AJ. Sociobehavioral aspects of periodontal disease. Periodontol 2000 2012; 60:54-63.
• ¹⁴
  Vettore MV, Amorim Marques RA, Peres MA. Social inequalities and periodontal disease: multilevel approach in SBBrasil 2010 survey. Rev Saúde Pública 2013; 47 Suppl 3:29-39.
• ¹⁵
  Borrell LN, Burt BA, Warren RC, Neighbors HW. The role of individual and neighborhood social factors on periodontitis: the third National Health and Nutrition Examination Survey. J Periodontol 2006; 77:444-53.
• ¹⁶
  Bower E, Gulliford M, Steele J, Newton T. Area deprivation and oral health in Scottish adults: a multilevel study. Community Dent Oral Epidemiol 2007; 35118-29.
• ¹⁷
  Celeste RK, Moura FRR, Santos CP, Tovo MF. Análise da produção ambulatorial em municípios com e sem centros de especialidades odontológicas no Brasil em 2010. Cad Saúde Pública 2014; 30:511-21.
• ¹⁸
  Goes PSA, Figueiredo N, Martelli PJL, Luvison IR, Werneck MAF, Ribeiro MAB, et al. Theoretical and methodological aspects of the external evaluation of the improvement, access and quality of centers for dental specialties program. Pesqui Bras Odontopediatria Clín Integr 2018; 18:3433.
• ¹⁹
  Cortellazzi KL, Balbino EC, Guerra LM, Vazquez FL, Bulgareli JV, Ambrosano GMB, et al. Variables associated with the performance of centers for dental specialties in Brazil. Rev Bras Epidemiol 2014; 17:978-88.
• ²⁰
  Souza MC, Araújo TM, Reis Júnior WM, Souza JN, Vilela ABA, Franco TB. Integralidade na atenção à saúde: um olhar da equipe de saúde da família sobre a fisioterapia. Mundo Saúde (1995) 2012; 36:452-60.
• ²¹
  Roncalli AG, Silva NN, Nascimento AC, Freitas CHSM, Casotti E, Peres KG, et al. Aspectos metodológicos do Projeto SBBrasil 2010 de interesse para inquéritos nacionais de saúde. Cad Saúde Pública 2012; 28 Suppl:S40-57.
• ²²
  Programa das Nações Unidas para o Desenvolvimento. Atlas do Desenvolvimento Humano. https://www.br.undp.org/ (accessed on 27/Oct/2021).
  » https://www.br.undp.org/
• ²³
  Celeste RK, Oliveira SC, Junges R. Threshold-effect of income on periodontitis and interactions with race/ethnicity and education. Rev Bras Epidemiol 2019; 22:e190001.
• ²⁴
  Holtfreter B, Albandar JM, Dietrich T, Dye BA, Eaton KA, Eke PI, et al. Standards for reporting chronic periodontitis prevalence and severity in epidemiologic studies: proposed standards from the Joint EU/USA Periodontal Epidemiology Working Group. J Clin Periodontol 2015; 42:407-12.
• ²⁵
  Celeste RK, Bastos JL. Mid-point for open-ended income category and the effect of equivalence scales on the income-health relationship. Rev Saúde Pública 2013; 47 Suppl 3:168-74.
• ²⁶
  Silva NN, Roncalli AG. Plano amostral, ponderação e efeitos do delineamento da Pesquisa Nacional de Saúde Bucal. Rev Saúde Pública 2013; 47 Suppl 3:3-11.
• ²⁷
  Goldstein H, Browne W, Rasbash J. Partitioning variation in multilevel models. Understanting Statistics 2010; 1:223-31.
• ²⁸
  Snijders TAB, Bosker RJ. Multilevel analysis: an introduction to basic and advanced multilevel modeling. London: SAGE; 1999.
• ²⁹
  Victora CG, Huttly SR, Fuchs SC, Olinto MT. The role of conceptual frameworks in epidemiological analysis: a hierarchical approach. Int J Epidemiol 1997; 26:224-7.
• ³⁰
  Pucca GA, Costa JFR, Chagas LD, Sivestre RM. Oral health policies in Brazil. Braz Oral Res 2009; 23 Suppl 1:9-16.
• ³¹
  Morris AJ, Burke FJ. Primary and secondary dental care: how ideal is the interface? Br Dent J 2001; 191:666-70.
• ³²
  Dalazen CE, De Carli AD, Bomfim RA, Santos MLM. Contextual and individual factors influencing periodontal treatment needs by elderly Brazilians: a multilevel analysis. PLoS One 2016; 11:e0156231.
• ³³
  Valente MIB, Vettore M V. Contextual and individual determinants of periodontal disease: multilevel analysis based on Andersen's model. Community Dent Oral Epidemiol 2018; 46:161-8.
• ³⁴
  Demmer RT, Papapanou PN. Epidemiologic patterns of chronic and aggressive periodontitis. Periodontol 2000 2010; 53:28-44.
• ³⁵
  Wang N, Marstrander P, Holst D, Øvrum L, Dahle T. Extending recall intervals - effect on resource consumption and dental health. Community Dent Oral Epidemiol 1992; 20:122-4.
• ³⁶
  Clarkson JE, Pitts NB, Fee PA, Clarkson JE, Pitts NB, Fee PA, et al. Examining the effectiveness of different dental recall strategies on maintenance of optimum oral health: the INTERVAL dental recalls randomised controlled trial. Br Dent J 2021; 230:236-43.
• ³⁷
  Thomson WM, Williams SM, Broadbent JM, Poulton R, Locker D. Long-term dental visiting patterns and adult oral health. J Dent Res 2010; 89:307-11.
• ³⁸
  Åstrøm AN, Ekback G, Ordell S, Nasir E. Long-term routine dental attendance: influence on tooth loss and oral health-related quality of life in Swedish older adults. Community Dent Oral Epidemiol 2014; 42:460-9.
• ³⁹
  Last JM. The iceberg: 'completing the clinical picture' in general practice. Lancet 1963; 2:28-32.
• ⁴⁰
  Morris JN. Uses of epidemiology. Br Med J 1955; 2:395-401.

Publication Dates

• Publication in this collection
  01 May 2023
• Date of issue
  2023

History

• Received
  27 Oct 2022
• Reviewed
  23 Feb 2023
• Accepted
  02 Mar 2023