Statins, Statistics and Stations

TV shows love a good controversy and where there isn’t one available it’s always handy to be able to generate one.

So ABC-TV show Catalyst did an excellent job last week running a two part episode on how statins – a family of tablets used to treat high cholesterol – are overprescribed, not very useful and that doctors don’t adequately think about (or warn patients about) their side effects.

There is enough truth in those claims that they can’t be rejected out of hand. But there is enough about the claims which bends, distorts or maims the truth that they have to be refuted in clear terms. Our depth of feeling can be gauged by the fact that it has drawn us back into blogging!

So what’s true and what’s not. Well …..it’s complicated.

Nobody ever died of high cholesterol.  People die of heart attacks and strokes. The controversy is whether cholesterol is a cause of those fatal conditions, and whether lowering cholesterol with tablets is worthwhile.

We know FOR CERTAIN that high cholesterol is associated with heart attacks.

We know ALMOST CERTAINLY that cholesterol is a risk factor.

We know FOR CERTAIN that statins lower cholesterol in most people.

We know FOR CERTAIN that people who have already had a heart attack, on average live longer if they are put on statins.

We are REASONABLY CERTAIN that people who have not had a heart attack but have high cholesterol are, on average, less likely to have one if they take statins.

We are REASONABLY CERTAIN how likely it is that statins will reduce the risk.

We are REASONABLY CERTAIN that in most individual people, taking a statin will not help – because most people will not have a heart attack, and statins don’t always work in people who do.

We have NO IDEA in a large group of people, which ones will be helped by the statins and who won’t.

We know FOR CERTAIN that a small number of people will get major side effects from statins and a moderate number of people will get mild side effects.

We have NO IDEA in a large group of people who will get the side effects till they’ve had them.

 

So ….should you take (or keep taking) statins?

Perhaps the most important concept to help understand whether statins are useful for you is an idea called Number Needed to Treat (NNT for short).   This tells us how likely it is that a statin will help.  And then YOU need to decide what degree of risk you are willing to accept.

In the following examples it’s not important to understand the individual risk factor numbers – it’s the overall risks that matter:

Imagine a 50 year old male, non-smoker, non-diabetic, with a blood pressure of 140/85 and a cholesterol ratio (Total cholesterol/HDL) of 6.  

His chance of a heart attack in the next 5 years is 5-10%. A statin might reduce that risk by about a quarter. Whether the patient thinks the statin is helpful depends on how his doctor presents the figures:

Doctor 1:   If you take a statin, we can reduce your risk of a heart attack by 25%  (ie from say 7.5% to 5.7%)

Patient 1: Wow Doc, that’s great …gimme a script

OR

Doctor 2: If you take a statin we can reduce your risk of heart attack by 1 or maybe 2%  (ie from say 7.5% to 5.7%)

Patient 2: And run the risk of side effects? Your kidding – why would I bother?

OR

Doctor 3: If we cloned 100 of you and we gave none of you statins, about 93 of you would not have a heart attack in the next 5 years, 7 of you would.  If we gave you all statins – 95 would not have heart attacks, 5 would.  So we can prevent one heart attack for every fifty people who take the tablet.  For the other 49 – about 46 WON”T have a heart attack whether or not they take a tablet. Three WILL have a heart attack whether or not they take a tablet and about 1 of you will prevent a heart attack by taking a statin tablet for five years.

Patient 3:  Well….I don’t want to take a tablet that’s useless, but the idea of a heart attack is pretty scary and I’d want to do whatever is reasonable to reduce the risk. Let me think about it……

 

Another example – a 60 year old diabetic woman who smokes and has a blood pressure of 160/100 and Cholesterol Ratio of 6. Now the NNT to prevent a heart attack is about 13 …ie 12 women out of 13 will get no benefit from the tablet but the 13th will avoid a heart attack.

For some patients, that screams “take the tablet”.  For others, that’s still not enough of a sure thing for them to want to take medication.

The problem with shows like the Catalyst program is that they suffer from the same deficiency as medical research – they deal with statistics, not the individual concerns of each patient and their own views on what they want for their health.

For that you need to see your GP.

We strongly urge patients already on statins NOT to stop taking them till they have had a proper discussion with their family doctor. We know FOR CERTAIN that is a worthwhile exercise!

 

PS … we cheerfully note that losing weight, controlling blood pressure and especially stopping smoking are usually more important than controlling cholesterol. Those issues are subjects for another day!

 

 

 

Statistically it’s much better when it’s exactly the same

It’s very easy to get confused over prostate cancer screening – doctors do it all the time! Currently in Australia there are SEVEN distinct “official” recommendations regarding who should be screened for prostate cancer and how often.  It’s no wonder nobody actually knows what to do.

The big debate of course, is whether prostate cancer screening should be done at all.  Both sides of the debate present statistics, and they can be twisted to defend any position at all.

Here is a great example we came across.

FACT – Survival rates for prostate cancer in America are about 80%.   In England 43%.

FACT – the chance of dying of prostate cancer in America is virtually identical to the chance in England.

OK….how do we reconcile those two completely true, but apparently contradictory, facts?

Here’s how.

Americans love screening and they test lots of people for prostate cancer very often.

And they love treating prostate cancer, no matter how early they find it.

So in America – 136 out of 100,000 men are found each year to have prostate cancer. You can bet nearly all get some sort of treatment.

In England, where many fewer men are screened, only 49 out of 100,000 men are diagnosed.

Now, if all prostate cancers mattered, you would think that the diagnosis in lots of men in England is being missed, that their treatment is being delayed and that they will die as a result.
But the rate of death from prostate cancer in England is 28 per 100,000 men per year.  In America, its trivially better at 26 per 100,000.

In other words, your chance of being TOLD you have prostate cancer in America is more than double in England (136 vs 49).  But your chance of DYING of prostate cancer is basically the same (26 vs 28).

Since more men are diagnosed and treated in America, it looks like regular screening, early diagnosis and aggressive treament is more effective, even when the end-results are the same.

The conclusion?  Research statistics are very dangerous creatures and need to be handled with care.  Talk to your GP about YOUR personal risk.