Factors perceived by health professionals to be barriers or facilitators to caries prevention in children: a systematic review

The search strategy was designed in collaboration with a medical librarian. Searches were conducted using two major biomedical databases (PubMed and Web of Sciences), as well as a francophone database targeting publications in the humanities and social sciences (Cairn). MeSH terms were used on Medline, and free text terms were used on Web of Science and Cairn (see Additional file 2). The search was conducted with no initial time restriction to March 2021. Since all of the authors are native French speakers, francophone literature that has gone through a complete editing procedure has also been reviewed in addition to articles written in English. No search of grey literature was undertaken. The collection was completed with hand searches of the reference lists of all selected studies.

After duplicates were removed, two researchers (GL and EM) independently screened the titles and abstracts. Articles not considered relevant to the topic were eliminated, and studies that met the inclusion criteria were collected in full text (consensus of the 2 researchers). In case of disagreement, a third reviewer (MC) was consulted for arbitration. To be included, studies had to focus on barriers and/or facilitators to caries prevention in children and include health professionals as study participants. In this work, barriers and facilitators were defined as factors that help or hinder the implementation of caries prevention with children by health professionals. Qualitative, quantitative, or mixed methods could be included. Conversely, because they were deemed irrelevant to identify barriers and facilitators to caries prevention, guidelines, editorials, and protocols were excluded. To ensure that articles do not appear more than once in the analysis, literature reviews, meta-analyses, and systematic reviews were also eliminated.

In the clinical setting, oral health prevention does not refer to one behavior but to a set of behaviors that health professionals can implement in their clinical practice. It includes screening, risk assessment, counseling, fluoride varnish application, pits and fissure sealants and dental referral. In this regard, caries prevention consists of a comprehensive approach for children. It has been considered that specific prevention measures (fluoride varnish application, fissure sealant, etc.) could not be regarded as a comprehensive prevention strategy and, as such, do not match to the approach the authors wished to take on this issue. Also, they were concerned to include very specific factors that may conflict with those selected as part of a comprehensive approach to prevention. For these reasons, the research team excluded specific studies that covered only one aspect of oral health prevention.

The assessment of the reporting quality of the studies’ methodology was conducted independently by two reviewers (GL and EM) using two validated checklists. For qualitative studies, the authors used the Consolidated Criteria for Reporting Qualitative Research (COREQ) [33], a 32-item checklist organized into 3 domains. For quantitative studies, quality assessment was appraised with the strengthening the reporting of observational studies in epidemiology (STROBE) reporting guidelines for cross-sectional studies [34] using a 22-item checklist. The qualitative analysis did not influence study inclusion, but it provided a critical framework for the articles reviewed.

Data were extracted using the same method as for the study selection. Full texts were analyzed independently by two researchers (GL and EM) with arbitration by a third team member (MC) in case of disagreement. The data collection template included the year of publication, country of the study, primary and secondary objectives of the study, study design, sample size, profession of respondents and main factors identified as barriers or facilitators to caries prevention. For quantitative studies, a factor was considered a barrier when at least 10% of participants reported it as such. This threshold value was decided by consensus of the research team members who considered 10% to be a population-wide significant portion. For qualitative studies, data were collected from participant quotations. A factor was included when both investigators agreed that it was explicitly and unambiguously defined in the text.

Perceived barriers or enablers were classified according to a three-stage process. First, data were sorted based on the 14 domains of the theoretical domains framework (TDF) [35]. The TDF is a comprehensive framework that synthesizes 33 psychological theories related to behavior change and is designed to understand implementation problems. The TDF can be used to conduct various types of studies, including qualitative research, questionnaire studies, evaluations of randomized trials or systematic reviews. To ensure the proper use of the TDF in this systematic review, the authors decided to work according to Atkins and All’s guide [36]. Second, findings were classified depending on whether they were clinician-, patient- or organizational-related. Finally, for each domain, data were reorganized by theme into several subcategories. After data extraction, several meetings with the research team members were organized to synthesize the main factors of the initial analysis into a single framework. This framework was designed using a consensus method.

The database research found 1768 references (Fig. 1). After duplicates were removed, 1710 studies were screened based on the title and abstract. Of these, 31 were selected for full text examination. After hand searches of the reference lists of all selected studies, 3 additional articles were selected. Ultimately, full text screening was conducted for 34 studies. Eight were excluded (see Additional file 3), and 26 met the inclusion criteria, among which half were qualitative studies [30, 37‐48] and half were quantitative studies [24, 25, 49‐59]. Methodological quality assessment did not lead to further exclusion of any references.

Among the included papers, 25 were reported in English and one in French [46]. All were published between 2003 and 2019 in nine different countries, including the USA [24, 38, 39, 42, 44, 47, 51, 54, 56, 57], UK [30, 37, 41, 45, 53], Canada [25, 58, 59], Australia [40, 43], Saudi Arabia [49, 50], France [46], Peru [55], Thailand [48] and Taiwan [52] (Table 1). Half of the articles were qualitative studies using individual interviews, focus groups or a combination of both. The other half were cross-sectional studies using a self-report questionnaire. The included studies involved a wide range of health professionals. The most frequently represented professions were dentists (n = 12), pediatricians (n = 11), nurses (n = 9), physicians (n = 5), and dental hygienists or dental nurses (n = 3).

