Uncategorized

A lesson in odds.

By Chris Rangel MD ⋅ February 19, 2006 ⋅  Email this post ⋅  Print this post ⋅ Post a comment

The birth control patch; The story behind the numbers.

As you know, 42.7% of statistics are made up on the spot! Seriously though, it is at times almost bizarre how the lay media slants coverage of the side effects of medications. For example, the once a week estrogen birth control patch is made by Ortho-McNeil and has recently been linked to increased rates of potentially fatal blood clots.

In 2005 the Associated Press reported that 12 young women had died in 2004 from thromboembolic complications (blood clots) thought to be related to the use of the patch. This was not a scientific study (the data was obtained by the AP from “Federal death and injury reports” using a freedom of information request) and no information was provided as to other associated factors such as smoking or autoimmune diseases that may have contributed to these deaths.

Last November the FDA and Ortho warned patients about the possibility of increased blood clot risk and the fact that users of the patch are exposed to up to 60% more estrogen than patients who take oral birth control (this is because estrogen from the patch is released directly into the blood steam and unlike birth control pills is not subjected to “first pass” metabolism by the liver).

This week Ortho-McNeil reported the results of two ongoing studies to the FDA which show a very mixed picture. While one of the studies (which I have not seen) found no increased risk, the other study suggested a risk of thromboembolic phenomenon of two times above normal for uses of the patch.

Every medication has risks and the problem here is one of perceived risk. Thromboembolic disease is very rare especially in the female population most likely to be on hormonal birth control. The risk of nonfatal blood clots in women on hormonal birth control is about 3-5 cases per 10,000 per year. If we assume that the patch increases the risk twofold then it is 6-10 cases per 10,000 per year.

But these are nonfatal cases! Let’s further assume that the Federal data obtained by the AP is true and that 12 women died in 2004 because of blood clots caused by the Ortho patch. According to CDC data from 1999 on maternal mortality there were 12 deaths per 100,000 live births. According to Ortho since going on the market in 2002 4 million women have used the patch. If we further assume that only 2 million were on the patch in any given year from 2002-2006 we can calculate that there were only about 1-2 deaths per 100,000 users from blood clots in any given year. What this means is that the risk of death from pregnancy is about 10-12 times more than the risk of death of someone using the Ortho patch!!!!

Since the purpose of the Ortho patch is ostensively to prevent pregnancy the benefits from a statistical context obviously still outweigh the risks and yet this fact is totally lost on the lay media (and the lawyers representing women who died or suffered blood clots). A big part of this misperception comes from the common misperception that prescription medications are supposed to be 100% safe. This is the problem with statistics. When analyzed in isolation they provide a very skewed picture of risk.

The criticism of this perspective is that this is simply a spin on the numbers. What about the fact that oral hormonal birth control is safer than the patch? In this case we have to consider the that there is evidence that use of the patch is more effective in preventing pregnancy then the pill (95%-99% for the pill compared to 98%-99% for the patch) and so the risks of the use of the patch should still far outweigh the risks of the pill by virtue of better compliance with dosing for the patch.

Yet even if the effectiveness was the same for the pill and the patch we can’t discount the value of the convince that the patch provides. This value in convince that we place in products is something that Americans do every day without thinking about it. Despite the fact that it is statistically three times more deadly to drive to work rather than walk we continue to be an automobile based nation. And yet we don’t commonly see news articles about the deadliness of the car.

No single medication comes without risks but you have to weigh the benefits with the risks. This simple concept is something that continues to elude most “journalists” and almost all trial lawyers and continues to cause the FDA fits. Statistics cannot be considered in isolation. To do so is intellectual laziness or manipulation of the story of which there is a-plenty in today’s society.