Ethics statement

The study was submitted to and approved by the Brazilian Ethics Committee and by the Ethics Committee for Human Research of the Universidade Federal de Minas Gerais (protocol numbers CAAE 02396512.8.0000.5149 and 31525514.9.0000.5149). Individuals signed the informed consent, the participation was voluntary and they could refuse to answer any part of the questions. Publicly available, de-identified data from the Brazilian Ministry of Health were also analyzed.

Study design

This is an explanatory mixed-method research. The mixed explanatory design is a study of two distinct and sequential phases, in which the quantitative phase is followed by a qualitative phase. The researcher first collects and analyzes quantitative data and then collects and analyzes qualitative data that will help explain the results found in the quantitative phase [14]. That is, the second (qualitative) phase is built from the first phase and the two phases are connected in the intermediate and final stages of the research. The justification for this approach is that the quantitative phase promotes a general understanding of the research problem studied, while the qualitative phase refines and deepens the statistical results by exploring the vision of the social actors involved, contributing to clarify the gaps in the previous phase [15].

In this case, the quantitative phase was developed from analyzing secondary data collected by the Brazilian Ministry of Health as part of an evaluation pay for performance program. This dataset was used to describe the work of 12,403 OHTs and Item response theory (IRT) [16] was applied to develop a construct called “OHT performance”. This quantitative phase showed initial findings that were explored further in a qualitative phase. Thus, we integrated this mixed method study in two moments: 1) selecting participants for a qualitative phase by choosing a purposefully extreme case sampling, i.e., OHTs from Belo Horizonte metropolitan region placed in the extreme quintiles according to the IRT scores, 2) discussing the results by mixing the quantitative findings and qualitative findings. This gave us a better phenomenon understanding, which was the performance of OHT.

The flowchart of this mixed methodology is illustrated in S1 Fig.

Quantitative phase

This first research phase used secondary data from a national survey performed by the Brazilian Ministry of Health in 2012, i.e., the Program for Improving Access and Quality of Primary Care (PMAQ-AB). PMAQ-AB is a performance payment program evaluation aimed to improve the quality of PHC services in Brazil, including oral health care services [17].

Data for 12,403 OHTs which answered the questionnaire were used. IRT [16] was applied, to evaluate the psychometric properties of the oral health questions used in the PMAQ-AB questionnaire [13]. The IRT model relates the likelihood of an individual’s response to an item and its latent trait (or construct). A latent construct is a characteristic that cannot be measured directly, such as attitude, satisfaction or proficiency, that is, a latent, unobservable variable that will be estimated based on the responses given to each of the items considered by the respondents participating in the study [17,18]. In this study, the latent construct derived from the application of IRT was the performance of OHTs, measured from 20 questions about dental procedures performed in primary care (Table 1) [19]. The IRT model provided a score, a rating grade, for each OHT according to the responses given to the PMAQ-AB questionnaire, called OHT performance (provision of primary dental care procedures) This score could theoretically vary from -4 (OHT with lowest performance) to +4 (OHT with best performance). Each oral health team received a score that can range from lowest performance (lowest frequency of primary dental care procedures) to highest performance (highest frequency of primary dental care procedures). More details have been published previously [13].

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Questions about dental procedures in PHC used for creating a rating grade–OHT performance.

https://doi.org/10.1371/journal.pone.0217738.t001

Qualitative phase

The objective of this phase was to understand aspects that lead to the different levels of performance of OHTs identified in the previous phase [20]. Integration of this mixed-methods study happened during participant selection for qualitative interviews: the quantitative results guided the selection of the qualitative sample. An extreme case sampling strategy was used (opposite results) [21]. OHTs were selected from scores measured by IRT, which originates the “OHT performance” construct. We chose to sampling OHT from the metropolitan region of Belo Horizonte. The metropolitan region of Belo Horizonte, with around 6 million people, was composed by 33 municipalities (Baldim, Betim, Brumadinho, Caeté, Capim Branco, Confins, Contagem, Esmeraldas, Florestal, Ibirité, Igarapé, Itaquara, Itatiaiuçu, Jaboticatubas, Juatuba, Lagoa Santa, Mário Campos, Mateus Leme, Matozinhos, Nova Lima, Nova União, Pedro Leopoldo, Raposos, Ribeirão das Neves, Rio Acima, Rio Manso, Sabará, Santa Luzia, São Joaquim de Bicas, São José da Lapa, Sarzedo, Taquaraçu de Lima, Vespasiano) cities and towns with a diverse socioeconomic development levels. Thus, IRT scores from all OHT that participated in PMAQ-AB in those 33 cities (n = 65) were displayed in a range from lowest to highest performance. All OHTs in the extreme quintiles (less than −0.40 and greater than 0.70) that had maintained the same dentist since quantitative data collection were included (n = 10).

Data were collected through semi-structured interviews. An interview guide was applied that contained interviewee’s demographics, and questions about access to oral health services, integrality of actions and coordination of care, especially with regard to work with oral cancer prevention and prosthesis supply in PHC. This interview guide was piloted before fieldwork in two interviews tests in order to adjusting and refining questions and probes. The average duration of the interview was 30 minutes. The interviews were recorded with the permission of the participants, and were transcribed verbatim.

