Regional differences in infection control conditions in a sample of primary health care services in Brazil

International guidelines have pointed out the importance of the physical environment of health care facilities in preventing and controlling infection. We aimed to describe the physical environment of dental care facilities in Brazil in 2014, focusing on characteristics designed to control infections. Exactly 16,202 dental offices in the Brazilian Unified National Health System (SUS) participated in this survey. Trained researchers extracted information about the infection control characteristics of health facilities by using a structured instrument. We used data from 12 dichotomous questions that evaluated the wall, floor, sink and tap conditions, and the presence and condition of sterilization equipment. We calculated a score by summing the number of characteristics handled appropriately for infection control, which could range from 0 to 12. Hierarchical cluster analyses were developed. None of the 12 criteria were met by all the oral health teams. Only 208 (1.3%) dental offices correctly performed all 12-infection control practices. Two clusters, with different frequencies of structure for infection control in dental offices, were identified. South and Southeast regions had the highest frequencies for Cluster 1, with better structure of infection control in dental offices. Dental care facilities of oral health teams were not typically meeting the infection control guidelines regarding clinic design and equipment. Adherence to the guidelines varied among the Brazilian geographic regions.


Introduction
International guidelines have pointed out the importance of the design, planning, construction, refurbishment, and maintenance of the physical environment of health care facilities for preventing and controlling infection 1,2 .However, reports about the physical environment of dental care settings are scarce in the literature; most of the research on infection control in dentistry has focused on the knowledge and practices of dentists and dental assistants.Lack of knowledge and low compliance to control infection practice in dentistry are some worrisome results in these surveys 3,4,5,6,7,8,9,10,11.Studies, carried out in five towns from São Paulo State 12 , and two towns, one from Espírito Santo State 13 and the other from Rio Grande do Sul State 14 , Brazil, have shown that physical environment of dental care had problems regarding dental office sink, tap not draining properly, personal protective equipment, equipment for sterilization, among others.
In Belo Horizonte, capital of Minas Gerais State, Brazil, dental offices had frequent problems with personal protective equipment and equipment for sterilization 15 .A study carried out in a dental clinic from a Brazilian public university in Minas Gerais State have shown that the lack of physical barriers among dental offices has led to real possibility of cross-infection 16 .
In 2014, the Brazilian Ministry of Health led a second cycle of a program to assess and enhance the quality of primary health care (PHC), National Program for Improving Access and Quality of Primary Care (PMAQ-AB).This national survey of oral health teams in the Brazilian Unified National Health System (SUS) included evaluations of the physical environment of dental care clinics (http://dab.saude.gov.br/portaldab/ape_pmaq.php?conteudo=2_ciclo).According to the World Health Organization (WHO), the structure of health care facilities should be monitored on a national level.They also advise that basic organizational structure and building infrastructure are key factors for preventing and controlling endemic health care-associated infections and outbreaks 17 .
Therefore, we aimed to describe the physical environment of dental care facilities of oral health teams in Brazil in 2014, focusing on characteristics designed to control infections.Secondarily, we described the frequency of adequate infection control practices in clinic design in different Brazilian regions and compared dental and medical clinic practices within the same health centers.

Ethics statement
The study was submitted to and approved by the Brazilian Ethics Research Committee and by the Ethics Committee for Human Research of the Minas Gerais Federal University (protocol n.CAAE 02396512.8.0000.5149).We analyzed publicly available and anonymized data from the Brazilian Ministry of Health.

Population
We analyzed the second cycle of the PMAQ-AB (http://dab.saude.gov.br/portaldab/ape_pmaq. php?conteudo=2_ciclo).This survey was based on the classical quality of care framework by Donabedian, in which quality is evaluated using structure, process and outcome parameters 18 .Considering the importance of analyzing control infection practices on a national level 17 , we focused on these criteria in this study.Of 23,251 oral health teams in Brazil in January 2014, 19,946 (85.8%) participated in this second cycle of PMAQ-AB.Of these, 2,433 (10.5%) teams were excluded because they did not follow PMAQ recommendations, such as an adequate oral health surveillance system, presence of a dentist in the PHC unit, and dental equipment.Some PHC units included more than one oral health team.Exactly 16,202 dental offices, which included 17,513 oral health teams, were evaluated.All but 10 of these provided information about both oral health teams and their general medical (not dental) care.

