Barriers to guideline implementation and educational needs of general practitioners regarding heart failure: a qualitative study.

OBJECTIVES: A clinical practice guideline (CPG) contains specifically developed recommendations that can serve physicians as a decision aid in evidence-based practice. The implementation of heart failure (HF) CPGs represents a challenge in general practice. As part of the development of a tailored curriculum, aim of this study was to identify barriers of guideline adherence and needs for medical education (CME) in HF care.
METHODS: We conducted a modified focus group with elements of a workshop of three hours duration. Thirteen GPs collected and discussed together and parallel in smaller groups barriers of guideline implementation. Afterwards they performed a needs assessment for a tailored CME curriculum for chronic HF. The content of the discussions was analysed qualitatively according to Mayring and categorised thematically.
RESULTS: Barriers of guideline adherence were found in the following areas: doctor: procedural knowledge (knowledge gaps), communicative and organisational skills (e.g. time management) and attitude (dissatisfaction with time-money-relation). PATIENTS: individual case-related problems (multimorbidity, psychiatric comorbidity, expectations and beliefs). Doctor and patient: Adherence and barriers of communication. Main measures for improvement of care concerned the areas of the identified barriers of guideline adherence with the focus on application-oriented training of the abovementioned procedural knowledge and skills, but also the supply of tools (like patient information leaflets) and patient education.
CONCLUSION: For a CME-curriculum for HF tailored to the needs of GPs, a comprehensive educational approach seems necessary. It should be broad-based and include elements of knowledge and skills to be addressed and trained case-related. Additional elements should include support in the implementation of organisational processes in the practice and patient education.

Note

The present study was supported by the Competence Network Heart Failure and funded by the Federal Ministry of Education and Research (BMBF-No 01G/0205).

Research Question

Chronic heart failure (HF) is a complex syndrome, which causes high individual and social strains (through limitation of quality of life, emergency hospitalisation, high mortality and costs for health system) [1].

Clinical practice guidelines (CPGs) contain care and treatment recommendations for physicians, which are based on the systematic collection of scientific evidence [2], [3]. Medical care according to the guidelines (using non-medicinal, medicinal and interventional therapy options) and the adherence of the patient to this care are seen as the key to less individual strain and to an improved prognosis [2], [3], [4]. GPs, however, do often not apply HF CPGs for the care of patients with heart failure as, for example, considerable room for improvement was identified in the field of pharmacotherapy [5], [6]. The IMPROVEMENT-study showed that the desired combination of an ACE-inhibitor (or AT1-blocker) and a ß-receptor-blocker (ß-RB) was prescribed in only 20% of all cases in Germany [5]. General reasons for the inadequate implementation of evidence have been sufficiently identified: they can be found, amongst others, in the fields of personal and professional experience of GPs [7]. A systematic review on barriers to the adherence/application of guidelines (physician adherence to clinical practice guidelines) analysed 76 articles (120 examinations), categorised the identified barriers and aligned the results to the sequence of behaviour change (knowledge – attitudes – behaviour) [8]. Seven generic categories on the part of the physicians were formed: lack of knowledge, lack of attention, lack of general and specific approval, lack of expectation of oneself, lack of self-efficacy and lack of motivation. Furthermore, external barriers were identified concerning the expectations of the patients, the quality of the guidelines and environmental factors with reference to the organisation of medical care. The authors drew the conclusion that measures to improve the implementation of guidelines require an accurate analysis of the barriers and the needs. In primary care in England, barriers to HF CPG implementation were identified in diagnostics and in management, which result, amongst others, from insecurities in its diagnosis and therapy [9], [10], [11], [12].

Interventions (in this case: strategies for guideline implementation) should be directed to the need for improvement and the barriers to implementation. Most of the implementation studies of the last two decades focused mainly on the “target group physicians”, on the “quality of guidelines or findings” and on the “form of imparting findings”. They mostly dealt with a modified special training for physicians and thus followed a “multiplier-model”. Most of these interventions offer - as they grow more complex and more intense, e.g. through a combination of additional trainings, reminders, feedback on prescription behaviour, management support - higher effects than the common, usually teacher-centred trainings; however, they are usually more expensive and time-consuming [6], [13], [14], [15].

Due to different frame conditions in a health care system of a country, such studies need to be carried out in their own context. The aim of our study was an analysis of barriers and needs according to Kern et al. [16] or, more precisely, the second of a total of six steps towards the development of a curriculum for medical education. We wanted to identify barriers to implementing HF CPG recommendations and the educational need of German GPs in this regard. The findings were part of the development of a tailored medical education intervention for GPs combining educational strategies with feedback elements [13] to be tested with respect to efficacy on physician (medical knowledge and medical care) and patient outcomes (ISRCTN08601529) [17].

