Identifying the Effect of the Tennessee Affordable Drug Act of 2005: Do mandatory generic substitution laws matter?

Through the Medicare, Medicaid and other public insurance programs, federal and state governments make over one-third of prescription drug expenditures.  With rates of prescription drug spending increasing faster than in other areas of healthcare, policy-makers are seeking ways to decrease these costs.  Because public and private spending on prescription drugs can be reduced through the purchase of lower-priced generic drugs, many states have passed laws mandating pharmacists to substitute branded prescription drugs with their generic equivalents.  Despite the large number of states with such mandatory substitution laws, their effect has not been identified.  The recent adoption of the Tennessee Affordable Drug Act provides a natural experiment that we use to determine the effect of a mandatory generic substitution law on generic-prescription fills.

During the time period of our study, pharmacists in eleven states were required to substitute a generic-prescription drug for a branded drug, while all other states had permissive substitution laws.  In addition, six states changed their drug product substitution laws:  Alaska, Hawaii, Maine, Nevada, Vermont, and Tennessee.  These states all switched to mandatory substitution laws.  We analyze the change of law in Tennessee because we have a large sample of prescription-level data from Tennessee and a comparable control state, Georgia.

This mandate could decrease prescription drug expenditures through increased generic fill rates, or it could have the opposite effect if physicians and patients use dispense-as-written indicators to avoid compliance.  We use the Thomson Reuters MarketScan Research Databases to analyze the prescription drug-fills during the six-month periods before and after the legislation, which enables us to determine the short-term effect of this law.  These databases have approximately four million individual prescription claims in both our treated and control states.

To identify the effect of Tennessee’s mandatory substitution law on generic-prescription-drug fills we use a differences-in-differences estimation procedure to compare Tennessee to Georgia. In our model, we include a complete set of demographic, employment-specific, prescription-specific, and insurance-specific regressors in addition to time trends, state dummy variables, and interaction terms for state and time variables.  We attempt to capture any pre- or post-treatment trend or other state-specific unobservable characteristics that could bias our estimates. 

In general, we find a very limited effect of the mandatory generic drug substitution law on generic-prescription fills in Tennessee.  The law results in a 0.54% to 1.5% increase in the probability of a generic-prescription fill; this effect is largest for non-chronic prescription drug users with Point-of-Service insurance plans.  It is important to note that the control state allows pharmacists to dispense generic drugs under permissive drug product substitution laws.  The lack of consistent and large effects of this law could imply that much of the substitution happened under the permissive law.  Therefore, mandating the substitution produces only limited increases in generic-prescription fills.  There is no evidence that patients or physicians are seeking to avoid this mandate by increasing their use of dispense-as-written directives.  On the contrary, we find patient-initiated requests for the branded drug decrease with the passage of this law.