Issue 2 / 28 January 2014

THE 2014 evidence-based guideline for the management of high blood pressure in adults opens with the line: “Hypertension is the most common condition seen in primary care and leads to myocardial infarction, stroke, renal failure, and death if not detected early and treated appropriately”.

As someone who has worked on guidelines, I have sympathy for the panel members appointed to the US-based Eighth Joint National Committee (JNC 8), who have been through a 5-year tortured process of getting these guidelines out, especially as their sponsor — the National Heart, Lung and Blood Institute — pulled out at the end. The long wait has been widely noted.

The guideline’s 2003 predecessor, “Prevention, detection, evaluation, and treatment of high blood pressure”, by JNC 7, was a 104-page report. In comparison, JNC 8 seems to have been cut off at the knees with a 14-page “special communication”.

However, the opening sentence of JNC 8 establishes that the authors are firmly in the camp of managing high blood pressure (BP) not as a major risk factor in disease but as a disease in itself: “hypertension”.

Treating blood pressure as a dichotomous variable (normotension or hypertension) rather than as a continuous variable (which it is after all) means forgoing benefit for some normotensive individuals at high absolute risk of the adverse events outlined by the authors in these guidelines. Trials that enrolled prehypertensive or normotensive individuals were excluded from their consideration.

As pointed out in one of three accompanying editorials this is not the approach taken by the recent US lipid guideline, which has embraced the more holistic absolute cardiovascular disease risk approach.

In JNC 8, the nine recommendations are all related to drug therapy based on the panel’s three “highest-ranked questions” on high BP management, which were identified through a modified Delphi technique — Does initiating therapy at specific BP thresholds improve health outcomes? Does drug treatment to specified goals improve health outcomes? Do different drugs/drug classes differ in benefits and harms?

These questions are likely important for clinical decision making but also reflect why the guideline does not match its title, which does suggest a more general guideline that includes non-drug management.

Four of the recommendations are entirely, and two partly, reliant on expert opinion. This is most likely due to the fact that rigorous evidence — randomised controlled trials — is demanded but there were limited numbers of studies available for consideration. The authors have come up with more conservative BP treatment thresholds and targets because of this.

They have given more prominence to diastolic blood pressure as the older trials had inclusion criteria based on this measure.

The guideline authors do admit to many limitations including the lack of assessing harm versus potential benefit.

The world view expressed by this document and its limited scope means it is unlikely to have a significant influence on clinicians or guideline writers.

Professor Mark Nelson is chair of general practice at the School of Medicine, University of Tasmania, and a professorial research fellow with the Menzies Research Institute Tasmania.

One thought on “Mark Nelson: Guideline mismatch

  1. Justin Coleman says:

    Professor Nelson rightly reminds us that a continuum is difficult to squeeze into a dichotomy—normal vs high blood pressure.
    He focuses on one side of the threshold (underdiagnosed normotensives), but equally concerning are problems on other side. The blunt instrument of a numerical threshold, creeping incessantly lower, sweeps in millions of overdiagnosed hypertensives who will not benefit from antihypertensive medication.
    Since the reduction of the cut-off point, almost one in three adults are now defined as having ‘prehypertension’ (BMJ 2010;341:c4442). Guidelines result in considerable pressure on GPs to prescribe: pressure from patients, specialist colleagues, the medical media and pharmaceutical-sponsored education and reps. Over-treatment causes patient harm and financial burden.
    Of the members of the 2003 JNC 7 panel cited, 82% declared ties with the pharmaceutical industry and the entire 104-page document contained not a single mention of any potential harm of their broadening of the definition of hypertension (PLOS Medicine 2013; 10.1371/journal.pmed.1001500).
    Perhaps things are improving with the new JNC 8: the 50-odd declared pharmaceutical conflicts of interest involve only a quarter of the panel members.
    In any physiological continuum such as blood pressure and lipids, treatments that are ‘no-brainers’ at the high end decrease in benefit as one moves towards the middle of the spectrum. Shifting the threshold too low—or preferring a simple dichotomy to complex absolute risk—can result in overdiagnosis, where patients, on average, are harmed.

Leave a Reply

Your email address will not be published. Required fields are marked *