Guidelines not for everyone.

I commend the editorial on electronic innovation in the implementation of clinical guidelines.1 While clinical guidelines ‘do not replace clinical judgement’ and ‘their application must be individualised to each patient, as they may not be appropriate for all patients’, the editorial highlighted that ‘only about half of all people with established cardiovascular disease are prescribed guideline-recommended treatments.’

What should be the expected rate of prescribed guideline-recommended treatments in a population? It varies with cultural, socioeconomic literacy rate and access to healthcare. Individuals have different outlooks or perceptions and consequently risk appetite which determines their actions. Others need time to deliberate on issues presented to them and may not decide immediately to take up offers of treatment. In shared decision-making, it is expected that some will not take up guideline-recommended treatment regardless of the quality of information provided. Given that compliance, defined as ‘the extent to which the patient’s behaviour matches the prescriber’s recommendations’,2 is nowadays regarded as paternalistic, expectations of near 100% uptake by patients of guideline-recommended treatment would be contentious and unrealistic. Most countries face similar issues in chronic conditions like cardiovascular diseases.3

Measuring the prescription rate of guideline-recommended treatment does not acknowledge any doctor–patient discussion which does not result in that treatment. This is particularly relevant if prescribing rates are used to judge the performance of health professionals regardless of electronic clinical decision support.

Beyond guideline-recommended treatment uptake lies the matter of adherence previously discussed in Australian Prescriber.4 Both issues present similar challenges. Not achieving a high uptake or adherence to guideline-recommended treatment should not be attributed predominantly to the clinical practice of doctors.

The author of the editorial comments:

I agree that we should not be aiming for 100% ‘compliance’ with guideline recommendations. Indeed, if that were obtained, there would undoubtably be concern about overtreatment and failure to individualise therapies. In general practice, multimorbidity is the norm and so clinicians take into account a number of variables, including patient preference, when considering their prescribing decisions. Taking these factors into account though, a translation of guideline-recommended care of only 50% suggests that there are significant barriers which may be attributed to the guidelines themselves, as well as the health professional, health system and patient factors mentioned in the editorial. The inclusion of shared decision-making aids within guidelines will hopefully facilitate discussion between healthcare professionals and patients to bridge part of this gap.

The terminology of compliance and adherence is not a helpful driver of change. Language is powerful. The diabetes community has led this discussion, suggesting that these terms should be avoided.1a I think we also need to consider the use of these terms for our health professional colleagues. Ensuring that health professionals and the broader community have access to high-quality information including guidelines and shared decision-making aids is important. Facilitating health professionals to interrogate their data to explore their practice relative to others and focusing on appropriateness rather than compliance may also be helpful drivers to assist in reflection and ongoing optimisation of clinical practice. Setting a broad-brush target for guideline ‘concordance’ in fact may not be helpful and may even be harmful.