Dr Terry McCormack revisits recent controversies about NICE guidance on statins and offers patient scenarios illustrating his own approach in clinical practice
Mired in controversy before publication, NICE Clinical Guideline (CG) 181 on Lipid modification-cardiovascular risk assessment and the modification of blood lipids for the primary and secondary prevention of cardiovascular disease1 was issued in July 2014. The media headlines had pre-empted the official announcement: '12 million Brits will be advised to take cholesterol-lowering drug' , stated The Daily Telegraph.2 Most of the major papers had similar announcements and there was much worry about overmedicalisation of the population. The public and medical professionals were equally confused. But why had this happened and what was the truth behind it all?
An 'open letter',3 entitled 'Concerns about the latest NICE draft guidance on statins', had been sent to Professor David Haslam, Chair of NICE, and copied to the Secretary of State for Health, Jeremy Hunt, on 10 June 2014.4 The letter declared six concerns:
- medicalisation of 5 million healthy individuals
- conflicting levels of adverse events
- hidden data
- industry bias
- loss of professional confidence
- conflicts of interest (real and perceived).
Nine doctors signed it, including Sir Richard Thompson, President of the Royal College of Physicians. There were also GPs, a cardiologist, a public health academic, a psychiatrist, and the Director of Clinical Research at Mount Sinai School of Medicine, New York, whose specialty is emergency medicine. One of the GP signatories was a namesake of the NICE Chair, a different David Haslam.
The impression given was that the NICE Guideline Development Group (GDG) had conflicts of interest, were too influenced by the pharmaceutical industry, and that there were hidden data that were not taken into consideration. The GDG could not personally respond as the letter was sent prior to publication of NICE CG181. Therefore NICE had to respond on their behalf.
The 10% threshold
The controversial statement in CG181 was the new recommendation at 1.3.18: 'Offer atorvastatin 20 mg for the primary prevention of [cardiovascular disease (CVD)] to people who have a 10% or greater 10-year risk of developing CVD. Estimate the level of risk using the QRISK2 assessment tool.'1
Recommendation 1.3.14 of the guideline1 had stated that lifestyle modification was to be the first option to consider, but that seems to have been ignored in all of this. The key word here is 'offer'. NICE allows its GDGs to use only two words in this context, of which 'consider' is the softer option. The stronger term, 'offer', must be used if the evidence supports it.5
The 10% threshold was based on a health economics analysis of what worked best in terms of the national economy—paying for statins or bearing the burden of heart attacks and strokes. The latter costs more and the GDG was conservative, in that the analysis cut-off threshold is really 7.5%, the figure the Americans included in their guidelines.6,7 The term 'consider' might have worked better in this context but because NICE uses an evidence-based guideline system, the GDG had to follow the rules and use 'offer'.
I personally intend to interpret this guideline in a pragmatic fashion and I prefer this to mean that I can offer patients the option of using this treatment, if they wish to do so, after careful explanation.
Some other changes
Apart from the part about the 10% risk, the guideline also recommends measuring non-high density lipoprotein cholesterol (non-HDL cholesterol), instead of low-density lipoprotein cholesterol (LDL cholesterol), as this is more accurate and only requires a software change in the laboratory.1
NICE recommends the use of QRISK28 for assessing people aged up to and including 84 years but accepts that the evidence regarding statin treatment in people aged over 80 years is extremely limited; indeed, they pose this as one of the areas for research.
NICE advises us to refer patients to a specialist if their total cholesterol (TC) is >9 mmol/l or non-HDL cholesterol is >7.5 mmol/l.1 Practitioners should consider patients who have an estimated glomerular filtration rate <60 ml/ min/1.73m2 and/or albuminuria to be already at high risk.1
Unlike the previous guideline (NICE CG67, published in September 2008), NICE CG181 sets a target of reducing non-HDL cholesterol by greater than 40% by the use of high-intensity statins.1 Very importantly, it says not to offer fibrates, nicotinic acid, bile acid sequestrants, or omega 3 in either primary or secondary prevention of CVD.1 This suggests to me that the GDG are not really in the pocket of the pharmaceutical industry. Furthermore, I cannot understand why the pharmaceutical industry would attempt to influence the GDG when the drug they advised us to use is a low-cost generic.
