|
BACKGROUND/RATIONALE:
Suicide is a devastating event that occurs most often in the context of a psychiatric disorder, usually a mood disorder. Reducing suicide is a national and a VA priority and may require comprehensive programs that specifically target high-risk groups, such as VA patients receiving treatment for depression. To develop effective programs to reduce risks, VA policy-makers and clinicians need information about the current suicide rates among depressed VA patients, risk factors for suicide that are germane to this population, and the impact of treatment practices on risks. Studying the risks and modifiers of suicide is challenging because of its low base-rate. However, researchers at the VA’s National Serious Mental Illness Treatment Research and Evaluation Center (SMITREC) have constructed a unique longitudinal data resource, the VA National Registry for Depression (NARDEP), that includes large enough numbers of patients followed for sufficient periods of time to allow the study of this tragic but rare event.
OBJECTIVE(S):
We will use NARDEP, supplemented with data from Medicare and the National Death Index to determine: 1) suicide rates among depressed VA patients and important subgroups of these patients, 2) patient-level factors associated with suicide within this population; 3) suicide rates at times of potentially greater risk (following hospitalization or antidepressant change); 4) levels of clinical monitoring provided to patients during high risk periods, and 5) the inter-relationships between patient-level factors, treatment practices, and suicide.
METHODS:
This will be a retrospective cohort design, using data from the VA National Registry for Depression, supplemented with Medicare data and data from the National Death Index (NDI). NARDEP currently includes detailed services and pharmacy data for over 1,300,000 patients diagnosed with depressive disorders in VA facilities from Fiscal Year 1997 forward. The NDI is the “gold standard” for US mortality databases and will be used to identify patients’ vital status and cause of death. Study analyses will use Cox-proportional hazards models with time-varying treatment exposures and covariates. The observation period will extend from April 1, 1999 through September 30, 2004, and patients will be followed until their date of death (from suicide or other causes), two years following their last VA services use, or the end of the observation period, whichever comes first. We will apply propensity scoring and instrumental variable techniques to address potential treatment selection biases that may affect the observed relationships between depression treatment practices and suicide.
FINDINGS/RESULTS:
We have completed analyes examining overall suicide rates among veterans receiving depression treatment in VA settings and patient level factors associated with higher suicide risks. The manuscript outlining these findings is under review.
We have also examined suicide risks during treatment periods following psychiatric hospitalizations, antidepressant starts, and dose changes, confirming that suicide risks are higher for veterans in the 12 week periods following these treatment events. A manuscript outlining these findings is under review.
We have also examined the frequency of clinical visits during these high risk periods and the services and costs implications of increasing the level of clinical monitoring during these periods. A manuscript outlining these findings is in preparation.
Current analyses are focusing on the impact of closer clinical monitoring on suicide risks.
IMPACT:
This project will produce important information about suicide rates, suicide risk factors, and the impact of treatment practices on suicide among depressed VA patients. Study results will be useful to VA clinicians who must make daily assessments of suicide risk for these patients and to VA policy makers who are responsible for developing effective interventions to reduce suicide risks for VA patients.
PUBLICATIONS:
None at this time.
DRA:
Mental Illness, Military and Environmental Exposures
DRE:
none
Keywords:
Depression, PTSD
MeSH Terms:
none
|
