How To Read Medical Research

There is plenty of free info on the web about Crohn’s and Ulcerative Colitis, but much of it is sketchy — at best. Meanwhile most of the good science is locked up in medical journals. How can patients read medical research to help better educate themselves and be more active in their care? Here are some tips, based on how I read articles to share on the Crohnology blog.

Finding Basic Info

Medical research isn’t the only good source of info, but where you get your info is important. For free basic info, I like the Mayo Clinic, HealthLine, the CCFA, and — of course — Crohnology. Most info on these sites will be written for average readers, not medical specialists.

One thing I watch for is whether the articles have been medically reviewed and how recently. For example, this Cleveland Clinic article on IBD genetics has not been updated since 2007; it’s good information, but there have been exciting discoveries in the last several years. I also avoid content-mill sites; some notorious ones include CancerFactSheet.org and Colitis.Ehoow.net.

I often find good info on health news services (e.g. MedCity News). However, many of these sites only report press releases, and the problem is that press releases are rarely written by medical professionals. For example, here’s one headline announcing a widely reported study: “Vitamin D Protects Against Inflammation in Crohn’s Disease“. I first read that to say the study showed that Vitamin D helped patients with Crohn’s feel better, but then found this in the (free) journal paper itself: “we observed no significant changes in markers of disease activity, namely CDAI, CRP, FC or QoL”. That is, there little evidence that Vitamin D supplements helped anyone feel better. The press release inflated the results of the study — which is pretty common.

Finding Medical Research

When I am looking for information to share with my doctor, the above sources usually do not fly. Many physicians, unfortunately, take a dim view of ‘Dr. Google’ — that is, patients researching their care online. What the doctors can’t argue with, however, is peer-reviewed research.

Two good resources for finding peer-reviewed research are Google Scholar and PubMed. Google Scholar looks at academic research generally, while PubMed focuses specifically on medical research. Both usually get me, at minimum, an abstract summary of any paper I want to read, or a good selection of papers on a given topic.

While some articles are free — in many cases required by the US government for research it funds — medical research is still too bottled up in expensive journals. I never buy the article — or, at least, only do so as a last resort. If PubMed and Google Scholar do not get you what you need, you can try to gain access through a university or hospital library system. Or, if you have an awesome doctor, they might download it for you.

But I usually like to email the author directly for an ‘offprint’. Although academic journals cost a lot, none of that money goes to the author. So the author has no incentive not to share their work for free, as long as I promise not to distribute it. Finding an email address can be tricky, but it almost always nets me a free copy of the article I want. As a bonus, if I have specific questions about the article, many authors will gladly answer them.

Reading Medical Research

Once you have the article, how can you make sense of it? Medical articles are written for people with extensive training in science and medicine, and often use lots of obscure words and acronyms. I have some pointless grad school behind me, but — fortunately — you do not need to know all the vocab to read a research article effectively.

Medical papers usually have a regular structure, and navigating that structure helps me read papers more quickly and effectively. Let’s go back to the Vitamin D paper — and to be scientific, we’ll call it Rafferty et al. 2015. — which has the following parts: an abstract; an introduction; materials & methods; results; discussion. This is pretty standard, but some papers will have an additional section called ‘conclusions’.

Here is the order I read those parts in:

  •  Abstract: I read this quickly, and don’t worry if it doesn’t make sense. Because abstracts have to be short, this is where the medical jargon gets the most concentrated.
  • Introduction: usually the last paragraph in the introduction is the most important, so I skip to there. For Rafferty et al 2015, it’s this part:

    We propose that vitD therapy may alter IP and LL-37 concentrations in CD, which may be associated with changes in disease activity markers. The aim of the current study was to investigate the effect of 3 mos. vitD supplementation on IP, LL-37 and markers of disease activity, namely C-reactive protein (CRP), calprotectin, CDAI and quality of life (QoL).

    That’s where they explain what the main goal of the paper is. Sometimes it’s buried more deeply, but I need a good idea of what the researchers are researching before I go any further.

  • Discussion/Conclusions: once I know what the goal of the project was, I skip to the end and find out whether they met that goal. Sometimes this info will be in the abstract, but not always — and usually the discussion/conclusions will lay it out more clearly than the abstract. The earlier quote about — “we observed no significant changes in markers of disease activity” — was from the discussion section.
  • Results: once I know whether they met their goal, it is easier to parse the results of a study. In this particular study, they report lots of p-values, which is a measure of whether a result is due to random chance; here’s a good clear explanation.
  • So where Rafferty et al. 2015 say that “no significant difference in disease activity markers, namely CDAI (p = 0.568)”, that doesn’t mean that 56.8% of patients saw a difference in CDAI, but instead that the result is only 56.8% likely to not be random (p-values of .05 — or 95% — are the usual minimum for ‘significant’ results). It’s a little frustrating because they don’t actually report how many patients saw an improvement in their CDAI, except in a graph.

    Another way that medical journals often report results is something called the odds ratio, or ‘OR’. This is also a source of confusion. If a study reports side effects with an odds ratio, then it means the odds that people taking the med will have the side effects, divided by the odds that people not taking the med will have the side effect. So for example, if the odds of vomiting in the placebo group are 10/100, and the odds of vomiting in the group taking the medicine are 16/100, then the odds ratio is (16/100)/(10/100) = 1.6.

    The trick to odds ratios is that they do not tell how likely the effect is, only how much more likely it. If the side effect only hit 32 in 10,000 people on the medicine, and 20 in 10,000 in the placebo group, that is the same odds ratio: 1.6 (sorry, lots of math). Often an odds ratio of 1.60 is reported as “a 60% higher risk” — but higher than what? And, unfortunately, in many papers they only report the odds ratio, which means I can’t answer that question. Again, it can be frustrating if you’re not looking for it.

  • Methods and Materials: This is the most science-y part, but it doesn’t really matter. Unless you work in a research lab, it’s really hard to evaluate or even use this information (unless it’s talking about survey research). I usually skim or skip this part entirely. Your physicians probably won’t read it; mine don’t.

I give myself plenty of time to process an article like this; it makes more sense the second or third time through. I usually spend some time googling words or acronyms that I don’t know to help unpack it.

If you can’t make sense of a paper, you can always email questions to the author. You can also take it to your physician — who should be happy to help you read it — or email us at social (at) crohnology.com , and I will try to help.

Share on FacebookTweet about this on TwitterShare on RedditShare on Google+Email this to someone