(PS7-92) Prevalence and Implementation of Psychiatric Advance Directives Among Veterans with Serious Mental Illness

A psychiatric advance directive (PAD) is a legal document in which people with mental health concerns express their preferences for future mental health treatment should they ever lack decisional capacity. This is an important tool to ensure self-directed care for individuals with serious mental illness (SMI), who are especially likely to undergo involuntary psychiatric hospitalization. PADs have well-documented benefits, including reduced compulsory admissions and improved patient-provider relationships, perceived autonomy, and empowerment. While the benefits of PADs are clear, a large body of research has established numerous barriers, such as the absence of a formal PAD completion process and misunderstanding of PADs, to implementation. Further, there is no current research on the prevalence of PADs among Veterans with SMI. The present study aims to determine the prevalence of PADs and assess evidence of implementation during psychiatric crisis visits among Veterans with SMI in the southwestern US.

We reviewed electronic health records for Veterans in the Southwestern US who have SMI, which was defined by participation in VA SMI specialty programs. The study had two phases. During the first phase, we collected data on the presence of PADs in the medical record, as well as what mental health preferences were included in each PAD. We also collected demographic and diagnostic data for each Veteran. In the second phase, which will be complete by August 2024, we are assessing evidence of implementation of the mental health preferences within the PADs by examining electronic health record notes of Veterans who have been admitted to a VA hospital for psychiatric reasons.

We reviewed the charts of 1,068 Veterans with SMI receiving care from VAs in the Southwestern US. Of these Veterans, 164 had PADs (15%). Veterans who had a PAD were mostly white (65%), male (75%), had more than one psychiatric diagnosis (82%), and had a mean age of 57.17 (SD=13.76). Veterans who did not have a PAD were mostly white (55%), male (85%), had more than one psychiatric diagnosis (76%), and had a mean age of 53.58 (SD=15.58). We conducted Chi-Square Tests of Independence and found a significant association between female gender and having a PAD, X²(4) = 11.42, p=.022. No other significant associations between demographics and presence of a PAD were found. The most common preference Veterans included in their PADs was the location at which they wish to receive care (n=69, 42%), followed by preferences related to medication (n=58, 35%) and preferences on who the surrogate decision maker should be (n=55, 34%). Some Veterans also stated preferences on other treatments, such as refusal of electroconvulsive therapy (n=13, 8%) or preferences for a certain provider (n=11, 7%). Implementation data collection is underway.

The rate of PADs among Veterans with SMI in the Southwestern US was slightly higher than rates in civilian settings, which are estimated to be 4-13%. One hypothesis for this is the Veterans in our sample were connected to outpatient mental healthcare, which could increase access to resources related to executing PADs. Additional research with a nationally representative sample of Veterans is warranted to further examine the prevalence of PADs across demographic groups.