1 INTRODUCTION

As the health sector uses patient-centered care models, managers and health professionals must understand user satisfaction and how they affect their practices.1 Identifying user satisfaction is essential to assess the quality of oral health services and dental treatment and health service users can point out the necessary improvements to the sector.2

However, user satisfaction and experiences are a significant part of identifying the quality of services rarely included in quality assessments.3 These analyses can provide information to be used as outcome indicators, pointing out the strengths and limitations of health programs.4

In Brazil, oral health care has historically presented an exclusive model, marked by low population coverage,5 in a restrictive political setting of low implementation of new oral health services.6 However, in 2003, the National Oral Health Policy, called “Brazil Sorridente”, improved the access to dental services and oral disease indicators, mainly due to the expanded PHC and access to specialized treatments via Dental Specialty Centers (CEO).7 The CEO are secondary care services that must offer, at a minimum, the specialties of periodontics, endodontics, services for patients with special needs, oral diagnosis, and minor oral surgery, to users referred by primary care.8

Aiming at improving the access and quality of CEO in Brazil, the Ministry of Health proposed the Program for the Improvement of Access and Quality for the Dental Specialty Centers (PMAQ-CEO) to collect information on CEO access and quality based on established standards. The PMAQ-CEO aimed to measure the effects of health policies, support decision-making, ensure the transparency of the Unified Health System (SUS) management processes, give visibility to the results achieved and strengthen social control and the SUS focus on users.9

Patient satisfaction perception surveys are significant and essential sources of information to identify gaps and develop an effective action plan to improve the quality of healthcare actions.10 Previous user satisfaction investigations were carried out, but they used only one variable to define user satisfaction with the service,11-13 or they used specific questionnaires to assess the perception of service quality by the user,14, 15 besides presenting satisfaction in gradations. However, the perception of satisfaction must be understood in its complexity with evaluating services through the tripod structure, process, and outcome.16

The development of a Service Satisfaction Index (SSI) that presents structure and process information is a valuable record that managers can use as a parameter to improve services.1 From the analysis of the SSI, one can identify users who are very satisfied with all the attributes investigated (maximum patient satisfaction) and possible factors associated with this condition that can contribute to the formulation of service improvement strategies. Given the above, this present study aimed to investigate the perception of CEO user satisfaction in Brazil through a Service Satisfaction Index (SSI) and identify the associated factors with maximum user satisfaction.

2 METHODS 2.1 General study design

This quantitative, cross-sectional, evaluative study used data extracted from the second cycle of external evaluation of PMAQ-CEO carried out in 2018, available at http://aps.saude.gov.br/ape/pmaq/ciclo2ceo. The external evaluation was held in 901 Brazilian municipalities that had implemented 1097 CEO. As the external evaluation is through census, 1042 of the total 1097 CEO were investigated, and 55 CEO were excluded as they were closed, disabled, or because they refused to receive the visit. Ten users participated in the interviews in each CEO unit investigated, corresponding to module III of the CEO quality assessment instrument, regardless of the CEO type. In total, 10391 CEO users participated in the study, selected by convenience sampling. Users’ anonymity was assured when completing the instrument. Procedures for collecting primary data have been published in a previous study.9

2.2 Service Satisfaction Index

The Service Satisfaction Index (SSI) was developed in this study to create a proxy variable regarding patient satisfaction with the dental service provided by the CEO. The variable consisted of an aggregation of the patients’ perception regarding (i) the assistance provided by reception personnel; (ii) the comfort during dental consultation, (iii) the infrastructure; and (iv) the assistance provided by the dentist. For each perception question (i to iv), patients’ answers were scored 1 to 5, where 1 was very unsatisfied and 5 very satisfied. Based on that, the SSI could vary between 4 to 20 points. A score of 20 would mean the patient was very satisfied with all attributes. The authors developed the SSI based on their own discussions regarding the theme and also on the dataset. From the theoretical perspective, its construction was based on the Donabedian quality assessment model16 and on previous studies that investigated the satisfaction of CEO users in the first cycle of the PMAQ-CEO.17