NICE CG181 in practice: some case studies
The guideline states (under heading 2.2):1 'The challenge is to ensure that statin treatment is presented as a patient choice in addition to lifestyle modification to avoid unnecessarily "medicalising" this group of people. To help patients make an informed decision, healthcare professionals will need an effective way of briefly but clearly communicating the pros and cons of the various options available.'
So how do I 'avoid unnecessarily "medicalising"' people and 'briefly but clearly' communicate 'the pros and cons'? I will use three examples of how I interpret and use the guidance for my patients (see Box 1, below). They are all based on real cases and I apologise in advance for using examples of non-diabetic men in all three cases. I do that to simplify the message.
Case one describes the typical scenario before CG181. Geoff is a 55-year-old social worker, who smokes twelve cigarettes a day, has a BP of 167/98 mmHg, a TC of 4.6 mmol/l and an HDL cholesterol of 0.88 mmol/l. I have used a desktop version of the Joint British Societies' JBS2 guideline computer risk algorithm since before 2005, when it was published.9 I still use it, as it is hard to teach 'old dogs new tricks'. It depicts a thermometer demonstrating the patient's relative risk (RR). The RR is the key point. In the Netherlands, it is common practice to base the risk assessment on both absolute risk (AR) and RR. Using both makes perfect sense.
In Geoff's case, I show him that his AR of having a heart attack or stroke in the next 10 years is 33%. I explain that Government policy is to recommend treatment with drugs such as statins in people with this level of risk. But then I focus the conversation on the thermometer and his RR of 92, which is highlighted in red. I explain that this is his risk compared with his peer group of 55-year-old men. I then tell him a statistical lie. I suggest to him that this represents a queue for the next heart attack or stroke and that the person at 100 is first in line. This is a lie on every level, in that the whole scheme is inaccurate and his risk is not linear but part of a bell-shaped curve. I admit this to Geoff but explain that using the lie helps me to simplify my explanation.
I then suggest to Geoff that we see what happens if he stops smoking, and I change his profile to that of a non-smoker and press the calculate button again. I explain that the drop in AR to 22% needs to be ignored as you cannot change this element and the focus of our conversation should be his personal RR, which has fallen to 63. Geoff is impressed by how far back in the queue he has gone through simply stopping smoking. My experience is that this delivers a strong motivational message.
Then I ask Geoff if he would like us to control his blood pressure to a target of 140/80 mmHg, which on recalculation reduces his RR to 51. The addition of a statin might get his TC to a target of 4.0 mmol/l that further reduces his RR to 32.
The choice is always Geoff's but in view of his high risk, I feel it is my job to try and sell the whole package to him.
Alternatively, we could have used the new JBS3 online calculator,10 which aims to reduce risk of CVD for the whole adult population rather than just those at high short-term risk and is considered compatible with recommendations in NICE CG181, see this Guidelines in Practice article covering JBS3. This calculator would have shown Geoff that his heart age is 69 years and that his average survival is likely to be to age 73.2, which we could extend to 76.4 if he applies all these measures, a gain of 3.1 years. Or I could use the QRISK8 smiley faces, although I do not personally find them as effective as a sales aid.
Case two, which becomes more relevant to CG181, uses the example of Jimmy, a 40-year-old overweight farmer. He is a non-smoker with a BP of 156/84 mmHg, a TC of 5.9 mmol/l, and an HDL cholesterol of 0.8 mmol/l. Using JBS29 again, we demonstrate an AR of 12% but a RR of 92. Jimmy's AR is low because of his age, but compared with other men of his age he is at relatively high risk. His TC to HDL cholesterol ratio is 7.2, which some would argue indicates the need for treatment anyway. I can show him that controlling his BP to 140/88 mmHg reduces RR to a modest 88 but getting his TC to 4.0 mmol/l would achieve an impressive reduction to 42. If things remain as they are, his AR at age 49 will be 20%. Using JBS3,10 we can show his heart age to be 46 and his average life expectancy age 75.1. Treating him will increase this to 80.4, a 5.3 year gain. Prior to CG181,1 I would not have been able to offer him treatment. Now I can, and I believe that treatment will be effective in his case because of his lifetime risk. I would try and convince Jimmy to accept the offer of treatment.
