The Nurnburg et. al. paper Spiesel discusses in the article doesn't explain the 'partner' factor in the sexual function questionnaire they used, but they do cite to the paper by the researchers who developed the questionaire. (Citation 27) I won't link to that paper because it's not available for free online and required me to log into the journal's site with my university access.

According to this second paper, by Quirk et. al., the Partner category addresses such issues as: "Worry partner will seek other relationship" and "Worry about partner’s negative feelings." It appears to be an attempt to measure female sexual function by indirectly assessing the impact of non-function on the woman's feelings about her relationship. Given the complex psychological factors that can contribute to female sexual disfunction, it seems like a useful factor.

It certainly doesn't involve women giving their blue pills to their husbands, as Spiesel spurriously suggests. Such behavior, if reported to the researchers running the study, would certainly have resulted in the women being removed from the study.

"Such behavior, if reported to the researchers running the study, would certainly have resulted in the women being removed from the study." What if the researchers were using an intent-to-treat design? In these studies, participants who are assigned to the treatment group are kept in the study even if they don't take their treatment properly. It is thought that this may reflect clinical reality better, since it often does happen in real life that patients do wierd things like throw up their meds, give them to the dog, or forget to fill their prescriptions. Thanks for the clarification of "partner." From your description, it does sound like a meaningful subscale.

I don't know what Spiesel was thinking with his foolish comment about the "partner" subscale. It should have been easy enough for him to look up the same reference you did. Also, Spiesel might have mentioned some of the caveats that the study authors raised, including the possibility that the women in the study might have been so highly motivated to obtain relief for their sexual disfunction that the favorable results might not be attributable solely to the medication.

Very disappointing article.

BTW, I believe the study did use an "intent to treat" design.

Yeah, I wondered why Spiesel didn't dig a little deeper; I assume he has the same widespread journal access I do.

And yes, the design is intent to treat. However, I suspect that while intent to treat would not remove a patient for, say, failing to attempt intercourse at least once in a week, they might still remove a patient who reported giving the medication to her husband, as a safety issue. Noncompliance is one thing, but giving a perscription medication to a person it has not be perscribed to could very well become a malpractice issue if the doctors became aware of it and continued to give the woman the pills.

The highly motivated factor might indicate that this treatment would only work for a subset of women who are being treated for depression and suffer from sexual disfunction, but the fact that women on the drug showed significant improvement over placebo suggests that for that subset the drug might be effective.

I'm not suprised by the failure to dig deeper. This is Slate; lazy journalism seems to be a specialty around here.