Skip to content

Advertisement

BMC Medical Education

What do you think about BMC? Take part in

Open Access
Open Peer Review

This article has Open Peer Review reports available.

How does Open Peer Review work?

Application of latent class analysis in assessing the competency of physicians in China

BMC Medical EducationBMC series – open, inclusive and trusted201717:208

https://doi.org/10.1186/s12909-017-1039-4

Received: 23 August 2016

Accepted: 2 November 2017

Published: 13 November 2017

Abstract

Background

The physicians’ competency is an important public health issue around the world. Several international organizations have taken the lead in examining the competencies required to be a physician. The purpose of this study is to identify subgroups of physicians’ competency based upon the importance results of competency evaluation and provide a scientific basis for the qualitative research of the competency of physicians.

Methods

A cross-sectional study was conducted on a large population-based sample in 31 provinces, autonomous regions and municipalities directly under the central government in China. The latent class analysis was performed to identify patterns of physicians’ competency using M-plus software.

Results

In this study, the latent class analysis was adopted to identify the appropriate number of distinct latent classes of physicians’ competency based on eight competency dimensions, and a four-class model best fit the data, which are excellent competency group, lack of professionalism competency group, individual competency driven group, and lack of competency cognitive group. Therefore, 6247 physicians can be divided into four latent classes based on the importance results of competency evaluation, and the number of each class is 5684, 284, 215 and 64, respectively.

Conclusion

These findings suggested that latent class analysis can be used to study the competency of physicians, and four distinct subgroups were identified. Therefore, we can effectively understand the patterns of physicians’ competency, and the health administrative departments could utilize more specific measures according to their different competency subgroups, and providing individualized training schemes in the future training and management of physicians.

Keywords

PhysiciansCompetencyLatent class analysis

Background

Along with an elevated level of economic development is the improvement of social security systems and advances in medical technology. As well as a diversified public demand for clinical medical services, the demand for clinical medical professionals has also been increasing. Clinical medicine is recognized as the science with dual characteristics of both scientific knowledge systems and practical activities, and the comprehensive ability of physicians is closely associated with the quality of health care. In 2010, the global medical education leaders published a report in the Lancet, which proposed the concept of the third medical education reform: based on the education system and health system, with competency-orientation, to establish occupational ability requirements with specific targets to improve the performance of the whole health system [1].

Around the world, several international organizations have taken the lead in examining the competencies required to be a physician. The United States, Canada and United Kingdom basically completed the competence index system and guidelines for physicians. The USA’s Accreditation Council for Graduate Medical Education (ACGME) has identified 6 domains of clinical competencies for all GME specialties [2, 3]. The Royal College of Physicians and Surgeons of Canada (RCPSC) developed Canadian Medical Education Directives for specialists (CanMEDS) framework, which was a guide to the essential abilities physicians need for optimal patient outcomes [4]. The General Medical Council (GMC) of the United Kingdom has published a version of Good Medical Practice (GMP) for each specialty [5, 6]. In May 2012, Chinese Ministry of Education and Health launched the “excellent physicians training program”, aimed to reform the training mode of clinical medical personnel, and to train high-level, international physicians who would adapt to development and social needs. This is of important strategic significance in deepening the medical system reform and improving the quality of medical professional training [7, 8].

In the research of competency, usually indirect measurement of the latent or implicit competency of the individual can be obtained through the observed and measurable behavior of the samples. Latent class analysis (LCA) is a person-centered statistical method, which assumes that individuals can be grouped into classes with similar patterns of some behaviors according to their response to a set of observed indicators [9, 10]. Therefore, the LCA based on statistical models has important methodological significance for the research of the physicians’ competency [11]. To date, some studies have utilized LCA as a way to describe and identify subgroups of medical students, patients and nurses based on their characteristics and behaviors. For example, the LCA identified three hypothetical classes in the unprofessional behaviors of medical students according to a template of 109 behaviors [12]. Another study suggested four distinct subgroups of complementary medicine users among patients with breast cancer using LCA [13]. Three subgroups were identified for head nurses’ competency based on characteristics of competency self-reported questionnaire [14]. Likewise, researchers have used LCA to explain co-morbidity or to study subtypes based on combinations of emotional and behavioral situations [1517]. However, these studies did not examine the classes of physicians’ competency. Thus, further research is needed to identify subgroups of physicians based on competency. As a first step in the process of developing competency model, different subtypes of physicians’ competency will have different training and evaluation programs. Therefore, the aim of this study was to explore whether distinct subgroups of physicians’ competency could be defined.

