The Evidence-Practice Gap in Minimal Intervention Dentistry: An International Comparison Between Dentists in Japan and Brazil

Abstract

Objectives:

This study was designed to: 1) evaluate and compare the evidence-practice gap (EPG) in minimal intervention dentistry (MID) in Japan and Brazil by measuring concordance between dentists’ clinical practice and published evidence; and 2) identify dentists’ factors associated with the EPG in both countries.

Methods:

We performed a cross-sectional study using a web-delivered questionnaire among 136 Japanese and 110 Brazilian dentists. The questionnaire consisted of three questions concerning “restoration diagnosis and treatment”, “deep caries diagnosis and treatment”, and “caries risk assessment” regarding MID. A chi-square test was used to analyze differences in concordance among clinical practice and evidence from the literature between Japanese and Brazilian dentists. Logistic regression analyses were performed to analyze dentists’ factors associated with overall concordance for all three questions.

Results:

Overall concordance was significantly higher in Brazil (56%) than in Japan (38%) (p<0.01). Concerning how evidence was obtained, textbooks, non-academic journals, and seminars and workshops were used as information sources more frequently by Japanese than Brazilian dentists (p<0.001), whereas scientific journal articles in English were used more frequently by Brazilian dentists (p<0.001). On logistic regression analysis, overall concordance was higher for Japanese dentists who frequently obtained evidence from scientific journal articles in English (p<0.05), whereas Brazilian dentists who frequently obtained evidence from the Internet were associated with lower overall concordance (p<0.05).

Conclusions:

Because overall concordance was significantly higher in Brazil than in Japan, Japan may have a greater EPG in MID practice. Specific characteristics of Japanese and Brazilian dentists showed significant associations with overall concordance.

INTRODUCTION

“Evidence-based practice” is “the conscientious, explicit and judicious use of latest and best evidence in decision making about the care for individual patients”.^1,2 Nevertheless, a gap between actual clinical practice and scientific evidence persists in a range of fields of healthcare; this is termed the evidence-practice gap (EPG), and it has received increasing attention.^3–7 Advantages of closing the EPG include decreased patient mortality, morbidity, and improved healthcare economics.²

In the dentistry field, minimal intervention dentistry (MID) was first defined in a policy statement.⁸ MID aims for the maximal preservation of healthy tooth structure and the lifelong retention of teeth function; this often involves using conservative techniques to treat caries lesions.⁹ The FDI World Dental Federation affirms this policy as currently the best option in managing dental caries.¹⁰ However, Innes et al.’s meta-analysis concluded that a large percentage of dentist continue to use a “drill and fill” methodology against enamel lesions.¹¹ International efforts to improve the EPG in MID are now a matter of priority.

Dental practice-based research networks (Dental PBRNs) have been set up to aid measurement and comparison of real-world procedures in routine dental clinical practice.^12–15 Dental PBRNs act to link practitioners with researchers, and aim to realize relevant and impactful clinical research that can propose improvements leading to positive changes in daily dental care. For example, the National Dental PBRN in the US (http://www.nationaldentalpbrn.org/) ^16–21 and Dental Practice-based Research Network Japan (JDPBRN) (http://www.dentalpbrn.jp/) ^22–30 have evaluated and revealed the EPGs in various domains of dental treatment. In Brazil, there are also reports recognizing the existence of the EPG in diagnosing and treatment of dental caries, using the same questionnaires as in previous studies performed at the National Dental PBRN in the US.^31–32 However, there has been no practice-based research using the same questionnaire to make direct international comparisons between Japan and Brazil.

Therefore, the objectives of this research were to: 1) evaluate and compare the EPG in MID in Japan and Brazil by measuring concordance among dentists’ clinical practice and evidence from the literature, and 2) identify dentists’ factors associated with EPG in both countries.

