3 comments

[ 2.5 ms ] story [ 59.6 ms ] thread
thought I'd repost given recent interest on HN in conflict of interest - rather than looking at impact of payments on specific medications, we wanted to see how much payments distort prescribing costs overall. The variation in impact across specialties was particularly notable (to me).
I'm glad the article called out that they can't prove a casual connection between payments and prescribing because the cause and effect could be the reverse.

When drug companies are looking for physicians to speak at conferences or speak to other physicians, they are looking for "true believers" in the product. Why else would you work with them.

As such, it's not surprising these physicians prescribe more branded products and that there is a higher cost per patient. I wonder whether or not that would be true without the payments.

Absolutely. Likely it's bidirectional - pharma identifies high prescribers (or recognized experts) and invite them to speak at conferences or dinner meetings. This increases their affinity for the product, and presumably their prescribing. Lather, rinse, repeat. One /could/ sort out causation a bit more by looking at lagged data (does prior-year payment influence subsequent year prescribing, eg) - but it was such a colossal pain to match in the first place that we never went back to look.

Of note: pharmacies sell docs' prescribing history to aggregators who sell it to pharma, in fully identified form. So, pharma sales folk know exactly how much of everything docs write.