Does the need for professional competencies change during the physician's career? – A Finnish national survey

Abstract

Background: Competence consists of a range of skills, knowledge and attitudes that physicians utilize in their work. Different models for defining physician competency areas have been used in medical organizations. The goal of this study was to explore how Finnish physicians perceive the need for different competency areas in their work.

Methods: The data for this study were collected in a national questionnaire administered by the Finnish Medical Association (response rate = 63%; N = 10,624). The competency framework was derived from the CanMEDS framework (seven areas) and detailed into 11 items focusing on different aspects of phycicians’ work. The participants were asked to evaluate how much they needed different items in their work.

Results: Factor analysis identified three broad competency dimensions: (1) medical knowledge, (2) management skills and (3) interpersonal skills. There were differences in the need for these competency dimensions according to work assignment and age, reflecting occupational status and amount of work experience and specialty.

Conclusions: The results were in agreement with the theoretical framework, but the factor analysis compressed the competency areas into three broader dimensions. This study suggests that different positions require different competencies from physicians. Therefore, if physicians are assessed, they should be assessed in accordance with their work.

Introduction

By definiton, competencies indicate an individual's ability to acquire skills in a given area. Within medicine, these include for example basic clinical skills, scientific knowledge and moral judgement (Epstein & Hundert 2002) and the ability to apply these skills in a clinical environment (Ten Cate et al. 2010). As treatment pathways become more complex and clinicians have to do more decisions based on economical limitations, the work environment is becoming increasingly challenging. Therefore, medical skills require the accompaniment of a vast number of other competencies such as economic reasoning, computing skills and leadership skills.

Health care is under constant evaluation including assessment of physicians’ performance and competence. Growing requirements have lead to long lists of important skills and the medical profession, among many others, have issued core competencies to summarize the central areas. These core areas form the higher order competencies, such as clinical skills in general, whereas subcompetencies include specific skills, for example, a certain operation (Swing & International CBME Collaborators 2010). Typically, core competencies consist of six to eight similar competencies like CanMEDS in Canada (Frank 2005), ACGME in USA (Accreditation Council for Graduate Medical Education 1999) and various listings of specialist boards in Europe (Palsson et al. 2007). All these models emphasize medical knowledge, communication skills, research, learning, professionalism and management skills. Medical knowledge rises from medical research and is essential to health care professionals, whereas other competencies, such as communication skills and management, are shared with many other professionals. Similarly, competencies are in constant evolution and organizations are eagerly creating long lists of needs for competencies.

This study examines what kinds of constructs can be found in Finnish physicians’ self-evaluations of skills that they need in their work, and how these constructs relate to currently used competency areas. Moreover, this study focuses on how self-perceived competencies differ between gender, medical specialties and physicians in leading positions and in performing positions. Third, it examines whether the need for competencies changes with ascending career.

Table

Practice points

Different models for defining physician competency areas have been used. The goal of this study was to explore how Finnish physicians perceive the need for different competency areas in their work. Three competency dimensions were identified with factor analysis: Medical knowledge, Management skills and Interpersonal skills. The need for competencies varied according to age, gender, specialty and position.

Methods

The survey

A questionnaire was mailed to all 18,918 physicians who were licensed in Finland in 2009 and who were not retired until March 2009. Of the 13,618 physicians (72%) who returned the questionnaire, 11,959 completed the questions about competencies, yielding a response rate of 63%. Information on age, gender, specialty and position (leading and performing position) was requested and collected from the register of the Finnish Medical Association. Distributions of age, gender and specialty of those who returned the questionnaire corresponded to the whole population.

Questionnaire

The questionnaire was constructed on the framework from CanMEDS-competency areas (Frank 2005). A list of physicians’ work areas was identified from different publications and translated into Finnish. The participants were asked to rate how much they needed different competencies in their work according to the following instructions: ‘How much do you need the following areas in your work assignments?’ A 7-point Likert-scale ranging from not at all (1) to constantly (7) was used. The items were as follows (CanMEDS areas are in parentheses):

• Medical knowledge (Medical expert)

• Managing information (Medical expert)

• Acquiring new knowledge (Scholar)

• Interpersonal skills (Communicator)

• Collaboration (Collaborator)

• Professionalism (Professional)

• Managing one's own work (Manager)

• Managing work processes (Manager)

