Statins for primary prevention of cardiovascular disease
BMJ 2019; 367 doi: https://doi.org/10.1136/bmj.l5674 (Published 16 October 2019) Cite this as: BMJ 2019;367:l5674
All rapid responses
Rapid responses are electronic comments to the editor. They enable our users to debate issues raised in articles published on bmj.com. A rapid response is first posted online. If you need the URL (web address) of an individual response, simply click on the response headline and copy the URL from the browser window. A proportion of responses will, after editing, be published online and in the print journal as letters, which are indexed in PubMed. Rapid responses are not indexed in PubMed and they are not journal articles. The BMJ reserves the right to remove responses which are being wilfully misrepresented as published articles or when it is brought to our attention that a response spreads misinformation.
From March 2022, the word limit for rapid responses will be 600 words not including references and author details. We will no longer post responses that exceed this limit.
The word limit for letters selected from posted responses remains 300 words.
In 2014, at the age of 61 years, I wrote in the BMJ that as a public health epidemiologist specialising in the prevention of cardiovascular disease and diabetes I found it near-impossible to make a decision on whether to start statins. (1) I had studied the literature and particularly NNTs closely and was given advice to start these drugs by cardiovascular physicians who were friends and colleagues, as well as my own cardiologist and general practitioner. I was aware that as a South Asian I was at about 60% increased risk even compared to my high-risk White Scottish counterparts. I declined to take them. I wrote then that the values and beliefs of the medical profession will determine the public's decisions more than the evidence.
Since then, I have scrutinised the literature with particular reference to coronary heart disease and type II diabetes in South Asians, and the modern approach of treatment based on risk prediction. This work has been published as a book by Oxford University Press (2). I feel little wiser as a result of this reading, particularly as there are no randomised controlled trials in the use of cardiovascular screening as a health technology to be applied at population level and evaluated as a screening programme (as opposed to individual trials in selected patients). It is a mystery why this medical technology is not been treated as a screening programme.
Nonetheless, I started taking statins a couple of years ago. Why did I change my mind? Firstly, as I was working on my cardiovascular/diabetes book I was conscious that there would be considerable reputational damage if I had a heart attack prior to publication. Secondly, according to about a dozen cardiovascular risk calculators that I tried out my ten-year risk varied from about 7% to 22% (the variation reflects the crudeness of the prediction). The high 22% risk resulted from doubling as a result of family history. However, the turning point in decision-making came when another family member had a serious heart attack.
I started statins and had a range of side effects including fatigue and emotional blunting. Of course, I told myself it was psychological and placebo effect. Nonetheless, I took medical advice and was given a different statin and the ostensibly psychological, placebo orientated side-effects disappeared. I am now one of millions of people on statins for primary prevention. Am I doing the right thing? I do not know. After all, for an individuals the risk of a heart attack is either 0% 100%. Scientifically, individuals can only base their decision on the experience of large scale populations. In practice, like me, I suspect people are taking decisions on personal experiences and personal relationships with their doctors.
References
1 Bhopal RS. Statins: numbers needed to treat and personal decision making. BMJ 2014;349:g4980.
2 Bhopal RS Epidemic of cardiovascular diseases and diabetes: explaining the phenomenon in South Asians worldwide. Oxford: Oxford University Press; 2019
Competing interests: No competing interests
We agree with Byrne and colleagues that more careful consideration should be warranted to statins use due to uncertainties about the benefits. 1 However, what should not be ignored is that statins use in the real world is not a true-false item. The situation differs significantly worldwide while the adherence to medication is alarmingly low in most countries, particularly in underserved populations.
Influencing factors other than risks predicted by clinical guidelines, such as socioeconomic status, may be more essential to determine outcomes. Prior epidemiologic research showed the substantial higher rate of major cardiovascular events in low-income countries than middle- and high-income countries. 2 The study also found that low-income countries had the lowest risk-factor burden, and the burden is lower in rural areas. 2
Typically, the strategy of chronic disease prevention is risk-based. People who have a higher risk calculated by clinical prediction algorithms are more likely to be recommended medication regimens. The disputes in regard to eligibility for statins have lasted for many years, and there is also no final conclusion whether a risk-based approach is applicable to underserved populations who face additional challenges (e.g., low awareness of medication or reduced access to health information 3) to follow the most correct recommendations. Meanwhile, it is difficult for health professionals in resource-limited settings to implement such a complex strategy due to the insufficient time of consultation, limited professional quality and health resources.
