Competency model for dentists in China: Results of a Delphi study

Objective With the increasing awareness of the importance of oral health, patients have an increasing need for integrated care from dentists. In China, the dentistry examination consists of two parts: a practical skills examination and a comprehensive medical examination; to date, no assessment methods that are based on specialized dentistry competencies, unlike the United States, Canada, and other countries, have been established. Therefore, the purpose of this study was to construct a competency model for dentists in China in order to guide the development, admission, training and assessment of dentists. Methods Using a literature review, focus group interviews and in-depth personal interviews, a dentist competency index was developed with an expert consultation questionnaire. A panel of 20 specialist experts was chosen from ten national medical universities to carry out two rounds of Delphi expert analysis, using the boundary value method to filter the indicators and the Analytic Hierarchy Process to calculate the weights of the primary indicators. Results Two rounds of Delphi results showed that the expert authority, enthusiasm, and coordination coefficients were high. Constructs of the competency model that included seven primary indicators and 62 secondary indicators determined the weight of each index. The seven primary indicators included the following: clinical skills and medical services, disease prevention and health promotion, interpersonal communication skills, core values and professionalism, medical knowledge and lifelong learning ability, teamwork ability and scientific research ability. Conclusion In conclusion, the use of the Delphi method to construct an initial model of Chinese physician competency is scientific and feasible. The initial competency model conforms to the characteristics and quality requirements of dentists in China and has a strong scientific basis. The dentist competency model should be used in the National Dental Licensing Examination in China.


Results
Two rounds of Delphi results showed that the expert authority, enthusiasm, and coordination coefficients were high. Constructs of the competency model that included seven primary indicators and 62 secondary indicators determined the weight of each index. The seven primary indicators included the following: clinical skills and medical services, disease prevention and health promotion, interpersonal communication skills, core values and professionalism, medical knowledge and lifelong learning ability, teamwork ability and scientific research ability.

Conclusion
In conclusion, the use of the Delphi method to construct an initial model of Chinese physician competency is scientific and feasible. The initial competency model conforms to the characteristics and quality requirements of dentists in China and has a strong scientific PLOS

Use of the Delphi model
Delphi was originally developed by the Rand Corporation. Some experts observe that when researchers lack scientific knowledge of the topic being investigated or face-to-face data collection is impractical, the Delphi method can be applied [7,8]. In this study, we used the Delphi expert consultation method, the threshold method and the analytic hierarchy process (AHP) to establish the initial model for "5 + 1" stage dentist competency [9,10]. Preliminary competency indicators. First, the framework of the Chinese University Hospitals' general model for clinicians was used for reference [11]. Second, a literature search and review were done using databases that included Wan Fang, CNKI, and PubMed to study the models of dentist competency from Canada, Europe, New Zealand, the USA and Australia. Third, six focus group interviews were performed. Finally, the initial physician competency model was constructed. The model included eight first-level indices and 81 second-level indices, resulting in a "dentist competency Delphi questionnaire." Design of the Delphi consulting questionnaire. The expert consultation questionnaire included the background, purpose, and expert basic information questionnaire, AHP matrix table, index table of all levels, degree of expertise quantification table, degree of expert opinion coordination and quantification table, and opinion column.
The index evaluation included three aspects: importance, feasibility, and sensitivity. The scores were between 1-10, where 1 means that the index is least important, least feasible, and least sensitive and 10 means that the index is most important, most feasible, and most sensitive. At the same time, experts needed to select the familiarity and judgment scores. Familiarity was divided into a scale of 1-5 using the Likert scale method, where 1 means that the expert is most unfamiliar with the index and 5 means that the expert is most familiar with the index. Judgment was based on four aspects: theoretical analysis, work experience, understanding of domestic and foreign counterparts, and insight. The judgment scores were used to determine the degree of influence, with scores of 1-3 points, where 3 = high, 2 = medium, 1 = small.
Selection of the experts. There are 20 experts in 10 national dentistry qualification examination areas, from 10 universities of stomatology and their affiliated dental hospitals: the Fourth Military Medical University, Lanzhou University, Sichuan University, Jilin University, Nanjing Medical University, Shanghai Jiao Tong University, Capital Medical University, Wuhan University, Sun Yet-sen University and Chongqing Medical University.
The specialties of the experts included management, education, and dentistry. All experts in their respective fields had high academic attainments; 20 experts had master's degrees or higher, and 75% had a doctorate or higher education. Information regarding the experts is shown in Table 1.
Implementing two rounds of Delphi expert advice. Two rounds of Delphi expert advice were implemented in writing to the 20 experts who were asked to evaluate each indicator and make comments. After the first round of the survey, SPSS 16.0 statistical software was used to analyze the questionnaire responses. The degree of coordination of the experts was calculated to include the experts' positive coefficient and degree of authority according to the boundary value method to reduce the competency index. The indicators were adjusted and supplemented according to the comments of the experts. A second round of expert consultation was made followed by further analysis of the questionnaire and revision of the competency index and calculation of the index. Finally, the competence model for dentists was constructed.

