Dr John Briffa's response to the BMJ's withdrawal of comments.

My reaction to the BMJ’s withdrawal of statements relating to the safety of statins

By Dr John Briffa on 21 May 2014

 

You may have noticed that there’s a bit of a ‘fight’ going on over the cholesterol-reducing class of drugs known as ‘statins’. I am simplifying here, but there are essentially two opposing camps. In one corner, there are those doctors and researchers who hold the view that the statins should be given to pretty much all adults from middle-age, and there is little to be concerned with regarding their safety. In the other corner, there are those who believe that statins do no good at all for the vast majority of people who take them, and that the side-effects are more common that ‘official statistics’ suggest.
The fight went up a notch last week when the editor of the British Medical Journal announced that remarks made in two BMJ articles about the side-effects of statins had been withdrawn [1]. One of the articles was written by Dr John Abramson from Harvard Medical School and colleagues, and principally questioned whether extending statin use to those at lower risk (as has been suggested by the National Institute of Health and Care Excellence) will save lives [2].  According to their analysis, it won’t. The other piece was written by UK cardiologist Dr Aseem Malhotra, and principally questioned the role of saturated fat in heart disease [3].

In both articles, the authors referred to a study [4] that found that 17.4 per cent of people taking statins had adverse effects as a result attributed to statins. In the articles this figure was expressed as 18 per cent or ‘about 20 per cent’ which, apparently, was the first error. However, other issues, according the BMJ, was that the authors claimed that the 17.4 figure related to the percentage of people who stopped statins (at least temporarily) due to side-effects – this is not correct (it represented the percentage of people who had adverse effects). Also, the study in question was ‘observational’ in nature, which means that we don’t know if the adverse effects are due to statins or are the result of the ‘nocebo’ effect (like a placebo effect, only with negative effects rather than positive). The Editor of the BMJ, Fiona Godlee, tells us in her editorial that the caveats concerning the observational nature of these findings should have been made.

Both the articles made it through the peer-review process, so how did these errors come to be picked up? Well, apparently, they were brought to the attention of Fiona Godlee by Professor Sir Rory Collins, who heads up what is known as the Cholesterol Treatment Trialists (CTT) collaboration based in Oxford, UK. The CTT holds data on statin studies and periodically assesses it to tell us that statins are safe and very effective.

However, the CTT has in the past made grand claims about the effectiveness of statins that are based on the extrapolation of data (rather than actual data). Plus, as some have pointed out, the CTT refuses to release its data for independent inspection. The CTT can basically say what it likes about the effectiveness and safety of statins, but no-one can challenge what it says because no-one else has access to its data.

Apparently, Professor Sir Collins was invited by Fiona Godlee to put his concerns in writing for publication but he refused. He and Fiona Godlee met, and this set the ball rolling which culminated in the withdrawal detailed above. However, apparently this is not enough for Professor Sir Collins: he is demanding retraction of both articles, even though he has not challenged the main points raised in the articles. Fiona Godlee has set up an investigation to determine whether or not the articles should be retracted.

I have written to the BMJ in the form of an online ‘rapid response’ that was posted yesterday. You can read my response here.

In my response I accept that errors were made, but ask if a simple correction might have done.
Professor Sir Collins bases his assertions regarding the safety of statins on data from randomised controlled trials, like the ones analysed by the CTT. However, there are many reasons why randomised controlled trials may not adequately identify and report adverse effects from statins. Fiona Godlee acknowledges this herself and lists some of the issues, and I add to that list in my response. In all, there are about a dozen reasons why randomised trials stand to ‘miss’ adverse effects.

I also raise the issue of a recent study which appeared to confirm the safety of statins [5]. The study also detailed many reasons why results from clinical trials are an unreliable assessor of side effects. The study was reported in the BMJ [6], but the report included none of the caveats listed in the paper (or any other ones). As I explain to the Editor, these omissions may lull individuals into a false sense of security, and put people at unnecessary risk of adverse effects of statins, some of which can be serious.

One might argue that the ‘crime’ committed by the BMJ is roughly equivocal to the ones perpetrated by Drs Abramson and Malhotra. I point this out to Fiona Godlee in my response, and request a meeting with Fiona Godlee to discuss my concerns. After all, why should I not be afforded similar privileges as Professor Sir Collins”

You can read the response in full here. If you agree with the sentiments expressed in it, please click on the ‘thumbs up’ symbol to the right of the response to log your agreement.

References:
1.    Godlee F. Adverse effects of statins. BMJ 2014;348:g3306
2.    Abramson JD, Rosenberg HD, Jewell N, Wright JM. Should people at low risk of
cardiovascular disease take a statin? BMJ2013;347:f6123
3.    Malhotra A. Saturated fat is not the major issue. BMJ2013;347:f6340
4.   Zhang H, et al. Discontinuation of statins in routine care settings. Ann Intern Med 2013;158:526-34
4.    Finegold JA, et al. What proportion of symptomatic side effects in patients taking statins are genuinely caused by the drug? Systematic review of randomized placebo-controlled trials to aid individual patient choice. European Journal of Preventive Cardiology 2014;21(4):464-74
5.    Wise J. Statins may have fewer side effects than is claimed, meta-analysis finds BMJ 2014; 348:g2151
=========================================================================================================
Read the complete article here.

