The proposed cohort study will use a large health care utilization data set and novel pharmacoepidemiologic methods to determine whether the risk of injury from falls, motor vehicle crashes, and suicidal behavior applies equally to all antidepressant (and antipsychotic) classes -- or whether there are particular regimens of antidepressants (and antipsychotics) with safety advantages that should be prescribed preferentially. Analyses will examine how the risk of injury depends on socio-demographic factors, co-morbidity, co-medication, dose and duration of use. The proposed study addresses many of the limitations of prior observational work, including failure to account for dose-related effects, duration of use, confounding by indication, physician channeling bias, and distortions resulting from prevalent user designs. Data for the current application comprise information about prescription drug use and both medical and psychiatric conditions (including unintentional injuries and suicide attempts) for an annual population of over 15 million beneficiaries aged 50 years and older who received prescription drug benefits from managed care plans throughout the United States, 1997-2009. Descriptive analyses comparing the risk of unintentional injury among patients taking antidepressants or antipsychotics to those not (aim1) will place in context the two key aims of this application: to compare the risk of injury due to falls, motor vehicle crashes, and suicidal behavior (the leading causes of injury deaths in older Americans) among initiators of 1) SSRI vs. TCA vs. SNRI antidepressants (aim 2), and 2) conventional vs. atypical antipsychotics (aim 3). The proposed research responds to the NIMH's Strategic Objective 4: """"""""Strengthen the Public Health Impact of NIMH-Supported Research"""""""". The proposed series of comparative studies on treatment emergent injury-related outcomes will advance understanding of the public health influence of the most common strategies used today to treat mental illness among older Americans. Because the therapeutic superiority of particular antidepressant (and antipsychotic) preparations have not been established for the conditions these agents are commonly prescribed to treat, 1-5 the potential clinical and possibly economic benefits of the proposed work could be large, since use of these agents is common and the consequences of injuries due to falls, motor vehicle crashes, and suicidal behavior are particularly untoward among older adults.
The current application will use novel pharmacologic methods and a large health care data set of Americans aged 50 and older to determine whether the risk of injury from a) falls, b) motor vehicle crashes, and c) suicidal behavior applies equally to all antidepressant (and antipsychotic) classes -- or whether there are particular regimens of antidepressants (or antipsychotics) with safety advantages that should be prescribed preferentially. Analyses will examine whether, and if so, how the risk of injury depends on socio-demographic factors, co-morbidity, co-medication, dose and time since starting an antidepressant or antipsychotic.
