Communication and mental health in general practice: physicians' self-perceived learning needs and self-efficacy

Abstract

Background General practitioners (GPs) often see patients presenting with mental health problems, but their training regarding mental health treatment varies. GPs' communication skills are of particular importance in these consultations, and communication skills training of GPs has been found to improve patients' mental health. To tailor a communication skills training by basing it on GPs' learning needs and self-efficacy, thereby maximising learning, we conducted a questionnaire study.

Objective To measure GPs' self-perceived needs regarding communication with and treatment of mental health patients.

Methods GPs in training for specialist general practice were given a questionnaire on communication and mental health in general practice. The questionnaire measured prevalence, referring practices, self-efficacy and self-perceived learning need regarding mental health treatment and communication skills, as well as interest in attending training.

Results A majority of GPs in our sample was interested in training on communication skills and mental health treatment. However, they reported moderate learning needs and high confidence on the different measures. GPs reported highest learning needs regarding specific communication skills and treatment of the most common mental health problems. At the same time, they reported highest self-efficacy in treating the same disorders. They also reported high confidence in communication skills.

Conclusion Despite being confident, GPs in this sample recognise the need for specific skills in consultations with patients with mental health problems, but may underestimate the importance of general communication skills. These results are informative when designing training for GPs in communication and mental health.

Introduction

At one time or another, about half of us will experience some kind of mental health problem, and many of us will turn to our general practitioner (GP) for help and support. GPs thus often see patients who do not primarily present somatic complaints. Studies find that, on average, about one third of GPs' patients experience mental health problems.^1,2

One of the challenges for the GP in treating patients with mental health problems is to develop adequate clinical communication skills.³ A GP's communication may influence a patient's mental health through a number of different pathways.⁴ In a much-cited review, Stewart showed that communication skills are associated with several positive patient outcomes such as higher patient satisfaction, better patient adherence and symptom resolution.⁵ Among the skills found to be positively related to outcomes were asking the patient about concerns, eliciting the patient's understanding of causes and expectations for treatment. There is evidence that communication skills training of GPs may have an effect on a number of these skills, but few communication skills training programmes are specifically directed towards the treatment of patients with mental health problems in general practice.^6–9

In spite of the documented importance of communication skills in treating patients with mental health problems and the large number of such patients in general practice, many GPs do not receive specific communication skills training with relevance for mental health after medical school.¹⁰ A number of studies actually indicate significant potential for improvement in GPs' communication skills.^11,12 This includes both general skills, defined here as skills relevant for all types of patients, such as asking open-ended questions and eliciting the patient perspective, and mental-health-related skills, such as handling denial and suicidal risk, defined here as skills applicable to dealing with mental health patients. Moreover, studies conducted in Australia and the United Kingdom (UK) found that many GPs indicated a need to learn better management strategies, but the emphasis in these studies was not to teach specific communication skills.^10,13 Assessment of learners' needs is important, since training based on needs assessment has been found to be more likely to produce changes in physicians' behaviour and patient outcome than training without a base in an assessment of learner needs.^14,15

There are a number of different ways to obtain direct and indirect indicators of GPs' perceived learning needs. One way is simply to ask about the GP's motivation to take part in a course or training programme. A second approach is to map the self-assessed need for increased competence in a specific area or regarding specific skills. A third, more indirect, indicator is to assess self-efficacy, defined as one's belief in one's ability to successfully perform a specific task.¹⁶ Self-efficacy has been applied as a dependent variable in communication skills training, but as far as we know, no studies have explicitly assessed GPs' self-efficacy regarding communication with mental health patients.¹⁷

In the process of planning and designing a communication skills training course for GPs focusing on mental health, we conducted a questionnaire study aiming to identify GPs' self-perceived learning needs regarding communication and mental health. In order to obtain a broad range of needs, we assessed the motivation to take part in a training course, self-perceived needs for increased competence and self-efficacy. We suggest that addressing these self-perceptions in designing a training programme could enhance motivation, adherence, enrolment, and ultimately learning and skill level.

Methods

Sample

All physicians (n = 273) attending a mandatory course for accreditation as a specialist in general practice in Norway were given the questionnaire. The course is five days long, given by a local chapter of the Norwegian Medical Association and focuses on different basic topics in general practice, including communication skills (one full day). The course is given twice each year. Data collection was conducted in 2011 (spring and autumn) and 2012 (spring). The questionnaire was administered and information about our research was given in one of the breaks. Completion was voluntary. Of the total 273 GPs, 220 (81%) responded.