Data collection
Data were collected from November 2013 to April 2014.The observation instrument was developed through a partnership between the Brazilian Ministry of Health and a committee of six academic institutions throughout the country.The evaluation of each criterion was made by trained researchers, all with higher education in the health area.A pilot study was performed prior to the fieldwork.The survey was conducted by using computer tablets and a program designed specifically for the PMAQ-AB.Survey questions included photographs indicating the health services to be evaluated and criteria for the evaluation of each item, according to the Ministry of Health guidelines 1,19 .The Ministry of Health assessed participation by telephone.
The questions were mostly dichotomous (yes/no).They evaluated dental office sink, sewage smell, tap not draining properly, personal protective equipment, sharps container, oven for sterilization, autoclave, floor and wall conditions, ventilation, and pack sealer.Some questions asked the number of pieces of a specific type of equipment; these responses were dichotomized (has at least one piece of such equipment -yes/no).Two originally separate questions about the presence of oven and autoclave were combined into one variable: "Presence of one oven for sterilization or one autoclave in good condition".There were two similar questions about the walls, floor, and ventilation, one regarding the dental office and the other regarding areas of the health center dedicated to general medical (not dental) care.For these two questions, we compared the proportion meeting the guidelines for the dental and non-dental areas of the clinic.

Statistical analysis
Each clinic setting received one point for each correct infection control criteria.Each clinic's score was the sum of the number of infection control criteria that were met (from 0 to 12).For example, if one clinic's score was 8, it meant that this service had met eight infection control criteria.Although the SUS may have geographical differences in the quality of its services 20,21 , for infection control, no previous results have shown any differences.Therefore, we used the multivariate agglomerative hierarchy technique and the between-groups linkage (average linkage) with squared Euclidian distances for the cluster analysis 22 .This exploratory data analysis technique, to organize observed data (in our case, from dental offices) into groups (clusters), is based on combinations of independent variables (in our case, structure for infection control in dental offices), and maximizes the similarity of cases within each cluster while maximizing the dissimilarity among groups.The clustering was based on the variables: presence of Personal Protective Equipment in sufficient quantity, presence of sharps container in sufficient quantity, presence of at least one oven for sterilization or one autoclave in good conditions, presence of at least one pack sealer in good conditions.The selection of these variables was based on the high risk of cross-infection of these parameters, as described in a previous study 13 .Data reduction made the management of subgroups easier.In our study, three sets of clusters (two to four) were formed from the 16,202 dental offices, and the choice of two clusters was based on improved understanding of the phenomenon (the characteristics of infection control) 22 .Finally, we compared the wall, floor and ventilation conditions between dental offices and non-dental areas in the health center, by calculating matched odds ratio (and 95%CI).All the analyses were performed using SPSS for Windows version 19.0 (IBM Corp., Armonk, USA).

