Lies, damn lies and statistics

[MrsMerryHenry]

I've been thinking about how powerful statistics can be since there are so many stats flying around about this, that and the other to do with the dreaded Swine Flu. So here's my 2 pence' worth. I call it When Good Stats Go Bad:

1 (a) If you are pregnant, you are more likely to have mild SF symptoms (NHS website) than severe ones.
1 (b) If you are pregnant, you are 10 times more likely to have complications (BBC news last week).

• Am I wrong in thinking that the NHS and BBC are contradicting each other? Actually I've just double-checked the NHS page - read under the first sub-heading and it appears to be sort of saying both!

• 200 people in the UK have died from Swine Flu. (But how many have not died...? Oh, look - they don't know ).

• 4 pregnant women in my local hospital are currently hospitalised with Swine Flu (but how many have had it with no complications? Oh, look - they don't know ).

It's not that I'm disputing the figures given, what I want is a BALANCED PERSPECTIVE (or a clear one in the case of the NHS website). Which clearly is not on the cards.

Anyway, I'm primed now to jump on any stats given about any subject under the sun - anyone have any other goodies, whether SF-related or not?

Just me, then?

Another thing has sprung to mind - the BBC reporting on the impact of the SF jab, said last week that 'most people' only had the mildest of side effects from the jab. So having reported extensively for several months on the extreme reactions of the minority to swine flu (complications, death, etc), all of a sudden they're talking about the mild reactions of the majority to the jab. Not exactly balanced reporting, is it?

No, it's not balanced reporting, you are absolutely right and it has bugged me too (although maybe for different reasaons than you )
There have also been no numbers published and rehashed in the media about how many "normal" seasonal flu deaths there are every year.

However, there have been many more pregnant women critially ill and dying from this virus than general population - and I will probably struggle to find any reliable statistical links to prove my point (preg brain).

The main concern about H1N1 Influenze is that it may change into something more virulent, so the vaccination campaign is largely preventing something that we do not know what it is yet, if that makes any sense? Obviously also means that the current vaccines will not work for whatever nasty it might turn into... Gah, I dispair!
For the record, I am up the duff and had my vacc 2 weeks ago.

No you are not alone.

Drives me mad the total abuse and misunderstanding of statistics on the News and by Govt Depts in general. It causes people to be frightened into a hysteria and Govt Ministers to jump to the wrong policy response in reacting to media preassure.

I cannot add any statistics of my own but looking dispassionately this swine flu is a lot less severe than the pandemic we were told would happen.

As you say, 200 people is not many in total but how many have actually had it and how many of those 200 were in vulnerable groups and how many of those were pregnant.

Scary stuff
A bit more

All NHS data, no original research, sorry, but there is pizza going cold here ..

There are no reliable stats about how many people HAVEN'T died because so many diagnoses have been made without actually testing people to confirm that it's really SF they have. So there are likely to be 000s of people who think they've had it in the past 6 months and even got Tamiflu, but who may still contract it later.

Mrs Merry, you are right that news/media reporting rarely represents statistics properly. One thing that really bugs me is when they say "shock horror, xx% of people with x cancer were found to have blah de blah chemical in their blood" or similar. That on its own means NOTHING and yet it's presented as if it's chemical blah is causing a problem. It only means something in context - how many people without the cancer have the same stuff in their blood? Maybe it's also 40%, in which case it's not significant. Maybe it's 70%, in which case chemical x may be having a protective effect. You just can't tell unless you get the whole picture and it's shocking that intelligent people who put the news together are so clueless about this.

Re your example though, these two facts are consistent:

1 (a) If you are pregnant, you are more likely to have mild SF symptoms (NHS website) than severe ones.
1 (b) If you are pregnant, you are 10 times more likely to have complications (BBC news last week).

Most people who get SF will have mild rather than severe symptoms - let's say 1 in 100 cases is severe (for sake of argument)

PG women are up to 10 times more likely than an average person to have severe symptoms so say 10 cases in every 100.

That still means if a pg woman gets SF, with a 10 in 100 (that is 1 in 10) chance of it being severe, she's still far more likely to have mild than severe symptoms.

A lot of the problem with terrible reporting is the poor numeracy levels among journalists (not to mention the general population). Lots of hacks just cream off a few eye-catching points from a study without making any effort to understand what they actually mean (or not), cf MmeduB's cancer/chemical example.

A very common example is when papers report a 50% rise in some kind of shocking but unusual crime. What they DON'T often point out is that this could be a rise of 2 cases last year to 3 this year: the actual numbers often matter, not just the rise or fall in incidence.

