Should a dying patient talking to a psychiatrist be diagnosed with adjustment disorder related to death?
Dare we ask: Could impending death not be sufficiently psychologically distressing for patients to routinely benefit from psychotherapy?
Patients aware that they are dying often wish to talk to someone about their predicament. Should psychiatrists and psychologists be the first line professionals for such discussions? After all, aren’t many dying patients experiencing substantial psychological distress? Specialty mental health professionals would have the skills to differentially diagnose this distress and offer appropriate targeted treatment. And what about their ability to identify and address issues of suicidality?
Or should discussions be left to clergy and pastoral counselors, especially for patients of faith?
Raising these questions could easily get us into philosophical and spiritual issues where we might feel excused from having to bring in evidence. But a relatively large-scale and well-designed study has given us a relevant answer. Maybe not the answer that the investigators hoped to find and they probably will not like my further interpretation. And, sure, it is only one randomized trial, but results seem to speak exceptionally clearly about a number of issues.
The study was negative, but demonstrates just how much a well-designed negative trial can reveal.
Chochinov, H. M., Kristjanson, L. J., Breitbart, W., McClement, S., Hack, T. F., Hassard, T., & Harlos, M. (2011). Effect of dignity therapy on distress and end-of-life experience in terminally ill patients: a randomised controlled trial. The Lancet Oncology, 12(8), 753-762.
The randomized trial was conducted with a large sample (441) of patients recruited from hospices and home-based palliative care in Australia, Canada, and the United States. Investigators compared their preferred intervention, dignity therapy (165 patients), to two comparison/ conditions – standard palliative care (140 patients) and client-centered care (136 patients).
This study represents the first randomized control trial of Dignity Therapy. We set out to determine if this novel psychotherapeutic approach would significantly outperform standard care, or Client Centered Care…on various measures of psychological, existential and spiritual distress.
Primary outcome. The primary outcome was distress, which sounds simple enough, but it was measured 23 different ways. No one variable was designated as the primary outcome for the purposes of evaluating the effectiveness or ineffectiveness of dignity therapy would be judged. There are also 22 different secondary outcomes, single items evaluating patients’ experience participating in the intervention.
The three conditions
Dignity Therapy provides patients with life-limiting illnesses an opportunity to speak about things that matter most to them. The conversations are recorded and transcribed to provide the basis for a document that patients can bequeath to individuals of their choosing.
Patients were shown the Dignity Therapy question framework Table 1] and asked to consider what they might wish to speak about during their Dignity Therapy session(s); this initial introduction to, and explanation of Dignity Therapy took about 30 minutes. Within a few days, or as soon as a second meeting could be arranged, the therapist used the question framework to help elicit the patient’s recollections, hopes, wishes for loved ones; lessons learned and things they wanted remembered by those they were about to leave behind. Dignity Therapy is flexible enough to accommodate participant preferences and choices regarding content, but the ethos of questioning targets those things that might enhance a sense of meaning, purpose, continued sense of self, and overall sense of dignity. All Dignity Therapy sessions were audio-taped; these sessions usually took about 60 minutes. Upon completion, the audio-recording was transcribed verbatim and the transcript edited, to provide a clear and readable narrative. This transcript or `generativity document’ was returned to the patient..,[and] read to them in its entirety to ensure that no errors of omission or commission needed to be addressed (this final session usually took about 30 minutes). .. The final version of the generativity document was then given to the patient, to be passed along to a designated recipient of their choosing.
Generativity is a key concept in dignity therapy.
Generativity or the ability to guide the next generation, encompasses how patients might find strength or comfort in knowing that they will leave behind something lasing and transcendent after death.
Some readers may recognize the similarity of the investigators’ concept of generativity and Erik Erikson’s life stage of generativity versus stagnation.
In Standard Palliative Care patients had access to a full range of palliative care support services, including specialist palliative care physicians and nurses with expertise pain and symptom management, social workers, clergy, and mental health professionals.
There were not, however, any components of standard palliative care directly comparable to the dignity therapy.
