We post a lot of links and references to research studies, but it can be difficult to interpret them. Even harder is determining how relevant those studies are. Here are a few things to consider as you look at research on endometriosis.
When you look at research studies, you first want to assess how strong of evidence it is presenting. One quick way to tell is by the type of study. The University of Alabama at Birmingham (2021) has a good list of the types of studies, going from strongest evidence to weakest. In short:
“The most scientific, rigorous study designs are randomized controlled trials, systematic reviews, and meta-analysis. These types of studies are thought to provide stronger levels of evidence because they reduce, but do not eliminate, potential biases and confounders.”
(University of Alabama at Birmingham, 2021)
Bias is any errors in design that might throw off the conclusions. Bias, whether intentional or unintentional, can occur in “data collection, data analysis, interpretation and publication which can cause false conclusions” (Simundic, 2013). Find out more about that here. Confounding factors are any variables not factored into the study that can muddle with the interpretation of the results- “they can suggest there is correlation when in fact there isn’t” (Statistics How To, n.d.). This is important when looking at the conclusions drawn. For instance, in last week’s newsletter about endometriomas, we noted that recurrence rates are hard to tell from studies because their definition of “recurrence” varies widely from study to study (was it pathology from surgery or recurrence of symptoms that might actually be from a related condition?). You also want to look how large the study was, did it include several different variables, was it limited to a specific group, how was the study funded and what ethical considerations are given, etc. One tool for helping to analyze such factors can be found here.
Most studies have a short summary at the beginning that can be helpful. For instance, let’s look at this study from Shigesi et al. (2019) found here. The title itself tells us it is a systematic review and meta-analysis: “The association between endometriosis and autoimmune diseases: a systematic review and meta-analysis”. So, we’re starting off good.
The study then starts with the summary called an abstract- which tells us the background of the problem they want to study, how they performed the study, the outcomes they found, and a discussion on what those outcomes mean. In the background, they state that “an association between endometriosis and autoimmune diseases has been proposed”, so that is what they are studying. Then they detail how they are going to study it- what databases they are using to look for information, what they are going to include or exclude from their study, and the statistical analysis they will utilize. They next present the outcomes: “the studies quantified an association between endometriosis and several autoimmune diseases, including systemic lupus erythematosus (SLE), Sjögren’s syndrome (SS), rheumatoid arthritis (RA), autoimmune thyroid disorder, coeliac disease (CLD), multiple sclerosis (MS), inflammatory bowel disease (IBD), and Addison’s disease”. But wait! There is a caveat they note- “the quality of the evidence was generally poor due to the high risk of bias in the majority of the chosen study designs and statistical analyses” and that “only 5 of the 26 studies could provide high-quality evidence”. Then they present their conclusions of how this information might be useful- “the observed associations between endometriosis and autoimmune diseases suggest that clinicians need to be aware of the potential coexistence of endometriosis and autoimmune diseases when either is diagnosed”. They end by giving suggestions for future research that might strengthen the body of evidence for this. In short, after looking deeper into the association, the evidence available is overall poor so more research is warranted.
This is by no means a comprehensive review of looking at research study but rather a brief intro. There are many available resources on the internet for looking more in depth about how to assess research and draw conclusions from it.
References
Shigesi, N., Kvaskoff, M., Kirtley, S., Feng, Q., Fang, H., Knight, J. C., … & Becker, C. M. (2019). The association between endometriosis and autoimmune diseases: a systematic review and meta-analysis. Human reproduction update, 25(4), 486-503. Retrieved from https://academic.oup.com/humupd/article/25/4/486/5518352?login=true
Simundic, A. M. (2013). Bias in research. Biochemia medica, 23(1), 12-15. Retrieved from https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3900086/
Statistics How To. (n.d.). Confounding Variable: Simple Definition and Example. Retrieved from https://www.statisticshowto.com/experimental-design/confounding-variable/
University of Alabama at Birmingham. (2021). Study Types- Definitions. Retrieved from https://guides.library.uab.edu/ebd/evidencestrength
comments are closed .