Well said, Madame, that was what I was kind of trying to say .

Also fair point re relative and absolute riks/increases, jkklpu, often MUCH misunderstood.

Pacific, I looked at your links, thanks. I couldn't see any stats on mortality for asthmatics, have you seen any?

Madame, thanks for your explanation of the SF/ preg stats, that certainly makes sense.

I'm glad to see that I'm not the only one who finds this statistical mis-reporting an irritation; it's as if we're being taken for fools.

One question about the SF issue - PacificDogwood, you said that the concern is that it may mutate into something worse. Surely if we're all being vaccinated that will only increase the likelihood of this happening?

jkk: "There are no reliable stats about how many people HAVEN'T died because so many diagnoses have been made without actually testing people to confirm that it's really SF they have" - yes, it's a rather inadequate way of handling it, isn't it? How on earth does it make sense to treat it this way? Or to dish out vaccinations without testing first whether people are already immunie? Isn't that repeating the problem we had with overprescription of ABs?

Sorry - didn't want this to become a SF thread!

I'e recently had the SF jab (at 23 weeks pg) and yes I wished they could have just tested me first - I've had various flu/cold/cough things that could have been a mild version of SF. But I think the issue there is probably cost and time - getting people to be tested, getting the results, and then getting the right people to come back for the jab - it would be chaos and take ages.

I don't think (though not 100% sure) that vaccinations work the same way as ABs in terms of resistance. A germ can develop resistance to an antibiotic/other treatment through the process of evolution - if a few of the "fittest" germs survive and breed, the stronger, most resistant strains can make it through and become predominant. Vacc is about giving you a weakened/dead form of the germ so that your body develops resistance (ie antibodies) to it. It won't always work perfectly, but vaccinating more people than necessary shouldn't encourage the germ to evolve, it should just make it harder for it to spread and breed. (But I stand to be corrected on that)

Hmm...wish I understood more about epidemiology (if that's the right discipline) - your description of the way the fittest viruses (or those with the 'best' mutations) survive and breed still sounds very much like something that viruses could do under a vaccination programme. But I also stand to be corrected!

I see what you mean about the potential chaos - but I still wonder whether the money that's been diverted to produce this very inadequate SF response could have been used better had they relied on better diagnostic tools. Good planning could take potential chaos into account. Apparently the SF epidemic was predicted about 10 yrs ago - according to the New Scientist. So they've had plenty of warning...

I do not listen to any statistics quoted by any newspaper or news programme because they are generally non-sensical. However, I also hate predictions.......they are generally dangerous. A study ( the name escapes me for now, but know it was Phil someone) looked at predictions made by 'experts' on a variety of subjects over a 15 year period, and compared them to predictions made by laymen, and found the 'expert' status to be of no benefit whatsoever. In pop-psychology, it's the White Coat Syndrome.

(I bet you have all read the books, but Bad Science, Ben Goldacre and The Black Swan, Nicholas Naseem Taleb cover these beautifully - esp. the Black Swan)

It's similar to reports that say things like:

21% of road deaths in 200X were caused by drivers who were either drunk or under the influence of drugs

So using that statistic 79% of road deaths were caused by people who were not drunk or under the influence of drugs

Therefore, usung statistics, it's safer to drive while off your face on something

Ladybla - haven't read either of those, but now I will! Love the use of White Coat Syndrome here - how appropriate indeed.

Iklboo - madness, isn't it?

It normally takes a couple of years at least to gather and analyse epidemiology data sufficiently to give a clear picture.

It also relies on clear testing of the population for cases like swine flu, which hasn't happened as so many people have been 'diagnosed' over the phone without being confirmed, and a large proportion of those people probably haven't had swine flu at all.

Obviously with something like swine flu, people are looking for the information now. As there are no definitive published statistics, information is being taken from various sources and so it often ends up being speculatory or extrapolated from early estimates, which means different sources will be contradictory.

Also, there won't be any data on the exact number of cases and the proportion of those without complications because the total number of cases is not known. They can only report clearly on confirmed cases that are reported as having complications/end up in hospital, because these are the only hard and fast numbers they have.

It isn't poor performance on anyone's part (except maybe in terms of the fact that so many diagnoses are 'guesses') - it is simply impossible to know all the facts and figures at such an early stage.