Methods

Questionnaires

The questionnaire used was “Chinese physician’s competency questionnaire”, developed by North Medical Education Development Center of China Medical University. The physicians needed to rank the importance of competency items. The questionnaire used a 5-point Likert-scale [18], ranked the items’ importance from 1 (definitely not important), 2 (not important), 3 (neutral), 4 (important) to 5 (definitely important). The questionnaire divided the competency of physicians into eight dimensions, which are: clinical skills and patient care, professionalism, interpersonal communication, master of medical knowledge, teamwork, health promotion and disease prevention, information and management, academic research. There were a total of 103 items, due to the large number of questionnaire items, the current study performed latent class analysis on the average importance scores for each dimension in the questionnaire. The form of the questionnaire is presented in Additional file 1.

Participants

The approval of the survey was obtained from China Medical University Review Board. Each participant gave verbal and written consent. Participants could retreat from the research at any time. We performed the multi-stage stratified sampling research in 31 cities in all 31 provinces, municipalities and autonomous regions conducted from October 2012 to June 2013. Among the total of questionnaires distributed in the formal survey, 6247 valid questionnaires (89.0%) were collected.

Data collection

The data was first stratified according to the region and hospital grade, and further stratified according to the title of physicians within the hospital for random sampling. The studied subjects are mostly physicians who have completed 3-years resident standardized training, general rotation or with equivalent qualifications, covering internal medicine, surgery, gynecology and obstetrics, pediatrics and other clinical departments.

Before the investigation, the investigators were uniformly trained with standardized survey language and methods, and the survey questionnaires were numbered. During on-site investigation, the investigators would explain and guide the filling of questionnaires. After the investigation, the questionnaires were reviewed again to sort out incomplete ones. The data input was performed by two trained investigators, with 25% of the questionnaires being randomly selected for double checks to ensure consistency of data entry, and that the data information was true and reliable.

Data analysis

Descriptive analysis and test

The descriptive analysis was used to show the basic characteristics of the physicians in the survey. To compare the effects of different factors on the competency of physicians, the t-test and variance analysis were performed in our research.

Basic model of latent class analysis

The LCA was realized by a latent variable analysis model: the latent class model (LCM), which mainly explains and estimates the relationship between observed class variables through latent class variables, and to further maintain the local-independence of the explicit variables. The LCA is a potentially useful characteristics categorization technique for effectively mining class variable data information, which can also make up for the deficiency that category variables can’t be analyzed by other methods, such as factor analysis and structural equation model [19, 20]. It makes up for the deficiency of the structural equation model, which can only deal with the continuous latent variables, and more importantly, the research of the classification latent variables improves the analysis value of the class variables, so that the researchers can more deeply understand the latent impact factors of the class variables through probability [2123].

The LCA assumes that the relationship between any two observed variables can be explained by the latent variable. Assume that the latent variable X has t latent classes; A, B, C are three explicit variables with a level of I, J, K. The most basic latent class model is:
$$ {\pi}_{ijk}^{ABC}={\sum}_{t=1}^T{\pi}_t^X{\pi}_{it}^{AX}{\pi}_{jt}^{BX}{\pi}_{kt}^{CX} $$

In the equation \( {\uppi}_{\mathrm{ijk}}^{\mathrm{ABC}} \) stands for the joint probability of a LCA, \( {\uppi}_{\mathrm{t}}^{\mathrm{X}} \) is the probability of observation data belonging to a specific latent variable x, \( {\uppi}_{\mathrm{it}}^{\mathrm{AX}} \)stands for the conditional probability of a subject belongs to the t-th latent class reacted to the i-th A explicit variable.