2.3 Associated factors

Independent variables were: sex, age, education, participation in a government assistance program, CEO type (Type I - up to 3 dental chairs; Type II - 4 to 6 dental chairs; Type III – 7 or more dental chairs), waiting time for consultation, completed treatment, whether users’ dental treatment at the CEO has already been interrupted due to lack of material (inputs), whether the facilities are in good condition, clean and with a sufficient number of chairs, whether users would switch CEO. The theoretical model used in this study was based on a previous study17 and presented in Figure 1.

Theoretical model based on the Cavalcanti, Cardoso, Padilha, 2018.17 2.4 Statistical analysis

Data were analyzed using the Statistical Package for Social Sciences (IBM SPSS, v. 24, IBM). Firstly, data were analyzed descriptively. We employed the exploratory factorial analysis to check statistical correlations between the attributes and the SSI. Multiple correspondence analysis estimated the percentage of variance explained by SSI attributes. We adopted Cronbach's alpha to estimate internal consistency, and the Kaiser-Meyer-Olkin (KMO) measure to estimate sample adequacy. A binary logistic multiple regression model was built according to theoretical model presented in Figure 1. The logistic regression model estimate the chance of independent variables influencing maximum patient satisfaction (SSI=20). A hierarchical approach was adopted to check the influence of sociodemographic profile, the previous experience within the CEO, and the opinion regarding the local infrastructure on SSI. The model was adjusted using the stepwise backward Wald method. Within each step, (Variables with p < .05 were considered statistically significant. Odds ratio (OR) and 95% confidence intervals (95% CI) were used to report the models.

3 RESULTS

The convergence between patients’ perceptions and the SSI is presented in Table 1. The attributes of SSI explained 98.7% of the total variance. Internal consistency was considered satisfactory (Cronbach's alpha = 0.817), as was the sample adequacy (KMO measure = 0.761). The SSI descriptive frequency is presented in Table 2. The SSI had a mean of 17.7 and a standard deviation of 1.8. The maximum satisfaction frequency was 21.1%.

TABLE 1. Bivariate Spearman's correlation matrix between the Service Satisfaction Index (SSI) and its attributes Service Satisfaction Index (SSI) Perception regarding the assistance provided by reception personnel Perception regarding the comfort during dental consultation Perception regarding the infrastructure Perception regarding the assistance provided by the dentist Service Satisfaction Index (SSI) Spearman's rho 0.721 0.726 0.782 0.754 p-value <.001 <.001 <.001 <.001 Perception regarding the assistance provided by reception personnel Spearman’ rho 0.721 0.411 0.369 0.475 p-value <.001 <.001 <.001 <.001 Perception regarding the comfort during dental consultation Spearman’ rho 0.726 0.411 0.406 0.425 p-value <.001 <.001 <.001 <.001 Perception regarding the infrastructure Spearman’ rho 0.782 0.369 0.406 0.463 p-value <.001 <.001 <.001 <.001 Perception regarding the assistance provided by the dentist Spearman’ rho 0.754 0.475 0.425 0.463 p-value <.001 <.001 <.001 <.001 TABLE 2. Frequency distribution of Service Satisfaction Index (SSI) scores in Brazil (N = 10391) Service Satisfaction Index scores Absolute frequency Percent (%) Cumulative percent (%) 4 1 0.0 0.0 5 1 0.0 0.0 6 3 0.0 0.0 8 3 0.0 0.1 9 5 0.0 0.1 10 7 0.1 0.2 11 15 0.1 0.3 12 69 0.6 1.0 13 118 1.1 2.1 14 268 2.4 4.7 15 637 5.7 10.8 16 1784 16.1 28.0 17 1579 14.2 43.2 18 1694 15.3 59.5 19 1861 16.8 77.4 20 2346 21.1 100.0 Total 10,391 100.0 100.0

Table 3 shows the distribution of individuals according to the satisfaction score. SSI distribution was analyzed concerning the sociodemographic profile, the previous experience within the CEO, and the opinion regarding the local infrastructure.