Questionnaire

The questionnaire was designed by the authors. GPs, psychiatrists and psychologists with expertise in clinical communication were consulted during questionnaire construction. We conducted a pilot using a convenience sample of eight physicians to test the questionnaire design and made changes according to their comments.

The questionnaire first covered demographic information, prevalence of patients with mental health problems seen in the GP's practice and referring practice. Respondents were asked to indicate how frequently they meet patients with the different mental health difficulties during 1 week. The list of mental health problems contained nine categories (anxiety/phobias, depression, pain disorders, other complex or psychosomatic disorders, psychosis or other severe psychiatric disorders, grief/loss or relational conflicts, addiction, trauma/abuse and personality problems). Respondents were also asked to indicate how often they refer each patient category to specialist care or others (on a scale from 1 to 5), as well as why. Eight different answer categories were given: severity of disease, time, lack of interest, lack of competence, treatment tried without effect, better treatment elsewhere, patient's own wish and request of the patient's family.

Needs and motivation for further training and competence development were then assessed in five ways:

• Self-efficacy regarding treatment of the nine different mental health patient categories (see above) was measured on a scale from 1 to 10, according to Banduras guide for constructing self-efficacy scales, where 1 was ‘not at all’ and 10 was ‘very much’ (α = 0.86).¹⁸

• Perceived learning need regarding how much they would like to improve their treatment of the same nine categories of mental health problems was assessed on a scale from 1 to 5 ranging from ‘to a small degree’ (1) to ‘to a large degree’ (5) (α = 0.89).

• Perceived need to improve different communication skills was measured by rating each of 20 items on a 5-point scale ranging from ‘to a small degree’ (1) to ‘to a large degree’ (5) (α = 0.90). The list is based on earlier research at our department and contains both general and specific skills (see Introduction for definition).¹⁹

• Self-efficacy regarding six communication skills on a 12-item scale (two items cover each skill) (α = 0.92) was measured on a scale from 1 to 10. These six skills have been defined by our research group in connection with communication skills training. They are developed on the basis of 4habits and GRIP, which are two communication skills programmes used at our department which have been found effective on physicians' communication skills.^20,21 4habits has also been found effective on physicians' self-efficacy.²²

• Motivation to take part in a communication skills training course was measured by answering yes or no.

The questionnaire is available from the first author.

Statistical analysis

Independent samples t-tests were applied to calculate group differences. Paired t-tests were used to test the significance of differences between single item scores. An acceptable level of statistical significance was set to 0.05. Because of missing data, degrees of freedom (df) vary from test to test. Descriptive statistics were applied to describe findings. SPSS 20.0 was applied for all analyses.

Results

Ninety-two (42%, N = 220) of the respondents were males. The mean age was 35 years (SD = 5.9), while mean number of years of experience was 5 (SD = 4.5).

In an average week, GPs' mean number of patients presenting a mental health problem was 33 (range = 5–171, SD = 23.5, median = 28). The mean number of patients within each mental health category is presented in Table 1.

Table 1.

Number of patients seen per week and how often they are referred, perceived learning need and self-efficacy on each mental health category

		Number of patients Mean (SD)	Referring frequency Mean (SD)	Learning needs Mean (SD)	Self-efficacy Mean (SD)
Most frequent	Pain disorders	7.62 (6.32)	2.18 (0.68)	3.92 (0.89)	5.75 (1.70)
	Depression	6.34 (6.24)	2.26 (0.63)	3.86 (0.91)	6.81 (1.52)
	Other complex or psychosomatic disorders	5.70 (6.40)	2.07 (0.72)	3.84 (0.88)	5.44 (1.66)
	Anxiety/phobias	5.38 (6.32)	2.15 (0.70)	3.91 (0.86)	6.16 (1.74)
Least frequent	Grief, loss, relational conflicts	3.00 (3.26)	1.53 (0.59)	3.48 (1.02)	7.21 (1.60)
	Addiction	2.37 (2.51)	2.66 (0.82)	3.52 (1.06)	4.65 (1.98)
	Personality disorders	2.16 (3.02)	2.25 (0.90)	3.52 (1.11)	4.50 (1.99)
	Psychosis or other severe psych. disorders	1.17 (1.40)	3.39 (0.91)	3.49 (1.09)	4.05 (2.11)
	Trauma/abuse	1.00 (1.23)	2.66 (0.96)	3.71 (0.99)	4.79 (1.94)

The severity of the problem influenced the degree of referring. Patients suffering from psychosis and other severe disorders were most often referred (mean = 3.39, SD = 0.91), while patients presenting with sorrow, loss or relational conflicts were least often referred (mean = 1.53, SD = 0.60; t210 = 26.15; P < 0.001). The most listed reasons for referring the different patient categories were ‘treatment tried without effect’, ‘severity of disease’ and ‘better treatment elsewhere’. The least frequent answer was ‘lack of interest’. The most frequent answer was given ten times more often than the least frequent one.

