The Value of Guidelines in Everyday Clinical Practice

Anthony H Gershlick BSc MBBS FRCP


Eur Heart J. 2018;39(13):1046-1048. 

Clinical guidelines have become an essential tool in the provision of evidenced-based patient care and exist for many medical specialities. They are particularly valuable when the clinical presentation is heterogeneous and/or when the discipline comprises various sub-specialities. Thus, in cardiology, we have guidelines for valve disease, heart failure, arrhythmias, ischaemic heart disease, and others. Guidelines should be both broad, and deep where appropriate. They are there for the specialist and non-specialist and should be written specifically with both groups in mind. The best guidelines provide reference to evidence-base when a patient presents, for example to a non-specialized junior doctor so that they can understand (perhaps through the pocket series), what is the current best management for that particular patient (e.g. indications or otherwise for NOACs). For the specialist, the guidelines present the available evidence to support the use of a strategy or therapy at the point of guideline publication.

Questions around the guidelines do exist. What happens for example if a physician decides to treat a patient without the guidelines? Are they medico-legally binding? The quote is that they are guidelines not tramlines, but it would be an interesting test if a patient suffers and the guidelines were not followed (perhaps in the mind of the physician for good reasons).

Secondly guidelines are only as contemporary up to the time that the last set were published (sometimes some years apart). Furthermore, they can only add to treatments not prune them since negative/comparative trials are far less common and less likely to be published, not least since each guideline-approved strategy becomes the next standard of care (aka the control group). As a result, the cumulative pharmacotherapy list generally gets longer. There are some exceptions to this, such as the attempts to trial reduction in the number (e.g. no aspirin) or length of duration of dual anti-platelet therapy in interventional cardiology. Guidelines can be confusing and overlapping (the revascularization options or anti-platelet strategies come to mind). Finally, one needs to ask whether guideline-recommended treatments that have been proven effective in trials, really work in the target population in the real world, and over time. Follow-up registry data are generally selective and poor.

For acute coronary syndrome, guideline-recommended strategies and therapies have flourished. What to do about the STEMI non-infarct-related artery (N-IRA) is a classic example. Initial registry data indicated that treatment of the N-IRA was harmful. Several small to medium sized randomized studies however and their meta-analyses[1–6] suggest treating the N-IRA may be an effective therapeutic strategy, with intervention attenuating the risk of subsequent adverse clinical consequences. While there continues to be ongoing debate at major meetings and in the literature about the best approach to this clinical issue, the data so far do support a view that 'significant' N-IRA lesions should probably be treated around the time of the P-PCI (i.e. maybe not at the time of the procedure but during that hospital admission).

While we await the larger trials[7–9] to help determine whether there is an impact of prophylactic management on hard endpoints this intervention is supported by the most recent ESC guidelines (Class IIa LoE A)[10] and that allows the physician the choice—initially harmful (early AHA/ACC guidelines—now changed) to 'you can consider doing it if you think it is the right thing to do' (recent ESC guidelines).

There are however additional, more subtle issues around N-IRA disease, including its development, progression, clinical presentation and the factors that influence these phases. The patient is very unlikely to have been fully cured by intervention to the presenting infarct-related artery, irrespective of whether any N-IRA disease is treated or not. Even if the N-IRAs are 'completely normal', this patient has a propensity to developing coronary artery disease. Holistic considerations around secondary prevention have evolved into practical logistics such as mental, emotional, and physical rehabilitation, cessation of smoking campaigns and good, instructive communications with the primary care physicians. Assessing the efficacy of any risk factor management may unfortunately be less well done, and merely discharging patients on secondary prevention medication may not be enough. Regular review and drive towards evidence-based target levels may need to be proactive and prolonged throughout follow-up.

Recently there have been published longer term registry data which make it clear that patient risk does not end at the point of discharge, in the individual apparently 'fit and well' enough to go home and having 'got over' their heart attack.

These are a good example of guideline into practice and answer the question about the value of guideline-recommended therapies in the real world. But the registries must be very good.

For example, there is excellent registry data from the SWEDEHEART group,[11] who report on the impact of evidence-based, guideline-recommended therapies on outcomes over a 20-year period. These data appear to show a direct correlation between delivery of evidence-based therapies and improvements in outcome. Thus the exponential rise in use of P-PCI from 0% in the mid-1990s to 80% in 2013–14, and increase in secondary prevention measures from 40% in the mid-1990s to levels of 80–90% in 2013–14, correlate directly with falls in mortality (from 20% in the mid-1990s to 14% in 2015–16), in re-acute myocardial infarction (AMI) (from 12% to 5%) and in incidence of heart failure (from 7.5% in the mid-1990s to 6% in 2015–16) over the same time period (Figure 1). These direct correlations are indeed impressively noteworthy.

Figure 1.

Correlation between increase in use of secondary prevention measures post-acute myocardial infarction and falls in adverse event rates over same period (Reproduced with permission from the SWEDEHEART registry).

However while such data certainly make the case for implementation of guideline-directed therapies, they also tell us there is still work to be done. The hard endpoints event rates at 12 months at these levels (mortality 14%, 5% AMI, and 6% heart failure; Figure 1) require a continuing focus on improvement, not least since the absolute numbers involved mean residual outcomes are higher than most would consider important from just looking at the percentages.

Acute myocardial infarction occurs at a rate of about 50 per 100 000 in Sweden and 43 to 144 per 100 000 per year in other European countries. In the USA, the rate has decreased from 133 per 100 000 in 1999 to 50 per 100 000 in 2008. Taking an 'average' 50 AMI per 100 000 and an EU population of 742 073 853, then in 2017 there were approximately 375 000 STEMIs per annum. Extrapolation from the SWEDEHEART data above[11] suggests the absolute contemporary events at 12 months are therefore in the order of: 52 000 deaths, 20 000 repeat myocardial infarction (MI), and 23 000 repeat admissions for heart failure. That is a lot of mortality and knock-on morbidity, not even considering those with cardiogenic shock or heart failure at presentation. It would seem that there can be no place for the complacency that may be associated with the satisfaction of achieving TIMI 3 flow in the culprit artery.

It is clear that appropriate secondary prevention based on mandated guideline-recommended therapy is important, but monitoring of efficacy is even more important. Recording blood pressure, HbA1C, and lipid status and modifying therapy during follow-up should result in benefits, even 2–3 or more years after the STEMI. Things cannot end at time of discharge. It is a message that needs rolling out to primary care physicians and starts with those managing the acute event. Discharge letters and information to patients should emphasize not only the importance of compliance with secondary prevention measures, but that continued monitoring for their efficacy and appropriate dose modification is really important.

Guidelines have shaped the practice of medicine, in the main are a force for good and are here to stay. One is particularly encouraged by really good registry data that supports their value over time. Rolling online 6 monthly review by a dedicated team may make them more contemporary, but in essence they can only as good as the published data. Indicating where the research gaps are may help direct researchers to try and improve outcomes further.