METHODS

Study Design

We conducted a study with a cross-sectional design utilizing of a web-delivered questionnaire in Brazil and Japan between December 2021 and June 2022. The study complied with the Helsinki Declaration. The Japanese and Portuguese versions of the questionnaire can be accessed at http://www.dentalpbrn.jp/_p/acre/25660/documents/EPG_1st_Survey_Japanese.pdf and http://www.dentalpbrn.jp/_p/acre/25660/documents/EPG_1st_Survey_Portuguese.pdf. For reference, the questionnaire’s English version can also be accessed through http://www.dentalpbrn.jp/_p/acre/25660/documents/EPG_1st_Survey_English.pdf.

Participants

Participants were recruited from JDPBRN websites and emails in Japan, and emails and social networks in Brazil to dentists who reported performing restorative dentistry of some type in their practices. JDPBRN is a research network based around dental practice which is widely represented by dental clinics and practices that vary in terms of type of practice, characteristics of patients, and philosophy of treatment.²² To help ensure nationally representative samples, participants were recruited from the seven major areas of Japan (51 from Kyushu, 41 from Kansai, 15 from Kanto, 12 from Hokkaido, 9 from Chubu, 6 from Chugoku-Shikoku, and 2 from Tohoku) ³⁰ and five major areas of Brazil (85 from Southeast, 14 from Northeast, 5 from Midwest, 3 from South, and 1 from North). The response rate was 74% (136/185) in Japan. In Brazil, we could not discern a definitive number for the denominator because participants were recruited using social media. Informed consent was provided by all participants before study participation.

Questionnaire items

1). Measurement of EPG in MID in Japan and Brazil

We quantified the EPG in MID by assessing concordance of clinical practice with published evidence (henceforth, “concordance”) using an EPG measurement tool (Table 1).²¹ This tool consists of three questions about the EPG in MID in relation to several clinical areas: “Restoration Diagnosis and Treatment: Q1. Lesion depth for permanent restoration instead of only preventive or non-surgical therapy (proximal caries)”, “Deep Caries Diagnosis and Treatment: Q2. Treatment options for excavation of caries deeper than anticipated for a patient with deep occlusal caries in the mandibular right first molar and perhaps involving the mesio-buccal pulp horn (deep caries)”, and “Caries Risk Assessment: Q3. Assessment of caries risk for individual patients in any way.” These questions were determined with reference to the National Dental PBRN’s Practice Impact Questionnaire.²¹ An original version of the questionnaire is downloadable in English at https://www.nationaldentalpbrn.org/wp-content/uploads/2020/05/Study-22-DPBRN-CONDOR-PIRG-questionnaire.version-of-July-20-2011.pdf.

Table 1.

Categorization of concordance between clinical practice and published evidence (modified from Norton et al.²¹)

Clinical Area	Clinical Question	Response option(s) classified as inconsistent with evidence	Response option(s) classified as consistent with evidence
Restoration diagnosis and treatment	Q1. Lesion depth for permanent restoration instead of only preventive or non-surgical therapy (proximal caries)	Radiograph #1 or 2 (lesion in enamel only)	Radiographs #3, 4, or 5 (lesion into dentin)
Deep caries diagnosis and treatment	Q2. Treatment options for excavation of caries deeper than anticipated for a patient with deep occlusal caries in the mandibular right first molar and perhaps involving the mesio-buccal pulp horn (deep caries)	Continue and remove all the decay; Perform endodontic treatment or refer to an endodontist	Stop removing decay near the pulp horn and remove it elsewhere

2). Primary outcome

We assessed concordance using coding of responses for each of the three questions in the questionnaire (Table 1), namely consistent (“1”) or inconsistent (“0”) with evidence in the literature. The criteria for assessment of concordance of actual practice with evidence in the literature followed those used for categorization in an earlier US study (Table 1).²¹ “Overall concordance” was then set to be the primary outcome: namely, the proportion of dentists who responded consistently with the evidence for all three queries.

3). Dentists’ characteristics

The questionnaire also asked for information about the following three personal characteristics: 1) characteristics of the individuals (no. of years since graduation, gender), 2) characteristics of the practice (type of practice, busyness), and 3) information sources for evidence (colleague, textbook, non-academic journal, Internet, scientific journal articles in non-English language, scientific journal articles in English, clinical practice guidelines, as well as seminars and workshops).