Eighty-nine percent of GPs responded that they might be interested in attending a communication skills course for dealing with mental health patients. GPs' reported interest in course participation was not correlated with previous attendance at such courses (r = −0.06, ns).

Self-efficacy and learning needs on mental health treatment

Reported self-efficacy in handling the different mental health categories as well as self-perceived need of improving treatment of the same problems are displayed in Table 1. Sum scores of learning needs (mean = 3.70, SD = 0.71) and self-efficacy (mean = 5.48, SD = 1.23) were not correlated (r = −0.13, ns). Females (mean = 3.83, SD = 0.73) expressed higher learning needs than males (mean = 3.52, SD = 0.66; t203 = 3.12, P = 0.002). There were no gender differences on total self-efficacy score. However, compared with males, females reported higher confidence in handling sorrow/loss/relational conflicts (mean females = 7.47, SD = 1.45; mean males = 6.88, SD = 1.70; t212 = 7.25, P = 0.006). Compared with females, males were more confident in handling serious diseases (mean males = 4.42, SD = 2.25; mean females = 3.76, SD = 2.00; t213 = 2.27, P = 0.024) and addiction (mean males = 4.99, SD = 1.91; mean females = 4.41, SD = 2.02; t213 = 2.11, P = 0.036). Age was not associated with the degree of self-efficacy, but the number of years of experience was positively correlated with self-efficacy (r = 0.18, P = 0.008) and negatively correlated with learning needs (r = −0.16, P = 0.027).

To reduce the number of comparisons on the data and to look for general trends we grouped the mental health problems into two categories based on frequency (Table 1); most frequent mental health problems (mean = 25.40, SD = 20.42) and least frequent mental health problems (mean = 9.89, SD = 8.02; t195 = 11.94, P < 0.001). GPs reported lower referring rates of the most occurring mental health problems (mean = 2.15, SD = 0.48) than the least occurring mental health problems (mean = 2.48, SD = 0.57; t186 = 6.97, P < 0.001). In addition, they reported higher self-efficacy (mean = 6.03, SD = 1.33; mean = 5.04, SD = 1.40; t211 = 12.05, P < 0.001), but also higher learning needs (mean = 3.88, SD = 0.74; mean = 3.55, SD = 0.84; t208 = 6.53, P < 0.001) on the most occurring compared with the least occurring mental health problems.

Self-efficacy and learning needs on communication skills

Table 2 shows physicians' self-perceived need of improving communication skills. Across the 20 items the mean score was 2.57 (SD = 0.52). The highest learning need was reported for ‘facilitate coping’ (mean = 3.37, SD = 0.91) and lowest for ‘being empathic’ (mean = 1.97, SD = 0.89; t215 = 19.38, P < 0.001).

Table 2.

Perceived need for improving communication skills; in descending order

Communication skills	Mean (SD)
Facilitate coping	3.37 (0.90)
Handle denial	3.23 (0.90)
Handle suicidal risk	3.10 (1.04)
Handle withdrawal	3.03 (0.91)
Tell/give bad news	2.91 (1.09)
Handle anger	2.90 (1.00)
Ask about psychosocial issues	2.89 (0.97)
End the consultation	2.88 (1.15)
Focus on patient's strengths and resources	2.78 (0.98)
Elicit the patient perspective	2.72 (0.83)
Elicit information	2.55 (0.82)
Be open with patient's family	2.50 (0.84)
Be sensitive to cues about worries	2.44 (0.87)
Ask open-ended questions	2.26 (0.93)
Explore concerns	2.17 (0.85)
Be sensitive to emotions	2.16 (0.88)
Avoid being dominant	2.15 (0.92)
Communicate own feelings	2.12 (0.89)
Listen	2.01 (0.90)
Being empathic	1.97 (0.89)

The sum score for learning needs was not associated with age or years of experience as physicians. Females (mean = 2.64, SD = 0.51) expressed higher learning needs than males (mean = 2.48, SD = 0.53; t194 = 2.18, P = 0.031) on total communication skills score.