TABLE 3. Frequency distribution of maximum satisfaction patients (SSI = 20) and less satisfied individuals (SSI < 20) according the sociodemographic profile, the previous experience within CEO, and the opinion regarding the local infrastructure in Brazil (N = 10,391) Independent variables Less satisfied (SSI < 20) Maximum satisfaction (SSI = 20) N % N % Sex Male 2594 25.0 782 7.5 Female 5451 52.5 1564 15.1 Age Up to 40 years-old 3851 38.6 961 9.6 Over 40 years-old 3845 38.6 1307 13.1 Education level Up to high school 6902 66.4 1925 18.5 Superior education 1143 11.0 421 4.1 Beneficiary of an income transfer program Yes 2070 20.2 437 4.3 No 5869 57.4 1857 18.1 CEO type I (up to 3 dental chairs) 3189 30.7 991 9.5 II (4 to 6 dental chairs) 3819 36.8 1050 10.1 III (7 or more dental chairs) 1037 10.0 305 2.9 Waiting time until first consultation Up to one week 2390 23.6 808 8.0 More than one week 5458 53.8 1487 14.7 Conclusion of dental treatment Yes 2669 25.7 874 8.4 No 5376 51.7 1472 14.2 Discontinued the dental treatment Yes 1305 12.6 158 1.5 No 6740 64.9 2188 21.1 Feels that CEO has good conditions of use Yes 6597 63.5 2303 22.2 No 1448 13.9 43 0.4 Feels that CEO has good cleaning Yes 7294 70.2 2280 21.9 No 751 7.2 66 0.6 Feels that CEO has adequate number of chairs Yes 5505 53.0 2003 19.3 No 2540 24.4 343 3.3 Would rather be assisted in another CEO Yes 1133 10.9 63 0.6 No 6912 66.5 2283 22.0

The sociodemographic profile, the previous experience within the CEO, and the opinion regarding the local infrastructure have influenced the likelihood of maximum satisfaction (SSI = 20), according to the binary logistic multiple regression model (Table 4). Factors that expressed a more significant influence on the SSI were feeling that CEO has good conditions of use (OR = 7.05; CI 95%: 5.15–9.65), not wishing to be assisted by another CEO (OR = 4.17; CI 95%: 3.12–5.57), and not discontinuing the dental treatment before (OR = 2.05; CI 95%: 1.70–2.47).

TABLE 4. Binary logistic multiple regression model that estimated the chance of being maximum satisfaction (SSI = 20), according the sociodemographic profile, the previous experience within CEO, and the opinion regarding the local infrastructure in Brazil (N = 10,391) OR 95% CI Lower Upper Age (up to 40 years-old) 1.31 1.18 1.46 Education level (superior education) 1.30 1.14 1.49 Beneficiary of an income transfer program (No) 1.28 1.13 1.46 CEO typea I (up to 3 dental offices) 1.02 0.87 1.20 II (4 to 6 dental offices) 0.92 0.79 1.08 Waiting time until first consultation (up to 1 week) 1.17 1.05 1.31 Conclusion of dental treatment (Yes) 1.08 0.97 1.20 Discontinued the dental treatment (No) 2.05 1.70 2.47 Feels that CEO has good conditions of use (Yes) 7.05 5.15 9.65 Feels that CEO has good cleaning (Yes) 1.38 1.04 1.85 Feels that CEO has good number of chairs (Yes) 1.74 1.52 2.00 Would rather be assisted in another CEO (No) 4.17 3.12 5.57 Note OR: Odds ratio; 95% CI: 95% confidence interval. 4 DISCUSSION

This study proposed to evaluate the CEO's user satisfaction through an SSI indicator, one of the differentials of this research. The construction of the SSI indicator was based on the satisfaction of individuals with reception, comfort, infrastructure, and care provided by the dentist. All these attributes were strongly correlated with the indicator created by the study, which suggests high internal consistency. Furthermore, an SSI score of 20 was considered maximum satisfaction with the CEO’s service, and the factors associated with this maximum satisfaction were investigated.

