Don't take Magazine articles on Antidepressant Drugs Too Seriously

Home > Drugs / Therapy >
5 July 2011

If you are manic depressive (with bipolar disorder) or depressed, it is hard to find reliable sources of information that would allow you to understand what is happening and what you can do about it.

But some information is worse than others and the public media - newspapers and magazines - generally fall into the worse category. I came across the article below and I was horrified at how misleading the information was. The article was published in 2000, but the quality of reporting on drugs has not gotten better over the years.

The Article and My Comments

The article is called "Difficult Mood Disorder Successfully Treated; Largest Study Proves Effectiveness Of Lamotrigine For Rapid Cycling Manic-Depression" and was published on the website of Science Daily at this reference. On the surface, the article sounds good but let's look at it in detail.

Note: Text is italics is quoted from the article. Plain text are my comments.

"ScienceDaily (Dec. 14, 2000) — According to the largest and only prospective placebo-controlled study in patients with rapid cycling bipolar disorder (manic depression), a drug called lamotrigine is safe and effective in preventing episodes of mood swings. The study, published in the current issue of Journal of Clinical Psychiatry, was authored by Joseph R. Calabrese, MD, Director of the Mood Disorders Center at the University Hospitals of Cleveland, Professor of Psychiatry at the Case Western Reserve University School of Medicine, and the lead researcher. "

Right, so we are told it is safe and effective in preventing mood swings. This is the first line in the article and the one that is supposed to make the most impact. So you'll remember it. Safe and effective - all those big names and universities say so. Still can't pronounce the name of the drug yet, but that's less important than safe and effective.

“There is a clear need for more information in this disease area,” notes Dr. Calabrese. “Patients with rapid-cycling bipolar disorder are constantly moody and difficult to treat. This illness devastates personal relationships, as well as employee/employer relationships. Up until now, there were no medications that were systematically evaluated in large-scale studies.”

Awful disease, but there were no medications until now. And a large scale, systematic study was done, so it really must be safe and effective.

"Bipolar disorder (also called manic-depressive illness) affects one to two percent of the general population and is characterized by distressing and disruptive mood swings. This disorder usually begins in adolescence or early adulthood, although it can sometimes start in early childhood. Nearly one-fifth of bipolar patients also suffer from rapid-cycling disorder, in which at least four times a year, they experience any combination of manic, hypomanic, mixed or depressive episodes."

The information for the laymen reading the article, not us bipolar persons. Nothing wrong here.

"In this study, 324 patients who met the criteria for rapid-cycling bipolar disorder were enrolled in an initial stabilization phase, during which lamotrigine was added to the patient’s current treatment."

There were 324 people in the study. They were on lamotrigine and other medications at this point. Hold the number 324 in your head.

"Patients who displayed early signs that their moods were stabilizing were taken off any other medications. If they remained stable on lamotrigine, they were randomized into a new phase of the trial:"

Wait. What? Of the 324 people, at least some were on other medications. Some stabilised. If they did, they were taken off everything other than lamotrigine. There is no proof yet that the lamotrigine made them stable - any of the other medications they were taking could have done that too. NOW that a group of people are stable, the trial starts.


"Some patients received ongoing treatment with lamotrigine; others received a placebo. Of the 177 patients who entered this second phase of the study,"

So there weren't actually 324 people in the study, there were only 177 people in the study. That's quite a difference isn't it.

"41 percent of the patients who received lamotrigine remained stable without relapse into depression or mania for six months; only 26 percent of the placebo group remained stable for that same period."

Forty one percent compared to 26 percent. Big change in numbers. Sounds good, doesn't it. There are three things that aren't stated here.

1» No information is given on whether this difference in percentages could have occurred purely by chance. That's a real possibility and one that researchers constantly have to look out for.

A bit more on the chance thing. Generally, the smaller the chance of the different percentages happening (getting 41% as compared to 26%), the more likely it is that the drug actually does something. If there is a big chance of the getting the different percentages by fluke, then the drug may not be doing anything at all.

So if you are reading an article, you want some indication that there was a small chance of having 41% of the people in one group staying stable compared to 26% of the people in another group staying stable.

It turns out this chance can be calculated, and the article could have said what the chance of getting the two different percentages were. But it didn't.

2» Related to the chance thing, it would have been useful to know how many people were in the group which took the drug and the group which took the placebo.

Generally, the greater the number of people in each group, the more reliable the results that you get. If the groups are small, then the results are less reliable. So it is handy to know these numbers.

The author does not give these numbers, nor does s/he say that half of the patients took lamotrigine and half took a placebo. All we know is that 177 people started the research study. So we really have no idea what the 41% and 26% is based upon. Anything you read into the numbers are pure assumptions that you make.

Strangely though, the author did manage to include the number 324, which has nothing to do with the purpose of the study, but which does tend to make you assume that the study was pretty large. Odd, eh?

So although the 41% compared to the 26% sound great - if you were trying to determine if lamotrigine might be a good thing for you to take, these numbers are useless.

3» The third issue, and this is a biggie, is that we are so busy comparing 41% to the 26% that we forget what the 41% means. What it means is that for 59% of the people who took lamotrigine, it didn't keep them stable for 6 months.

To phrase in another way, according to these results, if five people took lamotrigine, three would destabilise over a 6 month period. Would you take a medication if a doctor told you that it had a greater than 50% chance of NOT working?

Wouldn't you expect a responsible magazine article to point out clearly that the medication, at best, still DIDN'T work for more than half the people?


Back to the article...

"The greatest degree of effectiveness was observed in those patients with bipolar II disorder (mostly depressions with only mild highs). Of these patients, 46% of patients on lamotrigine were stable without relapse for the entire study as opposed to only 18% on placebo."