• Computer skills (Manager)

• Economical skills (Manager)

• Health advocacy (Health advocate)

Statistical methods

Missing value analysis revealed no missing values concerning age, gender, specialty and position. One variable, professionalism, was dropped out because 6.9% of respondents had not provided an answer to this item. The high number of missing values in one variable resembled intentional non-responding and there is no reliable method of handling such missing data properly (Donders et al. 2006). After this, cases with more than one missing value (N = 105) were deleted, resulting in a total number of 11,854 participants. The remaining single missing values were replaced with an expectation-maximization algorithm. The amount of replaced observations was 268 adding up to only 0.24% of all the observations. The variable that had most missing values was ‘Managing work processes’ with 116 (1.0%) missing values.

The Leading position category (N = 2114) was formed from chief physicians and deputy chief physicians regardless of their employer. Rest of physicians were regarded as the Performing position category (N = 9740) including specialists, senior physicians, general practitioners, physicians in specialist training and others. The Eurostat (2010) classification for specialties with eight categories was used to compare specialties. Those without a specialty were grouped under non-specialists.

Data analysis was conducted in three stages. First, an exploratory factor analysis was carried out on the 10 items in order to sort out the structure and number of competencies. A principal axis factoring with varimax rotation was used and factor scores for each factor were used for further analysis. Second, a multivariate analysis of variance (MANOVA) was carried out to test for the main effects of sex, specialty and position (leading and performing position). Third, one-way analysis of variance (ANOVA) were conducted to find out whether the physicians differed in their evaluations of their competencies in terms of sex, specialty and position. Post hoc analysis for individual differences between specialties was made with a Tukey's B. Effect sizes were presented as partial η² for the MANOVA and as η² for the ANOVA analysis. A p < 0.01 was used as level for significance in all analyses.

Results

Factor structure

The 10 items concerning evaluations of competencies were submitted to a principal-axis factor analysis with a varimax rotation specifying three factors. An eigenvalue of 1.00 or more was used as a criterion for selecting the number of factors. As shown in Table 1, the first factor reflected Management skills, with strong loadings on Managing work processes, Managing one's own work and Economical skills. The second factor reflected Interpersonal skills and had strong loadings on Interpersonal skills and Collaboration. The third factor indicated Medical knowledge and had strong loadings on Managing information, Acquiring new knowledge and Medical knowledge.

Table 1.  Loadings for the three factor-solution and correlations between three factors

	Factor
	Management skills	Interpersonal skills	Medical knowledge	Communality
Managing work processes	0.81		0.22	0.74
Managing one's own work	0.59	0.31	0.28	0.52
Economical skills	0.51			0.27
Computer skills	0.46		0.40	0.39
Interpersonal skills		0.94	0.21	0.95
Collaboration	0.32	0.69	0.29	0.65
Health advocacy	0.30	0.37	0.25	0.29
Managing information	0.35	0.20	0.73	0.70
Acquiring new knowledge	0.21		0.69	0.53
Medical knowledge		0.22	0.46	0.26
Eigenvalue	4.4	1.1	1.0
Percentage of variance	43.6	11.4	10.2
Correlations between factor scores	1
Interpersonal skills	0.037*	1
Medical knowledge	0.175*	0.063*	1

Note. Primary loading values are given in bold; the secondary loadings are shown only if they exceeded 0.20.

*p < 0.01.

The 11-item three-factor solution fitted the data fairly well. The three factors accounted for 65% of the variance. After the factor analysis, the factor scores for each variable were used in further analysis.

Need for competencies in different groups

Differences for the experienced need for competencies in different groups were investigated with a MANOVA. The model considered gender, specialty and position (leading versus performing) as independent variables and Management skills, Interpersonal skills and Medical knowledge as dependent variables. With the Wilks’ criterion, the combined dependent variables were affected by sex F(3, 11,820) = 30.5, p < 0.01, partial η² = 0.008 (99% confidence interval, CI from 0.004 to 0.012), specialty F(21, 33,941) = 17.3, p < 0.01, partial η² = 0.010 (99% CI from 0.008 to 0.012) and position F(3, 11,820) = 31.3, p < 0.01, partial η² = 0.008 (99% CI from 0.004 to 0.012). Multivariate interactions between the independent variables were not statistically significant.

