Talk to the Veterans Crisis Line now
U.S. flag
An official website of the United States government

VA Health Systems Research

Go to the VA ORD website
Go to the QUERI website

SHP 08-163 – HSR Study

SHP 08-163
A Modified Delphi Consultation: Implementing the VA Psychiatric Advance Directive
Jennifer L Strauss, BA MS PhD
Durham VA Medical Center, Durham, NC
Durham, NC
Funding Period: April 2008 - September 2008
VHA is developing its own psychiatric advance directive (PAD) document. This will enable veterans with psychotic or bipolar illnesses, who on relapse often lose decision making capacity and the ability to express treatment preferences, to state preferences for future acute care at a time when they retain capacity, thereby avoiding future coercive treatment. PADs can also lead to constructive discussions with outpatient clinicians that enhance the therapeutic relationship and provide emergency room and inpatient clinicians with valuable information about what has or has not worked for a consumer before, and what treatment they will accept. Several surveys show high levels of consumer demand for PADs, and implementation studies show that the majority of content can be followed by clinicians once accessed.

However, there is substantial evidence that implementation of PADs is difficult, leading to low uptake (excepting research interventions), providers reservations about them, inconsistent trial data regarding their effectiveness, and very low rates of access following completion, problems that apply equally to directives for end of life care. While it is possible to create a valid PAD and have it followed, for many consumers there are numerous barriers both to completion and use. One survey found that psychiatrists, psychologists and social workers all perceived operational barriers to PADs (e.g. lack of access to the document, lack of communication among staff) at a higher rate than clinical barriers (e.g. inappropriate treatment requests). The randomized trial of Facilitated PADs (F-PADs) at Durham VA Medical Center found that while the majority of participants welcomed the intervention, there was little evidence to suggest that provider exposure to F-PADs led to any behavior change, despite in-services from the study team. This trial highlighted two important aspects of implementation which will be addressed by this study. First, the important distinction between barriers to completion, which all trials of advance statements have largely been able to overcome, versus their use during acute mental health care. Second, the need for training and other aspects of implementation to address providers reservations about the use of PADs, especially those of psychiatrists.

Our aim is to produce a set of expert guidelines on the implementation of a VA PAD. As provider training is an essential aspect of implementation, we shall also elicit content that such training should cover as part of this exercise. We consider the legal status of a VA PAD to be beyond our remit but will consider all other aspects of implementation. Specifically, we shall seek the views of four expert stakeholder groups to establish the existing consensus on a number of questions, derived from our previous work and review of the literature:
1. What questions are veteran service users, their caregivers and their mental health care providers likely to have about PADs?
2. Who should be involved in helping veterans complete a PAD?
3. How should VA ensure that veterans are offered a PAD in a way that allows them to make an informed choice?
4. How should VA ensure that all veterans who wish to make a PAD are given the help they need to do so?
5. How should VA ensure that a PAD clearly states the veterans care preferences?
6. How should VA ensure that veterans understand the implications of their PAD content?
7. In what care settings should PADs be completed?
8. How should VA ensure that PADs are kept up to date and do not conflict with other plans such as the treatment plan and Wellness Recovery Action PlanTM (WRAP)?
9. How should the accessibility of a PAD during an emergency be ensured?
10. How can VA ensure that PAD content is followed as far as possible?
The groups to be consulted are (1) Clinicians in leadership positions with an interest in or experience with PADs, both within and outside VA; (2) researchers who have undertaken work on PADs or other types of advance statements for mental health care; (3) consumer advocates, both veterans using VA services and consumers who use non-VA services; (4) caregiver advocates, including those with experience of VHA either as caregivers or services users.
The proposed work will also inform: (1) the development of a provider training package on advance statements, to include vignettes to guide role plays of the completion or execution of a PAD, and a DVD with simulated discussions between veterans, their caregivers and providers about their PAD content before and after it is completed; (2) a research proposal to study implementation of the VA PAD including the use of the training package described in (1).