Client Centred Care was a supportive psychotherapeutic approach, in which the therapist guides patients through discussions focusing on here and now issues. These might include their illness and symptoms and what is being done to address symptom distress. However, in order to keep this condition distinct from dignity therapy, the therapist did not encourage discussion of issues of meaning and purpose. If these topics came up, the therapist redirected the conversation back to illness-related issues.
The therapist for client-centred care was a research nurse.
What was found.
The 23 items representing secondary outcomes had only a few differences between the three groups, no more than would be expected by chance. We should be careful about the statistically significant results that were obtained, but perhaps patients were expressing appreciation that they had been randomized to the specialized treatment, as well as having the document to leave for family members. With only five items dignity therapy was found to be superior to both standard palliative care and client-centred care.
The investigators’ interpretation.
Despite the beneficial effects of Dignity Therapy, its ability to mitigate outright distress, such as depression, desire for death or suicidality, has yet to be proven. However, there is currently ample evidence supporting its clinical application for patients nearing death, as a means of enhancing their end-of-life experience.
While floor effects precluded our ability to demonstrate significant differences across study arms, our secondary outcomes revealed substantive benefits of Dignity Therapy. Using a post-study survey, patients who received Dignity Therapy were significantly more likely to report benefits, in terms of finding it helpful, improving their quality of life, their sense of dignity; changing how their family might see or appreciate them, and deeming it helpful to their family, compared to the other study arms.
The investigators were obviously passionate about their intervention and looked exhaustively for evidence of the efficacy of dignity therapy. They really did not find any.
What I liked about this trial.
I like the unusually large sample size and the inclusion of two different control/comparison groups the allowed answering different kinds of questions. This is certainly not an underpowered pilot study being passed off as a full randomized trial. The Standard Palliative Care group allowed determination of whether dignity therapy offered anything beyond routine care. And a reader definitely gets the sense that routine care in this study was neither no care or inadequate care, as it is with so many studies. The Client Centered Care condition pitted the investigators’ preferred intervention against a lower intensity intervention that provides support, but without an element that the investigators considered key to their intervention.
The intervention followed a structured standardized protocol. Standardized training was provided, along with group supervision and review of transcripts of recordings of actual sessions to ensure fidelity of delivery.
The study recruited from a variety of settings and had an excellent uptake from patients who were offered an opportunity to participate.
Patients were randomized with the investigators blinded as to group assignment.
What I like less
The investigators administered a full battery of potential outcome measures including total and subscale scores of
The Functional Assessment of Chronic Illness Therapy Spiritual Well-Being Scale, Patient Dignity Inventory, Hospital Anxiety and Depression Scale, items from the Structured Interview for Symptoms and Concerns, Quality of Life Scale, and modified Edmonton Symptom Assessment Scale.
These measures tend to be moderately to highly correlated and so having this battery available for assessment of outcome represents considerable redundancy. The investigators not designating one or two measures as the primary outcomes set the stage for the investigators selecting a primary outcome after they knew the results, risking confirmatory bias and capitalization on chance.
The investigators had to contend with substantial lots of patients, not unexpectedly, because these are palliative care patients, but they did not avail themselves of a number of ways to provide estimates for the missing data and so analyses were not intention-to-treat.
This little reason to believe that this changed the results, however, because of the floor effect that they noted.
I did not like the positive spin put on the null findings of this trial. Confirmatory bias was clear in the abstract and extended to the discussion. The structured abstract conceded a lack of effect on distress, but cherry picked positive findings to emphasize out of an overall null set of secondary outcomes. In the interpretation section of the structured abstract offered:
Although the ability of dignity therapy to mitigate outright distress, such as depression, desire for death or suicidality, has yet to be proven, its benefits in terms of self-reported end-of-life experiences support its clinical application for patients nearing death.
The laudatory accompanying commentary.
The commentary on this article is an embarrassingly obvious attempt to spin and refute results the commentator did not want to accept.
