1 INTRODUCTION

Since 1988, Brazil established a Unified Health System (‘Sistema Único de Saúde’—SUS), which was based on the principles of universality, integrality, and social participation and conceived to assure access to health actions and services, including oral health. Fifteen years later, the National Oral Health Policy was instituted and proposed a reorientation of the healthcare model, supported by an adaptation of the working system of oral health teams so that they include health promotion actions, treatment provision and rehabilitation. The intention was to rationally increase access to integrated oral health care, where ‘care lines’ (since childhood through adolescence, adulthood and old age) may have a centralized flow that includes the stages of welcoming, information giving, attendance and referral (including referral and contra-referral), in order to reach resolutive outcomes for dental problems.1, 2

The implementation of the dental speciality centres (CEO) is one of the actions of the National Oral Health Policy that are financed by the Ministry of Health. The CEOs offer dental clinical treatment complementary to procedures provided in primary care units, including periodontal surgery, endodontic treatment, minor oral surgeries, diagnosis and treatment support of oral lesions, and dental treatment for individuals with different types of disabilities.1, 2 According to the current legislation that regulates CEOs,3, 4 there are three models of CEO. CEO type I (three dental chairs); CEO type II (four to six dental chairs); and CEO type III (seven dental chairs or more). The CEOs must offer dental care 40 hours a week where the number of professionals vary according to the type of CEO. The CEOs must achieve the following production goals per month according to the type of CEO: Type I—35 endodontic procedures, 60 periodontal procedures, 80 oral surgery procedures and 80 basic procedures for people with disabilities; type II—these goals are 60, 90, 90 and 110 procedures respectively; and type III—95, 150, 170 and 190 procedures according to the listed specialties.

There has been an improvement in access to specialized dental treatments in Brazil in recent years though remarkable inequalities in dental services use persist between cities.5 Access to dental care at CEOs is immediate in some cities though the waiting time can reach more than 1 year in other cities.5, 6 Currently, there are more than a thousand CEOs throughout the country7 and endodontics is the most in demand specialized service in the dental speciality centres. This is because dental pain is one of the main reasons for seeking dental treatment.8 Endodontic treatment is expensive in private dental practice, and this type of treatment was not offered in public services in the majority of cities until recently. Therefore, the great pent-up demand for endodontic procedures has generated long waiting lists at CEOs. Even so, there are large dropout rates of those patients undergoing endodontic treatment.8 CEO organizational factors such as availability of dental instruments and equipment, waiting time, patient absenteeism and professional qualifications have been studied.5, 9, 10 Also, contextual characteristics, such as Human Development Index (HDI),6, 11-16 per capita GDP17 and organization of primary health care (PHC)12, 13, 15, 16 have been pointed out as relevant factors associated with the performance of the CEOs. For example, municipalities with low HDI and small populations were less likely to achieve the goals set by the Ministry of Health. This was probably due to the difficulties related to the scarcity of resources faced by these municipalities. Likewise, municipalities with higher average income showed higher rates for dental procedures, demonstrating the inequalities in access to dental services in Brazil.13

The identification of the factors that influence CEO's productivity may generate relevant information favouring the reorganization of secondary dental care in Brazil. Thus, the aim of this study was to investigate organizational and structural factors, socioeconomic indicators, and CEO and PHC work processes that influence the achievement of CEO endodontic goals recommended by the Ministry of Health.

2 METHODS

The Program for the Improvement of Access and Quality (PMAQ-CEO) is a national program with all CEOs in Brazil instituted by the Ministry of Health whose CEO census is conducted by partner universities. It is a national study of evaluation of dental units of secondary care aiming to induce the expansion of access and the improvement of quality in the dental speciality centres through the standardization of quality in these establishments so that they all become comparable nationally, regionally and locally. This program has four phases: agreement, development, external evaluation and re-contractualization.18 In our study, data from the external evaluation carried out in 2018 were used. Eighty-five trained external evaluators interviewed health professionals of the CEOs (CEO coordinator, dentists or the municipal health manager) using standardized questionnaires and made on-site observation in the Brazilian CEOs that adhered to the proposal (N = 1042) distributed by the regions of the country as follows form: 66 (6.3%) in the North, 410 (39.3%) in the Northeast, 364 (34.9%) in the Southeast, 131 (12.6%) in the South and 71 (6.8%) in the Midwest.

