Projected Spending on Psychotropic Medications 2013–2020

Abstract

Spending on psychotropic medications has grown rapidly in recent decades. Using national data on drug expenditures, patent expirations, future drug development and expert interviews, we project that spending will grow more slowly over the period 2012–2020. The average annual increase is projected to be just 3.0 % per year, continuing the steady deceleration in recent years. The main drivers of this expected deceleration include slower development of new drugs, upcoming patent expirations which will lower prices, and payers’ growing ability to manage utilization and promote generic use. The slowdown will relieve some cost pressures on payers, particularly Medicare and Medicaid.

Appendix

Detailed Methodology

This appendix provides detail on various assumptions built into the projection approach.

Generic Entry Assumptions

After reviewing various possible spending drivers, we determined that generic entry would be a major determinant of future spending.

To identify the exact impact of generic entry, we distinguished between two groups of psychiatric medication classes. First, we identified five therapeutic subclasses where the generic share of prescriptions was below 70 % in 2012, and classified these as “low generic share” (Appendix Exhibit 4). For these subclasses, patent expirations in the next few years could significantly alter the rate of spending growth, as lower-cost generics enter and capture market share. The five subclasses are: antipsychotics (other than combinations and phenothiazine derivatives); serotonin-norepinephrine reuptake inhibitors (SNRI antidepressants); stimulant ADHD treatments; non-stimulant ADHD treatments; and addiction treatment medications. All other therapeutic subclasses had a generic share that was already above 90 % in 2012 (and were classified as “high-generic share”). For these subclasses (e.g., anticonvulsants, tricyclic antidepressants), any future patent expirations are unlikely to have a significant effect on spending, because the products losing patent protection have only small shares of the retail market. (No subclass had a generic share between 70 and 90 %).

Exhibit 4 Classes and subclasses of psychotropic medications

Generic Price at Introduction

During the first year after introduction, the generic price is assumed to average 60 % of the price for the brand version at the time of generic entry (i.e., the initial brand price). During the second year, the generic price is assumed to average 30 % of the initial brand price. These rates were set at levels somewhat lower than those observed in the literature (e.g. (Berndt and Aitken 2011; Saha et al. 2006), based on consultation with experts and on evidence in the IMS data used for this study. The different pricing levels reflect several factors. First, there is the influence of time-limited marketing “exclusivity” often granted by the FDA to the initial generic entrant into the market. Second, if the generic version enters midyear, it is only available for half of the year. Third, it typically takes time to switch patients from branded to generic products, as they may have 1–3 month prescriptions to consume before a refill is needed.

Price Growth

The basic projections were calculated in real terms and then adjusted to nominal terms by applying the consumer price index (CPI) for pharmaceuticals. For each of the years in our forecast period, the CPI increase is projected to be between 3.0 and 3.6 %. All results are presented in nominal dollars.

Growth in Generic Share

It is assumed that in the first year after generic entry, generic versions will capture 70 % of prescriptions. For the second year after generic entry, this share is expected to increase to 95 %. These figures are higher than observed in published studies such as Saha et al. (2006), reflecting a view that generic entry is now occurring more rapidly than it was at the time covered by that study. This view is based on expert advice from Dr. Ernst Berndt, recent experience in the IMS data, and a recent National Bureau of Economic Research study (Aitken et al. 2013). Insurance plans now can change formulary content quickly, which promotes the rapid acceptance of generic versions through favorable pricing incentives.²

New Brands

We did not explicitly model the introduction of new branded products for the period 2013–2020, based on the expert opinions and literature reviews described in the main article earlier. Trends from earlier periods are captured in our volume projections, which include the introduction of some limited-impact new products, but which also reflect the slowdown in introductions (since the volume projections are mostly based on 2009–2012).

Competition Across Molecules

For simplicity, our approach assumes that patent expiration affects prescriptions only for the molecule losing patent protection, and not for other (potentially competing) molecules. There has been evidence of cross-molecule substitutions after some recent patent expirations [e.g. Lipitor (atorvastatin) in 2011] although not after others [e.g. Prozac (fluoxetine) in 2001 (Druss et al. 2004)]. There is not yet strong evidence on the size of these effects.

Impact of Rebates

The base-case projections of prescription drug spending are pre-rebate. The effect of rebates (and the ACA) on these projections is taken into account in a later step. Growth rates from this modeling effort were used to project prescription drug spending from 2012 through 2020. The 2012 estimates, part of the SAMHSA spending estimates, are estimated using information from IMS and payer information from the Medical Expenditure Panel Survey. Once the payer amounts are estimated, a rebate proportion is applied to Medicaid, Medicare, and private insurance spending estimates. These negotiated rebates lower the costs for the insurer. Once 2012 levels are established, the growth rates developed in this study are applied to the spending levels in 2012 that already account for rebates.

Impact of the Affordable Care Act (ACA)

The projections for spending on MH/SUD medications presented in this paper were also used to develop baseline projections of drug spending in the SAMHSA spending estimates. Baseline projections began in 2010 and extended through 2020, supported by actual data from IMS through 2012. After baseline projections were created, the impact of the ACA was estimated. Projections of the impact of Medicaid expansion included only those states that had agreed to implement the expansion and those states that had not yet made a decision. The ACA impact was based on state-level estimates of the population eligible for enrollment in Medicaid and the Marketplace; estimated insurance take-up rates; the percent of users among those newly insured; and a factor to take into account the increased spending and use rate among people who become insured for the first time (moral hazard of insurance). See (SAMHSA 2014) for a complete explanation of the methods used to project spending and the impact of the ACA.