Ortho Evra and [Clueless] Media Reports

Back in September, ABC aired a report about Ortho Evra, titled
Do Users of the Birth Control Patch Know Enough About Its Potential Dangers? Basically, ABC implied that using the patch increases your risk of death. The report mentioned 17 deaths, allegedly caused by blood clots in the lung (PE), in women using the patch. In my first post, I speculated that the report was inaccurate, and that ABC is either clueless, or is priming the jury pool in anticipation of a lawsuit against the patch's manufacturer. I have more information now, and it appears I was right [at a minimum, about the report being inaccurate]:

Ortho-McNeil Pharmaceutical, Inc., manufacturer of the patch, has been informed of only six deaths in patch users since its introduction in April 2002. The role of the patch in these deaths is undetermined, and several of the deaths were reported multiple times, accounting in part for the erroneous number of 17.

Little is known about the circumstances of 3 deaths. This is the information on the other 3 deaths:

One of the six deaths was due to myocardial infarction [heart attack] in a woman with both Down syndrome and Eisenmenger syndrome; her health had been deteriorating before starting contraception. Another death in Germany was due to suicide. One death in New York City was attributed at autopsy to a pulmonary embolism [lung blood clot]; the woman was not wearing a patch at the time of death. Whether she had been using the patch previously has not been determined.

Since apparently ABC couldn't be bothered with, you know, factual reporting, let alone useful context, let's review how one assesses risk and calculates incidence rates. First, you have to determine if an alleged death actually occurred. Then, you have to evaluate the potential association between a given exposure (wearing the patch) and outcome (death due to PE). This evaluation includes:

confirming the exposure--Did the woman who die use the patch at all?

identifying the temporal relationship between exposure and outcome--When was the woman who died using the patch, and for how long? Was she wearing it at the time of death, and/or for a period of time immediately preceding that?

determining the cause of death--What exactly was the cause of death?

identifying alternative causative factors--Was the patch wearer involved in a car accident, did she commit suicide, etc.?

Once you have all this information, you can calculate the risk of the outcome (the risk of PE death in patch users). So, let's assume that all 6 reported deaths were caused by PE, that they occurred in current patch users, and that there were no other alternative causes of death involved.

We know that the patch has been used by about 4 million women x 2 years (this gives us 2.2 million woman-years of patch use).

The risk of an outcome is determined by dividing the number of documented cases of that outcome (in our example, 6 PE deaths in patch users) by the number of individuals exposed per unit of time (in our example, 2.2 million woman-years); our result is 2.7 per million woman-years. What does this mean? Here's a practical way to understand this: the death rate per year from a blood clot is about 1 per million for women aged 15 to 24 years who don't use hormonal birth control; 3 per million users for women who use the combination birth control pill; and ~13.2 to 20 per 100,000 live births for pregnant women.

So, patch users have a slightly lower risk compared to Pill users, and a considerably lower risk compared to pregnant women. Women who don't use the patch or the Pill, and who aren't pregnant have the lowest risk.

Bottom line:

Case reports of deaths need to be fully investigated by formal epidemiological studies. Until this is done, a casual relationship remains questionable. At the present time, no evidence suggests that the transdermal patch is associated with an increased risk of death compared with combination oral contraceptives.

All I have to say about the ABC report is: what a sorry state of affairs!


Grimes DA, Mishell DR Jr. Assessing Rare Event Reports: A Numerator in Search of a Denominator. Dialogues in Contraception. Fall 2004;8(7):7.