Off-Script

As mentioned in “Good and Bad Arguments Against the Obamacare-Opiate Effect,” one reason to question whether Obamacare caused the divergence in overdose rates between Medicaid expansion states and non-expansion holdouts is that prescription rates do not seem to have risen in tandem with overdoses. Instead, prescriptions for medical opioids appear to have peaked around 2012, even as overdose deaths have continued to rise rapidly.

Peaked

You can confirm using CDC’s data that while the rank order of which states and counties prescribed more opiates did not change much, in both states and counties average quantities prescribed declined from 2010 to 2015:

 

In fact, the Medicaid expansion states already prescribed fewer opiates in 2010, and reduced their prescription rates more rapidly from 2010 to 2015:

staterates20102015

You can also confirm that while prescriptions for opiates correlate with overdose deaths.

However, this correlation is actually less close at a state level in 2015 than it was in 2010:

And also less close at a county level, particularly for Medicaid Expansion states:

 

In fact, states and counties that reduced their prescription rates more in Medicaid expansion states had higher overdose rates in 2015:

prescribechangeprescribechangestate

An other way of looking at it is that at both the state and county level, particularly for Medicaid expansion states, 2015 overdose rates are positively correlated with 2010 prescription rates and negatively correlated with 2015 prescription rates, all else being equal.

(1) (2)
VARIABLES State-Level Overdose Rate County-Level Overdose Rate
State-Level Prescription Rate 2015 -0.312**
(0.126)
State-Level Prescription Rate 2010 0.385***
(0.102)
Medicaid 5.372*** 6.732***
(1.584) (0.941)
County-Level Prescription Rate 2015 0.0966***
(0.0311)
County-Level Prescription Rate 2010 0.0599
(0.0371)
Constant 5.592 3.389**
(3.358) (1.527)
Observations 51 556
R-squared 0.474 0.239

Standard errors in parentheses

*** p<0.01, ** p<0.05, * p<0.1

So far, this looks roughly in line with the idea put forth by this RAND paper that while in earlier stages of the crisis, legal and illegal opioids were acting as complements, at this point they are more likely to act as substitutes: attempts to cut down on legal opioids can result in more use of incredibly deadly fentanyl, heroin, and other substitutes. (The New York Times reported earlier this week that more people died in 2016 from fentanyl overdoses than heroin.)

Unfortunately, these patterns of prescription in of themselves do not explain the Obamacare effect- note that controlling for prescription rates in the regression above increases the coefficient on Medicaid expansion from our 4-5 per 100,000 baseline estimate. It still looks to me like insurance is the important channel for the difference between Medicaid expansion and non-expansion states (controlling for 2015 uninsurance rates does reduce the Medicaid expansion coefficient by about half), but how insurance does this is unclear.

Is it that a subset of the population with very high propensity to abuse got access to prescriptions, even as the overall number of prescriptions declined? Are there completely unscrupulous “pill mill” docs like those described by Sam Quinones in Dreamland who have managed to work around prescription reporting guidelines while facilitating addicts using Medicaid to pay for the drugs? Is there some other, unrelated channel; Medicaid recipients would simply have more disposable income, thanks to lower medical bills, and a subset might also be less likely to remain employed full time thanks to getting insurance without work. Neither are especially satisfying explanations to me, although the huge rise in overdoses among Native Americans earlier in the 2000s  during the time of increasing disbursements of casino income and declining labor force participation rates among recipients makes me think it’s not impossible:

native-american-death-rates-by-age-group-and-year-2 (1)

screen shot 2013-03-08 at 7.22.04 pm (1)

20150117_USC725 (1)

It still seems possible that we can find methods within the medical system to contain the crisis. Moreover, as Senator Ron Johnson pointed out in his letter to HHS about the issue, a substantial portion of the Medicaid-financed purchase of opioids appears to be out-and-out fraud that wouldn’t be picked up by prescription statistics correctly. But for ordinary doctors and hospitals trying to make things better and reduce abuse, it looks like there’s no easy answer, including cutting down on prescriptions.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s