GPs reported higher learning needs regarding specific skills than general skills. For example, reported learning need on the specific skill ‘handle denial’ (mean = 3.23, SD = 0.90) was found to be significantly higher than the highly ranked general skill ‘give bad news’ (mean = 2.91, SD = 1.09; t214 = 3.93, P < 0.001).

The mean self-efficacy score on the 12 items list measuring confidence in six communication skills was 7.06 (SD = 1.10). Table 3 shows mean score on each skill (items a + b/2). Paired t-tests conducted on means of all six skills showed significant differences in confidence between all skills, except between skills 5 and 6 (tests not shown). Sum score on self-efficacy of communication skills (12 items) was negatively correlated with sum score on learning needs of communication skills (20 items) (r = −0.44, P < 0.001).

Table 3.

Mean self-efficacy score on each communication skill (items a + b/2)

	Skill	Item a	Item b	Mean (SD)
1	Assess emotions	Initiate communication about emotions or worries	Be aware of hints to emotions	7.20 (1.34)
2	Respond to emotions	Respond with empathy	Show the patient that you care	7.84 (1.28)
3	Assess cognitions	Assess the patient perspective	Show interest in the patient perspective	7.35 (1.26)
4	Respond to cognitions	Explain clearly cause of problem	Enhance the patient's understanding of cause–effect	6.85 (1.35)
5	Assess resources/coping	Assess the patient's strength and resources	Enhance patient's own awareness of strengths and resources	6.56 (1.40)
6	Respond to resources/coping	Find strategies for health promotion together	Communicate about coping	6.57 (1.38)

There were no gender differences on sum self-efficacy score. However, at the item level, females (mean = 7.93, SD = 1.55) expressed higher self-efficacy than males (mean = 7.53, SD = 1.31; t212 = 2.03, P = 0.044) on ‘Respond with empathy’. Both age (r = 0.15, P = 0.025) and years of experience (r = 1.14, P = 0.047) were positively correlated with self-efficacy sum score.

Discussion

In this study, we investigated GPs' self-efficacy and self-perceived learning needs regarding communication and mental health treatment applying different methods. At first sight, our findings seem paradoxical. On the one hand, the respondents of the questionnaire generally expressed high interest in attending a course focusing on communication and mental health, however, the reported learning needs varied and were, in general, moderate, whereas self-efficacy was found to be high. GPs reported the highest learning needs regarding specific communication skills and treatment of the most common mental health problems. At the same time, they reported highest self-efficacy in treating the same disorders. They also reported high confidence in communication skills. However, looking more closely at these findings they may not be that paradoxical after all.

Nearly all GPs who answered the questionnaire were interested in attending training. GPs reported a learning need regarding the most common mental health problems, such as depression, anxiety and pain disorders. At the same time, these problems are less frequently referred to secondary care, it therefore seems as though GPs acknowledge the importance of their behaviour in these common consultations. Interestingly, GPs reported high confidence in treating the same disorders. We assumed to find the contrary; an inverse relation between learning needs and self-efficacy. By contrast, our results did not show any significant association between self-perceived learning needs and self-efficacy on mental health treatment. Festinger's theory of cognitive dissonance might explain this finding.²³ According to the theory, cognitive dissonance is a result of two conflicting cognitions, e.g. when what you do (your actions) conflicts with what you think is the right thing to do (your perceptions). This psychologically uncomfortable state will influence us to change one of the cognitions in order to be similar to the other, because that will reduce the dissonance and create consonance. Since the GPs in our study treat most of the mental health patients themselves and report low referring rates, it could create cognitive dissonance if, at the same time, they reported low levels of self-efficacy in treating these patients. We therefore interpret the divergent findings to be the result of an action to avoid cognitive dissonance. However, it is also possible that the GPs' interest in learning about the most common disorders is a result of their seeing these disorders so often. We also believe that GPs find it easier to accept some learning needs than to report low self-efficacy. Because there is always room for improvement, reporting moderate learning needs will probably not create cognitive dissonance among the GPs.

Referral frequencies might explain why GPs reported moderate learning needs as well as low self-efficacy on the most severe mental health problems. These are most seldom seen in general practice, and since GPs often refer these patients they have no skills or experience for improved treatment, or a need to develop themselves. At the same time, the most reported reason for referral was in fact ‘severity of disease’. Low self-efficacy would therefore not create a cognitive dissonance in this case, explained by the high referral rate.