MrsMerryHenry, as Madame said, vaccination work entirely different from antibiotics: in the case of seasonal flu vacc or Swine flu vacc an inactivated from of the virus is used to make your own body produce antibodies which kind of just lie in wait until you are exposed to the real thing and then pounce. Most antibiotics stop bacteria (not viruses) from multiplying (there are only very few meds that actually kill bacteria) which again gives a different part of your immune system a chance to fight those that are left. If these bugs are repeatedly exposed to an AB that kills say 90% of them then the remaining 10% get a chance to mutate into something that is resistant.

The worry about H1N1 is that it has already shown it can mutate v readily as it changed from a bug that made pigs ill, to one that could affect humans in contact with pigs, to one that can be spread from human to humann (no pigs required ). So even having potentially even just partial protection is in my eyes worth it.

You are right, a lot of predictions are worst case szenarios and nobody without a crystal ball can predict whether they are going to happen or not (likely not worst case, but governments like to prepare for the worst..). Also bear in mind that Public Health people and epidemiologists look at the bigger picture of a nation or even an economy surviving; the health of an individual like you or I or our kids are not their primary concern.

Also, nobody really knows why the Spanish Flu in 1918/19 killed millions and millions of mainly young people. It was an Influenza A virus like seasonal flu and like Swine Flu but obviously turned into something really nasty. Apparently amongst the French it was kown as the "3 day fever" in that it came on suddenly, did not last long and at the end of it you were either ok or dead . There have been attempts to get viral DNA from victims of the Spanish flu by taking samples from dead bodies that were buried 90 years ago, yeuch, and no success.

Again you are all right you cannot confirm the diagnosis of Swine Flu without a test as the symptoms are similar/the same as any other Influenza. As anybody who has ever had "proper" flu will tell you, symptoms are far worse than even a bad cold. On the balance of probabilities if you have flu-like symptoms at the moment you have an 80% chance it being H1N1 (this is according to numbers from England in October I think, when people were tested in certain sample areas).

Right, I am going to stop. It is all very complicated. I struggled to get my head around it and quite likely still have things wrong.
I just despise conspiracy theories along the lines of "the government is trying to kill us all and our babies with these untested vaccines", gah!!

If you got this far, here have a medal .

Toosh - yes, I know all that already - but don't forget, even where they do have full data (e.g. with my earlier jab side effects example), the information presented is still partial and biased.

They knew we were due for a flu pandemic but that doesn't make it straightforward to deal with it when it arrives. I don't think it's true to say we were unprepared. We were prepared but turning a new strain of flu into a vaccine and testing it to a reasonable degree is going to take time - and the authorities can't win because if they had managed to come up with a more immediate response, the cries of "dangerous untested vaccine" would have been even louder.

I suppose what I'm trying to say is that yes, the media often misrepresent science and statistics by dumbing them down to a point where they don't make sense any more...

BUT - we need to remember that in the case of swine flu or similar, the basic facts are not yet there, because it is too early to measure/too ambiguous with the difficulty of confirming swine flu cases, so even among the raw data, there are bound to be lots of contradictions and inconsistencies at this stage.

It's unrealistic to expect anyone to have all the answers at this stage in the swine flu story.

On the subject of other data, I agree the media are very guilty of over-dramatising statistics and taking them horribly out of context.

Remember the headline that "chocolate is actually good for your teeth" a while back? All because they found some ingredient in coccoa powder that alledgedly fights plaque, but were conveniently forgetting the amount of sugar in chocolate...

I also think something the media/newsmakers do is jump on bandwagons all the time. Something happens that is considered news, then they will report on loads of other instances of the same thing happening and make it look as if it's suddenly a massive issue when it wasn't before. When in reality they just weren't reporting it before because it wasn't deemed "news". Bad reactions to vaccinations is a classic example. They have always happened and for years that wasn't news in the slightest - now it is because concern about vaccinations has become a bandwagon issue.

I wonder how many people who have let themselves be whipped up into a frenzy about the dangers of SF vaccine/cervical cancer vaccine etc have been on holiday to thailand or india and had their holiday vax without a second thought. But those jabs can cause a bad reaction too - it's just not bandwagon issue to worry about them.

(I'm not btw saying that there's nothing to worry about - just that how much we worry and what about is often influenced by media whims.)

Leningrad FWIW - mine's a bit sore but only affects me when lying on that side in bed.

Madame - I certainly wasn't saying I'd expect them to have prepared a vaccine in advance - not unless they already knew the nature of the virus, of course. No, I'm saying that they should have planned their strategy better. Sorry, how foolish - that would require governments to take a long-term approach to governing.

Re your subsequent post - too blardy right.

Rofl at chocolate being good for your teeth! A beautiful example of us being treated like asses!

Good luck with your jab, Lenin!