Could NICE guidance on the choice of blood pressure lowering drugs be simplified?BMJ 2012; 344 doi: https://doi.org/10.1136/bmj.d8078 (Published 13 January 2012) Cite this as: BMJ 2012;344:d8078
All rapid responses
I don't know if the medieval theologians really did discuss how many angels you could fit on the head of a pin, probably an attempt to discredit them. But as someone involved daily in the managemnt of hypertension, I feel much of this debate is in the same category. The discussion at the last meeting of the British Hypertension Society was definitely of the heat not light variety. There are perhaps some points worth making:
1. We now have a large repertoire of treatments for hypertension and radically new ones are unlikely in the foreseeable future.
2. The choice of diuretics is open to endless debate, but we are unlikely to get a head to head trial of say chlortalidone and bendroflumethiazide in thousands of patients over 5 years. But it has to be noted that the diuretics now proposed are even more likely to cause hypokalaemia than the thiazides and monitoring for this is mandatory, and drugs like co-amilozide acn be very useful in this context.
3. White-coat hypertension is certainly not benign, but we do not have any evidence-based approach to dealing with it.
4. Even if we did patients may not be enthusiastic about following our advice. Many people fail to share their doctor's keenness for medication.
5. However we start our patients on medication the majority will ultimately take two or more drugs to reach anything like target BP. The concept of starting treatment with low doses of 2 drugs is very appealing, and now likely to be quite economical.
6. On the whole the demotion of conventional beta-blockers is probably justified. Sadly we don't have enough data on the vasodilator beta-blockers (carvedilol and nebivolol) to know whether they should be exempt from this, and probably never will.
As a cynical Frenchman put it 300 years ago, disputes would not be so acrimonious if only one side was in the wrong. In this case probably everyone is to some extent, and may well remain so indefinitely .Meanwhile a lot of people need treatment, though many may dispute the need for it when we tell them so. There are likely to be lots of occasions when we will need to use professional judgement, before of course that sort of behaviour is banned.
Competing interests: No competing interests
Response to Reecha Sofat’s article in the BMJ of 28/1/12, ref 2012; 344: d8078.
This is an intelligent and informed analysis of the evidence for the pharmacological treatment of hypertension. The high-lighting of ‘the two new comprehensive meta-analyses’ by Fiona Turnbull (ref 15) & Malcolm Law (ref 16) is to be applauded. Both these analyses concluded that all major classes of antihypertensives were broadly equivalent in preventing serious cardiovascular events. Reecha Sofat & her colleagues then remarked on the only exceptions to this that the latest and largest meta-analysis by Malcolm Law found ie the slight improvement in stroke prevention with calcium channel blockers & the slight decrease in cardiovascular events with beta-blockers given to newly diagnosed ischaemic heart disease patients. The latter point has additional supporting evidence from a cohort study by Julia Hippisley-Cox and Carol Coupland (ref BMJ 2005; 330: 1059-63) which compared different drug regimes in newly diagnosed ischaemic heart disease and showed that the best combination was aspirin, a statin & a beta-blocker but adding in a fourth drug (or substituting the beta- blocker) from the angiotensin converting enzyme inhibitor group decreased survival rates.
It is worth stressing that both these new comprehensive meta-analyses included a beta-blocker group. The only minor issue I would have with Reecha Sofat’s article would be the quote wrt a specific beta-blocker meta-analyses when it was asserted after mentioning the findings of Lindholm (ref 9) that ‘two other meta-analyses reached similar conclusions’. This is certainly not true for Khan and McAllister (ref 11) who conclude that ‘importantly, younger patients randomly assigned beta-blockers exhibit similar rates of cardiovascular death, MI or stroke to those assigned other agents’. They also pointed out several methodological flaws in Lindhom’s meta-analysis. It appears well founded that beta-blockers are less effective in the elderly at preventing serious cardiovascular events.
There are a few other issues to raise. Regarding 24 hour monitoring and as a practitioner who regularly performs such analyses, there are some pitfalls. Undoubtedly the use of 24 hour monitoring can save patients being falsely given life long treatment for a condition they haven’t got. However, there is good evidence that oscillometric devices can falsely over-estimate blood pressure. So my practice is to check raised blood pressures with an oscillometric device and aneroid device on alternate arms, doing up to eight readings, to find the means comparing machines and arms. Almost always it is the machines that vary most and usually it is the oscillometric device which reads highest. This saves both the need for a 24 hour analysis and unwarranted treatment.
Finally, it is surely no longer tenable to say that we haven’t got enough evidence (bar one proviso – see below) to make informed choices about drug treatment of hypertension. At times I have felt that the drive towards using the rennin-angiotensin system agents first line has been commercially driven. The evidence is that there is no reason for using these drugs first line in the absence of a specific indication. My opinion is that the first line drug for younger patients is still a beta- blocker, other patients a calcium channel blocker and just possibly angiotensin type drugs in the elderly. The reason for my choice in the former two groups is the lack of the need to check renal function after initiation and that will save alot of time and money. If there is any yawning gap in our evidence base it is the effectiveness of these agents long term. The mean age of our current evidence base is about four years. It is an extrapolation to suggest that we get the same proportional benefit at ten years: we need evidence not conjecture.
Dr. Adrian Smith, GP Leeds.
Competing interests: No competing interests