Case three is Kevin, a 66-year-old retired general practitioner. A lifelong non-smoker and keen tennis player, his BP is a respectable 138/82 mmHg. When we enter his TC (5.7 mmol/l) and HDL cholesterol (1.86 mmol/l) into the JBS2 calculator,9 it reveals an AR of 15% and RR of 3. I demonstrate that a statin used to a TC target of 4.0 mmol/l will reduce his RR to 1. He is not impressed at all. JBS310 reveals a heart age of 66; it cannot be lower than your actual age. His cardiovascular survival age is already 83.5 and can only move up 0.74 of a year. The main driver of risk in Kevin is his age. In theory, he meets the CG181 target and I can offer him a statin. He sensibly refuses the offer and remains medication-free. I agree with his decision and I am sure most of the members of the NICE CG1811 GDG would concur.
CG=clinical guideline; BP=blood pressure; TC=total cholesterol; HDL=high-density lipoprotein; RR=relative risk; AR=absolute risk; CVD=cardiovascular disease; GDG=guideline development group.
The question was: should everyone with a 10% risk of CVD take a statin? My answer is not everyone, but certainly some should.
Research grants—Amgen, Bayer, Boehringer Ingelheim, Galderma, Daiichi-Sankyo, NIHR, Servier
Advisory boards and speaker fees—Alere, Amgen, Astellas, AstraZeneca, Bayer, BMS/Pfizer, Boehringer Ingelheim, Daiichi-Sankyo, Lundbeck, MSD, Roche Diagnostics, Sunovian
Member of the Guideline Development Group for NICE CG127 on Hypertension: clinical management of primary hypertension in adults (2011)
- NICE. Lipid modification: cardiovascular risk assessment and the modification of blood lipids for the primary and secondary prevention of cardiovascular disease. Clinical Guideline 181. NICE, 2014. Available at: www.nice.org.uk/guidance/cg181
- Donnelly L. Statins: 12 million Brits will be advised to take cholesterol-lowering drug. The Telegraph, 11 February 2014. Available at: www.telegraph.co.uk/news/health/ news/10632047/Statins-12-million-Brits-willbe- advised-to-take-cholesterol-lowering-drug. html (accessed 16 July 2015).
- NICE website. Concerns about the latest NICE draft guidance on statins. 10 June 2014. www.nice.org.uk/Media/Default/News/NICE-statin-letter.pdf
- Wise J. Open letter raises concerns about NICE guidance on statins. BMJ 2014; 348: g3937
- NICE website. Developing NICE guidelines: the manual. 9: Developing and wording recommendations and writing the guideline. Strength of recommendations. Available at: www.nice.org.uk/article/pmg20/chapter/9- developing-and-wording-recommendations-andwriting- the-guideline (accessed 16 July 2015).
- Goff D Jr, Lloyd-Jones D, Bennett G et al. 2013 ACC/AHA guideline on the assessment of cardiovascular risk: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines.Circulation 2014; 129 (suppl 2): S49–S73.
- Stone N, Robinson J, Lichtenstein A et al. 2013 ACC/AHA guideline on the treatment of blood cholesterol to reduce atherosclerotic cardiovascular risk in adults: a report of the American College of Cardiology/American Heart Association task force on practice guidelines. Circulation 2014; 129 (suppl 2): S1–S45.
- University of Nottingham and EMIS. QRISK2-2014 risk calculator website. www.qrisk.org (accessed 8 June 2014).
- JBS2 Board. Joint British Societies' guidelines on prevention of cardiovascular disease in clinical practice. Heart 2005; 91: v1–v52. Available at: heart.bmj.com/content/91/suppl_5/v1.full (accessed 16 July 2015).
- JBS3 Board. Joint British Societies' consensus recommendations for the prevention of cardiovascular disease (JBS3). Heart 2014; 100: ii1–ii67. Available at: heart.bmj.com/ content/100/Suppl_2/ii1.full?sid=f6401a46- fe7f-4e55-85de-7d5f30d65e0f? (accessed 16 July 2015).