After the MANOVA, univariate ANOVA with F-tests for the individual effects was selected. The results of the analysis are shown in Table 2. ANOVA was selected because, as can be seen in Table 1, the correlations between the three factors (dependent variables) were very low. In addition, the MANOVA analysis did not indicate any significant multivariate interactions between the independent variables. As a result, it was concluded that testing for individual effects with ANOVA would produce the results in a reliable and readable form.

Table 2.  Mean (99% confidence interval), ANOVA significance analysis for differences and effect sizes (η²) for the experienced need for the three competency factors in different groups

				Management skills			Interpersonal skills			Medical knowledge
	N	% male	Mean age	Mean		η^2	Mean		η^2	Mean		η^2
Sex
Male	4835		48.8	−0.01	(−0.03 to 0.03)		−0.19	(−0.23 to −0.15)		−0.14	(−0.17 to −0.10)
Female	7019		44.6	0.00	(−0.02 to 0.03)		0.13	(0.10 to 0.15)		0.09	(0.07 to 0.12)
Difference between sexes						0.000			0.026*			0.018*
Speciality
General practice	1348	41.5	51.3	0.19	(0.14 to 0.24)		0.15	(0.10 to 0.21)		0.01	(−0.05 to 0.06)
Paediatrics	444	31.8	49.9	0.13	(0.03 to 0.23)		0.19	(0.11 to 0.27)		0.05	(−0.06 to 0.14)
Obstetrics and gynaecology	461	24.9	50.9	0.06	(−0.05 to 0.16)		0.11	(0.01 to 0.21)		0.01	(−0.09 to 0.12)
Psychiatry	901	32.4	50.7	0.05	(−0.02 to 0.12)		0.38	(0.32 to 0.44)		−0.32	(−0.40 to −0.24)
Medical group of specialities	2446	49.8	49.7	0.06	(0.01 to 0.10)		−0.15	(−0.21 to −0.10)		0.08	(0.03 to 0.12)
Surgical group of specialities	1422	61.0	49.1	−0.01	(−0.07 to 0.05)		−0.13	(−0.20 to −0.06)		−0.12	(−0.18 to −0.06)
Other specialities	587	42.8	51.5	0.39	(0.32 to 0.47)		0.010	(0.01 to 0.18)		−0.01	(−0.09 to 0.08)
Not specialised	4245	32.8	39.3	−0.17	(−0.21 to −0.14)		−0.04	(−0.08 to −0.00)		0.06	(0.02 to 0.09)
Difference between specialities						0.034*			0.025*			0.017*
Position
Leading position	2117	58.7	52.3	0.28	(0.24 to 0.32)		−0.03	(−0.08 to 0.02)		−0.08	(−0.12 to −0.03)
Performing position	9737	36.9	45.0	−0.06	(−0.08 to −0.04)		0.01	(−0.02 to 0.03)		0.02	(−0.01 to 0.04)
Difference between positions						0.022*			0.000			0.002*

Note. Highest values (mean >0.2) are given in bold, and the lowest values (mean <0.2) are given in italic.

*p < 0.01.

As can be seen in Table 2, females evaluated the perceived need for Interpersonal skills and Medical knowledge higher than males, but there were no differences concerning Management skills. The specialty exhibited largest differences in the need for Management skills and Interpersonal skills. Post hoc analysis with Tukey's B indicated that the non-specialists differed from others with a low need for Management skills, while physicians in the group of Other specialties had higher values than others. Further on, physicians in Medical group of specialties, Surgery and Not specialized had lower need for Interpersonal skills, while Psychiatry stood out with high need. Concerning Medical knowledge, the post hoc analysis separated Psychiatry as a group with low perceived need. As expected, the physicians in Leading position evaluated the need for Management skills higher than the physicians in performing position.

While the need for Interpersonal and Medical knowledge remained on the same level, the need for Management skills increased substantially with age, as can be seen in Figure 1.

Figure 1. Factor score means by ages (99% confidence limits).

Discussion

The goal of this study was to explore how Finnish physicians perceive the need for different competency areas in their work. Three broad competency dimensions were identified with factor analysis: Medical knowledge, Management skills and Interpersonal skills. The need for competencies varied according to age, gender, specialty and position.

Do three areas make sense?