Results
For the 12 infection control criteria, the frequency with which criteria were met varied from 4.8% to 99%.The criteria met in > 95% of clinics pertained to the sink, e.g., if it was unclogged and the tap draining properly.The rarest item among the recommended equipment in these Brazilian dental offices was a non-touch tap (Table 1).
Only 208 (1.3%) of dental offices met all 12 infection control criteria.Most met at least half the criteria (Figure 1).Thirty-four clinics had scores lower than five, each clinic had scores of zero and one, two clinics had a score of two, and 12 and 18 clinics had scores of three and four, respectively   (not visible in Figure 1).The first quartile and the median of this score were 9.The third quartile was 10.When we analyzed the dental offices with scores 9 and 10 (n = 9,768; 60.3%), we identified that among these, 99.8% had no clogged sink, 99.6% presented no useless sink, 99% had no sewage smell, in 98.5% the tap was not draining properly, 96.1% had Personal Protective Equipment in sufficient quantity, 95.2% had sharps containers in sufficient quantity, 93.7% had no mold in the sink, 87.6% had Cad. Saúde Pública 2017; 33 (11):e00072416 at least one oven for sterilization or one autoclave in good condition, in 89.6% the dental office had good ventilation, in 74.6% floor and walls of the dental office were smooth and had washable surfaces, 19.1% had at least one pack sealer in good condition, and 2.6% had non-touch tap.Two clusters, with different frequencies of structure for infection control in dental offices, were identified.Cluster 1 (n = 15,124) included the OHTs with better structure compared with Cluster 2 (n = 1,078).Of the Brazilian regions, South and Southeast regions had the highest frequencies for Cluster 1, with better structure of infection control in dental offices (Tables 2 and 3).
Compared with dental areas, non-dental areas in the health centers had 1.5 greater odds of floor and wall surfaces that were not smooth and washable (95%CI: 1.4, 1.6).Non-dental areas in the health center also had 2.3 greater odds of having poor ventilation (95%CI: 2.2, 2.5; Table 4).

Table 2
Presence of equipment for infection control accordingly to two clusters.Brazil, 2013-2014.