Methods

As research method we chose a focus group with elements of a workshop. It was chosen in order to detect barriers regarding the implementation of guideline recommendations and the need for special training of GPs. According to Kern and Kitzinger et al. focus groups are an efficient method for disclosing attitudes, opinions and needs of those concerned [16], [18]]. Thus, large barriers and optimisation approaches, which could not be reproduced by using a standardised questionnaire, can be made more transparent. One feature of facilitated focus groups is that the participants have the opportunity to discuss certain problems more intensively. As a start, we directly addressed a heterogeneous sample group of 13 GPs from the Heidelberg region for this qualitative study. These GPs had a background of different experiences. Some of them had already taken part in projects of the section “General Medicine” of the department for General Clinical and Psychosomatic Medicine or were, because of their working in an academic teaching practice, more familiar with medical education than others. All of the 13 GPs addressed took part in the discussion in July 2004 and received an allowance of 50 Euros. The description of the sample can be seen from Table 1 (Tab. 1). All in all, the discussion of the focus group took 3 (2x1,5) hours. The whole group was briefly introduced to the topic of the study. Afterwards it was divided into three (for answering the first three key questions) and later two small groups (for answering the fourth key question, see below).

Table 1

Features of the participating GPs (n=13)

The four key questions were:

Which are the five most important barriers to implementing chronic heart failure CPGs?

Name five suggestions of how to improve the care of patients with heart failure.

Name five training methods from which you would profit the most.

What should a two-day training curriculum for your trainee contain?

This question was posed because the identification of one’s own need for training might be difficult [19] and the articulation of one’s own deficits in front of a group could be shame-connected.

The participants were allowed to give several answers, but they should prioritise them. The purpose of the question was to identify the most important barriers and areas for improvement in terms of a hierarchy. As a start, the participants worked individually on the first three questions and wrote their answers on small cards, which they exchanged within the group and discussed the results (in the presence of SL, TMT and FPK). Afterwards, one member of each small group presented the results to the whole group using the metaplan method. Those results were discussed with the help of a facilitator (SL). The discussion was recorded in the minutes (TMT, NL, FPK). Furthermore, the principle of rigour of Flick [20] was applied, i.e. only the relevant aspects should be recorded. We then proceeded with the summarizing qualitative content analysis according to Mayring. Initially, the statements of the physicians were paraphrased and equivalent statements were deleted (IN, FPK und JJ). Afterwards, main and sub categories were formed [21], [22]. A system of categories could be developed on the basis of the text material. In order to make the elements and key points of the future training intervention visible and predictable, the system of categories was tested: In how far could it be related to the physician, the patient (or the interaction of both) or other fields? The traditional dimensions “knowledge, skills and attitude” could be established as sub-categories. The developers of the training intervention (FPK, TMT, NL and JJ) applied a synopsis of the results to the revised taxonomy of educational objectives of Bloom, i.e. according to Anderson and Krathwohl [23], [24], in order to answer the question of how a medical training should be organised. The use of such a categorisation was the concretisation of intended learning outcomes, while taking the learning perspectives more into account. It should also help to clarify a possible broader use of the educational methods of the intervention to be developed. The required level of learning targets can be shown in a two-dimensional way using the abovementioned taxonomy: Accordingly, the dimension “continuum of cognitive processes” (based on Bloom) will be divided into knowledge (to recall relevant knowledge from memory), comprehension (identify importance or relevance of knowledge), application (choose a specific course of action in a given situations), analysis (classification of material), evaluation (judging according to certain criteria) and creation (organise a new, working pattern). The second dimension can be divided into “different types of knowledge”, i.e. factual, conceptual, procedural and metacognitive knowledge. The taxonomic construction implies that categories of higher levels include proficiency in the category below. That means that we are talking about a cumulative hierarchy here. The writer of the minutes carried out the categorisation on the basis of the moderated presentation of the participants’ results. The persons responsible for the development of the training intervention (FPK, TMT, NL, JJ) gave their consent at a later date.

Results

The results can be divided into two fields: barriers and areas of improvement and training contents for residents or GPs.

