If a person is diagnosed with a serious illness like dementia, heart failure, or cancer, they may begin to receive a specialized type of medical care called palliative care, which focuses on improving the quality of life of both patients and their caregivers.
Each year, an estimated 56.8 million people are in need of palliative care, the World Health Organization said in 2021. However, despite a growing demand for palliative care due to an aging population, there are gaps in research and evidence within the field, according to a 2020 article published in the Journal of Aging Research.
The need for further research on palliative care is clear — but how can investigators best collect and share their data? That’s when Kathryn Colborn, PhD, MSPH, enters the conversation.
Since 2016, Colborn, an associate professor in the Division of Health Care Policy and Research at the University of Colorado Department of Medicine, has co-directed the Data, Informatics, and Statistics Core (DISC) of the Palliative Care Research Cooperative Group (PCRC), an interdisciplinary research community made up of more than 600 members across the country.
“People do really robust science in palliative care,” Colborn said. “It’s rigorous research but it’s hard because the outcomes can be difficult to define and measure. There are reliable tools for measuring quality of life, but there are a lot of things that happen between an intervention and its impact on quality of life that should be addressed. We often refer to these as mechanisms.”
“Finding those intermediate outcomes on the causal pathway are really important,” she added. “We need to really think about what is causing the effect and how we can best measure that.”
For example, cancer patients experience a lot of pain toward the end of life, she said. A palliative care intervention might first target improving patients’ knowledge of pain coping skills, which would impact the patients’ daily pain scores, and eventually, their quality of life.
Applying her skills as a biostatistician, Colborn co-leads the PCRC’s data repository and helps mentor palliative care investigators on how they can best set up their research and clinical trials to collect meaningful data. As a result of her leadership, Colborn was asked to lead the data coordinating center of a roughly $19 million palliative care research project funded by the Patient-Centered Outcomes Research Institute (PCORI) that will continue through 2028.
“We’ve made a lot of progress on engaging patients to understand what they think is important, and caregivers as well. We’ve figured out how we can design real-world, pragmatic research studies in robust ways and some advances in statistical methods in order to analyze those in robust ways,” said Colborn, who also directs the Biostatistics and Analytics Core at the Adult and Child Center for Outcomes Research and Delivery Science (ACCORDS).
“I think our field has really changed over the last decade, and I believe that the PCRC has been a big part of that,” she added.
More than 14 years ago, a group of senior palliative care investigators, methodologists, and clinicians came together to form the PCRC to help build an evidence base that supports the clinical practice of palliative medicine.
The group officially launched with a meeting at the CU Anschutz Medical Campus in January 2010, according to an article published in the Journal of Palliative Medicine. A host of the meeting was University of Colorado Hospital Chief Medical Officer Jean Kutner, MD, MPH/MSPH, whose research focuses on palliative care and hospice.
“Palliative medicine has made rapid progress in establishing its scientific and clinical legitimacy, yet the evidence base to support clinical practice remains deficient in both the quantity and quality of published studies,” the article said.
The PCRC aimed to address this by applying a cooperative group framework to “help overcome many obstacles to palliative care research, such as lack of standardized definitions and processes, small populations in clinical trial sites, and uneven distribution of expertise,” according to the article.
The PCRC has facilitated or supported many multi-site trials over its 14 year existence, Colborn said. It also developed a library of standardized data elements, which was provided to investigators across the cooperative group to encourage collection of important measures using common definitions, such as self-reported gender, primary patient diagnosis, and self-reported race and ethnicity. The PCRC also hosts a library of serious illness and palliative care patient and caregiver outcome measurement tools that are valid and reliable.
Since its formation, the cooperative group has grown to have more than 600 members and a national network of 220 locations.
“We really help palliative care researchers … jumpstart their careers, because it can be hard when you’re spread across the country and there aren’t a lot of people in that space at your own institution,” said Colborn, who joined the PCRC in 2015.
“We’re providing a resource to help people build their careers in palliative care, do robust, reproducible research, and network with other people working in the field,” she said.
The PCRC offers some grants for pilot studies, gives consultations to investigators who need help developing and/or designing their trial or study, and helps investigators set up multi-site trials.
It also offers clinical trial intensives, which are short workshops in which members of the PCRC teach palliative care investigators how to design their clinical trial and write a grant.
“The mentorship is one of my favorite parts, going to these intensives and answering people’s questions about the design and analysis of their studies,” Colborn said. “It’s a great way to give back and help others who are early in their careers.”
Given the various roles the PCRC serves, it is structured into several different centers and cores, including the Data, Informatics, and Statistics Core (DISC), which Colborn co-leads.
The DISC hosts a de-identified quantitative data repository, which is a hub of valuable information that palliative care investigators can access through the PCRC’s website.
“We have a large data repository of palliative care and serious illness clinical trials, and other types of studies, where people can go online and request to use these data for secondary research purposes,” Colborn said.
Highlighting the value of these types of databases, Colborn said she thinks palliative care investigators should try to use them more often when conducting research.
“In palliative care, there are a lot of studies doing primary data collection — which is great, but it is expensive. It also takes longer to get an answer to a research question,” she said. “There are studies that can be done using our repository data … that are cheaper to do and quicker.”
Colborn’s leadership in the PCRC is a big part of the reason why she was asked by the principal investigators of a roughly $19.2 million research project to lead the project’s data coordinating center.
The research project aims to compare two ways of providing palliative care to improve the quality of life of patients who have acute myeloid leukemia, a type of blood cancer. The multi-year project involves 1,150 adults with high-risk acute myeloid leukemia who are receiving intensive chemotherapy, as well as their caregivers.
“They reached out to us because they wanted the PCRC DISC that I lead — the data, informatics, and statistics core — to support this large trial,” Colborn said.
The study is just one of many Colborn has aided throughout her time at the PCRC. Although it can be time consuming, helping palliative care investigators conduct meaningful research is something Colborn is passionate about.
“It’s such a great population of investigators,” Colborn said about the people she works with. “It feels like a family.”
In addition to hosting the data repository, the DISC offers guidance to palliative care investigators, helping ensure they are appropriately tracking their clinical trials, implementing strong data documentation, and conducting reproducible research, Colborn said. This type of mentorship is valuable, especially given the complexities of the field.
“There are a lot of statistical methods challenges in palliative care research,” Colborn said.
For example, if a palliative care investigator is researching the effectiveness of a certain intervention and wants to look at patients’ quality of life at 12 weeks following the intervention, there is a possibility that some of those patients may die before the investigator gets their final outcome measurement, she said.
“You have to have sophisticated statistical models to try and handle this bias that occurs, because if you just analyze the people who’ve made it to that time point, then you’re only analyzing the healthiest people in your study,” Colborn said.
Another challenge can be designing large, pragmatic cluster randomized trials when palliative care investigators face limited resources, she said. To help investigators develop a statistical analysis plan for randomized trials in palliative care, Colborn helped co-write a paper, published in the Journal of Pain and Symptom Management, that walks them through the process.
These types of challenges underscore the value and importance of having consultations with biostatisticians like Colborn.
“I think we need more collaborative biostatisticians in this world — there aren’t enough of us,” she said. “It'd be great if we could find more people who are interested in going into this field to support these studies.”
SPECIAL NOTE: Much of this work has been possible thanks to Jean Kutner, who has been the multiple principal investigator (MPI) of the PCRC since its formation. The National Institute of Nursing Research has offered several funding renewals for the PCRC to continue its work (funding citations: 1UC4NR012584, 5U2CNR014637, 5U24NR014637.) While the focus of this article was on biostatistics in palliative care research, we want to recognize her essential leadership of the PCRC.