Systematic reviews made simple

BMC Cancer recently published a systematic review of previous systematic reviews with statistical analyses and meta-analyses regarding the effects of exercise on cancer-related-fatigue in adult cancer patients and survivors.
Professor George Kelley, Director of the Meta-Analytic Research Group at West Virginia University, discusses this more recent approach to reviewing the literature, the need for this work, the challenges associated with it, and makes further suggestions for conducting this type of study in the future.

Why did you examine the effects of exercise on cancer-related-fatigue in adults?

Cancer-related-fatigue is a major problem reported by both cancer patients and survivors (please see here for a review). While one may intuitively believe that exercise might increase cancer-related fatigue, it may be that exercise actually reduces it. In addition, numerous other benefits may be derived from exercise.

Professor George Kelley, Director of the Meta-Analytic Research Group, at the West Virginia University

What is a systematic review of previous systematic reviews with meta-analyses on exercise and cancer-related fatigue?

Multiple systematic reviews with meta-analysis on exercise and cancer-related fatigue now exist. As a result, there is a need to systematically review these previous reviews to aid decision-makers with the “state-of-the evidence”, provide researchers with direction for future research, and avoid unnecessary and redundant systematic reviews, with or without meta-analysis, on the same topic. One of the first articles written about this approach appeared in BMC Medical Research Methodology.

What did you find?

Of the 16 systematic reviews with meta-analyses that included as many as 3,245 participants, we concluded that a lack of certainty currently exists regarding the benefits of exercise on cancer-related fatigue in adult cancer patient and survivors. This conclusion was based partially on the fact that slightly more than half (52.7%) of the meta-analyses reported statistically significant improvements in cancer-related fatigue as a result of exercise.

While additional research is needed to see what specific subgroups, if any, may possibly benefit, exercise does not appear to increase cancer-related fatigue

While additional research is needed to see what specific subgroups, if any, may possibly benefit, exercise does not appear to increase cancer-related fatigue, something that is probably important given the numerous other benefits that can be derived from exercise.

What were some of the challenges associated with conducting this type of research?

Like any systematic review of previous systematic reviews with or without meta-analysis, one of the major challenges we faced were the subjective decisions that needed to be made with respect to categorizing and interpreting findings. For example, while we talked globally about cancer-related fatigue, different scales consisting of different scoring components and scoring methods were included, thereby potentially confounding findings.

What are some suggestions for best practices when conducting this type of research?

A systematic review of previous systematic reviews, with or without meta-analysis, has also been described as an “umbrella review”, guidelines of which have been previously suggested. However, I prefer to describe this approach as a “systematic review of previous systematic reviews,” conduct analyses above and beyond the aforementioned guidelines, and follow the PRISMA guidelines, where applicable, for the conduct and reporting of such.

Finally, and more broadly, I believe the following stepwise approach to systematically reviewing the literature currently makes sense given the proliferation of information in today’s world:

(1) You have an idea about a systematic review with or without meta-analysis so you develop a protocol to conduct a systematic review of previous systematic reviews, with or without meta-analysis, and submit it to a trial registry such as PROSPERO (see here).

(2) After you receive a registration number from a trial registry such as PROSPERO, you submit a detailed protocol to conduct a systematic review of previous systematic reviews, with or without meta-analysis, for publication consideration in a journal such as Systematic Reviews. I think it is also important for journals such as BMC Cancer to consider publishing such protocols in the future.

(3) After the protocol is published, you conduct a systematic review of previous systematic reviews, with or without meta-analysis, submit it for publication consideration in a journal such as BMC Cancer, and include a decision on whether or not a new or updated systematic review with or without meta-analysis is needed on the topic you’re interested in. This decision could be based on one’s own subjective assessment based on the review, the application of existing formulas, previous guidelines, or some combination.

If a decision is made that a new or updated systematic review, with or without meta-analysis, is needed, you would then:

(4) Develop a protocol to conduct your own systematic review with or without meta-analysis and submit it to a trial registry such as PROSPERO for a registration number.

(5) After you receive a registration number from a trial registry such as PROSPERO, you submit a detailed protocol to conduct your own systematic review, with or without meta-analysis, for publication consideration in a journal such as Systematic Reviews.

(6) After the protocol is published, you conduct your own systematic review, with or without meta-analysis, and submit it for publication consideration in a journal such as BMC Cancer.

The proliferation of information in today’s world now requires new approaches to synthesizing the literature. I am hopeful that others interested in cancer research will be motivated to utilize this approach in their own research endeavors to help advance their respect field(s) from both a research and applied “real-world” perspective.

 

Dr. George A. Kelley, DA, FACSM, is a tenured Full Professor and Director of the Meta-Analytic Research Group in the Department of Biostatistics at West Virginia University (WVU).  He came to WVU in 2002 from Massachusetts General Hospital’s Institute of Health Professions in Boston, MA. Dr. Kelley’s research includes, but is not limited to, use of the meta-analytic approach for examining the effects of exercise and physical activity on health-related disease. He has a consistent record of external funding as a Principal Investigator since 1998, including R01 funding. To date, he has authored more than 100 publications in peer-reviewed journals, more than 80% as first author, i.e., person who did the most work. In addition to his research, Dr. Kelley currently teaches a graduate course in Applied Biostatistics as well as an Introduction to Meta-analysis course. He has received the Distinguished Alumni Award from his Alma Mata, Boise State University as well as the Dean’s Award for Excellence in Research from the WVU School of Medicine. Dr. Kelley has also been a nominee for Teacher of the Year on several occasions as well as the WVU Benedum Distinguished Scholar Award and WVU Chancellor’s Award for Outstanding Achievement. He currently serves as a meta-analytic expert on six different journal’s editorial boards and in addition to his 80+ hour work week and with a belief in modeling the behavior of a Public Health professional, runs 3-7 miles per day, 3 times per week, cycles 20-50 miles per day, 3 times per week, resistance trains 3 times per week, and walks 5-7 times per week. He has been tracking his exercise and nutrition intake since 1980.

View the latest posts on the BMC Series blog homepage

Comments