The participating physicians identified barriers to implementing guidelines, which involve the influenceable categories knowledge, skills and attitude (of the physician) (see table 2 (Tab. 2)). The participants particularly mentioned the lack of access to “neutral procedural knowledge” (without any influence by other’s interests) and the lack of access to “practicable guidelines”, which are not taught at traditional CME events. Furthermore, individual problems in the organisation of everyday life (“lack of time”, “stress”) and in practice management were disclosed. The attitude of the physicians hinted at potential barriers, too (“idleness”, “to have one’s own guidelines in mind”). The participants saw other barriers, which are rather caused by patients than physicians, in individual problems (“multimorbidity” or “insufficient adherence to medical advice”). “Personal truth”, “rejection of conventional medicine”, “expectations” or “missing distress” were further barriers to the implementation, which could be assigned to the attitude or the subjective perception of patients. One of the main problems in implementing guidelines were difficulties in the doctor-patient-interaction, e.g. because of psychological co-morbidity or insufficient adherence of patients.

Table 2

“barriers to and areas of improvement in implementing guidelines” from the perspective of the participants

The participants’ suggestions for improvement referred mainly to the development of a CME for GPs, in which not only knowledge is imparted, but also the training of skills like, for example, physician-patient communication in difficult situations and efficient practice organisation with regard to individual problems is integrated. Furthermore, the GPs wished corresponding patient education to foster self-care behaviour (“empowerment”). In addition, the participants considered media (“information material, patient ID”), “early interdisciplinarity”, the inclusion of family members and Public Relations to be helpful (see table 2 (Tab. 2)).

The postgraduate training programme for GP trainees, which was developed by the participants, is contrasted with suggestions for an additional training for GPs (see table 3 (Tab. 3)). From the participants’ perspective, GP trainees should complete a two-day training, which contains medical, communicative and organisational elements of knowledge and skills with emphasis on peculiarities for GPs, e.g. individual practice-prevalence or availability of specialist expertise (in ambulatory and stationary sector). “Modern didactics and multimedia-based teaching material” should be used. The participants’ wish for a consistent CPG, for practicing difficult situations in physician-patient-contact and for the acquisition of a practical time and quality management became evident. For more in-depth training, the group wished for interdisciplinary quality circles in order to discuss difficult aspects in CPG implementation with the help of concrete cases.

Table 3

“Preferable content for GP trainees” from the perspective of the participants (with categorisation according to revised taxonomy of Bloom [23;24])

Discussion

In this qualitative study of the perceived barriers of and educational needs for optimal guideline-oriented care of heart failure (HF) patients participating GPs saw barriers to the implementation in the fields of knowledge, communicative and organisational skills and attitudes on part of the physicians. The patients with their mindset and their expectations as well as the interaction between physician and patient pose problems as well. Most of the mentioned barriers were taken up again by the participants when they talked about the areas for improvement (see table 2 (Tab. 2)). The mindset and the expectations of the patients were not mentioned again, which may therefore be seen as dimensions that can hardly be influenced.

While the participants of the group “GP trainees” favoured a comprehensive training, which should offer a broad range of knowledge and skills elements, they saw a need for training of the group “GPs” in the organisational field and in difficult communicative situations. Additionally, they wished for interdisciplinary quality circles with case-audits.

The aim of this survey was to identify the essential improvable areas for primary care of patients with heart failure in Germany. The survey should also serve as a component for the development of an innovative medical education intervention to be implemented and evaluated later. As there is scarce German literature about barriers to implementing HF CPGs, and the educational need to be derived as part of a development of an innovative CME intervention for GPs, we will compare in the following discussion our results with the results of international studies about barriers to the physician adherence to (HF) CPGs in primary care. Finally, we will compare our findings with the generic review about barriers to physician adherence to CPGs we summarised at the beginning [8], [9], [10], [11], [12].

Knowledge, skills and attitudes of physicians

A lack of knowledge and of awareness with regard to HF CPGs could be found in our, but also in international studies [9], [10], [11], [12].

Deficits in knowledge and insecurity of GPs concerning diagnostics as well as evidence-based medicinal therapy and its observation were mentioned in all studies:

Although the participants of our study did not explicitly mention any barriers with regard to the diagnostic process, they discussed the topic “step-wise diagnosis while taking low prevalence into account” with relation to the design of a training course instead. The value of anamnestic, clinical signs and additional examinations like echocardiography were addressed internationally [9], [10], [11], [12]. The focus was on the availability of diagnostics (echocardiography), which – from the participants’ perspective – also plays a role in rural areas in Germany [9], [11], [12].

Unsureness in the initiation, dosage and follow-up of an evidence-based pharmacotherapy with ACE-inhibitors [9], [10], [11], [12] and beta-blockers [10], [11] were found, which is in line with our results (not shown).

