critical appraisal

"I do my own research"

Something that I see a lot in on-line debates about alternative medicine is phrases like “I did my own research” or “people should be allowed to do their own research and make their own decisions”

However, I don’t think that the vast majority of people are able to do their own research. Now, that’s probably a pretty unpopular opinion. It’s patronising, paternalistic, and it flies in the face of patient choice. Who am I to question the intelligence and abilities of other people? Why do I think I'm so clever compared to anyone else out there? Allow me to explain myself.

I've been a pharmacist for a very long time now. From uni, through pre-reg, to my own revision at work, I've been taught critical appraisal skills. Yet to this day, it’s something that I actually find really hard work. It’s a skill that requires continual honing, and every time I use it I feel like I am fighting with my brain. 

Even in the last two weeks, I've been revisiting my critical appraisal skills to make sure they are up to date. I've done some in-house work, three on-line courses, and a one to one training session. Yet I still find myself sat here at my desk for several hours, if not days, looking over the same study with a furrowed brow, desperately trying to make the numbers and statistics tell me their story.  If I find it so hard, then how on earth is someone without any medical background or critical appraisal training supposed to do any of it? 

There’s hazard ratios, odds ratios, confidence intervals, numbers needed to treat, event rates, absolute risks and other confuddling terms to deal with. I naturally struggle with numbers at the best of times; like most people, I much prefer narratives. That means that I have to constantly argue with myself to keep looking at the results page, rather than just flicking to the discussion. Because if I did that, I'd be relying on what the authors, with all of their possible biases and agendas, say their numbers say. Then, when I eventually manage to squeeze the swimming mass of figures into some sort of order in my head, I find out that these numbers aren't the full story, and I need to dig even deeper into other analyses of the same figures to find out what’s really going on.* 

A quick and very simplistic visualisation of all the layers of interpretation that might lead to information found on your common or garden health information website. That's a whole lot of bias.

A quick and very simplistic visualisation of all the layers of interpretation that might lead to information found on your common or garden health information website. That's a whole lot of bias.

It’s not a pleasant task by any stretch of the imagination. It really does feel like a mental marathon. I often question whether I am even up to the task- I can end up feeling stupid, and confused. But in order to really figure out whether or not a drug works I need to strip away all the levels of other peoples’ interpretation and start from scratch, with the cold, hard, impersonal numbers. That way I can build my own narrative, uninfluenced by what the study’s authors or sponsors want me to think, by what newspapers want me to believe, by what campaigners want me to know. The only way to know the truth is to start right at the bottom, in a dark dank pit of statistics, then to slowly start building yourself a ladder until you emerge, blinking, into the pleasant knowledge that you've worked out what on earth is going on.

This sort of raw data is not only extremely hard to deal with once it’s in front of you, but its also pretty difficult to come by. Finding it in the first place includes searching multiple medical databases- and these things aren't just a quick free text search like you would do on Google. Constructing a search can in itself take an hour or so, and then you have to trawl through the results to decide which are relevant to what you are specifically looking for. For me, most of the time, a question is structured like this:

What is the evidence that [drug/ group of drugs] works for [disease] in [patient group

 So, in my poorly drawn Venn diagram below, I need to find those holy grail papers that reside in the pink area:

I am truly terrible at MS paint, but you get the idea.

I am truly terrible at MS paint, but you get the idea.

What a typical EMBASE search looks like. This is for a new drug with few synonyms so its a fairly straightforward one. Others can have forty odd lines of searches.

What a typical EMBASE search looks like. This is for a new drug with few synonyms so its a fairly straightforward one. Others can have forty odd lines of searches.

Some of these papers might be pay-walled, so it’ll take me a week or so to get my hands on them. Some of them might initially look promising, but once you start to dig down into the figures you see that there might actually be problems with how they were undertaken or reported, or they might turn out to not quite fit in some way- perhaps the dose they used in the trial is different to the licensed dose in the UK, or the people enrolled into the trial don’t quite fit the population you want to know about, or perhaps the trial just didn't recruit enough people so any results from it are invalidated.

I've been doing this job for years, and I really do still struggle with all of this stuff. That’s not because I'm poor at my job, or because I'm stupid, or because I haven’t put the effort in to understand it. It’s because, when it comes down to it, this stuff is really bloody hard. It’s time-consuming, boring, and unintuitive.