Model fitting and parameter estimation

The LCA mainly uses the maximum likelihood method to estimate the parameters. The evaluation methods for model fitting include the Pearson test, likelihood ratio chi-square test and information evaluation criteria. AIC (Akaike information criterion) and BIC (Bayesian information criterion) are the most widely used evaluation criteria in LCA selection, both are built based on the likelihood chi-square test and can be used to compare models with a different limitation on parameters, in both of which smaller results mean better fitting. Some researchers have pointed out that the BIC index is more reliable when the sample size is in several thousands. Therefore, the subgroup evaluation in our study is mainly based on the BIC index [24]. In order to more easily explain and understand the LCA, the current study re-categorized the importance evaluation into two relevant competency groups: unimportant and important groups. The unimportant group included the importance evaluation for 1 (unimportance) and 2 (less importance). The important group included the importance evaluation for 3 (general importance), 4 (important) and 5 (very important).

Latent classification

After determination of the optimal model, the final step is to assign the observed values into the appropriate latent classes, and to explain the posterior classification properties of the observed value, which is the latent clustering analysis. The LCA was conducted using M-plus V6 [25]. The classification is based on the Bias theory, with the calculation formula as follows:
$$ {\pi}_{tijk}^{XABC}=\frac{\pi_{tijk}^{XABC}}{\sum \limits_{t=1}^T{\pi}_{tijk}^{XABC}} $$

Results

Demographic characteristics and comparison of physicians’ competency

The average age of the physicians was 38.98 years (SD = 8.71). The demographic characteristics of physicians were shown in Table 1.
Table 1

Demographic characteristics of physicians

Characteristic

 

N (%)

Gender

Male

3523 (54.5%)

Female

2724 (43.6%)

Age

20–30

1349 (21.6%)

31–40

2255 (36.1%)

41–50

2112 (33.8%)

51+

531 (8.5%)

Work experience in equivalent-years

0–4 years

1293 (20.7%)

5–9 years

1056 (16.9%)

10–14 years

1006 (16.1%)

15–19 years

1106 (17.7%)

20 years and above

1786 (28.6%)

As shown in Table 2, the results suggested that no significant differences were observed in different gender except for competency of interpersonal communication. For the different age groups and years of working experiences, significant differences were found in the importance evaluation of eight dimensions among physicians.
Table 2

The comparison of physicians’ competency in the importance evaluation of eight dimensions

 

Gender

Age

Work experience

 

t

P

F

P

F

P

1.Clinical skills and patient care

−2.10

0.35

5.59

0.001

4.37

0.002

2.Professionalism

−1.52

0.12

3.03

0.02

2.86

0.022

3.Interpersonal communication

−2.52

0.01

10.10

<0.001

6.90

<0.001

4.Master of medical knowledge

−0.71

0.47

5.90

0.001

3.79

0.004

5.Teamwork

−0.70

0.47

10.46

<0.001

7.84

<0.001

6.Health promotion and disease prevention

−1.71

0.08

14.78

<0.001

11.46

<0.001

7.Information and management

−1.21

0.22

10.75

<0.001

7.57

<0.001

8.Academic research

0.42

0.67

11.39

<0.001

8.66

<0.001

Exploratory latent class analysis

Based on the importance results of the 103 competency items by physicians in China, the current study selected the average importance scores for each dimension in the questionnaire. The mean and standard deviation of each dimension of physicians’ competency were shown in Table 3. The first three dimensions of importance evaluation were clinical skill and patient care, professionalism, and interpersonal communication.
Table 3

Importance evaluation of physicians’ competency in eight dimensions

 

Mean

Standard deviation

1.Clinical skills and patient care

4.54

0.42

2.Professionalism

4.50

0.45

3.Interpersonal communication

4.46

0.50

4.Master of medical knowledge

4.42

0.50

5.Teamwork

4.43

0.51

6.Health promotion and disease prevention

4.34

0.57

7.Information and management

4.35

0.53

8.Academic research

4.38

0.58

Table 4 lists the LCA results of the competency indicators for importance evaluation of physicians. By exploratory LCA, one to eight cluster models were fitted. According to the model fit parameter and theory, BIC gradually reduced from the baseline model to the four-class model, and then began to rise with five-class model. Among the eight models, the four-class one is the best model with the smallest BIC value, suggesting satisfactory fitting of the original data. Therefore, the four-class model was chosen as fitting the final model.
Table 4

Fitness indicators of different latent class models

 

BIC

AIC

df

P-value

1-Class

20,258.18

20,204.26

247

<0.001

2-Class

17,428.45

17,313.87

238

<0.001

3-Class

17,275.14

17,099.91

229

<0.001

4-Class

17,238.98

17,003.08

220

<0.001

5-Class

17,241.09

16,944.54

211

0.020

6-Class

17,293.57

16,936.36

202

0.094

7-Class

17,353.99

16,936.12

193

0.180

8-Class

17,418.60

16,940.07

184

0.250

Latent classification probability and conditional probabilities of the selected model

The four-class LCA chosen as described above was utilized to calculate the corresponding latent classification probability and conditional probabilities. The competency of physicians was evaluated based on their importance evaluation of the eight competency dimensions.

