Tag Archives: Surrogate endpoint

Evaluating health care claims

In some ways this post is a continuation of my previous one titled “…because it’s natural”. In a lot of diseases such as DM-2, Alzheimer’s, and heart disease, there are multiple mechanisms that contribute to either the disease itself or to it’s complications. Two mechanisms of disease that seem to get a lot of space on TV, print and on the internet is that of inflammation and oxidation. Whether it is someone promoting a “super food” that has a lot of antioxidants in it (or a lot of anti-inflammatory activity), or a pill that has plant extracts in it that reduce oxidation or inflammation, the claims should be  taken with a grain of salt. For example in some cases, there may be multiple good studies that show taking said supplement does act as good antioxidant. However this doesn’t mean that the supplement will improve one’s health or lengthen lives. This is a problem with using what’s called a surrogate endpoint. Don’t get me wrong, using surrogate endpoints can be useful when the more significant and relevant endpoints are things to be avoided (death or disability for example) or might not happen for years  – I don’t think a study that would take 30 years to start showing something works would get funding – or finding enough people to study would be practical. It helps if a change in the surrogate endpoint has already been shown to be related to reduction in a particular disease’s morbidity or mortality.There are also observational studies that show certain things (high vitamin A levels, higher beta-carotene intake) are associated with lower levels of a particular disease state. Sometimes using a surrogate endpoint  (or  noticing an association between two things such as high levels of vitamin A and lower rates of a particular disease) ends up leading to people doing negative studies. For example, many deaths after a heart attack are related to arrhythmias. A study called the CAST (short for Cardiac Arrhythmia Suppression Trial) showed  higher mortality in people who were on anti arrhythmia drugs. It doesn’t mean that the drugs didn’t have a role in other disorders. I have a feeling it means that we were just asking the wrong question (even though it needed to be asked and answered) about treating arrhythmias after heart attacks. I use the example of the CAST study to make the point that showing something changes the level of something (either up or down) that is thought to be involved in causing disease – be it inflammation, oxidation, arrhythmia or otherwise – doesn’t mean it affects the disease it is said to. It may be that to treat a disease with anti-inflammatory medications (or substances), one also needs to block other pathways of that disease as a well for any to be effective. This is why most cancers are treated with more than one drug. This is why people are often on more than one blood pressure medication. As I mentioned in my last post, any claims should have data supporting that they work. The results should be reproducible, hopefully other by other researchers. Getting back to my original assertion from the start of this particular blog post, if someone is touting a “superfood” for health, It is appropriate to ask if it actually improves health, decreases complications, etc. In my opinion, just to say something is a “super food”  because it is a ‘super anti oxidant’ is blowing smoke.  That isn’t to say that anti-oxidant rich foods don’t promote good health. There are too many studies that are negative that to look at one thing and say “this is the holy grail of food”. This also undersimplifies things too greatly. It’s better, in my opinion, to say “these are the types of foods/eating (or habits if one is talking about other aspects of lifestyle) that are associated with good health. For examples, many of the studies that show that olive oil intake is inversely related to cognitive decline are more agnostic about what role each component in olive oil plays, despite showing benefits. Is it the kinds of fatty acids in olive oil? Is it the polyphenols that act as antioxidants?  Is it the anti-inflammatory chemicals in olive oil? I suspect the answer is yes, it is all three. Is it the answer to everything: just have olive oil and you’ll live to 100? I doubt it. My bottom line on health care claims:

Be critical but open minded: ask the following questions: “Does it actually affect or prevent disease? Does it decrease complications of the disease and not just something thought to be associated with disease or complications thereof?”. If the answer is “yes, it does reduce _fill in the blank_ and there is a proportional reduction in the amount of deaths/strokes/people going on dialysis/etc then you have a winner. IF the answer is “it does reduce _fill in the blank_” but there is no reduction in _fill in this blank as well_” it may mean that the answer is more complicated than we think. It may mean that the wrong question(s) were asked, or the right ones hadn’t been asked.