People might well feel like they've done their own research. They might spend several hours digging about on the internet and feel empowered by any decisions that they make. But what they don’t realise is that what they've been researching isn't just the information- it’s the information with many, many layers of interpretation (and therefore bias) added. For a choice to be truly informed, you need to go right back to the start, to those terrifying tables of numbers and statistics. That’s simply not realistic for the majority of people.

Far better, then, to learn how to decide on whose interpretation you’re going to rely on. Will it be those that take the media reports at face value, or who have an agenda or a product to sell you? Or will you go with those that have years of training in how to pull apart complicated data and disseminate it in understandable ways?


*I thought I’d give you a quick real life example here, but I thought it best to asterisk it because I've probably bored you enough already. I'm currently looking at a drug called edoxaban and its use in reducing the risk of stroke in patients with atrial fibrillation. It’s the newest in a series of novel oral anticoagulant drug- they’re supposedly like warfarin, but less faffy. So I find and look at the main trial, and spend days unpicking the stats. It looks like both strengths used in the trial are no worse than warfarin, and the higher dose might even be a little better. Great, right?

Well, that’s not quite the end of the story. Because it turns out- and this isn't reported in the trial at all, but instead is contained in the FDA’s briefing document- that in people with fully working kidneys, edoxaban is actually worse than plain old warfarin. In people whose kidney’s aren't quite at full capacity though, it might work better than warfarin. So the overall trial results are kind of skewed, and if we didn't dig deeper, we might have been giving a whole group of people a more expensive drug with worse outcomes than warfarin. Even the FDA findings are borderline- some of what they describe doesn't reach statistical significance.

When real science gets left out in the Coldzyme

There’s no getting away from it, folks. Its sniffle season. For the next 6 months or so, the sounds of sneezes, coughs, and millions of noses being blown will echo throughout the nation.

We all know by now that the common cold is a virus. We all know that there is no cure. We also all know that, although you feel like crawling into a small dark warm cave and dying at the time, its usually much better after a few days, and it goes away of its own accord. Cold and flu remedies do nothing to actually get rid of your cold- they are there to make you feel better during it, although many of them are actually irrational combinations of products in shiny boxes with a redonkulously high price.

It is often said that if someone did come up with a cure for the common cold, they would be millionaires. I was, therefore, surprised to read this week in Chemist + Druggist magazine that indeed, the first ever product to not only treat the symptoms but to act on the virus itself was winging its way to pharmacy shelves as we speak. Really? Because blimey charlie, if that's the case, then this product should be Big News. 

The product is ColdZyme, a mouth spray that costs £8.99 for 20mLs. Seems a pretty fair price to pay for a product which claims to cure the most prominent infectious disease in the western hemisphere. It seems odd, though, that instead of this marvellous scientific breakthrough being plastered all over the media and medical literature, the article announcing it is tucked away quietly in a barely read corner of a trade journal.

What is this breakthrough, miracle product that will powerfully break down viruses? Well, an enzyme called trypsin. An enzyme that already merrily and plentifully kicks about in your digestive system, breaking down proteins. An enzyme which, for the purposes of this product, is inexplicable being derived from cod (which has meant that I have had to resist the urge to refer to it as somewhat fishy.) An enzyme which should be stored at temperatures of between -20 and -80 degrees Celsius, to prevent autolysis. Now, I've seen some fancy medicine packaging in my time, but never a simple mouth spray bottle that can manage such cold chain storage feats. So, if trypsin really is present in this product, then it seems fairly likely that its going to be inactive, unless the manufacturers have found a way of warping room temperature. Or you happen to be in Winnipeg in the middle of winter.

Medicine vs. Medical Device

The manufacturers make some really very extraordinary claims on their website, including one textbook example of special pleading. Their product, they state, isn’t a medicine. It’s a medical device, because it has no systemic effect. They then of course go on to helpfully tell us about the systemic effect it has:

“The medicines currently on the market only treat the various symptoms of a cold. ColdZyme treats the cause of the symptoms – the virus itself – and thus works both preventively against the common cold and shortens the duration of illness if you have already been infected.”

Right. So in the same breath, they are claiming that the product only forms a barrier, no more. But then they are also claiming that this barrier affects the ability of the virus to produce illness if you are already infected- viruses which are already through that barrier and inside your body. Come on, Enzymatica, you can’t have it both ways.

