Why don't physicians adhere to guideline recommendations in practice? An analysis of barriers among Dutch general practitioners

Clinical practice guidelines are commonly regarded as useful tools for quality improvement [1]. However, their impact on clinical practice is not optimal. Several reviews have shown that guidelines have only been moderately effective in changing the process of care, and that there is much room for improvement [2‐6]. For instance, general practitioners (GPs) in the Netherlands do not prescribe drugs according to the national guidelines in about one-third of cases, and this figure has stayed fairly constant during the last few years [7, 8]. In addition, levels of adherence vary largely between practices and between diagnoses [7‐9].

To improve adherence to guidelines in practice, an analysis of barriers to implementation of guidelines among target users is advocated [10, 11]. A large number of potential barriers have been identified operating at different levels, such as the level of the practitioner, the level of the patient, the organisational context, and the social and cultural context [10‐14]. A recently conducted review and synthesis of qualitative studies [15] identified six themes of barriers to the implementation of guidelines among GPs: the content of the guidelines, the format of the guidelines, GPs individual experience, preserving the doctor-patient relationship, professional responsibility, and practical issues.

Few studies have focussed on a set of guidelines considering the variety of barriers that should be addressed to improve guideline adherence [12]. In addition, guideline studies often focus on barriers regarding the guideline as a whole, rather than on barriers operating at the level of the individual recommendations within the guidelines [16‐19]. As different recommendations within the same guideline can have different barriers, it might be more useful to focus on barriers of individual recommendations to optimize the strategies needed for implementation of guidelines in practice.

The aim of this study was to identify the perceived barriers towards the use of national guidelines for general practice by focusing on the key recommendations within the guidelines. By analysing multiple key recommendations from a set of guidelines, we aim to identify which barriers occur most frequently across the selection. These findings may be useful for guideline developers as well as for professional organisations in designing tailored implementation strategies.

Our study revealed a broad spectrum of barriers that Dutch GPs perceive in applying the key recommendations of a set of nationally developed guidelines. Although the focus of the barriers differed across guidelines, each key recommendation had a unique combination of barriers. As a consequence, multiple interventions tailored to the specific barriers of the key recommendations are needed to improve the implementation of guidelines in practice.

The most prominent barrier was lack of agreement with guideline recommendations. GPs often disagreed with recommendations because they argued the underlying evidence provided or felt that it was not clear why they should apply them. In addition, they perceived some recommendations not being applicable due to heterogeneity of patient populations. Other studies also demonstrated that lack of applicability is an important barrier to guideline adherence, particularly to patients with comorbidity [18, 28, 29]. Evidence-based guidelines focus on patients with single diseases and often exclude complex patients, which limits the applicability in practice [30‐33]. Further research and efforts are needed on methods to address comorbidity in guidelines in order to improve the applicability of guideline recommendations [31, 32, 34].

Environmental barriers, particularly organisational constraints, were the second most often perceived group of barriers to implementation. These constraints mostly referred to logistic problems within the own practice or within out-of-hours healthcare services. Moreover, lack of collaboration with other types of healthcare professionals was perceived as a barrier in our study, which is consistent with other studies [17, 35‐38]. Improvements can be made by better organising care and by improving multiprofessional collaboration. Standardisation of processes and procedures, and inter-professional agreements on referral and follow-up might be useful.

Dutch GPs are generally aware of the guidelines because they are a fundamental part of the postgraduate training and continuing medical education. This is a strong feature of the professionalisation of GPs that is rooted in the 1980s when the guideline program of the NHG started. Nevertheless, GPs did not know the content well for almost half of the key recommendations in the guidelines selected in our study. GPs might be confronted with too many guidelines, as each year eight to ten new guidelines or updated versions are produced. To improve knowledge on guidelines, it may be useful to regularly conduct sessions among GPs, because the participants in our study appreciated the focus group sessions and considered these as an innovative medium for guideline education and implementation. The effectiveness of interactive education with active involvement and participation has been demonstrated in other studies as well [39‐41].

In our study, we found that guideline factors were a relevant barrier to implementation, which is consistent with previous studies [12, 42]. GPs prefer short guideline recommendations that are easy to understand. The challenge is to produce simple and clear guideline recommendations that also address the complexity of problems seen in daily practice. Presenting guideline recommendations in multiple formats, such as algorithms, one or two page summaries, and electronic web-based versions with hyperlinks to more detailed information might serve the varying needs of physicians and patients [42, 43].

We used an existing framework of barriers to guideline adherence from Cabana et al. [12], and explored whether it covered the full range of barriers perceived by GPs in our study. We suggest that lack of applicability should be a more prominent category, including different reasons such as that the benefits may not outweigh the harms or patients with comorbidity who need special attention. In addition, the external barriers could be extended with some subcategories, as presented in Table 2. Finally, organisational constraints could be subdivided into organisational constraints within the own organisation or practice, those outside the organisation and those between organisations. Other studies also suggested additions to the framework [44, 45].

One of the strengths of our study is that we examined a large set of guidelines produced within one longstanding guideline program. Most qualitative studies have focused on a specific health topic, or studied only one or two guidelines [18, 19, 42, 46, 47], limiting the applicability of their findings. Secondly, we focused on barriers to key recommendations, rather than on barriers to guidelines as a whole. Our in-depth analysis of barriers provides detailed information on potential interventions needed to improve guideline adherence. This information can be used by professional groups or organisations, regionally and nationally, to develop multifaceted interventions, tailored to the individual recommendations in the guideline. For example, to improve the implementation of the guideline on urinary tract infections, it was suggested to develop local protocols for diagnosis in out-of-hours services, as the recommendation on diagnosis (i.e., the use of a dipslide method) did not apply well in these settings. Finally, the findings from our study may be useful for guideline developers in the process of updating the guidelines to raise the acceptance and implementability of the guideline recommendations.

