Health Affairs: Dr. Ashley Clayton reflects on care she received as a teenager

If my recent post on Patient and Family Centered Responses to Suicide Risk interested you–or if it didn't–I highly recommend an article by Dr. Ashley Clayton published this month in Health Affairs:  How ‘Person-Centered’ Care Helped Guide Me Toward Recovery From Mental Illness.

Thanks to Dr. Yeates Conwell for pointing me to the article.

Helping at-risk patients who rarely show up

From early on in my quest to understand the unique challenges mental health professionals face in working with patients at risk for suicide, I've wondered aloud about the things that make us the most nervous. I'm still working my way through a  list of questions I posted based on my notes from a series of trainings I delivered across New York State. One that has kept coming up since that time is this one:
How do we handle individuals at risk who are only marginally involved in treatment-they miss more appointments than they make, but still come enough that they remain on our caseloads?

Mental health care was not organized to fit the way many people at risk utilize services. Much of outpatient mental health is organized around a fantasy that most patients will (1) Make an appointment ahead of time for an evaluation; (2) come to that appointment at the specified time; (3) make another appointment; (4) come to that appointment; (5) work on a signed treatment plan in between.

Now, that model probably does work well for some people. But for many individuals at risk, that level of organization and consistency is not congruent with their lives. If it were, they wouldn't need us! The situations that make professionals the most nervous are the ones where the patient (1) Makes an appointment.  (2) Comes to the appointment later in the day or the next day crying and upset, hoping to be seen. (2) Misses the next appointment.  (3) Makes it to one appointment with the psychiatrist. (4) Misses 2 of the next 3 appointments with the primary therapist (5) can't recall the treatment plan when asked about it. Naturally, clinicians get frustrated and wish the person would either sign-up or drop out.

But of course it's not the fault of the individual at risk that we're set up the way we are, or that all the empirically supported treatments assume the patient is actually in treatment. I've addressed this issue here, here, and here when I've talked about how to manage patients who violate the "social contract" of being a patient. But there's something distinctly anxiety-provoking about someone who is loosely connected to mental health services, compared with someone who actively refuses plans we suggest.  It can be especially hard if the loosely-connected person only shows up when in crisis.

This won't be a surprise to those who have followed my posts, but my approach to the loosely connected person hinges on two core clinical tasks: connection, compassion, and documentation.

Connection means that we should have the bias of keeping people involved to the level they are able.  I know that this is difficult in the context of productivity demands and limited resources.  You can't have 80 people on your caseload who all miss 75% of appointments.  But I think we should condition ourselves (and set up services) to have a least a few people like that because the resources might be well spent in being a bridge to life for the person.

Compassion is always core. It can be hard to feel and show compassion to someone in crisis when they haven't taken the least first step to try ideas you've given. It's easy to find yourself thinking, "Maybe if you came a little more often, you wouldn't be in this position." It's especially hard to find compassion for someone who is at-risk for suicide and not showing up because it feels like they could take us down with them. All of those thoughts are natural, and it shouldn't end there. Instead, we have to summon the courage to enter in to the suffering one more time, offering the main thing we can offer under these circumstances: a caring commitment to living.

Documentation is one way I get to that place. One barriers to compassion is fear.  It takes courage to keep working with someone who has only one foot in the room--especially when they present with suicide risk. Part of every clinician's brain is occupied with worry under these circumstances--worry about losing a patient, worry about being blamed, worry about being sued. It's hard to have compassion for someone if you think they could ruin your life. Having an unassailable risk assessment and other supporting documentation can put the fearful part of the clinicians brain to rest and make way for the kind of compassionate connection the hurting patient really needs. In other words, one of the key purposes of documentation is to quiet fears that might interfere with caring--and with doing the right thing. Once you're confident about documentation, you can focus on doing what's best for the patient, instead of feeling cornered into defensive courses of action.

Connection, compassion, and documentation share at least one thing in common: they are easier said than done. Their importance is so obvious that it may be tempting for an experienced clinician to speed passed these concepts. Anyone NOT think connection, compassion and documentation are important? Of course not. But, as with most simple good ideas, there are complex barriers to implementing them. Each person has to identify his or her own barriers and find a way to build these principles in to every day practice.

