Plan B: Is the FDA Using Outdated, Inaccurate Information? [Serge]

An e-mailer recently asked about the fact that the FDA does list prevention of implantation as a potential mechanism of action of Plan B. You can see that in the FDA Q&A here as well as the package insert here. Does the FDA know something that we don't?

The FDA relies on information provided by the manufacturer for the mechanism of action of any medication (they do not do their own research). I responded to the information provided by the manufacturer in this post here. In short, the manufacturer has a responsibility to list any possible mechanisms of action for any medications. Since there is an open question in the literature regarding the mechanism of Plan B, and since post-fertilization events are listed as one possible mechanism, it is not surprising (nor particularly helpful) that interruption of implantation is included as a possible mechanism of action.

In other words, the fact that the manufacturer mentions this as a possible mechanism is evidence that the question is an open one that we have no definitive answers to. Therefore, the presence of this information by the FDA does not trump the most recent literature on the topic. In fact, the opposite is true.

I can illustrate this in another way that does not directly address the mechanism of action of Plan B. The manufacturer makes this claim in its literature, and its repeated on the packaging of Plan B that the FDA agreed to:

A double-blind, controlled clinical trial in 1,955 evaluable women compared the efficacy and safety of Plan B (one 0.75 mg tablet of levonorgestrel taken within 72 hours of intercourse, and one tablet taken 12 hours later) to the Yuzpe regimen (two tablets of 0.25 mg levonorgestrel and 0.05 mg ethinyl estradiol, taken within 72 hours of intercourse, and two tablets taken 12 hours later). Plan B was at least as effective as the Yuzpe regimen in preventing pregnancy. After a single act of intercourse, the expected pregnancy rate of 8% (with no contraception) was reduced to approximately 1% with Plan B. Thus, Plan B reduced the expected number of pregnancies by 89%.

The exact citation isn't listed, but clearly they were citing this study from 1998 (both studies had 1955 women enrolled), Lancet 1998 Aug 8;352(9126):428-33. This is the same study that I posted about here. They used a crude method of determining ovulation: they added 14 days to the last period. The primary author for this paper was H Von Hertzen.

In other words, the FDA is allowing data 8 years old that used a very inaccurate way of determining ovulation, which is essential in determining effectiveness in stopping pregnancy. In fact, the author of that study has more recently contributed to an article which acknowledges the inaccuracy of their data in estimating the expected rate of pregnancy(Contraception 2003 Apr;67(4):259-65.)

This article, in which Von Hertzen is an author, states in the abstract:

The expected pregnancy rate among typical users was 6.2% in the Population Council trial and 7.4% in the WHO trial based on conception probabilities by cycle day relative to the day of ovulation. Based on conception probabilities by cycle day relative to the first day of bleeding, the expected pregnancy rates dropped to 5.4% and 5.2%, respectively. The two trials yield conflicting evidence regarding whether effectiveness declines with treatment delay. CONCLUSIONS: Our results suggest that the absolute levels of effectiveness for the Yuzpe regimen of emergency contraception and the cost-effectiveness of this regimen have probably been overstated when based on conception probabilities by cycle day relative to day of ovulation.

Although the authors only mention the Yuzpe regimen the same pregnancy estimation was also used to determine the efficacy of Plan B. If the effectiveness of the Yuzpe regimen had been overstated, then the same can be inferred with Plan B. It should also be noted that even with the "more accurate" method of determining ovulation, there is a significant chance for error in determining the actual date of ovulation.

I hope I haven't lost anyone. Here's the bottom line: the FDA is allowing the manufacturer of Plan B to claim that it can be almost 90% effective based on a study from 1998. This study has been shown to be inaccurate due to its crude way of estimating ovulation. This fact was pointed out by none other than the author of the original study. Yet there it is on the packaging being readied for OTC sale.

Still confident in the FDA information now? There is significant reason why we should not be. In any event, it should be clear that the information given by the FTC does not refute more recent scientific studies.