A population-based strategy that shifts the entire risk distribution with lower cost and less adverse effects should be considered in preventing cardiovascular disease for underserved populations. Polypill, including the statins, has been regarded as a cost-effective strategy for primary prevention of CVD among LMICs. An example is the PolyIran study, of which 80% of participants were from rural areas. 4 In the 5-year follow-up, it had achieved a 40% reduction of major cardiovascular event risk in individuals without a history of cardiovascular disease by a polypill strategy (consisting of atorvastatin, aspirin and two antihypertensive drugs). The difference of 5-year change in LDL cholesterol between intervention group and control group is -19.54 mg/dl (95%CI -21·33 to -17·74, P<0.0001), whereas no significant difference of blood pressure and adverse events was found.
The population-based strategy does not necessarily run to counter trendy individualized treatment, 5 but healthcare providers need to adjust recommendations from guidelines to the local context. Despite some limitations of the polypill strategy, we suggest that stakeholders should consider the evidence provided by pragmatic trials like the PolyIran study and develop more feasible strategies to reduce cardiovascular disease burden in underserved populations.
Reference
1 Byrne P, Cullinan J, Smith SM. Statins for primary prevention of cardiovascular disease. BMJ 2019; 367: l5674. doi: 10.1136/bmj.l5674.
2 Yusuf S, Rangarajan S, Teo K, et al. Cardiovascular Risk and Events in 17 Low-, Middle-, and High-Income Countries. N Engl J Med 2014; 371(9): 818-827. doi: 10.1056/NEJMoa1311890.
3 Rosengren A, Smyth A, Rangarajan S, et al. Socioeconomic status and risk of cardiovascular disease in 20 low-income, middle-income, and high-income countries: the Prospective Urban Rural Epidemiologic (PURE) study. Lancet Glob Health 2019;7(6): e748-e760. doi: 10.1016/S2214-109X(19)30045-2.
4 Roshandel G, Khoshnia M, Poustchi H, et al. Effectiveness of polypill for primary and secondary prevention of cardiovascular diseases (PolyIran): a pragmatic, cluster-randomised trial. Lancet 2019; 394(10199): 672-683. doi: 10.1016/S0140-6736(19)31791-X.
5 López-Jaramillo P, González-Gómez S, Zarate-Bernal D, et al. Polypill: an affordable strategy for cardiovascular disease prevention in low–medium-income countries. Ther Adv Cardiovasc Dis 2018;12(6):169-174. doi: 10.1177/1753944718764588.
Competing interests: No competing interests
Recent guidelines have widened the criteria to prescribe statins for the primary prevention of cardiovascular diseases (CVD) (1). A very large share of the population becomes eligible for his treatment. Applying these criteria requires, however, a complex process of absolute risk estimation, which is not feasible in most clinical settings. Indeed, comprehensive information on cardiovascular risk factors are most often not routinely recorded in general practice files (2).
Maybe it is time to abandon this strategy of complex risk estimation; it could be wiser to apply simply the age criteria proposed by Wald more than 15 years ago (3) and start statins for everyone at the age of 55 years old. Age screening for future CVD was shown to be simpler than Framingham screening with a similar screening performance and cost-effectiveness (4). It remains however to determine at what age to stop treatment, as there is no clear evidence of benefits among very old and frail adults (5). Waiting for more evidence, stopping at 80 years old could be reasonable.
1) Byrne P et al Statins for primary prevention of cardiovascular disease. BMJ 2019;367:l5674
2) Turner LR, et al. Cardiovascular disease screening in general practice: General practitioner recording of common risk factors. Prev Med. 2017; 99:282-285
3) Wald NJ et al. A strategy to reduce cardiovascular disease by more than 80%. BMJ 2003;326(7404):1419
4) Wald NJ et al. Screening for future cardiovascular disease using age alone compared with multiple risk factors and age. PLoS One. 2011; 6(5):e18742
5) Cholesterol Treatment Trialists' Collaboration, Efficacy and safety of statin therapy in older people: a meta-analysis of individual participant data from 28 randomised controlled trials.