Statistical Analysis

Regarding busyness of the dental practice, participants chose one option from among ‘too busy to treat all people requesting appointments’; ‘provide care to all but the practice was overburdened’ (categorized as ‘busy’); as well as ‘provide care to all but the practice was not overburdened’ or ‘not busy enough’ (categorized as ‘not busy’). For information sources of evidence, we classified the answers ‘Frequently’ or ‘Sometimes’ as ‘Higher frequency.’ We also classified ‘Never’ or ‘Rarely’ in the category of ‘Lower frequency’.²⁹ International comparisons of concordance and frequency of use of information sources for evidence in Japan and Brazil were performed using the chi-square test. We used multiple logistic regression analysis to investigate associations among dentists’ characteristics (independent variable) and overall concordance for the three MID-related questions (dependent variable), using a separate regression for each country. We calculated the adjusted odds ratios (OR) and 95% confidence intervals (CI). Because the logistic regression model included five dentist characteristics as predictor variables, more than 50 event occurrences may have been appropriate according to the 1:10 rule (10 events per variable).^33–35 Assuming an incidence of one-half for the outcome (overall concordance) based on our previous study,³⁰ a total of 100 participants would be expected as the sample size. Missing data were handled by available-case analysis. Statistical analysis was done using IBM SPSS Statistics 23^® (IBM Corp., Somers, NY), using an alpha set at 0.05.

RESULTS

Demographic characteristics

Demographic characteristics of the enrolled dentists in Japan and Brazil are shown in Table 2. One hundred and thirty-six dentists in Japan and 110 dentists in Brazil participated. The mean (±standard deviation) number of years since graduation from dental school was 18.5±13.7 in Japan, and 16.1±11.3 in Brazil. Most Japanese dentists were male (84%, n=114), whereas most Brazilian dentists were female (71%, n=78). With regard to the type of practice, most of the dentists in Japan and Brazil answered “self-employed” with 52% (n=71) and 48% (n=52), respectively. Regarding the busyness of their practice, 56% (n=76) of the respondents in Japan answered “busy”, while 71% (n=77) of the respondents in Brazil answered “not busy”.

Table 2.

Demographic characteristics of participating dentists in Japan (n=136) and Brazil (n=110)

	Japan (n=136)	Brazil (n=110)
	Mean ± SD or Number (%)	Mean ± SD or Number (%)
Years since graduation from dental school	18.5±13.7	16.1±11.3
Gender
Male	114 (84%)	32 (29%)
Female	22 (16%)	78 (71%)
Type of practice
Employed by another dentist	44 (32%)	29 (27%)
Self-employed	71 (52%)	52 (48%)
Public dental care facility	21 (15%)	27 (25%)
Practice busyness
Busy	76 (56%)	32 (29%)
Not busy	60 (44%)	77 (71%)

International comparison of EPG on MID

We assessed EPG through three questions (Q1-3) concerning MID. International comparisons among the two countries using the chi-square test showed that concordance with the evidence based on Q3 (conducting caries risk assessment) was significantly higher in Brazil (91%, n=98) than in Japan (68%, n=93) (p<0.001). Overall concordance for all three questions was significantly higher in Brazil (56%, n=61) than in Japan (38%, n=51) (p<0.01) (Table 3).

Table 3.

International comparison of concordance in MID

Clinical Question	Japan Concordance (n=136)	Brazil Concordance (n=110)	p value*
Q1 Restoration diagnosis and treatment	63% (86/136)	73% (80/110)	0.149
Q2 Deep caries diagnosis and treatment	81% (110/136)	75% (82/110)	0.299
Q3 Caries risk assessment	68% (93/136)	91% (98/108)	<0.001
Overall concordance for all 3 questions	38% (51/136)	56% (61/110)	0.007

^*chi-square test

International comparison of information sources for evidence

The percentages of dentists who frequently obtained information from “Textbooks” (92% Japan: 72% Brazil), “Non-academic journals” (67% Japan: 21% Brazil), and “Seminars and workshops” (86% Japan: 58% Brazil) were significantly higher in Japan than in Brazil (p<0.001). In contrast, the percentage of dentists who frequently obtained information from “Scientific journal articles in English” (Japan 49%: Brazil 82%) was significantly higher in Brazil than in Japan (p<0.001) (Table 4).

