Comments on: Nefarious Nefazodone And Flashy Rare Side Effects

By: Douglas Knight

Douglas Knight — Sat, 23 May 2015 02:18:52 +0000

ask “would you rather have a 50% chance of these sexual dysfunction symptoms for one year, or a 1 in 300,000 chance of death per year?”

That is exactly what Scott did in this post. More specifically, he consulted a QALY database that was built by asking patients lots of questions like this and trying to smooth out their answers into coherence.

Report comment

By: AlexanderRM

AlexanderRM — Fri, 22 May 2015 18:20:21 +0000

I feel like comparing medical effects is an excellent case to introduce Rawls-Style calculations: Instead of asking “Which is worse – ruining ten million people’s sex lives for one year, or making one hundred people’s livers explode?”, ask “would you rather have a 50% chance of these sexual dysfunction symptoms for one year, or a 1 in 300,000 chance of death per year?”.

Better yet, rather than having the doctor or the regulatory agency ask this question, you can ask the actual patients this question. Admittedly, they aren’t likely to be hyper-rational about tiny chances of spectacular things either, but I’m not sure it’ll be any worse than the voting public or a regulatory agency; in fact, they’ll probably be significantly more rational overall.
That also has an obvious advantage of allowing customization for the individual patient- a patient who has little to no sex anyway might weight sexual dysfunction much lower, for instance, a patient in their 80s might weight a chance of death a bit lower (again not likely to be fully rational, but moreso than a voting public talking about someone else’s life), etc.

Furthermore, and most importantly, you could potentially use this method and then ask patients to sign off saying that they chose this chance over the other, which seems far more defensible in a lawsuit (“my client decided that a 1 in 300,000 chance of death was small enough to be worth it” rather than “I decided a 1 in 300,000 chance was small enough to be worth it”) and might help with regulation as well.

Report comment

By: GE 20 -“No estoy absolutamente seguro de nada, Bitch!” | Guía Escéptica

Thu, 14 May 2015 03:29:13 +0000

[…] Antidepresivos y efectos secundarios. […]

Report comment

By: Anonymous

Anonymous — Wed, 06 May 2015 17:45:45 +0000

No, I did not cite any particular problem. I really don’t think it is necessary to explain the cause of the numbers to assert that they are false (as opposed to “weird”). The problem you mention did not even cross my mind.

And even if you don’t like the common use of rape, that doesn’t make NCVS false.

Report comment

By: grendelkhan

grendelkhan — Wed, 06 May 2015 17:21:57 +0000

Anonymous: yes, those rape numbers are far too low (they’re the ‘NCVS’ tier, not the ‘Koss et al.’ tier), but I think the numbers they’re being compared to are… comparable, e.g., from the same tier? In any case, I don’t think the problem you cite exists for robbery, and that seems to come with a one-in-four-ish chance of risking being shot by your intended victim.

Report comment

By: grendelkhan

grendelkhan — Wed, 06 May 2015 17:02:52 +0000

I’d like to nominate this Wikipedia article (“List of SJS-inducing substances”) for possibly the worst, most useless thing I’ve ever damned well seen. It’s a table of chemicals with a “certainty of implication” column, each of which says “Certain”. Not a word about relative risk, not a word about the base rate, just… aargh. I’m not even a doctor and I know this is a goddamned awful table.

Report comment

By: Anonymous

Anonymous — Wed, 06 May 2015 16:48:12 +0000

Oh, come on. You know enough to say that the rape numbers are false. Here’s the chart.

Report comment

By: grendelkhan

grendelkhan — Wed, 06 May 2015 16:33:11 +0000

Some back-of-the-envelope checking here.

The National Crime Victimization Survey reported 6.1M violent victimizations in 2013, of which about a third (~2M) were due to strangers. (The category here faintly maps to the FBI’s UCR category for violent crime: “rape or sexual assault, robbery, aggravated assault, and simple assault”.)

The estimates for the prevalence of defensive gun use are high enough to, as this research brief says, make crime a “very risky business indeed”; see the graph on page 9. (More defensive gun uses against rape were reported than rapes, which means that prospective rapists are running worse than fifty-fifty odds of being possibly shot.) I know very little about this subject, but something is weird here.

Report comment

By: Douglas Knight

Douglas Knight — Tue, 05 May 2015 22:22:06 +0000

had to quit taking it those evenings

Did you take it in the evenings? Or do you mean take it in the mornings of those days?

Report comment

By: Winfried

Winfried — Tue, 05 May 2015 20:51:39 +0000

I’m a bit late to the party, but I am narcoleptic and have taken Provigil (modafinil) for years. Now I take Nuvigil (armodafinil). It’s much better for the already mild side effects I was having.

On very rare occasions, I would have days where I felt strongly compelled towards gambling, smoking, and drinking, activities that I seldom take part in. I played a friendly game of Poker with friends every week and had to quit taking it those evenings because I could not keep from incorrectly estimating odds and going broke (or winning quickly on a streak).

In addition, if I took it every day for more than a week or so, I would get headaches that felt like a caltrop behind my right eye. These headaches were untouchable by common OTC headache medications.

Nuvigil has yet to give me any of these headaches and I haven’t really gambled much on it, but from occasional forays I seem to do about as well as can be expected for basic strategy in games of chance.

I highly recommend Nuvigil over Provigil.

Report comment