Lancet 2019; 393(10170):407-415
Competing interests: No competing interests
To my knowledge, the major statin RCTs have all restricted the definition of myopathy by requiring a plasma CK at least 10 times the upper limit of normal. I have seen 2 cases of nearly fatal rhabdomyolysis from statins that would have met this criterion. However, it is much more common to meet people with disabling muscle weakness (and sometimes pain) who have little or no elevation of CK. My clinical experience has been that such people improve when statins are stopped, but this sometimes requires many months or longer. Some have tolerated statin re-challenge at a lower dose or using a different drug; others not. Along with many astute clinical colleagues, I continue to believe that the statin RCTs deliberately and inevitably underestimate the incidence of clinically significant drug-induced muscle problems.
Why won't the CTT collaborationists relinquish their monopoly over the individual patient data volunteered by tens of thousands of RCT participants? Were they to do so, is it possible that a more open-minded definition of "myopathy" might lead to different conclusions on its incidence?
Tom Perry MD, FRCPC
Vancouver
Competing interests: No competing interests
A key message from this Analysis is that "Uncertainty remains about the benefits of their use for primary prevention". This is a worrying conclusion as the implementation of NICE guidance takes up an enormous amount of resources for General Practice who are left the burden of implementation without any extra resources for the implementation.
I looked at the Lipid guidance in 2014 in some detail and was shocked by the false assumptions in the costing report accompanying NICE clinical guideline 181 "Lipid modification: cardiovascular risk assessment and the modification of blood lipids for the primary and secondary prevention of cardiovascular disease".
In essence, the authors of the costing report take a long hard look at the evidence of benefit on mortality and morbidity and then also look at the cost of drugs and blood tests to monitor. They then allow £34 per year for GP appointments to undertake the monitoring. However, the authors then discount this appointment cost altogether by stating in section 3.1.3 "It is assumed that any additional GP appointments will be managed within existing resources."
With this simple but fundamentally flawed assumption they are effectively saying that GPs can implement the guideline without the need for extra resources. This amounts to a kind of financial abuse of general practice goodwill to implement supposedly “evidence based” guidance.
There is a real lack of GP representation on NICE guideline groups and I would suggest this needs to be urgently collected. The costing templates should also be submitted to the BMA and LMCs who may actually understand how the GP contract operates.
I also think there is insufficient public and patient involvement in the guideline process. Patients will have views on what priorities they wish to have in the way we deliver care. They may prefer we do not take on A huge extra workload for marginal if nay gains in low risk patients which might free up a bit of time to actually see and listen to patients who are ill or who think they are ill.
In the words of Peter Drucker: The most serious mistakes are not being made as a result of wrong answers. The truly dangerous thing is asking the wrong questions.”
The more recent NICE Hypertension guidance 2019 NG136 has also perpetuated the urban myth that NICE seems to promote that GPs are able to undertake unlimited amounts of extra work without any extra resources. It actually states “A similar evaluation was recently undertaken as part of the NICE lipids guideline update and so it was agreed that it would be appropriate to take a similar approach for this guideline.” This means any false assumptions in the lipid guidance are now perpetuated in the hypertension guidance
In this sense NICE guidance has been contributing to the current crisis in General practice with ever increasing demands. In other health systems NICE guidance may contribute greatly to costs if Doctors are paid on a fee per consultation.
The CQC has then been brought in as a regulator on quality and safety and now demands that we undertake regular audits to ensure patients on medications such as blood pressure drugs have annual blood tests. Often CQC inspectors have limited clinical knowledge and assume that NICE guidance must be correct when sometimes it is actually flawed. Many GPs are too busy to have the time and energy to challenge CQC inspection reports.
Both CQC and NICE need to be held to account for reducing safety in General practice by increasing workload to unsustainable levels. CCGs and NHS England also put pressure on General practice to deliver these ever increasing workloads without any extra resources and in some cases are actually reducing GP budgets.
Big organisations within the NHS are guilty of a serious kind of “groupthink” and cant seem to think clearly that we need be very patient centric and listen to their concerns if we are to get ourselves out the current workload and manpower crisis. By listen I mean really listen and not have another “consultation” where the answer is already pre-determined.
References
CG181 costing report. https://www.nice.org.uk/guidance/cg181/resources/lipid-modification-upda...
Hypertension NG136. https://www.nice.org.uk/guidance/ng136?_ga=2.123184719.596474652.1571496...