For the qualitative studies retained, quality assessment using the COREQ checklist showed significant differences in terms of methodological quality (Additional file 4). Overall, the studies provided sufficient detail on aims, the participant selection process, data analysis and reporting. In contrast, more than half of the studies provided poor or no information on the use of a theoretical framework, interviewers’ characteristics, the relationship between the research team and the participants and data saturation. The assessment of quantitative studies showed good methodological quality since most of the items from the STROBE checklist were mentioned for all studies (Additional file 5).

A wide range of factors have been identified by health professionals as barriers or facilitators to caries prevention in children (Table 2). The factors were sorted into 3 main categories: clinician-related factors, patient-related factors, and organizational-related factors. For each category, factors were then classified according to the 14 TDF domains.

The main factors from the TDF analysis were extracted and classified into 5 categories: “Political and social context”, “Training and health systems”, “Organization of health care facilities”, “Team organization” and “Health professionals”. Figure 2 highlights the fact that barriers to and facilitators of caries prevention in children involve all stages of the health care system, from public policies to health professionals’ opinions.

In this systematic review of 26 studies, health professionals reported many challenges to caries prevention in children. The barriers identified in this systematic review are varied and systemic and involve all stages of the health care system: the political and social context, health system organization, health care facilities organization, health care team organization and health professionals’ skills and opinions. Health professionals frequently point to organizational barriers, particularly lack of time, poor material resources, inadequate funding or reimbursement, insufficient oral health training and difficulty accessing dental care. Parents would constitute another obstacle to children’s oral prevention. Due to their lack of knowledge, parenting skills, and health literacy, they may not recognize their child’s oral health as a priority. Health professionals are also questioned because of their lack of dental knowledge, lack of self-confidence, and unclear understanding of their role in promoting oral health.

To our knowledge, this is the first systematic review to investigate health professionals’ perspectives on the barriers to and facilitators of caries prevention in children. In 2017, a scoping review on a related topic was conducted by Harnagea et al. [60] to identify the factors that influence the integration of oral health into primary care. The main barriers identified in their study were very similar to those found in our research: a lack of political leadership and health care policies, lack of time, lack of staff, limited knowledge and competencies and insufficient oral health education. By applying a multilevel analysis theoretical framework [61], the authors also demonstrated, as we did in our study, that the various factors mentioned by health professionals involved all stages of the health care system (macro, meso and micro levels).

Previous quality improvement projects have been conducted to increase the delivery of oral health care and prevention within clinical practice. By improving payment for preventive dental services, some of these projects have sought to address one of the most widely cited barriers in our review [62‐65]. These studies show a positive but limited impact of funding measures on the provision of preventive dental care. Although more generous payment policies are needed, they are not sufficient to ensure the widespread implementation of preventive services at the organizational and practical levels. Other initiatives focus on improving health professional oral health education, which was also identified as a major barrier in our research [66, 67]. These studies report a moderate effect of oral health training on the provision of preventive dental services. Some authors [24, 51] even state that there is no connection between prior training or knowledge in the field of oral health and the delivery of preventive care. Health professionals’ training is also a necessary but insufficient factor for clinical changes. These findings indicate the importance of actions addressing multiple barriers. Several interventions [68, 69] to address various types of barriers have resulted in a significant improvement in the delivery of preventive dental services (fluoride varnish (FV) application and dental referral). In a few months, the FV application rate rose by more than 75% in facilities where the programs had been implemented. In addition to increasing reimbursement and professional training, these projects included hiring a project manager, developing education brochures and posters, providing an updated list of local dentists, and involving care assistants to share the workflow.

The organizational barriers identified in our study are not specific to oral health prevention. The same difficulties in prevention implementation are discussed in other systematic reviews focusing on different noncommunicable diseases (obesity [70], diabetes [71], mental illness [72, 73], cardiometabolic diseases [74], and asthma [75]).

Regardless of the disease, health professionals report struggling with time and workload, insufficient funding, lack of staff, shortage of materials, poor collaboration with specialists, inadequate training, confusion about roles and responsibilities, and a lack of leadership and management. The complexity of integrating prevention into clinical practice is not specific to dental caries and appears to apply equally to the prevention of a wide range of noncommunicable diseases. This topic could be further investigated through another systematic review studying barriers and facilitators shared by different noncommunicable diseases. This research provides a comprehensive view of the difficulties encountered by health professionals and encourages policy-makers to reconsider the health care system’s organization to better integrate prevention into patients’ care pathway.

Among all of the factors discussed in this systematic review, health professionals commonly mention parents as a barrier to effective oral health prevention for children. Parents of children with dental caries are described as lacking oral health knowledge, parental skills, motivation, and authority. Obesity research has shown that health professionals’ negative perceptions of their patients could affect disease management quality due to shorter consultations, less respectful communication, and a less patient-centered approach [76]. Therefore, it would be useful to undertake additional research to identify health professionals’ perceptions of children with dental caries and their families and how these perceptions could influence children’s quality of care.