Researchers pronouncing ‘statins are safe’ are undermined by their own observations - Briffa

Researchers pronouncing ‘statins are safe’ are undermined by their own observations

By Dr John Briffa on 13 March 2014

 

Listen to most ‘key opinion leaders’ talk about statinsand you will hear soothing reassurances about their safety. Yet, my experience as a doctor suggests that adverse effects such as fatigue and muscle pain occur more commonly than ‘official statistics’ suggest. However, a study published this week claims to provide evidence that, for the most part, statin side effects are ‘imagined’ [1].

In this research, the adverse effect rates from statins was compared with those seen in individuals taking placebo (dummy) pills in a total of 29 studies. The conclusion was that apart from increasing the risk of diabetes, statins don’t generally have any more adverse effects than placebo. The actual words the authors use in their conclusion are: “Only a small minority of symptoms reported on statins are genuinely due to the statins: almost all would occur just as frequently on placebo.”

This is confident, seemingly ‘evidence-based’ stuff, indeed. However, these findings do appear to me to be at odds with what I and many other doctors observe in real life: that a significant number of people who take statins have side-effects that resolve (sometimes slowly) on discontinuation of their medication. Of course, as the authors of this most recent study allude to, these side-effects may be nothing more than a negative placebo response – sometimes referred to a ‘nocebo’ response.
However, is there anything about the way statin trials may be designed and conducted that could jeopardise our ability to get accurate data on the adverse effects of these drugs?

Several explanations are possible. First, commercial sponsors of clinical trials may not be motivated to search exhaustively for potential side effects. One pointer towards this is that, although evidence of liver damage is documented in the majority of trials, diabetes diagnoses were only documented in three of the 29 trials assessed in the recent study.

Second, many trials do not state clearly how and how often adverse effects were assessed. Because of this, it far from certain that all adverse events were ‘caught’ and logged appropriately.

Third, some trials’ exclude patients with severe diabetes, kidney failure or high blood pressure. In reality, though, these individuals may come to be prescribed and take statins.

Fourth, trial volunteers tend to be enthusiastic, and may therefore be less likely to report side effects than patients in routine clinical practice.

Fifth, many trials have a ‘run-in’ period where individuals are given a placebo to help ensure adequate compliance with medication. This can cause studies to be ‘enriched’ with highly motivated individuals who, again, may be less likely to complain of side-effects.

Finally, many trials excluded patients on medication sharing the same liver metabolic pathway as statins (e.g. fibrates and macrolide antibiotics). Patients on such drugs, in the real world, might well suffer higher rates of pharmacologically mediated effects.

I make no secret of the fact that I think the benefits of statins are over-hyped and that the adverse effects are generally downplayed. As a result, a cynical observer might read my reservations here and think ‘well, he would say that’.

But, here the kicker: those six issues I detail above were plucked from the very same study that trumpets the safety of statin [1]. Much of what is written in this section of the post was actually lifted verbatim from the study.

So, by the authors’ own admission, there are many reasons why the adverse effect rates seen in statin studies may not accurately reflect the rates seen in the real world. But then how can the authors conclude that: “Only a small minority of symptoms reported on statins are genuinely due to the statins: almost all would occur just as frequently on placebo.”

The reality is the deficiencies of the studies do not allow the authors (or anyone) to conclude that at all. The authors’ pronouncement on safety is utterly undermined by their own admissions about the incompleteness and untrustworthiness of the study data.

The opening line of the study is this: “Patients and doctors need clear reliable information
about benefits and risks to make informed decisions.” The only clear thing about the risks of statins, to my mind, is that there isn’t much clarity. Making bold pronouncements on the safety of statins without us having the facts is potentially misleading, and may cause many to come to considerable harm, needlessly.

References:
1. Finegold JA, et al. What proportion of symptomatic side effects in patients taking statins are genuinely caused by the drug? Systematic review of randomized placebo-controlled trials to aid individual patient choice. European Journal of Preventive Cardiology March 12, 2014

Drug company survey suggests statin side-effects are common and the most common reason for people stopping their medication

Drug company survey suggests statin side-effects are common and the most common reason for people stopping their medication
Jul, Fri 13th, 2012  By : Dr John Briffa

Statins are a runaway success in modern medicine, though it should perhaps be borne in mind that for the great majority of people who take them, they do nothing to prevent disease or delay death. Plus, there is growing awareness that these drugs are not without risk, and have genuine potential to cause adverse effects that include diabetes and damage to the muscles, liver and kidneys.

The hazards of statins is not something that the pharmaceutical industry likes to shout about, of course. In fact, industry sponsored studies sometimes look as though they’ve been designed to ensure the real risks of statins remain unseen. One tactic, for instance, is to screen out those most susceptible to side-effects before the study even begins. Another is to detect and log damage to organs only once blood test results are several times the upper limit of normal. For more on this, see this blog post from earlier this year.