As shown in Table 5, in Class 1, importance evaluation of the physicians suggested eight competency dimensions were all important, which can be determined as “excellent competency” group. In Class 2, importance evaluation of the physicians suggested that seven competency dimensions were important, and the dimension of professionalism was evaluated as unimportant, which was determined as the “lack of professionalism competency” group. In Class 3, importance evaluation of the physicians suggested six competency dimensions were important, but the dimensions of teamwork and academic research were evaluated as unimportant, therefore these physicians were determined to be the “individual competency driven” group. In Class 4, importance evaluation of the physicians suggested the eight competency dimensions were all unimportant, which can be determined to be the “lack of cognitive competency” group.
Table 5

Latent classification probability and conditional probabilities of the eight competency dimensions of the physicians

 

Conditional probability

 

Class = 1

Class = 2

Class = 3

Class = 4

1

 Unimportant

0.0295

0.3964

0.0763

0.6370

 Important

0.9705

0.6036

0.9237

0.3630

2

 Unimportant

0.0325

0.5480

0.2370

0.9795

 Important

0.9675

0.4520

0.7630

0.0205

3

 Unimportant

0.0296

0.2672

0.1877

0.8568

 Important

0.9704

0.7328

0.8126

0.1432

4

 Unimportant

0.0077

0.1725

0.2484

0.8549

 Important

0.9923

0.8275

0.7516

0.1451

5

 Unimportant

0.0053

0.1309

0.3445

0.7875

 Important

0.9947

0.8691

0.6555

0.2125

6

 Unimportant

0.0113

0.0592

0.5162

0.8622

 Important

0.9887

0.9408

0.4838

0.1378

7

 Unimportant

0.0119

0.1815

0.5510

0.9034

 Important

0.9881

0.8185

0.4490

0.0966

8

 Unimportant

0.0065

0.0753

0.3154

0.6168

 Important

0.9935

0.9247

0.6846

0.3832

 Latent classification probability

0.8801

0.0657

0.0432

0.0110

Latent class categorization of the physicians’ competency

The final step of the LCA is to categorize all the individuals into appropriate groups of latent classification, and to calculate the probability of dividing the different combinations of eight dimensions into each latent class.

Table 6 lists the categorization results of different combinations of eight dimensions. For example, if a physician evaluated the competency dimensions as “important” in all eight factors, the probability of being categorized into the latent Class 1 is 0.9890, which is higher than the probabilities of other categories, and therefore grouped into latent Class 1. Based on the same theory, 6247 physicians can be divided into four latent classes based on their importance evaluation results of competency dimensions, and the number of each class is 5684, 284, 215 and 64, respectively.
Table 6

Individual categorization results of the latent class model

Combination of variables

Frequency

Posterior Probability

Class

{1,2,3, … 7,8}

 

C1

C2

C3

C4

 

{2,2,2,…,2,2}

4854

0.9890

0.0087

0.0023

0

1

{2,2,2,…,2,1}

41

0.7833

0.0858

0.1309

0

1

{1,1,1,…,1,2}

10

0

0.0007

0.0021

0.9972

4

{1,1,1,…,1,1}

12

0

0

0.0006

0.9994

4

Discussion

Currently, the selection, training, and performance evaluation of physicians in China are mainly based on their basic medical knowledge, skills and work performance. Knowledge and skills are easy to be trained, which can be evaluated through examinations. However, the real work performance of many seemingly capable people can turn out to be disappointing. The reason behind such phenomenon may be the ignorance of other competency elements. How to objectively evaluate the physicians, and how to effectively design specific training programs for the physicians, depends on the establishment of a physician competency model with Chinese characteristics [26]. The research on competency models of physicians in China is still at the early stage, which is far from being able to provide a full-range, objective and standard evaluation system for the selection, training and evaluation of physicians, the establishment of the evaluation competency model is a key issue to be solved in our current society. However, before the establishment of the competency evaluation model, we still need to estimate the subgroups of physicians’ competency.