Our study shows that the implementation of individual caries prevention in the medical setting is a global issue involving numerous, varied, and systemic barriers. Improving individual prevention will likely require a wide range of interventions addressing different types of factors. Therefore, the development of caries prevention in health care settings is likely to follow a lengthy and challenging implementation process. In this context, it appears critical that researchers and policy-makers continue to work on diversified prevention strategies, such as collective measures. While fluoride has long been used at an international level [77], other strategies are still underexploited and warrant further investigation. This is the case with several sugar-lowering measures recommended by the World Health Organization [2]: 1) taxation of sugar-sweetened beverages and foods with high free sugar content; 2) clear nutrition labeling about sugars contained in a product; and 3) regulation of marketing and advertising of food and beverages high in free sugars to children.

The use of the theoretical domain framework (TDF) in the development of the data extraction and analysis template is one of our study’s strengths. Many different behavioral change theories exist, and others could have been used as theoretical frameworks for this study. These psychological theories involve a wide range of constructs, and their complexity can sometimes make them difficult to apply in a research setting. In this context, selecting and applying a theoretical framework may be challenging for researchers. The TDF offers a reasonable answer to these challenges by providing a comprehensive and practical framework that synthesizes 33 psychological theories and 128 constructs. Developed in 2005 by Michi et al. [78], the TDF was modified and validated to strengthen its structure and content [35]. This model is now commonly used by researchers to assess health-related behavior and implementation problems [79]. In this review, factors were categorized following the 14 TDF domains and then sorted according to whether they were clinician-, patient-, or organization-related. This two-step method provided a clear understanding of the factors that affect oral health prevention in children. Additionally, the relevance of our approach is reinforced by its use in other recent systematic reviews studying implementation difficulties in the medical setting [71, 80, 81]. Another strength of our work is that it included studies that questioned all primary care professionals engaged in children’s dental health follow-up. Moreover, our analysis provides a well-distributed number of studies pooling dental professionals and others pooling nondental professionals. Our findings thus provide an overall view of the challenges that limit the implementation of pediatric caries prevention.

Regarding limitations, the search strategy was limited to 3 international databases with no search of the grey literature, and only articles written in English and French were considered for the analysis. The databases were chosen after discussions with a medical librarian. Prior to the investigation, more databases have been explored (especially Scopus). Because the findings from Scopus contained an extensive number of duplicates and articles irrelevant to the research issue, the study was eventually restricted to three databases (PubMed, Web of Sciences, and Cairn). The exclusion of grey literature may also be regarded as an important constraint, since the inclusion of unpublished data can reduce the effect of publication biais. However, the variability of this literature’s editorial process does not always ensure reliable data. Moreover, it has been found that unpublished studies rarely influence the results and conclusions of a review [82]. For these reasons, the research team and the professional librarian involved in the project agreed that the use of three databases, in combination with a rigorous manual search, would be sufficient to guarantee the quality of the research. Although a few references may have been overlooked, it is likely that the review included the most relevant references. Additionally, investigators did not use specific tools for the evaluation of the methodological quality of studies (such as the Critical Appraisal Skills Program criteria Checklists). However, a reporting quality assessment was consciously and independently conducted by the two main reviewers (GL and EM) using the COREQ and STROBE checklists. Two tables detailing the completion of these checklists are supplied, providing a good overview of each article’s reporting quality strengths and weaknesses (Supplementary material). In addition, the research team members decided that in quantitative studies, a factor was considered a barrier when at least 10% of participants reported it as such. This arbitrary choice of a 10% cutoff number may be considered a methodological limitation. Given its greater significance at a population level, a 20% threshold value may have seemed more reasonable. This issue was discussed extensively during the study design. The researchers finally decided to record, during the data extraction process, all the factors that 10 to 20% of the participants considered barriers. A small number of these factors were identified during the data analysis. The research team consequently chose to maintain a 10% cutoff since a 20% value would not change the results. Moreover, there is no consensus in the literature concerning this cutoff value. Finally, some critical components may be underrepresented in our findings. The methods used in the studies included in our analysis could have influenced the participants’ responses and led them to emphasize some factors more than others. The use of closed-ended questions in quantitative studies means that the factors discussed are suggested to the participants by the investigators. In these studies, participants are not given the opportunity to cite factors that are not mentioned in the questionnaires. As a result, some factors are heavily cited in our review, which may lead to the mistaken assumption that some factors are more important than others when they are simply overrepresented in quantitative studies. However, our study included a significant amount of qualitative research, which resulted in a wide range of factors being discussed. To strengthen the results of this systematic review and address this bias, future research may conduct a quantitative study using the TDF questionnaire [83, 84]. If tailored to the context of pediatric caries prevention, this questionnaire could be used to independently assess the influence of the 14 TDF domains on clinicians’ behaviors. This study could help researchers identify the most relevant levers for designing evidence-based interventions to improve health professionals’ integration of caries prevention in clinical practice.