Discussion
We evaluated compliance with 12 different recommendations regarding the physical environment and equipment for controlling the spread of infectious organisms in Brazilian dental health clinics in 2013-2014, the most recent survey available.Among this large sample of 75% of primary oral health care clinics in Brazil, compliance with these criteria was far from ideal, and the degree of non-compliance varied across the different Brazilian regions.It is true that the dental offices were self-selected, which may have biased the results.Highly motivated teams that were more attentive to infection control practices were more likely to have joined the program, compliance could be even lower than that we observed in the sample.Over 2,400 of the ~7,000 non-participating clinics were excluded specifically because they were not compliant with the general PMAQ recommendations, which may also be not compliant with the recommendations considered here.
According to international protocols and scientific literature, ventilation is required because routine dental practice generates aerosols that can be contaminated with microorganism and chemicals 23 , and a well-functioning ventilation system dilutes the concentration of infectious aerosols 2 .The floor and walls should be smooth and washable to facilitate their disinfection 1,2,24 .Routine cleaning is necessary to ensure that the environment is visibly clean and free of dust and soil.Ninety percent of microorganisms present visible dirt in health care settings 25 .In the Brazilian clinics evaluated the high proportion of dental offices with poor ventilation or floor and walls that are not smooth or washable urges for investment on refurbishment to ensure the adequate protection of patients and dental staff.
The items with better compliance among the dental offices were those related to access to water and sewer in the dental office.For the most part, dental offices have functional sinks that can be operated properly.These characteristics are relevant because hand washing is considered one of the most important measures of infection control in health care settings 7,26 .However, the presence of mold in the sink of clinics and the near-absence of non-touch taps causes concern.The presence of water inside the sink contributes to the growth of microorganisms.The use of non-touch taps has been recommended 24,27 , despite seemingly conflicting data indicating that non-touch taps are a probable risk factor for Pseudomonas aeruginosa biofilm formation and subsequent transmission to patients 28 .Thus, if health services included non-touch taps, they would also need to be properly maintained 27 .
The lack of sufficient personal protective equipment of dental offices also causes concern.Without sufficient personal protective equipment, the oral health team might opt not to treat patients rather than to put clinicians at risk.These conditions may limit the access of patients to the already insufficiently accessible dental treatment in the Brazilian health system 29,30,31 , or put professionals and patients at biological risks 1,24 , if they treat patients without proper protection.
A high percentage of the dental offices had sterilization equipment, yet there were offices that lacked any such equipment, and the majority lacked a pack sealer in good condition.It is possible that sterilization occurred outside the clinic in a centralized sterilization facility.Nevertheless, there are some advantages to providing sterilization on-site in health centers 32 , such as efficiency, safety and costs.Because we did not have data to discern this possibility, further study should be conducted to identify whether the lack of such an important equipment in dental settings is an indicator of low quality.
The final criterion assessed concerned waste management, and most health teams had sufficient access to sharps containers for disposal of this type of biohazard waste.
Only a small percentage of dental offices complied with all 12 recommended practices.This result calls for more investment in the physical infrastructure of PHC in Brazil.Both patients and dental staff appear to be routinely at risk of cross-infection in dental health-care settings 17 .Despite being not very big, the regional differences in the distribution of frequencies of clusters demonstrated the same trends identified in another previous evaluation of PHC in Brazil 20 .The planning of healthcare facilities should also consider these inequalities so that more resources are applied in the regions with worse conditions.
For the two criteria that could be assessed outside the dental offices, adequacy of the floors, walls, and ventilation were similar in the dental areas and for most health centers.When compliance dif-Cad.Saúde Pública 2017; 33 (11):e00072416 fered, it was much more common that the dental clinics were more compliant than other areas of the health center, probably because oral health care has just recently been integrated into primary health centers in Brazil, i.e., the dental facilities are newer and possibly more up-to-date on environmental infection control practices 33 .A similar integration between dental care and medical care in PHC has taken place in other countries 34,35 .Some important environmental infection control practices were not evaluated by this survey, such as access to ultrasonic cleaners, disinfectants, antiseptics and more detailed information about personal protective equipment (e.g., gloves, masks, etc.).It is important to highlight that the descriptive approach in this research does not allow an evaluation of factors associated with the identified problems.Moreover, we are not able to objectively define the reliability and reproducibility of each measurement made by the trained researchers in this survey.Although we do not rank the relative importance of each infection control criterion, we selected the variables for the cluster analysis based on their high risk of cross-infection 13 .Despite these limitations, this is the most comprehensive evaluation of the environmental infection control practices in oral health clinics conducted in Brazil, and it is likely that no other country has conducted a survey of this size.Our results are quite similar to the local evaluations of structure of infection control in dental offices made in Brazilian small towns 12,13,14 and Belo Horizonte 15 .As far as it was possible to review the scientific literature, we did not identify other similar large empirical studies.Even in countries where clinics are supposed to meet these criteria in order to operate, as it is the case of Brazil and other countries 1,24 , our results indicate the need for frequent audits 36 .
Despite the indisputable advances in dental health policies in Brazil recently 33 , this evaluation suggests that even more effort will be required to provide and maintain adequate infection control environments and equipment in Brazilian oral health clinics.The continued audit of these criteria is advised both in Brazil and internationally.

Contributors
M. H. N. G. Abreu was responsible for the acquisition of data and for organizing the dataset, conceived and designed the study, responsible for analysis and interpretation of data; contributed to the writing of the manuscript revising it critically for important intellectual content, and approved the final version.V. L. S. Resende and J. R. Starr conceived and designed the study, analysis and interpretation of data; contributed to the writing of the manuscript revising it critically for important intellectual content, and approved the final version.K. H. Lee and A. T. G. Matta-Machado were responsible for the acquisition of data and for organizing the dataset, analysis and interpretation of data, contributed to the writing of the manuscript revising it critically for important intellectual content, and approved the final version.

Figure 1
Figure 1Distribution of the scores on a survey of compliance with infection control recommendations conducted in Brazilian primary oral health care clinics, 2013-2014.

Table 1
Prevalence of appropriate infection control equipment and physical environment in primary oral health care clinics.Brazilian Unified National Health System, 2013-2014 (N = 16,202).
* The parenthetical responses indicate which response was given a point for meeting the given criterion.

Table 3
Proportion of the two clusters according to Brazilian geographical regions in 2013-2014.

Table 4
Comparison between dental and non-dental areas of health centers regarding structural infection control practices.Brazil, 2013-2014.