The Evidence

All these claims are backed up by evidence, right? Well, there is a tiny trial performed on only 46 people, which isn’t published anywhere. I can’t say whether or not it is a well designed trial, because I can’t see it in full, so to be honest, we pretty much have to just discount it. What we can do, however, if have a look to see if there is any other decent published information looking at the effect of trypsin on the cold virus. So I turned to the medical databases Medline and Embase, to trawl through the published medical literature. 

I did find one experiment which looked at the trypsin sensitivity of several human rhinovirus serotypes(1). And this appears to have found that viruses are only really susceptible to trypsin when there have been exposed to low pH, followed by neutralization- something which wont have happened to your common or garden cold viruses. I couldn’t find much else suggestive of a clinically significant antivirus action of trypsin.

The practicalities

This isn’t a simple, one-off- couple of sprays and away flies your cold sort of product. You have to use it every two hours, as well as after you brush your teeth and before you go to bed, and you have to continue this “until your symptoms are relieved”. That’s one hell of a regime. I have difficulty remembering to use medicines twice daily, never mind every two hours. I’ve never used this product, but I’d imagine that if it really does leave a “barrier” coating in your mouth, its a pretty unpleasant sensation. I can’t imagine many people sticking closely to these dosage instructions, and if the mechanism of action is as the manufacturer’s claim, skipping doses would cause the product to fail (if, indeed, it works in the first place)

We are also directed to “Start using ColdZyme® as soon as possible when you detect symptoms of a cold.”. Now, those of use who suffer with cold sores who have ever used aciclovir cream will know that this is often easier said than done- you probably haven’t got the stuff in the house, or at work, and by the time you’ve managed to get your hands on some, its already too late- your cold sore is out loud and proud, and using the drug will be pointless. Its likely that the very same thing will apply here. And remember that the incubation period for a cold is about 2 days- so the virus will already be cosily settled into your body before you even know about it. Its therefore completely ludicrous that this product claims to be able to reduce the length of a cold simply by forming a barrier.  

I know it can be used as a cold preventative, but how many people who feel completely fine are going to remember to use the product every two hours, every day, for the entirely of the cold season?

To Summarise

So, do I think there is scientific evidence to back up the extraordinary claims being made by ColdZyme? I might do when hell freezes over. Or at least when some decent trials are published, which might take just as long.  Do I think that this product should be sold through pharmacies? Absolutely not- this isn’t, if you ask me, real medicine. This is pure pseudoscience, trying its best to fool you into buying real medicine. Do I think lots of people will buy this, use it once or twice, then leave it to languish in their bathroom cabinet? Absolutely.

Here’s the problem though: this stuff will appear on the shelves of pharmacies all over. The pharmacists wont have a clue what this stuff is, and because they are really busy and probably quite tired at the end of each day, they wont be able to do the sort of evidence review I have managed to squeeze into a quiet moment. So they’ll get asked about it, and they’ll sell it. Some people will buy it and will feel better after a few days, and will think that the spray has made them better, forgetting that colds are self-limiting anyway. A customer might come back in the pharmacy one day, and say something like “hey, that new-fangled spray got rid of my cold!”, and the pharmacy staff will end up making recommendations on the basis of customer feedback and anecdotes, rather than on the basis of rational, scientific evidence. In my eyes, this really is a shame, and by selling this sort of nonsense, we really are cheapening our profession, and we're causing our customers to waste their money. 

If patients ask me about it, when I’m working behind the counter, I’ll tell them something along the lines of: “there’s no evidence or logical way that it works. It seems to be a bit of an expensive gimmick, with no decent basis to it. You’ll feel horrible with your cold, but it will start to go away of its own accord, I promise. In the meantime, you’d be much better off looking after yourself, having plenty of fluids and rest, and taking paracetamol according to the packet.”


A Miracle Migraine Machine?

Cefaly. No, it's not a village in Wales, nor is it a type of cheese (actually, it might be for all I know, but nevermind.) It is instead a new all singing, all dancing miracle cure for migraines, according to its manufacturers anyway. So, in our usual fashion, let's take a look at the evidence and see what on earth it is, and whether it is worth spending money on.

It's a medical headband device that you wear on your noggin, around your forehead. This means that you can easily pretend to be the Empress from the Never Ending Story. The downside is that you'll have to pay somewhere in the region of £250 to do so, plus electrodes and batteries. So, for that amount of money, you want to know that what you're getting is going to provide you with a bit more than simply cosplaying as a child-like film character.