"Trusting" a person at risk who agrees to transport self

A colleague  forwarded me the following excellent question posted to a listserv:
One question that has been raised is how to handle an individual who reports willingness to voluntarily go to the ER for psychiatric assessment.  Since trusting a questionably unstable and suicidal individual to present for treatment opens our agency and the patient up to considerable risk, I wanted to get input on this issue from professionals in the field.

My questions are: What is the process that you use when you genuinely believe an adult patient will voluntarily present her/him self for possible commitment/assessment?  How do you manage patient risk and your own liability in this instance? (e.g . requiring family members to be involved; or requiring a signed written contract to present at the hospital, etc) What do you do to ensure/confirm that they do indeed go to the hospital? What do you do when they do not go to the hospital, as agreed?

I'm looking for both a description of specific steps that you take and what variables you take into consideration as you decide what to do in this case.

This is an excellent question for several reasons:

  • the writer wants to support patient agency, preferring in principle to go with the patients voluntary wishes, rather than become unnecessarily coercive.

  • the writer is correctly concerned about putting all eggs in the self-report basket.   A person at risk who is not stable may not be able to follow through with intentions to get help.   It is central to both really listen people in our care AND understand that self-report has limitations and has to be considered in context

  • the writer has shared honestly her concern about both patient safety and his/her own liability—almost everyone who works with individuals at risk has these dual concerns and needs to think openly about how to balance them.


This is such a good question that I wish I had put it on my docket of questions to blog about.  And it relates as a follow-up question to one that is on that docket, "When it comes to hospitalizing isn’t it always best to “err on the side of caution”?

Here is my attempt to offer some principles for the assessment and documentation of a patient's agreement to voluntary actions:

  • Document assessment of reliability of self-report.  When assessing and documenting self-report the following should be considered and noted:

    • past experience, if any with the patient as a reliable reporter or not.   “Pt. has reliably followed through on medical decisions and plans in the past.”

    • evidence of psychosis, thought disorder, intoxication, extreme agitation or other factor that would put the person’s capacity to make decisions in question.   If not, “Pt’s thinking is logical, coherent, and reasonable.  Judgment is intact.   No indication of impaired capacity to make decisions and follow through with them.”

    • degree of patient cooperation.  Explicitly note that the person volunteered information, though s/he didn’t have to.   “The patient has been open, cooperative, and collaborative in the assessment and planning process.  There is little reason to doubt his/her sincerity in agreeing to seek help and additional evaluation.”

    • impulsivity.   The greatest risk here would be that the person would all of a sudden feel overwhelmed with pain and impulsively kill him/herself.  If there is not strong evidence of that kind of impulsivity (even if there has been some in the past) note that.  “The patient has a past history of mild impulsive behavior, however, this was several years ago and under very different circumstances..."



  • Show your reasoning.  The key to great documentation is to state each of the factors considered, then show that the plan came as a result of synthesizing these factors.  This can come in a statement like “In light of these factors, the plan for voluntary self-transport seems reasonable and prudent.”

  • Say what you didn't do.  Related to the previous point, it as as important to document the road NOT taken.   This shows that you took the situation seriously and considered the full range of options.

    • A corollary to this principle is to note risks associated with alternative courses of action.   “Involuntary transport has potential of violating the patient’s rights and of harming the patient’s opportunity to benefit from hospitalization.   The risk associated with involuntary transport outweigh the relatively small risk that the patient will precipitously abandon our plan and harm him/herself.”



  • Document consultation.   Much more consultation occurs in primary care than is ever documented.   This would be a time to do it.  One line is sufficient:  “Discussed case with Dr. X who concurred with the plan.”


Note that each of these points needs only a sentence or bullet-point, and not all will be applicable in every circumstance.

I'd be interested in feedback and ideas from who employs these principles in their documentation.  I would also be interested in reading comments with other ideas about how to address the situation the questioner described.

