A new county-by-county study found that the more a pharmaceutical company markets opioids to physicians, the higher the overdose death rate is.

40 million dollars. That’s how much money was spent by pharmaceutical companies marketing opioids to physicians over one and a half years. In lunches, speaking fees, consulting agreements. $40 million is a drop in the bucket, really, for Pharma. But this week, a paper appearing in JAMA Network Open links pharma payments to physicians to opioid deaths across the United States.

There’s no free lunch, folks.

The Data

The analysis used three main data sources.

First, the authors used Sunshine Act data to figure out how much pharmaceutical companies spent on marketing to physicians on a county-by-county basis.

If you haven’t played with this website, you definitely should.  Find your doctor. Find a pharmaceutical company. You can see here that Purdue Pharma, makers of Oxycontin, spent around 6.8 million dollars in 2017 marketing directly to physicians.

The authors then linked this data to a CDC database of opioid prescribing rates and, finally, to CDC data on opioid deaths.

With all the pieces in place for the analysis, the topline results were perhaps not surprising. The more a pharmaceutical company spent on physicians in a county, the more physicians prescribed opioids in that county, and the higher the opioid overdose rate was.

Putting some numbers to it, for every 4 extra physician payments made per 100,000 people in a county, there were 13,000 extra opioid prescriptions and an 18% increase in opioid overdose deaths.

The Caveats

But there are a couple of nuances here worth noting.

First is the question of causality. Do companies spending more money lead to more prescriptions, leading to more overdoses. Or perhaps do companies just spend money in places where physicians are already writing more prescriptions?

To address this, the authors linked prior-year spending to current year prescription rates, but given that prior-year and current-year spending rates are likely highly correlated, I don’t think this totally gets around the reverse causation issue.

The authors conducted a clever mediation analysis to determine how much of the relationship between pharma dollars and opioid deaths is mediated by physician prescriptions.

About 25%, according to their data, suggesting that curbing physician prescribing could be a reasonable strategy to reduce overdose deaths. It also means that 75% of the relationship is due to something else, and perhaps the solution is to limit pharmaceutical spending on marketing period.

Now, I should mention that there is no negative control in this study. It would have been nice to see if marketing dollars spent on, say, cholesterol medications, were associated with opioid overdoses. If they were, it would give a hint that there is some unmeasured factor here – maybe healthcare utilization – that the analysis isn’t accounting for.  So this is not an open and shut case.

The Conclusion

And yet, I can’t help but wonder what possible public good comes from this type of marketing. Perhaps that $40 million that was spent marketing opioids to physicians that already know about opioids would be better spent educating the public that these drugs are not often necessary.