اثر تبلیغات مستقیم تلویزیونی و مجلاتی برای مشتری روی استفاده از داروی ضد افسردگی

We examine whether exposure to direct-to-consumer advertising (DTCA) for antidepressant drugs affects individual use of these medications among those suffering from depression. Prior studies have almost exclusively relied on making connections between national or market-level advertising volume/expenditures and national or individual-level usage of medications. This is the first study to: estimate the impact of individual-level exposure to DTCA on individual-level use of antidepressants; estimate the impact of individual-level exposure to television DTCA on individual-level use in any drug class; consider the relative and interactive impact of DTCA in two different media in any drug class; and, consider the heterogeneity of impact among different populations in an econometric framework in the antidepressant market. There are also important limitations to note. Unlike prior market level studies that use monthly data, we are limited to aggregated annual data. Our measures of potential advertising exposure are constructed assuming that media consumption patterns are stable during the year. We are also not able to study the impact of advertising on use of antidepressants for conditions other than depression, such as anxiety disorders. We find that: DTCA impacts antidepressant use in a statistically and economically significant manner; that these effects are present in both television and magazine advertising exposure but do not appear to have interactive effects; are stronger for women than for men in the magazine medium, but are about equally strong for men and women in the TV medium; and, are somewhat stronger for groups suffering from more severe forms of depression. The overall size of the effect is a 6–10 percentage point increase in antidepressant use from being exposed to television advertising; the corresponding magazine effects are between 3 and 4 percentage points.

The United States and New Zealand are the only two countries that allow direct-to-consumer advertising (DTCA) of prescription drugs, and the practice is surrounded by substantial controversy. DTCA started in the U.S. in the early 1980s and since that time has been subject to evolving regulatory control standards by the Food and Drug Administration (FDA). Since a 1997 change in FDA regulations that clarified and relaxed restrictions on DTCA, especially in the medium of television, the practice has burgeoned into a multi-billion-dollar industry. This is particularly true of the period between 1997 and 2005, when spending on DTCA increased by 296.4 percent while advertising aimed at physicians increased by only 86 percent. According to the Nielsen Company data, overall DTCA spending increased 1.9 percent to $4.51 billion during the period 2008–2009, while television spending was up 0.6 percent to nearly $3 billion (PharmaLive, 2010). After more than two decades of DTCA, it is important to assess the effects of this practice, especially in the largest drug classes (such as antidepressants), using new methods of accounting for targeting of ad placement by marketers. Antidepressants are the second largest drug class in the U.S. (second only to statins), and depression affects close to 15 million American adults in a given year (National Institute of Mental Health, 2008a). Despite the high sales of antidepressants, 50–66 percent of those with depression are estimated to receive no treatment (Kahn and Meyer, 2006 and Kessler et al., 2003). Depression is currently the leading cause of disability in the U.S. as well as other countries (World Health Organization, 2005). It results in more absenteeism than almost any other physical disorder, costs employers in the U.S. more than $51 billion per year in lost productivity, and contributes to the high cost of medical and pharmaceutical bills (Cross, 2004).

In this paper we test the hypothesis that exposure to DTCA for a popular class of prescription medication—antidepressants—affects individual use of these medications. We also test whether the ads have a different effect depending on the mass communication medium in which they appear (magazines versus television), whether there is a differential effect for men versus women, for severe versus milder cases of the illness, and whether there is an interactive effect between magazine and television DTCA exposure. Ours is the first study to: consider the impact of individual-level exposure to DTCA on use of antidepressants; consider the impact of television DTCA in an individual-level study; consider individual-level impacts by severity of disease; and compare the effects (and interactive effects) of exposure through two different media. Most prior work has relied on market-level exposure data, which assumes that each individual is exposed to the same level of advertising. However, market level studies have been able to use monthly data, while our data is annual. Our identifying variation comes from a model that is suitable for producing causal conclusions as we are able to control for targeting by including TV and magazine category (or program) fixed effects and a large list of other variables known to marketers. While our strategy minimizes bias, we acknowledge that some forms of bias are still possible. That is, there could be two people with the same number of total TV hours who each watch Seinfeld and Law and Order, but one watches relatively more Law and Order. Since we do not control for how the total time is split between categories, there could be unobserved heterogeneity in depression severity or some other unobservable between the two types of people that is correlated with the ad exposure measure and also correlated with the dependent variable. We argue that by also including in our model detailed demographic controls used by marketers that would capture some of this remaining heterogeneity, we have minimizes such possibilities.