What I love about pictures like this is that it's always perfectly made-up women in them. As if anyone can be arsed to think about makeup when they're vomiting everywhere and their head feels like its being crushed. 

What I love about pictures like this is that it's always perfectly made-up women in them. As if anyone can be arsed to think about makeup when they're vomiting everywhere and their head feels like its being crushed. 

It is essentially a TENS machine, which applies an electric current to the middle of the forehead via self adhesive electrodes. Anyone who has ever used one of those godawful Slendertone thingies on their stomach is probably right now recoiling in horror at the idea of having to endure such torture right between their eyes- I know I am. But first I suppose we need to see if it works- after all, migraines are horrible things which can massively impact on the quality of life of sufferers. Those who are desperate may be quite happy to have their foreheads electrocuted.

Its been approved by the FDA, which is nice. What isn't quite so nice is the fact that this approval is based on one trial- the one and only trial in existence, despite what the manufacturers would have you believe.

This trial included 67 patients who suffered at least 2 migraine attacks per month. Although small, this trial is well designed, with an identical sham stimulator being used as a comparison to the test product. After three months of daily 20 minute usage, the mean number of migraine days in users of Cefaly was significantly reduced (6.94vs 4.88, p=0.023), but were not significantly changed in the sham group. But here's the thing: the difference between groups was not significant (p=0.054).

There was significantly higher percentage of responders (defined as ≥ 50% reduction in no of migraine days per month) in the Cefaly group compared to the sham group (38.24% vs 12.12%, p=0.023).

There was no significant difference in severity of migraine.

Although some of the results in this trial are encouraging, it is limited by its very small size. It is worth noting that the authors and manufacturers claim that this trial proves that the product is effective at preventing migraine, despite the lack of a significant between-group difference in the primary outcome of migraine days.
Other papers have been published in the literature regarding this product, and the manufacturers try their best on their website to make them look like they are real trials. However, these range from letters, conference abstracts, experiments in healthy adults, and case studies- not robust clinical trials.

An uncontrolled survey of 2313 Cefaly rental users found that roughly just over half of patients were satisfied with the treatment and would be willing to buy the device. The rest of the patients stopped therapy- that's a pretty high number of people. There are a number of methodological and confounding problems with this study, so the conclusions drawn from it should be considered unreliable.

Being a rental user is one thing- at least they were able to try it out before taking the plunge and handing over a rather large wad of cash. In the UK, though, it seems that the rental option isn't readily available. £250 is an awful lot of money to spend on a product, especially when, for roughly half of its purchasers, its going to be used a couple of times then lie in a cupboard, forlorn and forgotten about. 

Let's have a think about compliance. To get the best results, you are supposed to use it for 20 minutes per day. Now, initially that might not sound like too big a deal, but if you work, have a social life, go to the gym, or spend every waking minute building a house in Minecraft, finding 20 minutes a day for something that could be, in most cases, painful, is probably pretty unappealing, and impractical. I can't see too many people who will be able to religiously use this product exactly as intended in the long term. I'm guessing that in most cases its going to go the way of that bit of exercise equipment that you bought 5 years ago and that you've used twice and now only trip over on occasion.

So to summarise: there is a little bit of encouraging data, though it's not as compelling as the manufacturers would like us to think. It's extremely expensive, impractical, and probably pretty unpleasant to use. Its an interesting device, but one that I am placing firmly in the "Yet to be convinced by larger trials" pile.


Blowing a raspberry

Imagine there is a door in front of you. There is a person standing next to the door:

"Hey there, pudgy, do you want to be thin and beautiful without having to do any exercise at all? All you have to do is hand me some cash and step through that door there."

This person goes on to explain that if you step through the magic door, you will step out of the other side with a perfect, slim figure. You won't need to change your diet or do any exercise. You just need to walk through the door. They show you some glossy before and after pictures. There are pictures of beautiful, smiling people holding out the waistbands of their old elasticated tracksuit bottoms, so you can see just how much more skinny they are now.

There is a slight rumble from behind the door. You ask what is behind it.

"Oh, we don't know. But honestly, it'll work." 
"But is it safe?" you ask. 