Also be wary of claims that make a product or procedure seem that it’s THE ANSWER for a particular disease. It may be a piece in the puzzle, but in order to be considered as such, the answer to the question ‘where’s the proof’ should be along the lines of “here are the studies…”

If the person makes statements like “doctors are in the pockets of ‘big pharma’ and aren’t interested in curing disease”, then be wary. I think most doctors get into this business to make people better. If there was a pill that taken once or twice cured someone of his or her type 2 diabetes (and did not cause some other severe life threatening disease), I think most doctors would use that pill.

Also ask if this the first study of something? Often a treatment is found in a study to be helpful. The numbers of people may be small. Due to the nature of studies, the participants are typically more homogenous than the population as a whole. Once larger studies are done, the benefits of a medication/procedure, etc may not be as large as initially thought.

Another question ot ask is this better than what we have now? Though it could be asked of a new medication, I’m primarily thinking here of new surgical procedures (eg, robotic surgery for certain things). If offered ask: is the rate of complications less with the new procedure? Is mortality less? Is the recovery time quicker with the new procedure?

Related articles

… because it’s natural.

I think with this post I am going a bit off topic (or at least getting on a soap box for a bit). Occasionally a patient will say that they don’t want to take a medicine because they want to try other treatments first, like loosing weight or exercising, if,  for example, they were just diagnosed with adult onset diabetes (also known as Diabetes Mellitus, type 2). Now if their numbers (eg, a glycosylated hemoglobin) that is at or near goal for a treated diabetic, that might not be an unreasonable approach. After all, even if someone is put on a diabetes medication, anti-hypertensive, etc, diet and exercise don’t become less important in treating whichever disease they have.  What concerns me at times is if the reason someone doesn’t want medication for a treatment is because the medication is “not natural” and that they want to try something that is derived solely from natural products.

Now on one hand this might not be totally unreasonable in that many common medications are derived from natural products. Think of aspirin and penicillin (the latter produced by fungi known as penicillium).   Streptomycin is another drug initially derived from a natural source. Digoxin/Digitalis is produced by the Foxglove Plant (which was used to treat what we now call congestive heart failure.  Digoxin and digitalis are still used at times). Morphine is derived from Poppies. The list goes on. As I don’t work for a pharmaceutical company, I don’t know how many of these medications are 100% man made vs being derived from plant sources. Given that a pill has a known quantity of a known medicine, I’d rather take that than risk a stroke or some other complication. My biggest issue is someone wants to take a  supplement to treat a disease “because it’s natural” and because “it’s safe”. I want to see the proof that it works before I’d recommend it. Note, if we’re talking about a disease that won’t kill you tomorrow, or next week, I try to be open minded if a patient wants to try something else first.

Having said that, when someone’s reason for wanting to use an herbal treatment, a supplement, etc to treat a disease because natural products are safe/safer than medications, I am apt to give them a list similar to this: Carbon Monoxide, Snake venom, Radium, Ricin, Atropa Belladona (nightshade), Botulinum toxin, etc. Aside from being toxic, they are all natural products as well. It is my way of hopefully, and gently, pointing out that natural doesn’t always mean safe or non toxic.  If someone is hawking a natural product as a “cure” for something, it’s reasonable to ask the following questions:

1) Is it safe?

2) Does it work? What is the proof that it works? For medications it’s multiple trials that are reproducible and show (usually) a clinically significant difference in cure/length of disease/significant decrease in morbidity or mortality. In some cases it is a change in a surrogate end point. Do the natural products have the same level of proof? Of note, a celebrity spokesperson or the fact an infomercial exists isn’t proof. In my mind if someone is saying that “studies show…” then they should be able at some point to tell you where to find the studies, or provide the references themselves. Obscure or non peer reviewed journals don’t count. Nor should there only be one study showing benefit (it should be a really compelling article if there is only one).

3) What are the alternatives?

Related articles