Commentaries in peer-reviewed journals are typically not rigorously peer-reviewed. They tend to be written with an agenda, by reviewers who nominated themselves to write them. I have previously blogged about an unusual commentary in which the writer sought to destroy the credibility of an article she did not like, despite it having been accepted for publication. Most often, however, commentaries are laudatory and written with an awareness that the authors of the commentaries can get away with praise that would not survive peer-reviewed.
The commentator sought damage control in the face of an utter lack of significant findings. She tried to undermine the validity of an RCT for studying psychosocial interventions. It is true that clinical trials restrict participation to patients with sufficient cognitive intactness, but so do any talk therapies. Her criticism that the sample was not heterogeneous and representative was contradicted by demonstration most patients who were approached participated. Differences in baseline variables are a threat to the validity of results of an RCT, but one of the strengths of such a design is that it serves to minimize such differences. And the commentator gave no evidence that baseline differences between groups undermined the validity of this particular trial.
The commentator raised issues about the standardization of the treatment across settings, but ignored the efforts made to ensure fidelity of delivery. The commentator further ignored the efforts of the investigators to control for nonspecific factors by inclusion of both a standard care and a client centred/comparison control condition. Maybe the richness of standard care precludes finding any effect for the addition of dignity therapy, but that is a valuable finding. It sng that a specialized dignity therapy is not needed. Finally, the commentator’s suggestion that outcomes may not adequately measured is bewildering in the face of the investigators administering a battery of 22 measures for primary outcome and 23 measures for secondary outcome. What does she think is missing?
Given the effort to positively spin solidly null findings in both the article and the commentary, one has to ask under what conditions the investigators and commentator with and willing to concede that this is not a promising line of treatment development?
Why I consider this study important.
Although one does not make a definitive judgment on the basis of a single trial, there is little here to encourage further consideration of dignity therapy as an evidence-based therapy.
The investigators made the testable assumption that end-of-life is a time of considerable psychological distress. They designed an intervention to be delivered by mental health professionals to relieve that distress. They evaluated their intervention in an adequately powered randomized trial, but failed to find any evidence of an effect. The most likely reason is that this large representative sample of palliative care patients was simply not distressed enough to benefit from a mental health intervention.
One could argue that the intervention was not sufficiently intense or mental health oriented, but low baseline distress would preclude a more intense intervention succeeding. There is a floor effect going on, as the investigators recognize.
Basic pilot work would have revealed the surprisingly low levels of distress in this population. A specialty mental health oriented intervention may not be warranted with patients who are not specifically selected for distress. And furthermore, the minority of palliative care patients who show clinically significant distress may not benefit from an intervention like this. They probably would require something more intensive and specifically aimed at reducing distress, with evidence of the intervention having worked in other populations.
But the assumption that cancer patients suffer substantial psychological distress, and patients in palliative care particularly so, is so well entrenched that it is difficult to challenge. Certainly, if investigators applied for a NIH grant and stated that end-of-life is not a time of great psychological distress, they would have risked a low priority for funding.
Giving dying patients a chance to talk: Going from a feature of routine care to a treatment
The investigators failed to find that their structured intervention, dignity therapy offered benefits beyond routine care or a client centred care that had key elements removed: therapists for this condition were not allowed to discuss meaning or purpose.
Suppose investigators had found a therapy mainly provided by psychiatrists and psychologists yielded substantial reductions in distress. These findings would have reinforced existing views of palliative care patients as having high levels of distress. They would further have demonstrated the inadequacy of both good quality routine care and a client centred care modeled after what pastoral counselors offer in providing dying patients with opportunities to discuss their “recollections, hopes, wishes for loved ones; lessons learned and things they wanted remembered by those they were about to leave behind.”
Dignity therapy would be on its way to being an evidence-based treatment with demonstrated ability to reduce significant psychological distress.
And this would propel our moving what had been a feature of routine care to a formal treatment that is of necessity a billable procedure that ultimately requires a diagnosis psychiatrist and psychologist do not provide treatment without diagnosis of disorder. But with what diagnosis?
