Processes and standards for reporting of known suicide events were implemented in all VHA facilities in 2008. However, there is evidence of variability in reporting across VISNs. Results from an internal comparison of 2009 reported versus estimated (expected) deaths from suicide suggest that the SPAN database contains approximately 40% of all expected suicide deaths among Veterans receiving VHA services; with substantial variability across administrative units (range: 17-66%). There are two possible hypotheses that could explain the observed variability in suicide case ascertainment: H1: Population-level variability in patient characteristics and service utilization is associated with differences in the number of observed vs. expected suicide events within each VISN, H2: VISN- level variability in suicide event reporting is a function of differences in the processes associated with implementation of suicide event reporting. This project used a mixed-methods approach to answer these questions.
Despite evidence of increased risk for suicide among Veterans and variability in suicide event reporting across VHA facilities, the degree of underreporting and most effective strategies for enhancing the quality of case reports have not been determined. The primary objectives of the project were to identify sources of variability in suicide event reporting and to develop recommendations for enhancing current reporting practices to improve case ascertainment. Such information is critical as a foundation for the development of prevention programs and availability of reliable information on suicide incidents and their characteristics.
Data for this project were obtained from two sources - reports of suicide attempts among Veterans receiving VHA services (SPAN) and VHA clinical and administrative records. Data from SPAN is available through a data use agreement with the Office of the Mental Health Services Director, Suicide Prevention and the PI through the VISN 2 Center of Excellence for Suicide Prevention. To allow for comparability with other data sources and ensure an adequate number of events for analysis, suicide event data was obtained for the fiscal years 2009 & 2010 . Estimates of associations with facility level characteristics were calculated separately for each fiscal year and then combined for form a single analytic data file.
Using data collected during in-depth qualitative interviews (described below) we conducted case studies of key informants from twelve VA facilities. The purpose of these interviews was to describe the case reporting process and facility-level factors associated with variability in suicide event reporting. This information provides critical insight into suicide event reporting and explanations for why it is or is not effective within specific organizational contexts. The selected site were chosen based on the number of reports submitted by facility representatives and changes in reports over the study period. Facilities were divided into high (top 25%) and low contributors (bottom 25%) based on case reporting. A total of twelve facilities were selected. Participating sites included Dallas, Portland, Atlanta, Chillicothe, West Palm Beach, Little Rock, Las Vegas, Boise, St. Louis, Denver, Mountain Home and Minneapolis.
Suicide event reporting was significantly associated with several facility-level characteristics including patient volume, mental health case mix and number of mental health providers; specifically the number of psychiatrists at each facility.
Best practices to support suicide event reporting obtained from interviews with key informants included identification of a single person with responsibility for oversight, routine reporting to and involvement of upper level administrators and sharing of examples of how reporting can improve patient outcomes.
Standardization of terminology and clear criteria for reporting were identified as critical needs to support improved reporting. Challenges in staffing for Suicide Prevention Coordinator (SPC) and SPC turnover were identified as significant impediments to reliable reporting. Key informants also expressed concern about the use of VA's suicide event reporting system for surveillance and noted the clinical significance of the system for case management and supportive services for those at increased risk for suicide.
By implementing a process for suicide event reporting that is not solely dependent upon official cause of death determination and guidelines for the standardization of suicide event classification (nomenclature), the Department of Veterans Affairs is uniquely positioned to advance understanding of the impact of suicide prevention programs and identification of characteristics associated with increased risk. However, the ability to evaluate outcomes among those enrolled in these programs or quantify the burden and characteristics of risk is dependent upon reliable and consistent suicide event reporting across VHA facilities. Results from this project identified several factors, both qualitative and quantitative, associated with variability in suicide event case reporting and support the need for ongoing assessment of case ascertainment and factors associated with variability in case reporting as programmatic activities.
None at this time.
Mental, Cognitive and Behavioral Disorders