Organizational factors that support care of suicidal person


Wendi Cross, a gifted and innovative colleague in our department, presented at our Family Research Roundtable yesterday. One of the ways she is contributing to the field is to raise awareness about, and develop methodology to study, the factors surrounding implementation of an evidence-based intervention (be it training, prevention, or therapeutic intervention) that influence its ultimate impact. Right now her focus is on implementers of an intervention, but in the course of discussion she mentioned that there are other extra-intervention factors, such as institutional and organizational factors that influence how, how much, and how well evidence-based interventions are put into practice after someone is trained.


This got me thinking about what might be the institutional factors that enable (rather than hinder) adoption and competence in evidence-based risk assessment. I started making a list of these. In the course of doing so, I realized that, without having the language for it, this is part of what I have been working to build in to my trainings about risk assessment. I accompany training about risk-factors with documentation examples and templates and with a map of the procedures, options, and services available to a clinician if elevated risk is identified. It dawned on me that the reason participants find this part of it helpful is that it embeds the training in institutional resources.


Here is my initial brainstorm of potential institutional/organizational enabling factors. Whether these actually make a difference is an empirical question, of course, but these are the ones that occur to me based on my experience:




  1. Documentation prompts. Embedding risk assessment in required documentation templates can support risk assessment training by providing ongoing “booster training” in risk factors and risk formulation. In terms of risk assessment training this is "external" to the training, but can enable its implementation.

  2. Clear instruction about what to do when risk is detected. As I noted above, my trainings bring together risk assessment with specific information about resources and procedures at the local level. We have gotten feedback that this improves implementation of the risk formulation training.

  3. A non-blaming, supportive environment. Clinicians will distance themselves from patients with suicide risk if they fear they will be blamed (in any way) if injures or kills him/herself. There needs to be a consistent message from every level of an organization that understands how difficult clinical work is. There needs to be a healthy respect for the courage it takes to stick with a person at risk—especially if the person is not as cooperative as we would like (see my previous posts about patient choice). A non-blaming environment needs to be in place around all kinds of clinical issues and outcomes, not just suicide—otherwise it will not appear genuine.

  4. A culture that emphasizes patient choice, informed consent, and meeting people where they are. A “comply or bye-bye” approach will probably discourage thorough and honest risk assessment. In a previous post on risk and patient choice I gave an example of how to handle a patient’s refusal to comply with a recommendation to attend a partial hospitalization program.  I have a suspicion that clinicians sometimes downplay risk or avoid asking too many questions about suicide when we suspect that a patient would not comply with the recommendation that higher risk would compel us to make. For example, if a clinician knows that the patient will refuse an E.D. for evaluation, the clinician might de-emphasize or not ask about risk factors that could signal need for such an evaluation, thereby avoiding appearing negligent for not referring.
    In contrast, we might promote risk assessment if the clinician knows his her her job is to make an assessment, inform the patient of options, present risks and benefits, and work with whatever the patient decides. I think this would be a significant mindset shift for many systems, and one that I think could be one of those enabling factors.
    We can take a lesson from our colleagues in family medicine about how to continue a supportive, therapeutic relationship with patients who do not follow our recommendations (who do not abide by the "social contract" as discussed in this post). Family physicians are accustomed to recommending treatments, medications, and interventions, and knowing that some proportion of patients will not end up following them. This is not without frustration, to be sure. But, except in rare cases, family physicians continue to work the patient, meeting the patient where he or she is, and hoping to move them slowly toward change. As one of our clinical leaders pointed out to me yesterday, that is where a motivational interviewing (MI) approach would come in handy. And, in fact, one of ther research fellows at our institution is studying on an MI intervention with suicidal individuals in the emergency department.


There are lots of others, but these are the ones that first come to mind. I welcome comments with other ideas about factors, outside of training and the individual clinician, that would enable good care and assessment for patient with risk for suicide.

Clinician response to violation of the "social contract"

I had a stimulating conversation with a senior colleague in the CSPS yesterday. One part of the conversation centered around what happens for us, as clinicians, when the patient does not fulfill his/her end of the "social contract" that is implied when someone goes to a mental health professional. The assumed contract is that the professional gets to ask all kinds of personal questions, make recommendations about intimate details of a person's life, and the patient is expected to accurately, honestly, and completely answer our questions, cooperate with recommendations, and be appreciative in the process.