Data analysis
Positive coefficient of experts. The positive coefficient of experts is the effective recovery rate of the expert consultation questionnaire and can reflect positive input from the experts.
The American sociologist Earl Babble argued that a 50% recovery rate was the minimum acceptable rate for analysis and reporting, 60% could be considered good, and 70% achieved a very good standard [12].
Degree of expert authority. The degree of expert authority was determined by two factors: the expertise to make a judgment on the program, and the familiarity of the expert with the problem [13]. In this study, Ca represents the coefficient that affects the expert judgment. The experts used the terms "practical experience," "theoretical analysis," "understanding of peers," and "insight" as the basis for judgments. Large, medium, and small judgments were made to determine the extent of influence. When Ca = 1, the degree of influence of expert judgment is medium, and when Ca = 0, it has no effect on expert judgment according to the valuation criteria in Table 2.
Familiarity was expressed in terms of Cs. This study used the Likert scale method to classify experts into five levels of familiarity: very familiar (5 points), more familiar (4 points), generally familiar (3 points), less familiar (2 points), and not familiar (1 point) ( Table 3). The degree of familiarity of each expert was calculated statistically.
The degree of expert authority was expressed by Cr: Cr = (Ca + Cs) / 2; values greater than 0.7 were considered to be acceptable [14].
Degree of expert coordination. The degree of expert coordination is an important index for judging the consistency of the indicators among the experts, and included the Kendall W coordination coefficient and the coefficient of variation of each index; the coefficient of variation is an important basis for index deletion [15]. The smaller the coefficient of variation, the greater is the degree of coordination of the experts. Boundary value calculation methods and results. The method of calculating the cut-off value of the full frequency and arithmetic mean was as follows: the mean value = mean-standard deviation; a score higher than the cut-off value was chosen. The calculation results of the boundary values are shown in Table 4.
Guidelines for the selection of indicators. There are three indicators, where the first is the importance of the three indicators does not meet the requirements; the second is two boundaries of the importance, feasibility, and sensitivity do not meet the requirements; and the last one is deleting or modifying the indicator according to the expert opinion.
Index weight calculation method. In order to reflect the importance of the dentistry competency index, another AHP software was used to calculate the weight of the primary index in the second round. The weight of the secondary index was calculated by the percentage weight method. In this study, we used the scale matrix judgment method to allow each of the experts to compare each indicator in each row and column and to complete the results in the corresponding form.

Positive coefficient of experts
In the first round of Delphi expert consultation, 20 questionnaires were sent out, and 20 were effectively recovered, resulting in an effective recovery rate of 100%. In the second round, 20 questionnaires were issued, and the effective recovery rate was also 100%. Some experts modified and improved a number of indicators.

Expert authority
The expert authority coefficients were 0.74 and 0.75 in two rounds of consultation; both of these were greater than 0.7, indicating that the expert consultation results were accurate and credible (Table 5).

Expert consultation coordination
The P-value of the coordination coefficient was less than 0.01 ( Table 6), indicating that the expert opinion was consistent.

Index screening
Screening results of the primary indicator. Statistical analysis of the experts' scores after the first round of consultation showed that the boundaries of "information and management capacity" and "scientific research capacity" did not meet the study requirements ( Table 7). The results of the second round of expert consultation showed that the "information and management capacity" had the lowest weight (0.0668), followed by "scientific research" (0.849). Finally, combined with the views of the experts, the decision was made to delete "information and management capabilities" and retain "scientific research capabilities." Screening results of the secondary indicator. Using the boundary value method combined with expert advice in the first round, we deleted a total of 10 secondary indicators. In the second round, we deleted nine secondary indicators (Table 8). Additionally, four secondary indicators were revised after considering expert advice (Table 9).
Index system and weighting results. Using the analytic hierarchy process and the percent weight method to determine the weight of the primary and secondary indicators, obtaining the weight of all indicators resulted in the development of the China dentist competency model (Table 10).