"Oh yes, its perfectly safe" is the reply. "All you need to do is go through that door, and you will definitely lose all of that stubborn weight and absolutely nothing bad will happen to you whatsoever. Now if you'll just hand me your money there, that'll be great."   

Sounds great, right? You excitedly take a step towards the door.

But before you step through the door, you take a moment to consider what might be behind it.  You don't know, and the person telling you to go through the door doesn't know what's in there either. No one has studied what's behind the door. There are no photos and  no videos of what's in there. All you have is this person's word that it is safe, and it will work. 

Broadly speaking, there are three possibilities of what's behind the door:

  1. Nothing. Its just a room. You go through the door, and nothing happens. You don't lose any weight, but you have lost your money. 
  2.  It works. You step through the door and emerge as a smiling, happy, beautiful, skinnier you. You quickly run home to find some horrible grey tracksuit so you can smugly show everyone how marvellous you are now you're skinny.
  3.  Something harmful is in there. Maybe there is a big pile of dog poo right behind the door, and you are destined to step in it, which will be mildly unpleasant for you. Or the door could be perched on top of a cliff, below which is a mesh of razor blades that will cut you into little pieces. Worse than that, perhaps Piers Morgan is through there.

Do you step through the door? 

Now, I really do want to lose weight. But I also don't like being deceived, and I really don't like being harmed. I also really don't like Piers Morgan. 

All of which brings me nicely on to this season's most fashionable weight- loss aid, Raspberry ketone. Its the chemical contained in raspberries that makes them smell nice. Its widely used as a flavouring and fragrance agent in foods, which begs the question: if its so effective, why don't we already lose weight when we eat food that it is in?

You've probably heard about it already. By which I mean that you've probably seen someone on Twitter tweeting about how they lost 3000 kg using raspberry ketone, along with a handy link to a website where you can buy it. You've also then probably seen the follow up tweets, that go something along the lines of "Urgh, my account has been hacked. I've changed my password now, sorry."

And there's the first warning sign. People who make real medicines which work don't tend to need to resort to hacking people's social media accounts in a desperate bid to get people to buy their wares. You don't see Pfizer or Glaxo, for all their faults, hacking into random peoples' accounts and mass tweeting "My cancer is in remission thanks to Drug X. Buy it here!!!!". Its not a particularly ethical way to sell healthcare products.

Putting all of that aside, is there any evidence that raspberry ketones work as a weight loss aid?

In short: no, except for 18 mice and a difference of about 5 grams at most. Which also means there is no evidence that it is safe. There are no human studies out there at all. All of this hype about it is based on the fact that its chemical structure looks a bit like two other chemicals (capsaicin and synephrine) which might have some effects on weight loss.

That's a bit of a stretch. Human bodies are complicated things, and we can't always predict how they will react to medicines. Even when we theoretically think that something might work for good, plausible reasons, there's no guarantee that it will (COX-2 inhibitors come to mind as an example). Marketing raspberry ketone as a foolproof weight loss aid just because it looks a bit like some other chemicals is like finding a random key on a street and expecting it to work without fail on your front door, just because it looks a bit similar to your front door key. 

If you were to decide to take raspberry ketone, you would essentially be walking through that door. You might lose a bit of weight, or you might not. You might be harmed, or you might not. There's just no way of knowing right now.


The Society of Homeopaths and what they pass off as evidence

So today has seen some great news for rationality, science, and above all patients. The ASA has announced this ruling, leading to the Society of Homepaths taking down a rather large chunk of their website- the bit about what homeopathy can be used for.

However, using their search function, you can still find some of the nonsense they are promoting. I stumbled across this article, for example, entitled "Homeopathy Offers Alternative Relief for hay fever sufferers". I'd be very surprised if this article doesn't get taken down soon also, to be honest. It really should, given part of the ASA's ruling relates to their claims over the efficacy of homeopathy for hay fever. 

That use of the word alternative (as opposed to complementary) is interesting. That in itself suggests that the Society of Homeopaths are advocating patients not using conventional medicines in favour of their homeopathic products.

One thing that I have learnt about homeopaths is that, despite the fact that they often claim that randomised controlled trials (and indeed science in general) can't explain their wondrous treatment because of its individualised nature and quantum nanoparticles blah blah all the other words that they're clinging onto, they like to cite trials. A lot.