Table 4.

International comparison of information sources of evidence

Information sources of evidence	Percentage of dentists with higher frequency of use
	Japan (n=136)	Brazil (n=110)	p value*
Colleague	82% (111/136)	82% (90/110)	1
Textbook	92% (125/136)	72% (78/109)	<0.001
Non-academic Journal	67% (91/136)	21% (22/106)	<0.001
Internet information sources (e.g., websites, blogs)	62% (84/136)	59% (63/107)	0.745
Scientific journal articles in non-English language	66% (90/136)	57% (61/107)	0.184
Scientific journal articles in English	49% (66/136)	82% (88/108)	<0.001
Clinical Practice Guidelines	79% (107/136)	74% (79/107)	0.464
Seminars and workshops	86% (117/136)	58% (61/106)	<0.001

^*chi-square test

International comparison of dentists’ factors associated with EPG in MID

Multiple logistic regression revealed two dentists’ factors (one in each country) significantly associated with overall concordance. In Japan, a higher frequency of obtaining evidence from scientific journal articles in English was significantly associated with higher “overall concordance”, with an OR of 2.44 (95%CI: 1.13–5.29) (p<0.05). By contrast, in Brazil, a higher frequency of obtaining evidence found in the Internet was significantly associated with a lower “overall concordance”, with an OR of 0.39 (95%CI: 0.16–0.92) (p<0.05) (Table 5).

Table 5.

International comparison of factors associated with overall concordance in MID

Variable (reference: lower frequency)	Japan (n=136)				Brazil (n=104)
	OR*	95% CI		p value	OR*	95% CI		p value
		Lower	Upper			Lower	Upper
Colleague	0.87	0.35	2.18	0.770	0.99	0.35	2.82	0.988
Textbook	7.18	0.85	60.4	0.070	1.00	0.40	2.53	0.998
Non-academic Journal	0.90	0.40	2.02	0.792	0.46	0.16	1.35	0.158
Internet information sources (e.g., websites, blogs)	0.96	0.45	2.05	0.909	0.39	0.16	0.92	0.032
Scientific journal articles in non-English language	1.44	0.66	3.15	0.358	0.53	0.23	1.26	0.149
Scientific journal articles in English	2.44	1.13	5.29	0.024	1.70	0.58	4.95	0.333
Clinical Practice Guidelines	1.22	0.51	2.92	0.664	1.06	0.41	2.73	0.904
Seminars and workshops	1.69	0.54	5.27	0.366	1.39	0.57	3.39	0.468

Multiple logistic regression analysis

CI: confidence interval

Dependent variable: Overall concordance for all three MID-related questions or not

Overall predictive accuracy is 64.7% for Japan and 68.3% for Brazil

^*Adjusted for years since graduation from dental school, gender, type of practice and practice busyness

DISCUSSION

This international collaborative study is the first to administer the same questionnaire to dentists in Japan and Brazil concomitantly. The results showed that overall concordance for all three questions was significantly higher for Brazilian dentists than for Japanese dentists, indicating that the EPG in MID may be larger in Japan than in Brazil. Logistic regression analysis to reveal variables related with the EPG showed that Japanese practitioners who frequently obtained evidence from English-language scientific journal articles showed associations with higher overall concordance, whereas Brazilian dentists who frequently obtained evidence from the Internet were associated with lower overall concordance. With regard to gender, the majority of Japanese participants were male, whereas the majority of Brazilian participants were female; there was no significant difference between gender and overall concordance in univariate logistic analysis for either Japan or Brazil (data not shown).