In principle, core competences must be general, whereas skills should be defined as specific and measurable. Though core competencies can be listed using the common sense, perhaps empirical input into the framework might be useful. Some practical reasoning lie in this as core comptetencies are seen as five to seven areas, but available assesment tools cannot measure the competencies independently. They either produce a single dimension of general competency or two to four areas that lack simple correspondence to the core competencies (Lurie et al. 2009). So, there is a tendency that empirical data simplifies the framework rather than complicates it.

In this study, the physicians’ evaluations of the need for different competencies are reduced into three categories: Management skills, Interpersonal skills and Medical knowledge, which could also be termed process, communicative and intellectual abilities (Epstein & Hundert 2002). Similar results of condensation of the competency categories have been found in research focusing on evaluations of residents’ performance. Factor analyses have usually identified two to four factors, which reflect clinical skills and professional behaviour (Verhulst et al. 1986) or clinical knowledge, patient data gathering and interpersonal skills as one (Arnold et al. 1984) or two separate factors (Dowaliby & Andrew 1976). We did not come across studies testing competencies among specialists in the manner we did, as most studies focus on the efficacy of assessment of these domains.

In their review article focused on ACGME competencies in graduate education assessment, Lurie et al. (2009) concluded that most evaluation methods were unable to separate different competencies. The only evaluation method that separated competencies consisted in global rating forms, which are summaries of rating forms that evaluate trainees’ abilities over multiple occasions. These did not, however, produce the six separate competencies, but two dimensions; medical knowledge and interpersonal skills (Silber 2004) or four; medical knowledge, interpersonal skills, patient care and professionalism (Brasel et al. 2004). The four factors (Brasel et al. 2004) correlated rather strongly, suggesting that these factors overlap and are reflections of a general competence. One must bear in mind that the previous studies focused on undergraduate students, making it understandable that management skills did not appear among dimensions.

Different physicians have different needs

Medical specialites can be seen as a spectrum of professions from craftsmanhip to humanities. Specialties such as general practice or paediatrics involving long patient relationships and determining care pathways, report a larger need for management skills than those purely dealing with patients with a specific problem. This is reasonable, though we did not measure their performance, but rather the perception of needs. Physicians in the group of other specialties reported a very high need for management skills. Typically, these are small specialties with one or two physicians per hospital in charge of managing the local care pathways. Psychiatry obviously requires good communication skills, and accordingly psychiatrists report a high perceived need for those as well. When developing definitions and assessments of competencies within organizations, we should recognize that different specialties will require these competencies in a different manner. A more general approach in core competencies and agreed expectations could foster coherence as well as accountability (Leach 2008).

Non-specialists are mostly young persons in specialist training or entering it. This group reports low needs for managament skills and communication skills but high needs for medical knowledge. Their position in the work environment does not require holistic organizational planning, but rather patient work in a successive manner. Balmer et al. (2008) found that physicians in postgraduate education perceive medical skills and knowledge as fundamental, but learning to improve their patient care practices or to comprehend the health care system as less important. Working experience, changes in work environment and climbing the organizational ladder will push individuals to perceive the need for more general competencies as seen in this study: the need for management skills increases with age and position. Alas, in Finland, neither graduate training nor specialist training includes sufficient skills for process management or economical reasoning. This needs to be taken into account when assessing physicians: work environment and career development phase are major external driving forces in setting the detailed targets for competencies.

The questionnaire used in this study does not describe competencies extensively, and is based on the CanMEDS framework. The results are descriptive and should not be considered as precise measures. This study, however, introduces new perspectives to the professional competencies of physicians. First, it widens the results of studies on residents’ evaluations by revealing that the same dimensions, medical skills and interpersonal skills emerge from with practicing physicians. However, on the basis of this study, a factor of management skills can be added to those two. It should be noted that because management was addressed with four items, which was more than with other CanMEDS areas, it was likely to produce a separate component in the factor analysis. This does not, however, reduce the reliability of that factor. This study also shows that the need for different competencies differs in different specialties, age and task. The evaluations and support given to the physicians’ continuous professional development should therefore be adjusted along with the assignments.

Conclusions

There is a constant strain in professional assessment between practicality and comprehensiveness. Practicality prefers less core competencies with less strict subcompetencies. Research from other professional fields indicates that assessment of core competencies could include three to four items, which is supported by this study.

Declaration of interest: Kristiina Patja has received royalties from publishing house of Medical Society Duodecim from books.