The demographic data of the professionals interviewed in the metropolitan region of Belo Horizonte (n = 10) and the performance of their respective OHTs are shown in Table 2. All OHT were due to provide community and clinical procedures, such as surgical, restorative and preventive procedures and dental emergencies at PHC level.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Demographic and health service data of interviewed’ dentists.

https://doi.org/10.1371/journal.pone.0217738.t002

Data analysis was conducted using deductive and inductive thematic analysis proposed by de Braun and Clarke [22]. Two authors (CR and MHA) coded those initial two interviews to check and adjust the initial codes in order to reflect upon the initial results [23]. Then, one of the authors (CR) immersed herself in the data analysis in an inductive way, seeking what emerged most from the interviewees’ speech After this first analysis, a second coding of the interviews happened in a deductive way, keeping in mind the theoretical reference of PHC [2] attributes and the objective of the qualitative phase of this mixed project, which was to respond in such a way that the qualitative results deepen the quantitative results.

Barriers to accessing oral health services in PHC

Challenges to accessing oral health services in PHC emerged in the interviews. It was observed in all OHTs, regardless of the level of performance, that there is a separation between users who access the service with urgent dental needs (they need immediate care, usually pain relief), and those who come with other complaints, whose consultation can be scheduled for another time. High-performance OHTs reported strategies to schedule treatment for users who come with urgent needs.

“I always catch him on the urgency treatment because it’s very rare I do not convince him to return to the treatment.” (T2H)

On the other hand, low-performance OHTs use other forms of access to dental treatment, such as waiting lists and telephone consultations. In some cases, patients wait up to 7 months to be called for care; in other cases, this appointment happens after years of waiting.

“It takes about 6, 7 years for us to call people in the wait list. And we are having a lot of troubles, because sometimes the telephone number changes and there is no contact with the patient.” (T5L)

In one case, the professional reported performing a risk classification as a way of organizing the first dental appointment.

“It has welcoming services every Tuesday. People come, we already look and sort by code 0, 1, 2, 3, 4, 5 and put the name in the notebook. There is a notebook for each code.” (T9L)

However, when the dentist was asked if he uses this risk classification criterion to organize the appointment and prioritization, the answer was negative. The risk classification is done because it is required by the municipality.

Welcoming services have been reported differently, and are often misunderstood as user triage in both high-performance and low-performing OHTs.

“We always make the welcome service on Tuesday. People come, get the examination, the classification, and their name goes in the notebook.” (T1H)

“The person arrives at the reception, the girl evaluates how many teeth she has to deal with, places her name in the notebook and the person waits for the vacancy.” (T9L)

Understanding of the welcoming service was also diverse. Some understand it as part of user triage, others as part of humanization strategies, and others as a way to involve all members of the PHC center in providing good information and welcoming users.

Another barrier to access was the exacerbated number of users for each OHT. This was a constant complaint in low-performing OHTs. This reality causes tensions between health staff and users.

Continuity of care

Another theme that emerged from interviews was longitudinal care. OHTs (high and low performance) reported that scheduling return appointments is used to track their users and prevent new oral diseases. Following users is seen as a way to prevent future oral problems.

“There’s a notebook. I’m going to write down, and every 6 months I call the patient back, which for me this return is sacred, and it works. I was surprised because very quickly I was able to control the oral health of the people here.” (T2H)

A high-performance team reported working with care throughout the life cycles, creating a link between professional and patient. It started following pregnant women and their babies.

“The baby is very easy, they come. I have treated a lot of mothers, so babies really come, it is spontaneous demand.” (T6H)

Actions to oral cancer care

In general, prevention and early detection of oral cancer are emphasized by high- and low-performance OHTs during campaigns to vaccinate the elderly against influenza. High-performance OHTs reported talking about oral cancer prevention throughout the year in the primary care center waiting room and in community centers.

“Two to three times a year we talk about oral cancer in the community centers. The primary care center reception has posters on oral cancer, and we teach self-examination.” (T8H)

In general, the network of services for monitoring suspected cases of oral cancer, which involves consultation with a specialist for biopsy, is deficient. The reference service for secondary care, in many cases, does not exist in municipalities. In the few municipalities that have the reference service, the OHT only directs the patient, and the specialist proceeds to monitor cases in isolation.

“Here, I used to follow the patients before, make this control … but now with the CEO (reference center), the stomatologist is requiring follow-up of those patients.” (T4L)

In high-performance OHTs, even when the municipality does not provide secondary care in the field of stomatology, dentists use their personal network to follow up on suspected cases of oral cancer.

“Especially edentates, I check to see those in question. If I suspect, I’ll send them to PUC (“Pontifical Catholic University of Minas Gerais”).” (T6H)

“… as I have connections in Belo Horizonte, I already direct them to Odilon Behrens. It’s informal, it’s just because I know it. I have acquaintances in Belo Horizonte.” (T7H)

Prosthetics in PHC

The provision of prostheses to patients in PHC is concentrated in high-performance OHTs.

“Every dentist makes his prosthesis as soon as he finishes the treatment. He goes to the social worker, he takes the referral he needs for the income.” (T6H)

In a high-performance OHT, the dentist provided a prosthetic service in a rural area, avoiding displacement of patients for long distances.

“Before everything went to the DSC, I do not know if it is because the dentist did not perform them, so now, as the material comes here, I will continue to make them here, so I do not need to be moving the patient to headquarters.” (T2H)

However, in another OHT, prosthesis provision does not happen in PHC, but in the DSC or other reference service of the municipality. However, in some municipalities, the prosthesis provision does not exist.

“It has a very large demand, but it still does not have the service. We feel bad, you do the treatment, the mouth is ready for the prosthesis, but then the patient comes back at the next appointment in 6 months, 1 year, and he’s still without anything, he’s toothless still.” (T3L)