Here the difference in percentages is greater, but so is the importance of providing the actual numbers of people involved, and the chance number. Remember, it is implied that this is a subset of the 171 people who were in the trial, so that the actual numbers of people involved are smaller and so the results may be less reliable.

Again, we don't have the information which is needed to provide a true feel for how effective the lamotrigine is. As in item g, this line of the article is of little value at best.

"In the study, lamotrigine was well-tolerated and the side effects associated with it during the randomized phase were comparable to placebo. Most side effects, including headache, nausea, rash, infection, were mild or moderate, and no significant weight gain was reported. There were no serious side effects and no serious rashes."

Nothing wrong with this, until you come to the end. Then you have to wonder why they explicitly mentioned serious rashes? What's so special that it deserves mention separate from headache, nausea, etc.. Of course if you read the instructions for taking lamotrigine, you'd realise that patients are warned that they might develop a severe rash (though to be fair only between .3 and 1.3 persons per thousand affected).


What the Article Tells You

So, what exactly would you know if you read the magazine article?

If you read it carefully, you would know that (1) Lamotrigine has been studied as a mood stabiliser, and (2) there is some evidence that it might work as a mood stabiliser, but the evidence is not provided to us, and (4) even at its most effective, if you took it, there is at least a 50% chance it would not work for you. Oh, and there is something about rashes you should be wary about.

That's not a heck of a lot of useful information, except for item 4, which allows you to measure the likelihood you will be disappointed if your doctor prescribed lamotrigine to you. The article could have been made more substantial by providing the additional data - the author apparently has sufficient ability to use numbers.

An argument could be made that is was "dumbed down" for the average reader. My response is that (a) a good journalist can learn how to write sufficiently well so that the pertinent information is provided and easily understood, and (b) the target audience for this information is not the average reader, but a specific set of people who either have manic depression (bipolar disorder) or know someone who is and who are willing to spend the extra time to read the article carefully.

The article as written, and placed under the breathless headline "Difficult Mood Disorder Successfully Treated; Largest Study Proves Effectiveness Of Lamotrigine For Rapid Cycling Manic-Depression" implies that rapid cycling mood swings were successfully treated and you should be taking lamotrigine.

That's just misleading at best. At worst, it can (a) raise false hopes for us in a situation where raising false hopes can cost us months of lost productivity and functionality as we try the new drug, (2) can make us believe that the drug is safe and effective so we don't pay attention to danger signs such as rashes and (3) can have us waste incredible amounts of time and effort explaining to well meaning family why the drug doesn't actually work as the article claims.


What You Should Look For

If you are reading and article on a new drug, look for the following information:

1» What were the researchers trying to find out.

Make sure the article says this in the article text. Journalists are very good at getting little details like this wrong. If you can't figure out what the study was supposed to be about, discard the article.

2» How many people were in the study.

Make sure the article says how many people actually took part in the part of the study that looked for what the researchers wanted. Do not assume that the numbers mentioned are the true numbers - remember the 324 and the 177 above.

3» Was the study Double Blind.

Double Blind means that neither the patient nor the doctor knows who took the drug and who took a placebo. The reason for this is best explained in this article. If a drug study is not explicitly stated as double blind, then assume that the results are more suspect, less valuable, or more likely to be unreliable.

4» How many people took the medication, and how many people took the placebo.

Make sure the article includes the numbers, not just percentages. Percentages alone don't mean much.

5» What was the success rate of the drug. What was the success rate of the placebo.

If you get the success rate of the drug alone, you don't have enough information to decide if the drug works. You need the success rate of the placebo also.

6» What are the chances of the difference in success rates of the drug and the placebo.

If this chance is not stated, then you have no idea if the apparent success of the drug is just a fluke. This should also be explained in plain English.

7» How many people were not helped by the medication.

Well, I'm not sure if any article will ever print this, but I can hope. But you check the success rate of the drug. The rule is: 1 minus the success rate of the drug will give you the percentage of people who will not be helped by the drug.

8» What are the dangerous side effects.

Not all drugs have dangerous side effects, but the article should mention those that do.

It sounds like a lot, but all of this could have been written in the same space as the magazine article above.


Other Things to Do

Apart from reading the articles carefully, also follow these guidelines

1» If someone tells you about the new wonder drug, or if an article promises a solution, be skeptical. A good rule of thumb is if a drug was so great, it would have made the front page of all the newspapers, not just be a science or lifestyle article. So chances are that the drug isn't as great as it sounds.

2» Always ignore an article headline. There are headlines about Elvis being an alien currently living on the moon, for god's sake.

3» Use the list above to see if the article has the all the necessary information you should look for. Remember percentages are not a substitute for raw numbers and can mislead.

4» Get information from reliable websites like Crazy Meds.

5» I'd like to suggest that that you find the original published research article. But I won't. Finding the research paper can be difficult, and is usually on a pay website. Research articles also are meant for other researchers and are usually incomprehensible to the average person. Really.

Automatically throw out articles like this one which fails pretty much all the rules for responsible journalism on new drugs.


1» I have no personal complaints against Science Daily or the author. I happened upon this article by chance and the use of this article versus some other article was arbitrary. I expect I could do a similar analysis on many other articles on drugs used by bipolar persons published by other companies or authors.

2» My analysis was of this article only. Nothing I say here is an implied criticism of any other article on the Science Daily website or written by this author.

3» I have used lamotrigine and I am one of the 1.3 persons per thousand who got a rash. This has not bothered me - I simply chalked up Lamitcal as one of the drugs I can't take. I'd like to think that some of the other people who can take the drug will find it useful in managing their moods.