The dependent variable was the number of endodontics goals fulfilled per month by each CEO in the year 2018. The following thresholds regarding the number of endodontic procedures per month were used to ascertain whether the CEO reached the goals regulated by the Ministry of Health4: ≥35 endodontic procedures (CEO type I), ≥60 endodontic procedures (CEO type II) and ≥95 endodontic procedures (CEO type III) endodontic procedures. The list of endodontic procedures related to the goal is presented in the Appendix. The endodontic production per month in the year 2018 at each CEO was initially tabulated on TABWIN 4.14 Program. The endodontic production of each month was transformed into a dichotomous variable on whether each type of CEO was able to meet the pre-established goals. Finally, the number of months in which the goals were achieved was added up, resulting in the outcome measure ranging from 0 to 12.

The independent variables are described according to two levels and the respective dimensions in Table 1. The level 1 measures included 12 variables organized into three dimensions: access, organization and structure. The level 2 measures referred to eight contextual variables of organization of primary health care and sociodemographic indicators of the municipality where the CEO is located.

TABLE 1. Dimensions and variables of independent variables Dimensions Variables Analysis Data source 1st Level – Dental Speciality Center (CEO)-level Access CEO area of coverage Municipal or Regional PMAQ-CEO Access to CEO dental appointment Demand-led model and mixed-demand model or Exclusive referral model PMAQ-CEO Organization Protocol for endodontic procedures available Yes or No PMAQ-CEO Endodontic retreatment offered Yes or No PMAQ-CEO Percentage of patients' absenteeism Up to 20%; More than 20%; Didn't know PMAQ-CEO Workload for endodontic procedures

Number of hours available for endodontic procedures

Up to 40 h per week; More than 40 h per week.

PMAQ-CEO Waiting time for endodontic treatment

Number of days waiting for endodontic appointment.

Up to 45 days; More than 45 days

PMAQ-CEO Sessions for endodontic treatment Number of days to complete endodontic treatment PMAQ-CEO Structure Type of CEO

Type I; Type II; Type III

Based on the number of dental chairs and endodontic goals

PMAQ-CEO Number of endodontic equipments/dentist Number of specific endodontic equipments (apex locators and rotating instruments) divided by the number of dentists in the endodontics specialty. PMAQ-CEO Number of endodontic instruments Presence of a sufficient number of instruments and with good quality. Sum of the answers yes for the following instruments: arch for absolute isolation, endodontic cement for filling, gutta percha cone, staples for endodontics, rubber sheet, endodontic files, sheet perforator rubber, clamp holder). Yes or No PMAQ-CEO Proportion of qualified dentists in Endodontics (%) Proportion of dentists with training or specialization in endodontics in relation to the total number of dentists providing endodontic treatment. PMAQ-CEO 2nd Level—Municipality level Organization of Primary Health Care Coverage of first dental appointment

Proportion of the populational coverage of the first dental appointment in primary care. Greater coverage indicates a higher access to primary dental care.

Up to 15%; More than 15%

DATASUS Proportion of tooth extraction Proportion of tooth extraction in relation to primary dental care procedures. Greater proportion indicates a less preventive approach. Up to 8%, More than 8% DATASUS Coverage of supervised toothbrushing

Average supervised toothbrushing (expresses the proportion of people who had access to toothbrushing with fluoride dentifrice under guidance / supervision of a health professional aiming at the prevention of oral diseases, mainly dental caries and periodontal disease)

Up to 5%, More than 5%

DATASUS Sociodemographic indicators Human Development Index (HDI) Composite index using life expectancy, education and income to assess social development. Ranges from 0 to 1. PNUD Gini Index Assess the level of income inequality. Ranges from 0 to 1. PNUD Gross Domestic Product (GPD) per capita Total market value of goods and services per person IBGE Population Number of inhabitants IBGE Region of the country North, Northeast, South, Southeast, Midwest IBGE Dependent variable Number of achieved goals in endodontics in 2018 Ranges from 0 to 12 DATASUS