The low levels of psychological distress do not indicate the likelihood of substantial psychiatric disorder in this population. There is already a suitable diagnostic category that is loose and limited in its validity, but that serves administrative purposes, adjustment disorder. Mental health professionals seeking to document the diagnosis of patients being treated with dignity therapy, but lacking any other formal psychiatric disorder, could always bill for adjustment reaction related to death. That mental health professionals wanted to treat the patient and the patient wanted treatment would in itself be an indication of distress and clinically significant impairment. And the cause of the disorder the dignity therapy targeted is impending death.
Making a diagnosis of adjustment disorder often is a successful resolution of acute compensation crisis of mental health professionals. They are not offered treatment in the absence of the diagnosis, they are treating adjustment disorder.
Psychiatrists and psychologists offer treatment in “sessions” that are of necessity of limited duration typically half hour to 50 minutes. They do so with a certain time pressure. They follow schedules allowing only small time intervals between patients and no running over, even if the discussion is intense and productive.
In contrast, pastoral counselors do not require diagnoses for what they provide to patients. While there may be heavy demands on their time, they do not typically operate with rigidly timed “sessions” and will stay with the patient as long as needed, within practical limits. This particular trial of dignity therapy fails to find any evidence that what they do is less efficacious than a psychologist or psychiatrist.
Much of the content of dignity therapy will differ according to the religious faith of patients. It would seem that pastoral counselors could address these issues with greater authority and knowledge.
As an agnostic, I approached working with pastoral counselors in cancer care with skepticism as to whether they would take advantage of patients’ dying days to win them back to their faith. I never saw any of that happen. Rather, all the pastoral counselors I have ever seen have a deep respect for patients’ level of commitment to their faith – or lack thereof. I would characterize what they do is provide a presence for patients’ talking about things that matter to them, nondirectively guiding the conversation forward, but without imposition of their own views.
How do we evaluate what pastoral counselors do? If what they provide is not considered treatment, there is no issue of whether it is evidence-based treatment. It does not require that patients be distressed in order to be eligible for a discussion, nor that the distress be resolved for the discussion to be “successful.”
An example from primary care illustrates some of the “crisis” of medical care in the United States and in many countries that extends to cancer care, dignity therapy would be implemented.
As detailed in a CBS news story, a woman sought a physical examination from her primary care physician, after not receiving regular care for some time. She came prepared with a long list of health questions for which she would seek answers in her appointment. The physician obtained reimbursement from insurance for the physical by billing for all of the specific tests and procedures that had been done. However, he billed the woman by the minute for the discussion. A commentator in the news segment discussing this said, in effect, ‘physicians get paid for doing procedures, not for whether they solve problems and not for talking about it.”
Cancer care is expensive, which means lucrative for both for-profit and not-for-profit settings that can obtain reimbursement. There is increasing emphasis on procedures that are billable to third-party payers and for efficient use of providers’ time. Provision of basic emotional support and time for patients discussing their concerns are endangered features of routine cancer care. But we should be careful about efforts to preserve these features by making them billable mental health procedures. That entails an inevitable need for a diagnosis and for rationing talk time, restricting it to those patients having a mental health related disorder.
More generally, psychotherapy intervention trials in cancer care do not typically attract patients with clinically significant distress in great numbers. Across trials, only about a third of the patients are clinically distressed. That is about the same as what you find in primary care medical population waiting rooms.
Correspondingly, RCTs of psychotherapy for cancer patients often failed to show a benefit because the patient samples that are recruited are overall insufficiently distressed. Why are the patients there, then?
It is likely that with the increasing scarcity of talk time in routine care, patients are simply seeking a safe place where they will be listened to, and can express and reflect on their feelings, not necessarily solve problems or reduce distress. That can be an entirely valid goal in itself. But problems arise when these discussions are of necessity provided as treatment with mental health professionals. Issues of cost effectiveness and efficacy arise, for which formal evidence is required. And such treatment is typically in short supply, with long waiting lists.
This RCT of dignity therapy came about because a psychiatrist passionate about what he provides for palliative care patients developed and evaluated a structured mental health treatment. Maybe it is not all bad that the trial was negative. We now have a greater chance of preserving supportive elements of palliative care, including time for patients to talk about their concerns, and can hold off rationing them.