The association of maximum satisfaction occurred for reports of individuals aged up to 40 years with higher education, in agreement with investigation carried out in the United States18 in which those with higher education reported greater satisfaction with the service, and with research carried out in Bangladesh, which demonstrates greater satisfaction among users over 20 years of age.19 Other authors20 evaluated satisfaction indicated by the Oral Health Survey SB2010, through an ordinal qualitative variable, and observed very good satisfaction or good satisfaction in the perception of adults and older adults. Thus, demographic and socioeconomic variables can influence user satisfaction with health services.

The migration of patients using private services to public services in Brazil may have contributed to the findings of this research. The satisfaction observed by users may represent the improvement in oral health services in secondary care in Brazil, pointed out by the majority as having positive similarities with private services.21 In a study carried out in Bangladesh, the authors observed that patients from the private sector perceive better performance in health services than patients from the public sector.19

There was an association between maximum satisfaction and shorter waiting time, also observed in other national15, 22 and international23, 24 studies. In a study carried out in Italy, waiting time and accessibility barriers harmed user satisfaction.25 Moreover, users whose dental treatment was not interrupted due to lack of material increased twofold the likelihood of maximum user satisfaction. The record that the CEO’s facilities are in good condition for use implied a sevenfold likelihood to find maximum satisfaction with the service and is the factor with the most significant association.

It is suggested that dental treatment delay contributes to lower users’ satisfaction. Frequently, treatment delay is generated from several appointments to complete the dental treatment, and when treatment is interrupted due to a lack of material. The influence of those factors on users’ satisfaction was not reported before. Additionally, a clean environment, with enough chairs and in a good condition, reduces the stress of waiting for the service on the scheduled day and contributes to an understanding of a favorable environment, which also suggests that the equipment is under constant maintenance, favoring the belief in the continuity of care provided.18, 26-29

Moreover, when reporting not wanting to switch CEO, most users showed maximum satisfaction with the CEO they attended and reinforced those points related to maximum satisfaction influence the desire to remain assisted in that unit. Personal bonds generated by health care in the professional-user relationship, and explanations or guidance on treatment, resoluteness, and care guarantee, can positively influence satisfaction.30-35

A study limitation could be the definition of the target population. The inclusion criterion was to be in the CEO at least for the second visit, and the interview took place at the health unit itself. This condition can direct responses of greater satisfaction through gratitude bias, mainly because use is the most critical factor, and the fear of losing it can influence responses. Furthermore, the sample was chosen by convenience, so the findings may not represent all secondary care users. Despite the limitations, this study presents information on a large number of users. Its findings reaffirm associated factors found by previous studies,15, 23, 24, 26 as they present the SSI indicator and new associated factors that can support further studies and contribute to the planning and decision-making of managers and professionals.

The index SSI developed in this study identified user satisfaction with the service, with waiting time of up to one week having a twofold likelihood to achieve maximum satisfaction. Not having the treatment interrupted, as well as considering the CEO in good cleanliness and use conditions also influenced the opinion of maximum satisfaction. Maximum satisfaction was associated with the age of up to 40 years and higher education level. Therefore, these findings point to managers and other actors engaged in developing efficient actions in scheduling appointments, making chairs and materials available, keeping the environment clean, providing maintenance of equipment, supporting human resources, and linking professionals to patients, to obtain maximum user satisfaction.

ACKNOWLEDGEMENTS

Thanks to the Collaborating Center of the Ministry of Health - Federal University of Pernambuco - CECOL / UFPE. Funding was provided by the National Health Fund (FNS).

CONFLICT OF INTEREST

The authors declare that there is no conflict of interests regarding the publication of this paper.