Cost effectiveness analysis. https://www.nice.org.uk/guidance/ng136/evidence/costeffectiveness-analys...
Competing interests: No competing interests
There seems to be a sort of selective blindness with respect to the comparison of demographically similar groups in France and in Britain.
The figures cited below come from the British Heart Foundation and the BMJ as well as several individual commentaries
CLASSICAL RISK FACTORS vs. HEART DISEASE
(UK & FRANCE MEN AGED 45 - 64)
Average total cholesterol level
France 6.1mmol/l
UK 6.2mmol/l
Average HDL ‘good cholesterol’
France 1.3mmol/l
UK 1.3mmol/l
Average systolic blood pressure
France 150mmHg
UK 148mmHg
Average body mass index (BMI)
France 26.6kg/m2
UK 26.6kg/m2
Rate of smoking
France 33%
UK 29%
Consumption of saturated fat % of calories
France 25.7%
UK 27.0%
Rate of type II diabetes (amalgamation of a number of studies)
France ~2.0%
UK ~2.0%
Death rate from CHD ICD 414 classification
France 128/100,000/year
UK 487/100,000/year
How does one explain the fact that two groups with near identical risk factors have such a disparity in CVD? Perhaps cholesterol, “good” or “bad” makes no difference except in familial hypercholesterolemia, the genetic condition for which statins were first created.
(Reference: .( Heart. 2004 Jan; 90(1): 107–111.
The French paradox: lessons for other countries
Jean Ferrières)
Competing interests: No competing interests
Studies of statins report similar adverse effect rates for both active and placebo groups. This has been repeated in studies of statin intolerant patients. This result has been interpreted to mean statins have minimal true adverse effects and those reported are mostly due to the nocebo effect. It is unsurprising that when any benefit for statins is demonstrated it will outweigh the zero risk and lead to a call for more widespread prescribing.
This logic is fallacious as it dismisses the nocebo effect as trivial and irrelevant as well as implying the affected patients are mentally delicate. If statin prescribing is extended to large numbers of otherwise healthy people the nocebo effect will cause 5-15% to have troubling adverse effects. Telling them that their symptoms are psychological and can be discounted just willfully ignores the significant harm caused by medicalising low risk populations. Any consideration of extending prescribing to low risk groups should consider the nocebo-attributed adverse effects as seriously as any direct pharmacological effects.
Competing interests: No competing interests
I share Dr Sarapinidis's view that statin results should be expressed in a more accessible way, and that when this happens the benefits at the individual level are not impressive (1). I have some doubts however on the meta-analysis he refers to. The heterogeneity in these studies is such (I^2 > 50%) that one is left wondering whether a meta-analysis, even analysed with random effect, was appropriate in the first place.
PN Trewby, AV Reddy, CS Trewby, et al. Are preventive drugs preventive enough? A study of patients’ expectation of benefit from preventive drugs. Clin Med JRCPL 2002;2:527–33
Competing interests: No competing interests
This analysis should have included in the references section the following systematic reviews, Level of evidence I:
A systematic review of randomized trials (RCTs) on patient all-cause survival after long term statin therapy revealed that death was postponed for only 3.2 days in primary prevention trials and for only 4.1 days in secondary prevention trials.
Reference
bmjopen.bmj.com/content/5/9/e007118
A systematic review and meta-analysis of randomized clinical trials (RCTs) involving patients prescribed statins for many years concluded that their survival was increased by only 10 days.
Reference
https://link.springer.com/article/10.1007%2Fs11606-019-05024-4
Competing interests: No competing interests
Re: Statins for primary prevention of cardiovascular disease. Please define your terms
Both Prof Bhopal and Dr Campbell-Taylor fail to define their terms. Which is essential if epidemiology is practised correctly.
Prof Bhopal talks of South Asians. Is he not aware that the Pakhtoons are, genetically speaking, as a group, totally different from Punjabees?
Even in India, the Tamils are different from the Bengalees.
Dr Campbell-Taylor must be aware that hundreds of thousands of Arabs from the former French Empire in North America are similarly different from the French born and living in France, and descended from the residents of the France of a hundred years ago? Of course there is some admixture but the populations are very largely distinct in terms of genetic make-up.
Maybe I am wrong. In which case, please explain how.
Thank you
Competing interests: No competing interests