In a previous post about risk and patient Choice, I offered ideas about how to approach a particular instance of patient patient choice (sometime known as "noncompliance"), including some initial documentation suggestions. But that post didn't address how to handle the emotions that accompany working with a patient who exercises his/her freedom by not answering our questions or by refusing our help.

Clinicians experience a range of emotions in these situations under these circumstances. Helplessness, anger, and fear come to mind first. From a family therapy perspective, this calls for what Murray Bowen (1978) called "differentiation"--the ability to remain engaged, present, and available while not becoming reactive, defensive, or distant. Easier said than done! Especially hard when clinicians often feel a duty to protect patients from harm.

The first step is to name what is going on. "This person is violating the 'social contract' and it is making me angry. I'm also afraid that this person could die while in treatment with me." Next is some cognitive work: "This person did not give up the right to direct his life, keep his privacy, and make choices when he sought professional help." Along with that, your best friend is good old fashion consultation. I plan to post more in the future about how to get consultation from colleagues about suicide risk, but for now, I think the main thing is for us, as clinicians, to explicitly frame the consultation in term of our emotions. "I am afraid and angry because a person I am working with, who has high risk of suicide, is not cooperating with my attempts to assess and intervene." That frames the consultation discussion as being about "how am I going to work with this person given my emotions?," rather than "how can I get this person to do what I want, given how difficult he is."

As always, there is much more to say about this. This discussion of the therapeutic social contract and patient choice is not specific to patients with suicide risk. But it deserves special attention in the context of suicide risk because the stakes are high, the issue is loaded, and medical-legal preoccupations kick in and complicate everything. More to come...

References

Bowen, M. (1978). Family Therapy in Clinical Practice. New York: Aronson Publishers.

Risk and patient choice

It's hard enough to assess for suicidal risk, interview, reach a formulation, and develop a plan that matches the risk level. But what do you do when the client does not agree to the plan? It is difficult to manage one's emotions, and difficult to know how to proceed. This is especially true when the recommended plan involves a higher level of care. Do you continue to work with the person in a setting that you have judged inadequate for the needs, or do you somehow insist or refuse to work with the person? How do you ddocument around these decisions?

Here are some initial thoughts I have about this unbelievably challenging clinical dilemma. I hope to articulate more principles and recommendations in future posts and papers, and would love to hear other clinicians' ideas.

1. Articulate and document the rationale for recommendation AND for the treatment decision you make. For example: "I recommended partial hospitalization because the daily therapeutic contact, extensive group work, and focus on stabilization fit Mr. X's needs in this case. I offered this recommendation to Mr. X and explained the potential benefits and risks of following this recommendation. I also explained the substantial risk I see in not pursuing this level of care at this time."

2. Respect the client's freedom, and avoid a power struggle. Most power struggles are born of anxiety. We sometimes do have totake coercive courses of action to protect a patient's safety, but many times we twist patients' arm to manage our anxiety more than their safety. This is not a black-and-white/either-or issue. We do need to honor our boundaries, our ethical/legal obligations, and have a level of comfort in order to work with someone. At the same time, clients with intact cognition (i.e, not psychotic or otherwise impaired from making rational decisions) and who are not at truly imminent risk need to have some latitude about the treatment they agree to participate in. Before drawing a line in the sand, we should be clear (and be able to articulate) why drawing that line is necessary for the patient and for the treatment.

3. If you decide to respect a client's choice that goes counter to your recommendation (a la #2), document the clinical reasons why. If Mr. X refuses my recommendation to go to a higher level of care, I might write: "Mr. X refused the offer of partial hospital; he wishes to continue weekly therapy but agrees to nothing more. I agreed to continue to see him outpatient because (a) our alliance is strong and is a protective factor, (b) Mr. X has benefited from outpatient treatment, even though his acuity level makes him more appropriate for partial, (c) he is not in imminent danger, and (d) Mr. X's cognition is grossly intact; his thought process is logical and coherent, and he is judged to have capacity to make decisions about the treatment he wishes." I'm sure this is not a perfect note (and I'd be happy to hear from others about what you might write), but it captures some of the key elements that I think should be documented.