2015 National Conference

1109 — Qualitative Content Analysis of VA Healthcare Records of Veteran Women who Died by Suicide

Denneson LM, Portland Center to Improve Veteran Involvement in Care (CIVIC); Cromer RD, Portland Center to Improve Veteran Involvement in Care (CIVIC); Jacobson LE, Portland Center to Improve Veteran Involvement in Care (CIVIC); Teo AR, Portland Center to Improve Veteran Involvement in Care (CIVIC); Dobscha SK, Portland Center to Improve Veteran Involvement in Care (CIVIC);

Objectives:
The suicide rate among Veteran women has increased by 35 percent in recent years, yet little is known about their experiences shortly before suicide. This project provides a rigorous examination of the VA healthcare records of a nationally-representative sample of Veteran women during the six months prior to suicide. We describe their psychosocial context, primary health concerns, and healthcare they received prior to death.

Methods:
In this qualitative content analysis, we examined complete VA medical record progress notes of a sample of 27 Veteran women from 11 states who died by suicide between 2007 and 2009 and received VA healthcare in the six months prior to death. We used a rapid review approach similar to those described by Beebe (2001) and Hamilton (2013) to identify commonalities and differences across cases, examine within-case phenomena, and compare these cases to commonly examined risk factors in the suicide prevention field. Our coding strategy was based on prior suicide literature and a social-ecological model. Three analysts coded the following categories: social, education/occupation, supports, stressors, health context, and healthcare received. Two psychiatrists conducted an additional clinical-focused coding.

Results:
The women were mostly white, non-Hispanic, and not married, with an average age of 44 (range 26 to 67). While many died by poisoning (12 women), almost as many died using firearms (11 women). We observed a strong connection between health concerns and psychosocial experiences, with an emphasis on disrupted, abusive, or a lack of supportive relationships. Trauma, substance abuse, and poor physical health were especially prominent health concerns. The healthcare they received was often extensive, yet we identified some potential instances of unmet needs shortly before suicide.

Implications:
Additional attention to non-military-related trauma and the capacity to develop meaningful, supportive relationships may be particularly important for Veteran women. In some cases, a better healthcare system response may be needed to more accurately identify and address patient needs.

Impacts:
This project provides a better understanding of the treatment needs of Veteran women at risk for suicide and enables us to elicit potential strategies to improve the quality of care for this high-risk population.