The variables were described through proportions and means (standard variation). First, Poisson regression was used to test the association between the independent variables at level 1 and the number of endodontics goals fulfilled per month using SPSS 23.0 (IBM Corp.). Multivariate multilevel Poisson regression analysis19 was used to estimate prevalence ratios (PR) and 95% CI of variables related to dental speciality centres (CEO) (first-level variables) and municipal factors (second-level variables) with the outcome. Initially, the null model was assessed to verify the relevance of using multilevel analysis. The first model included the first-level variables with p-values <.20 in the crude analysis. They were included and retained in the multivariate statistical modelling. The second-level variables were inserted in model 2 using the same statistical criteria (p-value <.20). The final model consisted of the CEO and municipal variables with p-values <.05 in model 2. The percentage of variance explained by the contextual level (partition coefficient of variance) was calculated by comparing the variance of the final model with the null model. Multilevel statistical analysis was performed on STATA 12.0 (College Station, TX, USA).

The PMAQ-CEO Project18 was approved by the Research Ethics Council of the Federal University of Pernambuco (Registration 23458213.0.1001.5208).

3 RESULTS

Overall, the dental speciality centres (CEO) achieved the endodontic goals in less than 4 months of the year (3.63 months SD = 4.14). Most of the CEOs provided dental treatment for the population of the city (68.6%), book dental appointments by referral (61.1%), have a protocol for endodontic procedures (85.7%), offer endodontic retreatment (78.3%) and have patients' absenteeism rate less than 20% (49.9%). Nearly 60% of the CEOs provided less than 40 h per week for the specialized endodontic treatment (59.7%) and their users waited less than 45 days for endodontic treatment (59.1%). The endodontic treatments were completed in 1.8 days (SD = 1.0) on average. Most of CEOs were type II (46.8%). The average number of endodontic equipment per dentist was 0.8 (SD = 0.8), and 89.4% of the CEOs had all listed instruments in sufficient number. In addition, a proportion of 0.8 (SD = 0.3) dentists registered at CEOs had training in endodontics (Table 2).

TABLE 2. Descriptive analysis N (%) PR (IC95%) p Independent variables (1st level) Access CEO area of coverage Municipal 715 (68.6) 1 Regional 327 (31.4) 1.10 (0.95–1.27) .190 Access to CEO dental appointment Demand-led model and mixed demand 405 (38.9) 1 Exclusively referenced demand 637 (61.1) 1.31 (1.13–1.52) <.001 Organization Protocol for endodontic procedures available Yes 893 (85.7) 1 No 149 (14.3) 0.79 (0.63–0.99) .037 Endodontic retreatment offered Yes 816 (78.3) 1 No 226 (21.7) 0.82 (0.68–0.98) .030 Percentage of patients' absenteeism Up to 20% 520 (49,9) 1 More than 20% 252 (24,2) 0.87 (0.73–1.02) .091 Didn't know how to answer 270 (25.9) 0.66 (0.55–0.80) <.001 Workload for endodontic procedures Up to 40 hours per week 622 (59.7) 1 More than 40 hours per week 420 (40.3) 1.44 (1.26–1.65) <.001 Waiting time for endodontic treatment Up to 45 days 577 (59.1) 1 More than 45 days 400 (40.9) 0.91 (0.78–1.05) 0.197 Mean (SD) PR (IC95%) p Sessions for endodontic treatment * 1.8 (1.0) 0.81 (0.69–0.95) .011 Structure N (%) PR (IC95%) p Type of CEO Type III 136 (13.1) 1 Type II 488 (46.8) 0.87 (0.70–1.10) .243 Type I 418 (40.1) 1.17 (0.94–1.46) .149 Sufficient number of instruments Não 110 (10.6) 1 Sim 932 (89.4) 1.21 (0.93–1.57) .166 Mean (SD) PR (IC95%) p Proportion of qualified dentists in Endodontics 0.8 (0.3) 1.00 (0.99–1.00) .371 Number of endodontic equipment/dentist 0.8 (0.8) 1.07 (1.00–1.15) .063 Dependent variable Mean (SD) Number of achieved goals in endodontics in 2018 3.6 (4.1)

