Continued deficits in the evidence base for palliative care in oncology

An unusual and sobering study on the state of the “palliative oncology literature” has recently been published online. Searching 6 bibliographic databases, this group from M.D. Anderson undertook a massive review of the palliative oncology literature, comparing from 2004 and 2009 the number of articles, proportion of all oncology articles, topics, and study designs.

The paper begins by highlighting the familiar barriers to palliative care research: limited research funding, few personnel trained in palliative care research, difficulty in recruiting and retaining patients/subjects, methodologic issues. They also pointed out that there still isn’t a consensus taxonomy and classification system for palliative care literature (can you tell they had professional research librarians on the teams?). An issue in designing this study is the great complexity and diversity of the palliative care literature. Here’s a sentence I loved: “Unlike other disciplines, palliative care transects numerous domains.” Ever seen ‘transect’ in a palliative care paper? The number of clinical issues, populations (multiple subsets), disciplines involved, and something they didn’t mention, the diversity of publication titles, makes the task of reviewing the whole of the literature for a specific major disease class pretty overwhelming to contemplate. They admitted, in a bit of understatement, that it was a “labor intensive” project.

I’ll skip over the study methods and jump to the results. Combining the 2 periods,

• Of over 6000 articles screened 1213 were reviewed. A significant proportion of studies were excluded from this review because they involved topics such as palliative chemotherapy with survival as the primary endpoint
• 70% of papers were original studies
• 42% of the studies were published in palliative care journals while 19% appeared in oncology journals
• Over 400 journal titles had at least one palliative oncology paper.
• “The palliative oncology literature is flooded with descriptive studies when we urgently need more practice-changing analytic studies.”
• Physical symptoms, health services research and psychosocial issues were the most common topics. There were many “orphan” (under-studied) topics even under physical symptoms).
• Pediatric palliative care was “largely unexplored.”
• Providers and lay caregivers are under-studied
  

Over the 5-year study period:

• The proportion of palliative care articles in the oncology literature remained below 1% and even decreased slightly, despite an increase in absolute numbers of papers.
• There was a 47% increase in original studies.
• The percent of randomized controlled trials decreased from 7% to 5%

Overall, the results show a disappointing lack of improvement in the overall quality and quantity of the palliative care articles focused on the oncology population. This is inherently limiting in addressing the total palliative care literature, but it would be time and cost prohibitive to do the same analysis over that larger universe. The barriers to doing palliative care research, especially improving the evidence base for treatments, haven’t changed and, despite increases in overall numbers of papers, methodology remains weak. Perhaps this paper will help to focus attention on improving the science of palliative care research.

By the way, articles in The Oncologist have recently been made freely available (again) with registration to individuals.


Hui, D., Parsons, H., Damani, S., Fulton, S., Liu, J., Evans, A., De La Cruz, M., & Bruera, E. (2011). Quantity, Design, and Scope of the Palliative Oncology Literature The Oncologist DOI: 10.1634/theoncologist.2010-0397