The LCA can be used to measure and analyze a number of abstract indicators that could not be directly observed, which can simplify and integrate complex human behavior experience and social phenomenon data [27]. To our knowledge, no previous studies have identified classes of physicians’ competency. The aim of this study was to identify subgroups of physicians’ competency based upon the importance evaluation of competency questionnaire. The latent clustering numbers were not pre-set before analysis, nor were specific limits set for parameters.

According to the importance evaluation results in eight dimensions, the competency of physicians can be divided into four classes: excellent competency, lack of professionalism competency, individual competency driven, and lack of cognitive competency. The physicians in Class 1, named “excellent competency” group, with competency evaluation of important in eight dimensions. The physicians in this group are equipped with good awareness of competency importance. The physicians in Class 2 mainly lacked awareness of importance of professionalism competency. The professionalism competency emphasized that physicians should regard serving for people’s health as the highest standards of ethics. In the medical career, physicians need to have the ultimate faith, compromising altruism, pursuit of excellence, sincere and trustworthy, strong sense of responsibility. Medical professionalism is the basis for the trust in patient-physician relationship. The view of improving health care through increasing medical professionalism has been gaining momentum among physician organizations [2830]. Especially in China, physicians find it increasingly difficult to meet their responsibilities to patients and society. In these circumstances, physicians should reaffirm the fundamental and universal principles and values of medical professionalism, which remain ideals to be pursued [31, 32]. Likewise, the medical universities must do more to ensure that medical professionalism among trainees is understood, discussed and practiced.

The physicians in Class 3 lacked awareness of importance of teamwork and academic research competency. The teamwork competency mainly includes good coordination with team members, willingness to help colleagues, and building good relationships with other departments. It is effective to develop a patient’s treatment plan in the form of teamwork, which improves patient safety, increase job satisfaction and maintain the stability of the medical team. Each medical unit has a complete consultation referral system, physicians should maintain an objective and impartial attitude in consultation with other departmental patients, and work out the treatment plan together [33, 34]. Academic research includes reading academic literature, cultivating innovation ability, participating in scientific research, and actively writing research articles. The physicians should take an active part in academic conferences, so as to broaden the professional academic vision and increase the breadth and depth of thinking. Therefore, the health administrative departments should highlight teamwork and academic research training to promote physicians’ competency for this subgroup in the future. In Class 4, the physicians lacked awareness of importance of eight dimensions. Few physicians belong to this subgroup, but in the process of clinical diagnosis and treatment, medical negligence will endanger the patient health [35]. The health administrative departments should emphasize the training and assessment of physicians’ competency to meet the demands of clinical practice especially for this subgroup. Moreover, the health administration could provide detailed training plans for this subgroup, and offering comprehensive assessment in the future training and management.

There are several limitations in our study. Firstly, the cross-sectional design was used in this study, which precluded the analysis of causality. Future longitudinal studies are needed to further explore the relationship between physicians’ competency and their associated factors. Secondly, this study relied on physicians’ importance evaluation data from competency questionnaires in China, although the instruments used were standardized and validated. Future research could benefit from using a more detailed data-collection method such as 360-degree assessment for physicians. Thirdly, the application of the research was not studied, and the hospital should provide individualized training schemes for different physicians’ competency subgroups [36].

Conclusions

The physicians’ competency can be divided into four distinct subgroups. The LCA approach can provide important information on how interventions could be targeted at or tailored for different physician subgroup. In future training of physicians, the health administrative departments and hospitals could utilize more specific measures according to their different competency subgroups, providing individualized training schemes. At the same time, the competency classification results can provide a categorical basis for evaluation and prediction of competency, providing a good strategic basis for the future training of modern medical talents with rich scientific knowledge, superb clinical ability and comprehensive professional qualities.

Abbreviations

ACGME: 

Accreditation Council for Graduate Medical Education

AIC: 

Akaike information criterion

BIC: 

Bayesian information criterion

CanMEDS: 

Canadian Medical Education Directives for Specialists

CMB: 

China Medical Board

GMC: 

General Medical Council

GMP: 

Good Medical Practice

LCA: 

latent class analysis

LCM: 

latent class model

RCPSC: 

Royal College of Physicians and Surgeons of Canada

Declarations

Acknowledgments

We thank the participants in the study.