Several limitations should be considered in interpreting our findings. First, we collected opinions from a small sample of GPs, with GPs working in group practices and in towns and small cities being slightly overrepresented [27]. However, the aim of our focus group study was to identify possible barriers qualitatively, rather than quantifying their relative importance among a representative group of GPs. Results from this study will be used as input for a survey to be conducted among a larger sample of GPs in order to quantify our findings. Secondly, we only included GPs and no other healthcare professionals in our focus group sessions. As some of the barriers were related to behaviour of the practice assistants or practice nurses, it might be useful to include these professions in focus group sessions as well.

In conclusion, we identified a wide range of barriers that Dutch GPs face when using national guidelines. Using the focus group method proved to be an effective method to collect information on barriers. Results from this study help explaining why GPs do not adhere to guideline recommendations in practice, and provide useful suggestions for improving adherence. Our study also illustrated that lack of adherence to individual recommendations is related to multiple barriers. A detailed, in-depth analysis of barriers, as conducted in this study, offers opportunities for professional organisations to develop multiple, barrier driven, and tailored interventions to improve adherence in practice.

LACK OF AWARENESS/FAMILIARITY

'Can I be really honest with you? I have never read the guideline, never looked at it, never...'

Guideline Cerebrovascular accident (KR 2)

'I did not know about 160 mg acetylsalicylic acid for the course of two weeks... I always start with 80 mg in patients with stroke.'

LACK OF APPLICABILITY – benefits do not outweigh the harms

Guideline urinary tract infection (KR 4)

'I usually prescribe ciprofloxacin for the course of 10 days, because Augmentin is badly tolerated according to my experience.'

LACK OF APPLICABILITY – not applicable to patient population

Guideline depressive disorder (KR1)

'In practice, you never see patients with depression only or anxiety disorder only. Both often overlap. Then, the management plan is unclear.'

INTERPRETATION/LACK OF EVIDENCE – lack of evidence

Guideline atrial fibrillation (KR3)

'I only do thyroid gland testing. I do not understand the need for testing Hemoglobin and glucose in patients with atrial fibrillation. What's the evidence?'

LACK OF SELF-EFFICACY

Guideline thyroid disorders (KR 2)

'I do not have experience in treating hyperthyroid patients and only see a few of them per year. I think this is not sufficient to build up expertise.'

LACK OF OUTCOME EXPECTANCY

Guideline sleeping disorder (KR 6)

'.as GP in training, I was motivated to stop long term use of hypnotics in patients with a sleeping disorder. But now, people tell me: don't do it, it demands a lot of energy, without any predicted result. Then you start thinking: hands off, leave it.'

INERTIA OF PREVIOUS PRACTICE

Guideline cardiovascular risk management (KR 4)

'The new guideline recommends using systolic blood pressure...in monitoring drug treatment in patients with hypertension. However, I am used to monitor diastolic blood pressure...and then I feel guilty if I see someone with 150...I think that's a big change.'

PATIENT FACTORS – Patient preferences and demands

Guideline rhinosinusitis (KR2)

'There is a tension between the recommendation and patient demands. Patients expect antibiotics. This sometimes causes friction...yes.'

PATIENT FACTORS – Patient ability and behaviour

Guideline asthma among children (KR2)

'Some children perform well in spirometry, but with a very large number the results are totally invalid. Well, with some children it is just not going to work.'

Guideline cardiovascular risk management (KR 7)

'Yes, I try to, but there are always patients that do not show up for follow-up. Always. Also with medication.'

GUIDELINE RECOMMENDATION FACTORS – Confusing/not clear

Guideline asthma among children (KR1)

'I read the recommendation [on allergy testing in children younger than six years] five times, and I still did not understand it!'

GUIDELINE RECOMMENDATION FACTORS – Incomplete/not up to date

Guideline cerebrovascular accident (KR1)

'This recommendation is based on obsolete opinions. You cannot keep patients with stroke at home. All of them should be immediately admitted to hospital.'

ENVIRONMENTAL FACTORS – Organisational constraints (outside organisation)

Guideline urinary tract infection (KR1)

'How to use a dipslide in out-of-hours services on Sunday? Then you need someone who reads the results on Monday. That is really bothersome.'

ENVIRONMENTAL FACTORS – Organisational constraints (within own practice)

Guideline eye inflammation (KR2)

'I would like to reduce antibiotic prescriptions in patients with red eye, but the practice assistant often deals with these patients who ask for a prescription by telephone. The bottleneck is mainly in prescriptions requested over the telephone. There is an important improvement to make there, yes! As the assistant thinks that at any time a prescription is necessary.'

ENVIRONMENTAL FACTORS – Organisational constraints (between organisations)

Guideline cerebrovascular accident (KR 4/5)

'It is unclear what the hospital arranges and what we need to do when stroke patients return to their homes. There should be a formal handoff between hospital and the GP.'

ENVIRONMENTAL FACTORS – Lack of time/time pressure

Guideline cardiovascular risk management (KR 1/2)

'It's great what we could offer in cardiovascular risk management, but it would need full weekdays to realize this in practice.'

ENVIRONMENTAL FACTORS – Lack of/unpractical resources/materials

Guideline sexually transmitted diseases (KR3)

'There are different media, which is unpractical in use...and the media used in cervix streams can only be shortly preserved.'