The Delphi method consists of the following steps: (1) Formation of a team to undertake and monitor a Delphi; (2) Selection of one or more panels of 15-20 members to participate; (3) Development of the first round questionnaire. (4) Questionnaire testing; (5) Transmission of the first questionnaire to panelists; (6) Analysis of first round responses; (7) Preparation of second round questionnaires; (8) Transmission of the second round questionnaires to panelists; (9) Analysis of second round responses. (10) Report preparation by the team. We shall additionally circulate the first draft of the implementation guidelines along with the second round questionnaire (at Step 6). This modified Delphi process [22] is a well established method which has the advantage of eliciting expert feedback on the product within the timeframe of the study.
The team to undertake and monitor the Delphi (Step 1) will consist of the PI, Co-PI, data analyst and research assistant. The PI and Co-PI will identify members of each group by asking colleagues who are members of these groups and who are based at a number of locations nationally and internationally to participate; we shall also ask them to suggest others they know, a process known as daisy-chaining (Step 2). Through this process we shall seek consultation from 15 members of each stakeholder group. The questionnaire will ask respondents to identify and rank what they think are the best methods of PAD implementation. For questions involving ranking, preselected choices will be offered as well as the opportunity for respondents to identify other answers. (Step 3). For questions asking respondents to agree or disagree with some aspect of implementation (e.g. that a third party facilitator is necessary), a 9-point Likert scale will be used. The questionnaire will be tested on colleagues who are not panel members but have some knowledge of PADs (Step 4). After sending the first questionnaire to panelists (Step 5) we shall analyze the results (Step 6) and feed them back in an individualized anonymized format (Steps 7 and 8) so that for each question respondents can compare their responses with those of others. To do this, for each question each respondent will be sent their own response along with the top ranked choice for questions involving ranking, or the median and percentile ranges (see Analysis) for questions using Likert scales. Respondents will be asked to either reconsider their outlying responses or justify sticking to their original response We shall then summarize the recommendations of the stakeholders and describe the level of consensus achieved (Step 9). We shall also use the first iteration of this process to solicit questions the experts think that mental health professionals will have about PADs, and what information veterans and caregivers will want about them. Since a measure of consensus on these is not needed (rather as full a range of content as possible is required), these will be collated for future use in the development of a training package on PADs for providers.
Analysis: For ranked choices, the Friedman test will be used to determine the top ranked choice. For questions using Likert scales, the distribution for each will be examined to determine the outliers. An appropriate percentile range will then be chosen to fit each distribution.

Respondents stated that providers need to know details about the process of completion, limitations of the document, and when and how it will be used. Strong consensus was achieved regarding the importance of provider training and that choice of who should be involved in creating a MHAD should be up to the consumer. Consensus was achieved that a mental health outpatient clinic is an appropriate place for completion. Strong positive consensus existed regarding the importance of periodic review; over 80% of respondents' preferred frequency fell in the 6-24 month range. Among four possible ways of alerting VA clinicians to the presence of a MHAD, no consensus was achieved. The top ranked option, a pop-up window to appear when the electronic chart was opened, was ranked top by 61%. We found strong positive consensus that lack of knowledge about MHADs among both ER psychiatric inpatient providers, and lack of communication between outpatient providers and ER/inpatient providers, present significant barriers to MHAD access. Strong consensus existed that VA should monitor all aspects of MHAD implementation. Views on the MHAD worksheets developed by the National Ethics Center were mixed. Many suggested improvements were provided, which will be summarized for the NEC.

As a national federal healthcare system, VHA is in a unique position to develop and implement its own advance directive for mental health care successfully. First, rather than being tied to the use of the existing available documents which vary by state, VA can introduce its own documentation. Second, the completed documentation can be made readily available to VA providers through the electronic medical record. Third, the existence of the website MyHealtheVet creates the potential for future access to VA-PADs on the part of non-VA providers of acute mental health care.

By providing implementation and educational content recommendations, the proposed work will allow VHA to make the best use of its unique position with respect to PADs, thus taking a lead in incorporating PADs into routine care.

Examination of stakeholders' opinions about VA MHADs may improve VA's ability to implement and maximize uptake of this new policy.

External Links for this Project

Dimensions for VA

Dimensions for VA is a web-based tool available to VA staff that enables detailed searches of published research and research projects.

Learn more about Dimensions for VA.

VA staff not currently on the VA network can access Dimensions by registering for an account using their VA email address.
    Search Dimensions for this project


Journal Articles

  1. Henderson C, Jackson C, Slade M, Young AS, Strauss JL. How should we implement psychiatric advance directives? Views of consumers, caregivers, mental health providers and researchers. Administration and policy in mental health. 2010 Nov 1; 37(6):447-58. [view]
Conference Presentations

  1. Strauss JL, Henderson RC, Jackson C, Slade M, Young AS, Strauss JL. How should we implement psychiatric advance directives? Paper presented at: Institute on Psychiatric Services Annual Meeting; 2009 Oct 15; New York, NY. [view]

DRA: Aging, Older Veterans' Health and Care
DRE: none
Keywords: none
MeSH Terms: none

Questions about the HSR website? Email the Web Team

Any health information on this website is strictly for informational purposes and is not intended as medical advice. It should not be used to diagnose or treat any condition.