Although overall concordance was significantly higher among Brazilian dentists than among Japanese dentists, when considering individual questions, there was a significant difference between the two countries only for Q3, regarding caries risk assessment. Previous studies reported that 26% of dentists in Japan (2011-2012) ²⁶ and 36% in Brazil (2014-2015) ³² performed caries risk assessments. However, the corresponding percentages in the current study (2021-2022) were 68% and 91%, respectively. Thus, not only was the percentage higher among Brazilian dentists when compared to Japanese dentists in the previous survey, but the percentage of dentists conducting caries risk assessments has increased over time, by 42% in Japan and 55% in Brazil. This improvement may be the result of an increasing amount of evidence available to clinicians regarding caries risk assessment.^36,37 Longitudinal and regular measurement of EPG using methods similar to those in our present study could help to ascertain and improve the quality of future dental care. It would also be desirable to develop and use a common, internationally comparable EPG measurement scale across countries.

We found that obtaining information from science journal papers in English showed a significant association with a higher overall concordance, which is consistent with the results of a previous study in Japan.³⁰ Our study results suggest that prevalent use of evidence-based information in English-language science journal articles may reduce EPG in MID among Japanese dental practitioners. However, the percentage of dentists who obtained information in this way was significantly lower in Japan than in Brazil (Japan: 49%, Brazil: 82%). Although the reasons for this finding cannot be deduced from the survey results, one possible explanation is the differences in language systems of different countries. The US Foreign Service Institute (FSI),³⁸ which has taught languages to US diplomats for more than 70 years, has calculated the time required to develop “professional working proficiency” in another language. Japanese is listed in Category IV (2,200 class hours), which includes “Super-hard languages - Languages which are exceptionally difficult for native English speakers.” By contrast, Brazil’s national language, Portuguese, is classified in Category I (600-750 class hours), as a “Languages similar to English.” This is a nearly threefold difference in language acquisition time. Clearly, it takes longer for Japanese to learn to read and understand English papers than for Brazilians. Therefore, it is possible that the “Language Gap” between Japanese and English language systems makes it particularly difficult for Japanese dentists to read papers in English, ultimately contributing to the “Evidence-Practice Gap” in dental practice. Further research in various languages could help determine if there is an association between these two gaps.

In Brazil, obtaining information from the Internet was significantly associated with a lower overall concordance. Regarding online dental information, the websites of the majority of German dentists are reportedly not sufficiently accurate with respect to periodontitis.³⁹ Similarly, Lotto et al. reported that Brazilian websites had a poor quality of toothache-related information, hampering decisions to seek dental care and leading to damage due to ineffective self-management of toothache.⁴⁰ Moreover, Oliveira et al. showed that one of the most important barriers that Brazilian dentists perceived in integrating scientific evidence into their dental practice was “the difficulty in determining whether the scientific content on the Internet is reliable or not”.⁴¹ This circumstance suggests that the EPG in MID in Brazil could be decreased by dental education including lifelong learning to support the strengthening of dentists’ critical appraisal skills as well as their ability to select reliable evidence from the Internet. Further research would be needed on specific methodologies to improve such skills.

Our present participants were from a wide range of dental practices in seven major areas in Japan and five major areas in Brazil. Their age and gender distributions were consistent with the real distribution of dentists in the two countries,^42,43 which increases generalizability of our results. However, a number of limitations of this study should be addressed. First, as the participants did not undergo random sampling, a degree of sampling bias may be present. Second, the data were obtained from self-reports, raising the possibility of information bias. Finally, the study’s cross-sectional methodology means that it is not possible to identify a causal relationship between specific dentists’ information sources and degree of concordance.

CONCLUSION

The results of this study indicate a significantly higher overall concordance in Brazil compared to Japan, suggesting that Japan may have a greater EPG in the implementation of MID. In Japan, the EPG might be decreased by creating an environment that facilitates clinician use of findings from English-language science journal papers. In Brazil, the EPG might be decreased by increasing continuing education opportunities, including training to enable them to select and critically evaluate reliable evidence from Internet information.

Clinical Relevance Statement:

In Japan, the EPG might be decreased by facilitating clinician use of evidence from English-language scientific journal articles. In Brazil, the EPG might be decreased by increasing continuing education opportunities, including training to enable them to select and critically evaluate reliable evidence from Internet information.