Funding

This study was financed by the National Science Foundation of China (Grant No.71473268), Foundation of China Medical Board (CMB#11–077), and Foundation of China Medical Board (CMB#11–047).

Availability of data and materials

The datasets were analyzed during the current study available from the corresponding author on reasonable request.

Authors’ contributions

ZL and YZ ran the statistical analysis and drafted the paper. LT and QC conducted the investigation and collected the data. YHZ and BZS had the original idea for this paper, made substantial contributions to the study design and interpretation of the data. All authors commented on and contributed to the final version of the manuscript. All authors read and approved the final manuscript.

Ethics approval and consent to participate

This study was approved by the Bioethics Advisory Commission of China Medical University. Written informed consent was obtained from each participant.

Consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.

Authors’ Affiliations

(1)
School of Public Health, China Medical University
(2)
Research Center for Medical Education, China Medical University
(3)
Department of Clinical Epidemiology, Shengjing Hospital, China Medical University

References

  1. Frenk J, Chen L, Bhutta ZA, Cohen J, Crisp N, Evans T, et al. Health professionals for a new century: transforming education to strengthen health systems in an interdependent world. Lancet. 2010;376:1923–58.View ArticleGoogle Scholar
  2. Accreditation Council for Graduate Medical Education (ACGME), Chicago, USA. http://www.acgme.org/acgmeweb/. Accessed 20 Aug 2016.
  3. Swing SR. The ACGME outcome project: retrospective and prospective. Medical Teacher. 2007;29:648–54.View ArticleGoogle Scholar
  4. Frank JR. The CanMEDS 2005 physician competency framework. Better standards. Better physicians. Better care. Ottawa: the Royal College of Physicians and Surgeons of Canada; 2005.Google Scholar
  5. Palmer KT, Harling CC, Harrison J, Macdonald EB, Snashall DC. Good medical practice: guidance for occupational physicians. Occup Med. 2002;52:341–52.View ArticleGoogle Scholar
  6. Irvine D. Doctors in the UK: their new professionalism and its regulatory framework. Lancet. 2001;358:1807–10.View ArticleGoogle Scholar
  7. Wai-Ching L. Competency based medical training: review. BMJ. 2002;325:693–5.View ArticleGoogle Scholar
  8. Carraccio C, Wolfsthal SD, Englander R, Ferentz K, Martin C. Shifting paradigms: from Flexner to competencies. Acad Med. 2002;77:361–7.View ArticleGoogle Scholar
  9. Muthen B, Muthen LK. Integrating person-centered and variable-centered analyses: growth mixture modeling with latent trajectory classes. Alcohol Clin Exp Res. 2000;24:882–91.View ArticleGoogle Scholar
  10. Ghosh JK. Latent class and latent transition analysis: with applications in the social, behavioral, and health sciences by Linda M. Collins, Stephanie T. Lanza. Int Stat Rev. 2010;78:449–50.View ArticleGoogle Scholar
  11. Hagenaars JA, Mccutcheon AL. Applied latent class analysis. England: Cambridge University Press; 2002. p. 304–41.View ArticleGoogle Scholar
  12. Mak-van der Vossen MC, van Mook WN, Kors JM, van Wieringen WN, Peerdeman SM, Croiset G, et al. Distinguishing three unprofessional behavior profiles of medical students using latent class analysis. Acad Med. 2016;91:1276.View ArticleGoogle Scholar
  13. Strizich G, Gammon MD, Jacobson JS, Wall M, Abrahamson P, Bradshaw PT, et al. Latent class analysis suggests four distinct classes of complementary medicine users among women with breast cancer. BMC Complement Altern Med. 2015;15:1–10.View ArticleGoogle Scholar
  14. Zhang H, Zhang GP, Li Y, Cheng SY. Current situation of competency distribution of nursing supervisors in secondary hospitals based on the latent class analysis. Chinese Journal of Health Policy. 2010;5:37–41.Google Scholar