Discussion
In recent years, the Delphi method has been widely used in health management and has become a mature indicator screening method in the field of health care [16][17][18][19][20]. This method combines qualitative and quantitative research methods to collect and screen expert advice. In the consultation process, the choice of experts is a key link. Inappropriate selection of expert members will not only increase the bias in the evaluation but may also lead to a decrease in the response rate. In general, there is a functional relationship between the accuracy of the consultation results and the number of participants in the consultation; between 15 and 50 experts is considered to be suitable [21]. Based on the principle of a combination of representation and authority, this study selected 20 experts and scholars involved in the training of dentists in 10 universities nationwide in China. The authority of the experts was greater than 0.7, and the chi-squared test results of the coordination coefficient showed that the P-values were less than 0.01. In the first round, the coefficient of variation was 0.05-0.50, and in the second round, it was 0.04-0.35, indicating that the participation of the experts was focused and detailed. Two rounds of expert consultation questionnaires had an effective recovery rate of 100%, showing that the experts were motivated and enthusiastic to participate.
In the index screening, we selected three indicators for the screening items: full frequency, arithmetic mean, and coefficient of variation. Each project determined the threshold. From the analysis of the screening results, they appeared to be reliable; the weight value of the dentist competency index reflected the importance of the indicator. Weight is the importance of indicators and the size of the quantitative performance; a reasonable weight setting is significant for establishing an index system. In order to quantify the importance of competency, we used the analytic hierarchy process to calculate the weight of the first-level index, used the experts' importance score, and combined the weight of the primary index to calculate the weight of the secondary index, so that the calculation of the weight of the evaluation index was more scientific.
In this study, among the first-level indicators, the top three weights were clinical skills and medical services (0.2309), disease prevention and health promotion (0.1564), and interpersonal communication skills (0.1343); these weights were related to the occupational characteristics and competence of the dentists. Dental practitioners are knowledge-based skilled personnel with solid professional theoretical knowledge and skill. In addition to a high level of practical skills, each dentist must be able to ensure harmonious dentist-patient relationships; these relationships require effective communication between dentists and patients, as well as good interpersonal and communication skills. Also, because prevention is becoming an increasing part of clinical practice, the dentist has a role in patient and community health education.
In this study, the weights of "scientific research ability" and "core values and professionalism" were analyzed through two rounds of expert consultation. We merged the second index of "information and management ability" into the "scientific research ability" index. The subjects of the study were dentists who had graduated from medical and dental colleges and universities and received a one-year standardized training in residency. The main goal for dentists at this stage is clinical skills training. From the dental student's point of view, the vast majority of students also believe that practical clinical teaching is helpful when supported by clinical research. Experts also believe that dental training should be based on clinical skills; some controversy regarding including research in the training process remains.  https://doi.org/10.1371/journal.pone.0194411.t008 Table 9. Indicator changes in two rounds of Delphi.

First round
Master oral surgery for local anesthesia and treatment-related complications.  In the treatment of patients, the dentist usually works independently. Because of the limited number of nurses, China's "four-hand" operation is not universal. Therefore, in this study, the "team cooperation ability" index weight value was low. In the interview, some experts noted that the practice of the dentist was uniquely independent and that the dentist's competency and the clinical physician's competency have some differences.
According to the competency "onion model," among the seven first-level indicators we obtained in this study, "clinical skills and medical services" and "disease prevention and health promotion" reflect knowledge and skills; "medical knowledge and learning ability," "interpersonal communication skills," "teamwork ability," and "scientific research ability" reflect selfawareness and training ability; and "core values and professionalism" reflect characteristics and motives.
There are 62 second-level indicators in the dentist competency model; this number is less than the 82 second-level indicators in the competency training for physicians and surgeons in China [11]. Because the dentists in this study had standardized training for one year after graduating from university, and the physicians and surgeons had standardized graduate training for three years after graduating from college, there are differences in the competency indicators. Furthermore, this study reflects the professional characteristics of dentistry and includes the evaluation of practical skills, such as tooth repair, surgical extraction, and traditional preparation of dentures, as well as management abilities including the rational use of drugs, avoiding unnecessary waste, and local anesthesia technology skills. Dentists also need to inform patients about oral health and actively participate in oral health education and promotion.
Compared with the competency model of the other five countries, we found the following commonalities and differences. We all have the following four competencies: "Professionalism", "Communication and Interpersonal Skills", "Health Promotion", and "Practice Management and Informatics". However, "Core Values and Professional Qualities of Doctors" is also important in this study and helps define the first competencies. In China, we place great emphasis on the cultivation of doctors' values and their enthusiasm for work. In addition, each of the competencies has a corresponding weight that helps us to apply our model to practical work, including the development, admission, training and assessment of dentists.
This study had several limitations. The use of a single model, the Delphi model, to construct the competency model for dentists in China was not verified with the use of alternative models. The constructed model has not yet been tested in clinical practice or postgraduate dental training and evaluation. Although the experts in the study had discipline, enthusiasm, and national geographical representation, their input was subjective and not objective [22]. Therefore, future follow-up studies on the use of the model will evaluate and validate the model with a behavioral event interview method in addition to being assessed nationwide in China to develop a dentist competency model that is suitable for our country.

Conclusion
This study describes the process of constructing the initial model of dentist competency using the Delphi method to develop an assessment method that contains seven primary indicators and 62 secondary indicators. The dentist competency model is scientifically based and clinically relevant and should be suggested for use in the National Dental Licensing Examination in China.