Guidelines: lost in translation

It is hard to imagine that clinical guidelines, in their current incarnation, will survive. Undoubtedly, doctors need high-quality information to guide clinical decisions, but the development and implementation of clinical guidelines is fraught with difficulty.

We have seen heated debate on this subject in the MJA, and two more articles in this issue add fuel to the fire. Williams and colleagues make a strong demand for the “Comprehensive disclosure of conflicts [of interest] . . . to safeguard the integrity of clinical guidelines and the medical profession”.

As they observe, and we know, compliance with guidelines is equated to delivery of high-quality care, and can affect doctors’ remuneration. Guidelines themselves, then, must be beyond reproach. Yet, according to Williams et al, only 15% of the 470-plus guidelines on the National Health and Medical Research Council portal contain a conflict of interest statement — a longstanding requirement for research papers. This is surprising, as guidelines have much more influence on clinical practice than a single research paper. A 2009 Institute of Medicine report (Conflict of interest in medical research, education and practice) outlined several examples of inappropriate industry influence on clinical guidelines development in the United States. Pharma ties to individuals and organisations still loom large as an important issue.

And there are many other concerns. Subtle influences from personal opinion, cultural mores and vested interests can influence the translation of evidence into clinical care — a step for which the methodology is less well defined than it is for the finding and grading of evidence (BMJ 2010; 340: c306). The role of the GRADE (Grading of Recommendations Assessment, Development and Evaluation) system in separating the strength of a recommendation from the strength of the evidence is hotly contested (MJA 2011; 195: 324-325). Because of the protracted process of development, guidelines are often out of date before publication, and so lose credibility and currency. Grol and Buchan (MJA 2006; 185: 301-302) lamented the high cost of development in time, labour and money and implored guideline developers to provide useful tools for practitioners and patients.

Also in this issue of the MJA, as an example of the discordance between guidelines and actual practice, Inam and colleagues report that disease-modifying antirheumatic drugs are underused in managing early rheumatoid arthritis compared with guideline recommendations. Instead of the usual explanations for such failure — “habit, lack of motivation, and external barriers such as lack of time, resources and organisational support” — they suggest it may reflect awareness by clinicians of the difficulties of translating a deficient evidence base into practice and practitioners’ sensitivity to individual patient issues, such as treatment cost.

Patient comorbidity is a real-life issue that makes guideline translation difficult. Generally, guidelines, as a result of development by specialist experts, focus on managing the disease, but not necessarily managing the whole patient, and commonly exclude non-drug treatments.

Perhaps new collaborative technologies may soon help to overcome many of the issues that currently plague this essential clinical tool.

We intend that the MJA will continue in its role as a repository of clinical guidelines. We already insist on a full conflict of interest statement for each contributor, and all guidelines published in the MJA are peer reviewed.