  15. Weich S, McBride O, Hussey D, Exeter D, Brugha T, McManus S. Latent class analysis of co-morbidity in the adult psychiatric morbidity survey in England 2007: implications for DSM-5 and ICD-11. Psychol Med. 2011;41:2201–12.View ArticleGoogle Scholar
  16. Hamza CA, Willoughby T. Nonsuicidal self-injury and suicidal behavior: a latent class analysis among young adults. PLoS One. 2013;8:e59955.View ArticleGoogle Scholar
  17. Rodgers S, Grosse Holtforth M, Muller M, Hengartner MP, Rossler W, Ajdacic-Gross V. Symptom-based subtypes of depression and their psychosocial correlates: a person-centered approach focusing on the influence of sex. J Affect Disord. 2014;156:92–103.View ArticleGoogle Scholar
  18. Sun BZ, Li JG, Wang QM. Establishment and application of clinical competency model in China. Beijing: People's Medical Publishing House; 2015.Google Scholar
  19. Neuhaus V, Ring DC. Latent class analysis. J Hand Surg. 2013;38:1018–20.View ArticleGoogle Scholar
  20. Rosato NS, Baer JC. Latent class analysis: a method for capturing heterogeneity. Soc Work Res. 2012;36:61–9.View ArticleGoogle Scholar
  21. Kaufman L, Rousseeuw PJ. Finding groups in data: an introduction to cluster analysis. DBLP. 1990;Google Scholar
  22. Meng C, Wu JQ, Li YY, Zhou Y, Li N, Zhang YF, et al. Principle of latent class analysis and their application for classification. Chinese Journal of Health Statistics. 2010;27:237–9.Google Scholar
  23. Vermunt JK, Magidson J. Latent class models for classification. Comput Stat Data Anal. 2003;41:531–7.View ArticleGoogle Scholar
  24. Lin TH, Dayton CM. Model selection information criteria for non-nested latent class models. J Educ Behav Stat. 1997;22:249–64.View ArticleGoogle Scholar
  25. Muthén LK, Muthén BO. Mplus User’s Guide. 6th ed. Muthén & Muthén: Los Angeles; 1998.Google Scholar
  26. Zhao L, Sun T, Sun BZ, Zhao YH, Norcini J, Chen L. Identifying the competencies of doctors in China. BMC Medical Education. 2015;15:207.View ArticleGoogle Scholar
  27. Fairley L, Cabieses B, Small N, Petherick ES, Lawlor DA, Pickett KE, et al. Using latent class analysis to develop a model of the relationship between socioeconomic position and ethnicity: cross-sectional analyses from a multi-ethnic birth cohort study. BMC Public Health. 2014;14:835.View ArticleGoogle Scholar
  28. Campbell EG, Regan S, Gruen RL, Ferris TG, Rao SR, Cleary PD, et al. Professionalism in medicine: results of a national survey of physicians. Ann Intern Med. 2007;147:795–802.View ArticleGoogle Scholar
  29. Breen P, Murphy K. Developing professionalism in our student clinicians. Perspectives on Issues in Higher Education. 2009;12:64–8.View ArticleGoogle Scholar
  30. Plochg T, Klazinga NS, Starfield B. Transforming medical professionalism to fit changing health needs. BMC Med. 2009;7:64.View ArticleGoogle Scholar
  31. Hu LY, Yin XY, Bao XL, Nie JB. Chinese physicians' attitudes toward and understanding of medical professionalism: results of a national survey. J Clin Ethics. 2014;25:135–47.Google Scholar
  32. Tsou AY, Creutzfeldt CJ, Gordon JM. The good doctor: professionalism in the 21st century. Handb Clin Neurol. 2013;118:119–32.View ArticleGoogle Scholar
  33. Patel MR, Wheeler JR. Physician-patient communication on cost and affordability in asthma care: who wants to talk about it and who's actually doing it. Ann Am Thoracic Soc. 2014;11:1538.View ArticleGoogle Scholar
  34. Ravitz P, Lancee WJ, Lawson A, Maunder R, Hunter JJ, Leszcz M, et al. Improving physician-patient communication through coaching of simulated encounters. Acad Psychiatry. 2013;37:87–93.View ArticleGoogle Scholar
  35. Oyebode F. Clinical errors and medical negligence. Med Princ Pract. 2013;22:323–33.View ArticleGoogle Scholar
  36. Lanza ST, Rhoades BL. Latent class analysis: an alternative perspective on subgroup analysis in prevention and treatment. Prev Sci. 2013;14:157–68.View ArticleGoogle Scholar

Copyright

© The Author(s). 2017

Advertisement