The multilevel Poisson regression (Table 3) showed the number of endodontics goals fulfilled per month was associated with variables related to the dental speciality center level (first level) and one variable of the organization of primary health care. CEOs with more than 20% of patient absenteeism achieved 26% lower number of goals than those with patient absenteeism rate lower than 20%. CEOs who have availability of endodontists for more than 40 h a week were 1.95 times more likely to reach the goals than those with less workload in endodontics specialty. CEOs with a waiting time for procedures greater than 45 days achieved a number of goals 31% lower. CEO type I and CEO type II showed 2.10 and 1.20 higher likelihood to reach the number of goals of the endodontics specialty than CEO type III. The number of endodontic instruments in sufficient number was positively associated with the achievement of goals.

TABLE 3. Poisson multilevel regression analysis for the number of achieved goals in endodontic treatment in Dental Specialty Centers, according to individual and contextual variables Variables Model 1 Model 2 Final Model

PR

(95%CI)

PR

(95%CI)

PR

(95%CI)

1st Level (CEO) Access CEO area of coverage Municipal 1 Regional 1.00 (0.84–1.21) Access to CEO dental appointment Demand-led model or mixed demand 1 Exclusively referenced demand 0.97 (0.81–1.17) Organization Protocol for endodontic procedures available Yes 1 No 1.20 (0.90–1.58) Endodontic retreatment offered Yes 1 No 0.88 (0.70–1.10) Percentage of patients' absenteeism Up to 20% 1 1 1 More than 20% 0.77 (0.64–0.93) 0.74 (0.61–0.90) 0.74 (0.62–0.90) Didn't know how to answer 0.57 (0.45–0.72) 0.61 (0.47–0.78) 0.61 (0.48–0.79) Workload for endodontic procedures Up to 40 h per week 1 1 1 More than 40 h per week 2.12 (1.77–2.54) 1.99 (1.64–2.43) 1.95 (1.61–2.37) Waiting time for endodontic treatment Up to 45 days 1 1 1 More than 45 days 0.78 (0.65–0.93) 0.70 (0.58–0.86) 0.69 (0.57–0.84) Structure Type of CEO Type III 1 1 1 Type II 1.10 (0.90–1.35) 1.20 (0.96–1.50) 1.20 (0.96–1.50) Type I 1.94 (1.50–2.50) 2.07 (1.56–2.74) 2.10 (1.60–2.77) Sufficient number of instruments Não 1 1 1 Sim 1.51 (1.14–2.01) 1.44 (1.09–1.91) 1.46 (1.10–1.93) Number of endodontic equipments/dentist 1.08 (0.97–1.22) 1.05 (0.93–1.19) 2nd Level (municipal) Organization of Primary Health Care Coverage of first dental appointment Up to 15% 1 More than 15% 0,83 (0.59–1.17) Proportion of tooth extraction More than 8% 1 1 Up to 8% 1.37 (1.04–1.81) 1.28 (0.98–1.68) Coverage of supervised toothbrushing Up to 5% 1 1 More than 5% 2.19 (1.02–4.73) 2.29 (1.07–4.93) Sociodemographic indicators HDI 0.15 (0.01–2.85) GINI index 1.79 (0.14–22.76) GDP per capita 1.00 (1.00–1.00) Population 1.00 (1.00–1.00) Region of the country Midwest 1 1 South 1.11 (0.62–1.98) 1.03 (0.59–1.80) Southeast 0.73 (0.44–1.21) 0.71 (0.43–1.17) Northeast 0.54 (0.30–0.95) 0.62 (0.37–1.04) North 0.89 (0.46–1.74) 1.00 (0.53–1.91) Random effects Null Model Variance (95%CI) 2.055 (1.749–2.413) 1.823 (1.565–2.157) 1.689 (1.408–2.026) 1.707 (1.424–2.046) Changes in variation (%)1 11.2 17.8 16.9 LR test (χ2. p-value) 2390.66 (<.001) 1950.87 (<.001) 1548.76 (<.001)