Homeopaths will often spout names of trials and provide links to PubMed abstracts with abandon, even when the trials say little about the clinical use of homeopathy in humans. In vitro or animal trials are favourites, and on the odd occasion where I have been sent a human trial, the result usually show that homeopathy is no better than placebo, and in some cases actually worse than placebo. At best, I'm guessing this is just ignorance- maybe they have misread the results of the trial? At worst (and more realistically), its a pretty obvious and petty method of obfuscation, and a pretty rubbish one at that. Presumably they think I will be so vowed by the fact that a trial exists that I wont bother to check the actual results of what the trial is saying.

This hayfever page overs a great example of this:

"A number of research trials have shown that homeopathic treatment can produce a significant improvement in hay fever symptoms,(4-7) but what does this involve?"

 Let's have a look at the "number of research trials", shall we?  

  • Reilly DT, Taylor MA, McSharry C, Aitchison T. Is homeopathy a placebo response?  Controlled trial of homeopathic potency, with pollen in hayfever as a model. Lancet,1986;2: 881-6.

This is a trial from 1986. Really, that's the best they can do, in 2013? The abstract of this trial appears impressive: "The homoeopathically treated patients showed a significant reduction in patient and doctor assessed symptom scores", but neglects to mention the most important part of a study like this: blinding. How can we assess the placebo effect in a study that isn't blinded? especially when the results rely on only reported outcomes. We can tick this one off the list as being a pretty rubbish effort at a trial. Must try harder. 

  •  Kleijnen J, Knipschild P, ter Riet G. Clinical trials of Homeopathy. Br Med J, 1991; 302:316-22.

Ahh, the early nineties. We're getting thoroughly modern and hip now, eh? This is a meta-analysis. hay fever isn't mentioned in the abstract at all, and the conclusion of the paper is that studies performed in homeopathy are rubbish, and better ones need to be done. Hardly a conclusive statement that homeopathy works for hay fever. We can tick this one off the list too.

We're now left with two trials to back up that statement above. To me, two trials is not "a number" of trials, even at this point. Even if these two trials were massive, robust, good quality randomised controlled trials, I still wouldn't be entirely convinced: I'd want to see the result replicated in as many different trials as possible. Anyway, we shalll soldier on, in the hopes of being dazzled by the brilliance of these two references. 

  • Launsø L, Kimby CK, Henningsen I, Fønnebø V. An exploratory retrospective study of people suffering from hypersensitivity illness who attend medical or classical homeopathic treatment. Homeopathy, 2006; 95: 73-80.

Oh dear. A retrospective study. So not a controlled trial at all then? The results? "The two groups of patients were similar in respect of their health at the start of the treatment, 57% of the patients who consulted a CH experienced an improvement of their state of health compared to 24% of the GP patients." well, that's all very well and good, but there is no blinding here whatsoever, and only 88 patients completed the study. means nothing at all, except for- as even the authors put it- as an exploratory study, maybe to try to find ways of how to conduct as more robust actual trial in the future.

That's it, down to the final trial. I'm expecting great things.  

  • Kim LS, Riedlinger JE, Baldwin CM, Hilli L, Khalsa SV, Messer SA, Waters RF. Treatment of Seasonal Allergic Rhinitis Using Homeopathic Preparation of Common Allergens in the Southwest Region of the US: A Randomized, Controlled Clinical Trial. Ann Pharmacother, 2005; 39(4): 617-24.

HURRAH!!! It's double-blind! We've gotten there! We've gotten some good, robust evidence tha- oh hang on, its only got 40 participants in it. It's just a wee ickle study that's far too small to draw any conclusions from.

So there you have it. This page is still up there on their website, using crappy references that don't back up their claims. The Society of Homeopaths- and quacks in general- need to realise that, no matter how hard they try, just trying to shoehorn poor excuses for studies in wherever they like isn't good enough.

Here's how it should go: you look at the evidence, you evaluate the evidence, and you make your claim on the basis of that evidence. Not: "I shall claim this, then try desperately to find something that vaguely looks like it backs me up, and I'll just hope for the best that no-one else bothers reading it. It seems the Society of Homeopaths are going in for the latter, and good on the ASA for pulling them up on it.

"We told the Society of Homeopaths not to discourage essential treatment for conditions for which medical supervision should be sought, including offering specific advice on or treatment for such conditions. We also told them not to make health claims for homeopathy unless they held sufficiently robust evidence of efficacy." -ASA ruling