Pilot Study of the Collaborative Assessment and Management of Suicidality - Group (CAMS-G)

Brief Summary

Detailed Description

A total of 60 patients from either the Robley Rex VAMC inpatient psychiatry unit or outpatient clinics will be enrolled in the study. If the Veteran is referred from inpatient psychiatry, study staff will meet with the Veteran to determine eligibility and consent the Veterans on the inpatient unit prior to discharge. A phone interview will be scheduled to complete the baseline interview after the Veteran has been discharged from the inpatient psychiatry unit. Outpatient referrals will be handled similarly. Whenever possible, to reduce participant burden, the study team will meet Veterans at a community based outpatient clinic (CBOC) to consent outpatients either immediately prior to or after a scheduled appointment. A phone interview will be scheduled to complete the baseline prior to the start of their group. Patients will then be randomized to either (a) an experimental group receiving CAMS-G or (b) a group receiving care as usual. All study participants will then complete telephone follow-up assessments at 1-, 3-, and 6-month time points administered by members of the research team. Dr. Gutierrez and his study team will rate the CAMS-G sessions for adherence to the treatment manual. Individual Veterans remain in the CAMS-G sessions until their suicidal crisis is resolved. The baseline assessment battery will take approximately 60 minutes and the 1-, 3-, and 6-month assessment batteries will take approximately 45 minutes. The Client Satisfaction Questionnaire will be administered at three- and six-month follow-up for each study participant. This will allow reflection on the acceptability and perceived utility of CAMS-G. Notes will be kept on patients who decided to drop out of the intervention and/or the study. The outcome of this aim is evidence that the final CAMS-G manual and study procedures that would be used in an efficacy trial are acceptable to participants. The investigators will examine the mean and standard deviation for the total score and individual items to determine if study participants report acceptable ratings of satisfaction. The investigators will review the responses to open ended questions regarding what Veterans perceived as being helpful about their group experience and what they could do in the future if they become suicidal. The investigators will tabulate average scores and standard deviations for the overall rating and each individual item on the CAMS-G Rating Scale to determine whether the facilitators maintained fidelity to the treatment manual over the course of the study. The investigators will tally the different services received by study participants in order to accurately describe what care as usual consisted of for all study participants, regardless of their assigned treatment condition. This will also the investigators to control for any major differences in care as usual between treatment conditions, for example more individual psychotherapy for one condition, when examining between group differences on outcomes of interest. At the termination of data collection, the investigators will complete a data cleansing process. Descriptive statistics for demographic and clinical variables will be tabulated. All scales will be scored and subscales described. This process will include examining the data for missing values, appropriate ranges, and outliers, with transformations to skewed data applied as needed. Although the randomization design should ensure balance between the two groups, it is essential to control for any known confounders in the design and analysis to prevent a biased assessment of the suicide assessment effect. Any baseline demographic or clinical variables found to be statistically significant in the analysis and significantly associated with the outcomes of interest will be included as covariates in all analytic models. Dropouts and missing data. The investigators will use an intent-to-treat sample to complete our analyses. That is, all participants who provide consent, are randomized to a treatment condition, and attend at least one group therapy session will be included in the analyses regardless of whether they drop out or complete a full course of group therapy. To understand whether treatment drops were completely random, random, or informative, the investigators will analyze our missing data and adjust our interpretation of the results accordingly. The investigators will use generalized linear mixed models (GLMM) to measure differences at 1-, 3-, and 6-months follow-up between veterans in the CAMS-G treatment condition and those in the control treatment condition on constructs of Hopelessness, Thwarted Belongingness, Perceived Burdensomeness, Overall Symptom Distress, and Reasons for Living. The investigators will use Poisson or similar variant of generlized linear mixed modeling (GLMM) for skewed and count data to measure differences at 1-, 3-, and 6-months follow-up between Veterans in the CAMS-G treatment condition and those in the control treatment condition on severity of suicidal ideation and incidence of self-directed violence.

Primary Outcomes