Palliative Care: An Approach for All Internists

In 2010 Jennifer Temel and colleagues published what many feel is the most influential article in palliative care since the 1995 publication of the Study to Understand Prognoses and Preferences for Outcomes and Risks of Treatment (SUPPORT).^1,2 Temel and colleagues randomized 151 patients at the Massachusetts General Hospital with metastatic non-small cell lung cancer to receive early outpatient palliative care or standard care. Patients randomized to palliative care had significant improvement in quality of life and lower rates of depressive symptoms compared to standard care. These findings confirmed benefits that many clinicians suspected from anecdotal experience. What shocked many was the finding that despite receiving less aggressive end-of-life care, patients randomized to palliative care had a longer median survival than patients in the standard care group.

The survival finding challenges the deeply-held assumption of a tradeoff between palliative approaches to care and aggressive “life-prolonging” treatments. This assumption is buried deep in the language clinicians use in everyday discussions about goals of care, as in, “All things considered, do you want to focus on quality of life and comfort, even if it means not living as long? Or living as long as possible, even if it may mean enduring more pain and discomfort?” In contrast, the survival finding suggests that, for certain patients, not only is there no evidence that palliative care shortens life, but for some patients, there is evidence that palliative care prolongs life. There may not be an “either-or” tradeoff between quality of life and length of life.

Evidence suggests a trend toward more aggressive end-of-life treatment. In a study of 215,484 Medicare patients who died with advanced cancer in the Surveillance, Epidemiology, and End Results dataset, rates of chemotherapy within 14 days of death increased between 1993 and 2000, as did rates of emergency department use and hospitalization in the last 30 days of life.³ Presumably, chemotherapy, emergency department use, and hospitalization are directed toward prolonging life. Yet there are two main hypotheses that might explain the survival findings in the Temel study. First, was increased survival due to improved recognition and treatment of depression, and improvements in quality of life? Or second, was prolonged survival due to lower rates of chemotherapy, emergency department use, and hospitalization in the last weeks of life? (See supplemental appendix)¹ If the latter is the case, then more aggressive treatment and hospital care presumed to prolong life may actually have the opposite effect for some.

To be sure, the Temel findings of improved survival need to be replicated. While there was some suggestion from previous literature that a palliative approach might lead to prolonged survival,^4,5 this was the first randomized trial to demonstrate this finding. The biggest question raised by the findings in the Temel study was: are these findings replicable outside of a single institution, the Massachusetts General Hospital?

A study by Yoong and colleagues (Temel senior author) in the current issue of Archives provides tantalizing hints about the answers to this question.⁶ In the original randomized trial, we have little data about what exactly the outpatient palliative clinicians did in the clinic. In the study published in this issue of Archives,⁶ the authors conducted a retrospective qualitative content analysis of the clinic visit notes of 20 patients in the palliative care group of the randomized trial. Clinic notes suggest that the outpatient palliative care visits followed a semi-structured yet flexible approach to outpatient encounters, and when the cancer and symptoms worsened, palliative care clinicians and oncologists documented a different focus to their conversations.

The progression of palliative care clinic visits from initial encounter to death are in some ways a model for how all good outpatient visits should proceed for patients with serious illness. Initial visits were focused on building a relationship with the patient, understanding the larger social context of the illness experience, and how the cancer affected their lives. Clinicians established preferences for decision-making and patient understanding of prognosis. Symptoms and coping were addressed across all visits. When the patient's clinical condition deteriorated, oncologists and palliative clinicians both addressed plans for end-of-life care. However, oncologists' notes emphasized discussion of medical complications of radiographic progression of disease, whereas palliative clinicians' notes emphasized the psychosocial impact of worsening health on the patient and family.

The study of these clinic notes is as remarkable for what the authors found as what they did not. Based on the survival findings from the randomized trial, one might have expected early outpatient palliative care encounters to introduce the idea of hospice and establish specific plans for end-of-life care. Notes do not indicate that such discussions took place early in the course of treatment, however. After a clinical deterioration, these issues were addressed by both the oncologist and palliative providers.

A number of issues limit the interpretation of these findings. The study did not include a comparison of outpatient charts of patients in the standard care group. Oncologists in the palliative arm were aware that their patients were in the palliative care group or standard care group, and this may have influenced the content of the oncologists' discussions. Most importantly, we do not know if what was documented in the notes is an accurate reflection of the discussions that occurred. Even if the notes are accurate, the subtlety, emotion, and non-verbal aspects of the encounters are not captured. As described by Atul Gawande, communication about goals of care is a skill that does not come easily to most clinicians, and like other aspects of medicine, requires intense training to learn to do well.⁷ Yet the study by Yoong does not suggest that the issues addressed are out of the realm of most internists. With proper training, perhaps any internist could learn to address these issues. This is an empirically testable hypothesis.

While we are waiting for the science in this area to progress, the general approach described by Yoong et al has such strong face validity that all internists should consider incorporating it in practice. The focus of palliative care is not browbeating patients into accepting hospice and avoiding resuscitation or hospitalization, as some might assume. Palliative care is at its heart a relationship-centered profession. Palliative care clinicians build strong relationships with patients and families in the hopes of finding a way through the symptoms and distress, a way to cope with the illness experience, and a way to make each day as good as it can be. Many patients with serious illness will not have access to an outpatient palliative care clinician, but they will probably have access to an internist. Patients might benefit if all internists incorporated this approach, seeking supplementary training in palliative care. And some benefits, like survival, might be surprising.