Impact of Food and Drug Administration pediatric antidepressant use warnings on treatment of depression, physician specialty practice patterns, and pharmaceutical stock prices

Background. The objectives of the dissertation were to investigate the impact of the FDA pediatric antidepressant warnings in February 2004 on: (1) the treatment of new-onset depression with antidepressants and psychotherapy; (2) the likelihood of psychiatrist prescribing an antidepressant relative to a primary care physician during a youth mental health visit; and (3) the magnitude and volatility of stock returns of antidepressant manufacturers.;Methods. Objective 1: Youth 2-17 years old with a new-onset depression diagnosis (NODD) from July 2003 through June 2006 constituted the study population. The data source was the i3 Innovus database containing administrative claims information of U.S. commercial insurance enrollees. A generalized estimating equations model was developed to estimate the warning effect on the treatment of new-onset depression. Objective 2: National Ambulatory Medical Care Survey data (2001-2006) was used to examine antidepressant use by physician specialty during youth (2-17 years old) mental health visits. A logistic model was fit with the antidepressant visits as the dependent variable and the year of visit (2004-2006 as the post-warning period) and physician specialty as the primary independent variables. Objective 3: CRSP stock price data (2003-2006) was used to examine the 2004 FDA warning impact on stock returns and volatility. A seemingly unrelated regression model was used with daily stock return as the dependent variable and warning period as the primary independent variable.;Results. Objective 1: Compared to youth with a new-onset diagnosis of depression (NODD) in the pre-FDA warning period, NODD youth during the post-warning period had: (1) a significantly lower likelihood of antidepressant use: children (2-12 years) [Odds Ratio (OR)=0.83 (0.75-0.92); adolescents (13-17 years) [OR=0.85 (0.82-0.89)]. When youth with the diagnosis of depression were separated into those with MDD and those with other depression diagnoses, only the latter had a significant post-warning antidepressant decline. (2) a significant increase in the odds of a psychotherapy visit [children, OR=1.31 (1.23-1.40); adolescents OR=1.19 (1.15-1.24)]. Objective 2: Combining all physician specialties, antidepressant prescribing visits decreased in the 2004-2005 period relative to the pre-warning period before picking up again in 2006. Antidepressant prescribing by each specialty did not change after the warning relative to the pre-warning period. In each period, the likelihood of an antidepressant prescription visit was significantly higher for psychiatrist visits compared to primary care visits (OR=3.78; CI: 2.72, 5.25). Objective 3: The warnings caused an insignificant negative impact on the stock returns. The average daily impact of the FDA warning in March 2004 and the black box warning in October 2004 on stock returns of the 5 firms was estimated to be -0.002% (95% CI -0.006, 0.002) and -0.001% (95% CI -0.007, 0.004) respectively.;Conclusions. Treatment patterns following the FDA warnings on pediatric antidepressant suicidality are more complex than previously reported. The FDA warning did not lead to a decrease in antidepressant use for youth with MDD and psychotherapy increased. National ambulatory trend data suggest that the FDA suicidality warning of March 2004 did not change the likelihood of an antidepressant prescription during youth visits to a psychiatrist relative to primary care physician. It appears that the warnings led to a modest increase in volatility of stock returns of the affected firms.