So, sometimes it can be difficult to gauge the true risk of taking statins. And one thing I have noticed quite commonly is that many doctors believe the side-effects associated with statins appear to be, in practice, much higher than ‘official statistics’ suggest. Plus, it’s well known that about 75 per cent of people who start statins stop again within a year. Could side-effects be a major reason why individuals default in such large numbers?

Recently, the drug company Eli Lilly issued a press release regarding a survey was called ‘Understanding Statin use in America and Gaps in Education’ (‘USAGE’). The USAGE survey was an attempt, on the face of it, to better understand the reasons for why so many individuals stop taking their statins. More than 10,000 people were polled, and the results are in.

It turns out that off all of the reasons individuals might stop their statin medication, ‘side effects’ was the most commonly cited reason. According to the survey, a full 62 per cent of respondents cited side effects as the reason for stopping their medication. The two next most common reasons cited were ‘cost’ and ‘lack of effectiveness’ at 17 and 12 per cent respectively. You can see from these statistics that problems with side effects was the standout ‘winner’.

One might ask what it was about this drug company that led to a moment of uncharacteristic candour regarding statin side-effects. Well, it turns out that Eli Lilly, in conjunction with Kowa Pharmaceuticals America, markets a statin by the name of pritavastatin (Livalo). This statin is not widely prescribed, but it is sometimes said that it puts individuals at lower risk of side effects.
It seems to me that Lilly is keen to grab a bigger slice of the statin market. However, in so doing, it has revealed to us evidence which supports the idea that, in the real World (rather than the more manipulated and controllable environment of clinical studies), statin side effects are a much bigger problem then some would have us believe.
==================================================================
Read the full article here.

Why statin side effects are likely to be much more common than official statistics suggest

 Why statin side effects are likely to be much more common than official statistics suggest

By : Dr John Briffa

In the UK, the most popular ‘drugs bible’ goes by the name of the British National Formulary (BNF). Within its pages is found a wealth of information about pills and potions that are available over-the-counter and by prescription, including indications and advice of dosages. A significant proportion of the pages in the BNF are taken up with information about contraindications (situations where the drug should be avoided or used with caution) and side-effects. This information is now to be found as part of the packet insert which comes with medication. I’ve known many, many people to read this information and decide that they’ll give the medication a miss.

One class of medication with a range of known side-effects are the statins. These cholesterol-reducing drugs are known to have the potential to cause symptoms such as muscle pain and fatigue, as well as cause damage to organs such as the liver and kidneys. About a year ago I was at a medical lecture, and one (doctor) member of the audience commented that he felt his patients experienced side effects from taking statins far more commonly than official statistics suggested. My own experience supports this observation.

Could there be an explanation for this phenomenon?

One explanation has to do with the design of statin studies. Quite often, individuals who are in poor health and perhaps at increased risk of side-effects are automatically barred from entering a study. Yet, in the real world, even people who are poor candidates in this respect may end up being prescribed a statin. Individuals with a history of problems such as muscular pain or damage to the liver or kidneys (all of which can be exacerbated by statins) are typically excluded from studies too, further reducing the chance that side-effects will arise.

Even those who make it through this screening process, however, may be subjected to what is known as an ‘run in’ period prior to the study. Here, individuals may be treated with a statin with idea being that individuals who are ‘non-compliant’ (do not take their medication as instructed) are weeded out. However, the run-in period also affords the researchers the opportunity to detect individuals who are susceptible to statin side-effects and stop them getting into the study proper.

In other words, in formal studies participants are often at a significantly lower risk of side-effects than those in the general population.

Another problem with conventional studies is how side-effects are defined. Muscle pain is a quite-frequent side-effect of statins. In extreme cases, statins can cause a break-down of muscle tissue known as ‘rhabdomyolysis’ which can have potentially fatal consequences. In some studies, the focus has been on rhabdomyolysis, which means less severe side-effects such as muscle pain or fatigue may ‘go missing’.

Another way in which the bar for side-effects can be set very high concerns the blood parameters used to detect damage. For instance, in a recent study muscle damage was only deemed to have occurred when muscle enzyme levels (a marker for muscle damage) were at least 5 times the upper limit of normal [1]. In this same study, liver damage (another potential hazard of statins) was only deemed to have occurred when liver enzymes were at least 3 times the upper limit of normal. In both cases, a more logical approach would be to regard a rise of any amount above the top end of the normal range as abnormal and significant. This would be more how it is in actual clinical practice.
The elimination of individuals prone to side-effects and the setting of the bar very high for abnormalities help explain why the side-effects from statins seem much more common in the real world than officially quoted statistics.

However, even in the real world, there might be under-recognition of the damage statins can do. That’s because, quite often, doctors will dismiss the idea that statins might be the cause for someone’s symptoms, even when scientific evidence supports such as link. For more on this, see here.

References:
1. Nicholls S, et al. Effect of Two Intensive Statin Regimens on Progression of Coronary Disease. NEJM 2011;365(22):2078-87
======================================================================================================
Read the complete article here.