Since the organisation of the practice is the responsibility of the office-based physician and since skills can in principle be supported by training, we added the points mentioned to the barriers on physician-level: lack of time or time management and the definition of organisational procedures in the practice are barriers or rather challenges that were identified internationally as well [10], [11]. In other studies they were sometimes categorised as external barriers or environmental factors [8] because the physician does not necessarily have barriers in either knowledge or attitude, depending on circumstances (e.g. physician is employee), and still there is a lack of care according to the recommendations in the CPGs. Therefore, Cabana et al. count patient factors (preferences), guideline factors (different, contradictory recommendations) and environmental factor (e.g. lack of time and resources, organisational shortages, insufficient remuneration for services) among the external barriers to guideline adherence of physicians.

Among categories of attitude we found motivation problems [8], [10], which can be a peculiarity of the physician (“idleness”), but they can also partly be caused by external factors (discontent of the physician with the relation of effort and effectiveness from an economic point of view) [8], [10]. High costs of pharmacotherapy and thus a higher recourse risk, which can be connected to the care by a GP, were not mentioned in our inquiry, but in literature [11], [12]. Whether this discrepancy is a phenomenon of social desirability, can not be conclusively assessed. Moreover, some physicians seem to give preference to their own experiences rather than to evidence from guidelines. Here, categories of knowledge seem to overlap with categories of attitude, maybe due to experiences of physicians with patients [8].

Patient level and physician-patient-interaction

From the physicians’ perspective, individual barriers were the willingness or ability of the patients to adhere to treatment recommendations, which could be improved by good education and guidance of the patient. The given barriers regarding attitudes were not mentioned in the specific qualitative studies, but categorised as “patient factor” in the generic systematic review [8]. Further barriers in communication, which were found in our study, can be supported by the interview study with patients of Rogers et al. [25]. The participants expressed their scepticism towards the transferability of guideline recommendations on the typical patient spectrum of a GP’s practice, which is characterised by elderly patients with co- or multimorbidity. They suspected side-effects of the therapy to occur [9], [10], [12].

Areas for improvement and training contents

The aspect of patient-specific organisation on practice-level (effective time management, information material, patient-ID, implementation of procedures in the practice programme), which was addressed in this study (see table 2 (Tab. 2) and 3 (Tab. 3)) was a point that did not directly match with the literature found. It provides, however, an answer to the explicitly posed question of improvement possibilities, which the barrier-studies did not ask for. Comparing the statements of the participating practice owners regarding the contents of further education and additional training (see table 3 (Tab. 3)), we see that they regard all knowledge dimensions and competence levels as important for GP trainees. For themselves, they prioritised application-orientated procedural knowledge and the training of skills. We interpreted this result as partly comprehensible and partly as a possible phenomenon of social desirability or the missing identification of a gap in personal knowledge. Therefore, we regard the statements concerning the GP-trainees as helpful for the conception of a medical education course. Furthermore, these results stand in contrast to the currently common frontal-oriented CME events that necessarily impart knowledge of facts [http://degam.de/fortbildung/dokumente/DEGAM-Fortbildungspapier2009-Langversion-Hintergrund-Papier.pdf].

The present analysis of barriers and needs served as a practicable instrument for the development and design of an innovative CME training for GPs, which was helpful to plan the next steps according to Kern et al. (definition of goals, strategies in teaching and learning) [16].

We chose a practicable approach that was to systematically record a formally still open, unstructured subject. It was also to enable a comparison of German conditions with findings from international literature. In this respect, open questions were posed, which should comprise the size of the subject area and be categorised systematically later on. The selection of the GPs, the small number of cases and the method of collecting data have to be mentioned as limitations. The participants were almost representative according to the criteria age, gender distribution and (not entirely) type of practice in Baden-Württemberg (cf. for example healthcare reports of 2007 and 2008: http://www.kvbawue.de/presse/publikationen/versorgungsbericht). But it was shown that in all three parallel groups almost the same results were achieved and therefore saturation with regard to the developed points was at least partially achieved. The congruence of our findings with the international literature supports the assumption that all important barriers have been identified. It cannot be ruled out that existing barriers, which are country-specific, were not identified. The purpose of this study in a project context was, all in all, fulfilled. Therefore, we consciously refrained from an extension of the sample in terms of another focus group.

Conclusion

The needs analysis for a CME programme for GPs regarding heart failure achieved numerous, systematically developed results, which, to a large extent, correspond to the international literature. Parts of the most important features of the desired training were patient- and practice-related integration of elements of knowledge and (communicative and organisational) skills, which should be trained in a user-oriented way.

Funding

Federal Ministry of Education and Research, Competence Network Heart Failure (BMBF-No 01G/0205).

Acknowledgements

I wish to thank all the general practitioners for their participation. I would also like to thank Mrs. Dr. phil. Dipl. Psych. Wilke and Mrs. Dipl. Psych. Nadja Ringel for their critical review of the first analyses and drafts.

Competing interests

The authors declare that they have no competing interests.