Our assumptions were supported regarding communication skills, as we found that higher confidence on the six skills was related to lower scores on learning needs. The discrepancy in self-efficacy between mental health and communication skills, where self-efficacy regarding communication was found to be significantly higher, is interesting. Because GPs reported higher confidence in communicating than treating mental health patients, it does not seem as though they acknowledge how communication is an important part of mental health treatment. As we found no earlier studies on this topic from general practice, this should be further investigated.

Most of the GPs conveyed a priority to learn more about specific communication skills, such as how to handle denial and suicide risk, skills relevant when dealing with mental health patients. They expressed a lower priority for learning general skills applicable with all types of patients. This is in line with earlier needs assessments. GPs in both Phongsavan et al and Khin et al reported an interest in learning specific skills related to mental health treatment.^13,24 Using a similar list of communication skills, previous research in our department also found a priority to enhance specific skills. That study was, however, done on cancer doctors and not GPs.¹⁹

There are at least two potential interpretations of this replicated finding. It is possible that GPs see specific skills as more advanced than general skills, and therefore perceive the specific skills as more interesting and important. If so, the GPs would be at odds with the research literature as general skills have been associated with positive outcomes.^8,25 Second, the physicians may recognise that general skills are important, but may perceive their own general skills as adequate, and therefore report a low learning need in this area. However, studies find that medical students and physicians are not very good at self-assessment of communication skills.²⁶

Interestingly, ‘facilitate patients' coping’ was the communication skill with the highest score on learning needs, and lowest score on self-efficacy. GPs appear to be uncomfortable focusing on patients' positive resources. One reason might be that the focus on negative emotions is often the predominant one in consultations. However, focusing on what makes us healthy and empowering the patient have become more common in recent years.^27–29 Promoting positive emotions, coping and patient resources have also been found relevant in general practice.^27,30,31

Empathy was the skill GPs reported the lowest need of improving, even though studies find that many patients prefer an empathic GP.³² The GPs also reported the highest confidence on the empathic communication skills, indicating that they do not need to enhance their empathic behaviour. Observational studies, however, find that patients could benefit from increased physician empathy.^33–35

Study limitations

The main limitation of this study is that the GPs answered the questionnaire in an educational setting. In spite of the fact that the accreditation course was compulsory, attending may have influenced their answers regarding motivation for a communication skills training course. It is also possible that the answers are influenced by the GPs' ‘social desirability’, meaning that they answer what they think we want to hear. It is a question of whether one can generalise from GPs under specialisation to GPs in general. The questionnaire is new and should be tested further. It was not possible to explain the huge variation in the prevalence of mental health patients among the GPs in this study. The cause of this large variation might, for example, be explained by personality factors in the physician or the patient, variation in detection rates, and more structural factors like availability of secondary health care, but it might also be a result of the questionnaire's design. Since comorbidity was permitted, some patients may have been counted in more than one category. Some GPs might be more aware of potential mental health problems as well. At last, subjective learning needs must not be interpreted as objective or actual learning needs.

Implications

Our results are interesting and potentially useful in planning training for GPs, however, we acknowledge that measuring self-efficacy alone will not give a reliable indicator of GPs' self-perceived learning needs or their motivation for attending training, and should therefore only be included as a part of several measures of needs.

Specific communication skills are reported as a learning need and should therefore be an essential part of training. Including specific skills will probably enhance learning motivation and satisfaction with training. However, that does not mean that we suggest exclusion of general communication skills from training. On the contrary, studies show both the importance of and physicians' potential for improvement regarding general skills.^11,12 This should challenge educators to specify the general skills to a higher degree, illustrating how important these skills are in handling mental health problems. Documentation of how general skills, like empathy, influence mental health is essential. It would be of great importance to teach GPs how to use communication with a therapeutic intention in consultations with mental health patients. Patient categories to target in training seem to be the most common ones, as GPs score those categories higher on learning needs and lower on referring rates. Such a mixed training, with more focus on specific skills, as well as coping, but with a main focus on how to use communication as part of treatment of common mental health problems would probably ease enrollment and enhance learning motivation among GPs.

Conclusion

GPs in this sample indicate confidence, especially regarding communication skills and treating the most common mental health problems. But at the same time, they recognise the need for specific skills in consultations with patients with mental health problems. The importance of general communication skills seems to be underestimated. Cognitive dissonance, referring frequency and lack of acknowledging how communication can be used in mental health treatment might explain the results. There are several implications of these results in the planning and execution of a communication skills training course focusing on mental health. It is, for example, important to show GPs in training how both general and specific communication skills can be therapeutically used in consultations with mental health patients.