Reading Primary Research

The promise of new treatments and cures excites our imaginations, especially when we’re looking to help a loved one or ourselves. Information on rare diseases is so hard to come by that any news on a potential discovery is welcome. But how accurate is that news? At a time when more information lies at our fingertips than ever before, healthcare professionals, patients, and caregivers have to be able to sift through all the claims to find the facts.

Where is this information coming from?

Read the research article behind a news report. Because a news story often focuses on what’s new and sensational, the conclusions of the research can be overstated. The results of a study are rarely conclusive on their own—interesting results that may lead to a treatment require many more studies before the information can be considered clinically relevant.

See who authored the original research. Note the first author’s name. In most cases, the first author is usually the graduate student, senior researcher, or study staff who did the most work between designing the study, performing the experiment, and writing up the results. Research also builds on the work done by others—reading citations at the end of an article may be helpful in finding related studies. Scientific leaders are often have their work cited in many articles.   

Look for research published in peer-reviewed scientific journals. Before publishing an article in a peer-reviewed journal, experts who work in the same field as the authors evaluate the research for quality. The most popular and influential medical journals include The New England Journal of Medicine, JAMA: The Journal of the American Medical Association, and BMJ: The British Medical Journal

Note any conflicts of interest the author might have. Who funded the study? Did the organization funding the research have something to gain from the results? Ideally, all research would be objective. However, personal bias can even influence the study’s design. Does it seem like the author had any underlying motivations that might have led them to a particular result? This may be an issue when looking at posters or abstracts submitted to conferences but is unusual in papers successfully published in peer-reviewed journals. Reading the abstract and the introduction of the paper tells you about the question the researcher wants to answer and any motivation for the study.

Also make sure the research is current by checking the date of its publication. Medicine is a rapidly-changing field, so an article can become outdated easily. Google Scholar is a useful tool to see if a study has been cited by more recent research. The technology used in an experiment determines how easily results may be out of date: techniques used in behavioral and pharmacological research from twenty years ago can still be relevant, but techniques used in genetic or fMRI studies just a few years ago are no longer appropriate.

Once you have determined that the research is credible, it’s time to dig deeper into the article. Not all studies are created equal.

How was the study designed?

The incorporation of research findings into clinical practice can only come after many studies, trials, and reviews have taken place. Each finding adds to a body of knowledge, ultimately resulting in proven recommendations for healthcare professionals.  Evidence-based medicine depends on strong, well-designed research that has been conducted using the scientific method.

More an ongoing cycle than a finite list, the scientific method gives researchers a procedure that helps them learn more about the natural world. When following the scientific method, researchers go through the following steps: 

  1. Asking a Question

A question serves as a prompt for research. It can be open-ended, or it can come from an observation. In this step, a researcher looks up background information and previous studies on the topic.

             2. Making a Hypothesis

After reading more about the topic, the researcher makes a hypothesis, an educated prediction that could answer the question. Statistical hypotheses, which focus on a population, are common in clinical research. A statistical hypothesis has two parts: a null hypothesis, H0, which assumes that any differences between the control and test groups are not due to the experimental manipulation, and an alternative hypothesis Ha, which guesses that a predicted, non-coincidental effect on the groups will be observed during the experiment. If the null hypothesis is found to be false, it doesn’t mean that the alternative hypothesis is true. Clinical research is based on determining whether phenomena do or do not fall under the null hypothesis.

3. Designing an Experiment

An experiment’s design is based on controlling some variables (independent variables) and measuring other variables that might be affected if the null hypothesis H0 is false (dependent variables).  Dependent variables depend on the controlled independent variables.

             4. Performing the Experiment

This is the study itself. Most clinical trials testing medical treatments are double-blind to add an extra element of control. In a double-blind trial, neither the researchers nor the experiment’s subjects know whether a patient was part of the control group, receiving a placebo, or the test group, receiving the treatment. The researchers only find out a subject’s group when they analyze the results after the study is over.

5. Analyzing the Results

Researchers collect data during an experiment that can be quantitative, which can be measured with numbers, or qualitative, which can be described. This data is combined and analyzed using statistical methods. The results of a study often lead to related experiments that expand knowledge on the topic incrementally.

When reading about a study, pay attention to how the researcher designed the experiment, why they did so, and how the experiment was performed. This information can be found in the “methods” section of the article. What were the controlled variables, and were they consistent? What was the independent variable, and how was it measured? Was the study a double-blind trial? Also, how did the researchers interpret the results? In order to draw strong conclusions from the results of an experiment, those results have to be analyzed for relevance and accuracy.

What do the different parts of a research article tell us?

Research articles generally follow the same structure. Each paper is divided into sections that explain different aspects of the research—these sections are:

1. Abstract

The abstract is a short paragraph or two that summarizes the entire paper.

2. Introduction

The introduction gives background information on the topic that the study addresses. A review of related research may be included here.

3. Methods

The methods section describes how the study—an experiment—was designed. This way, the study can be repeated to see if the results are reliable.

4. Results

The results section focuses on what was observed during the study. Exact facts and figures can be found here.

Note: Calculating a P value helps researchers determine whether or not there was an observable change in the experimental population as a group. P values are compared to a significance level α, which is an arbitrary number, and most clinical researchers call results “statistically significant” when the P value is less than a significance level α  of 0.05. When the P value is less than the chosen significance level, then H0 is usually rejected. That doesn’t immediately mean the difference between the groups is important, though—such a determination requires further context.

5. Discussion

The discussion section explores how relevant the results are to the question at hand as well as any implications that arise from these findings. This section also addresses how the results relate to existing knowledge and how they can inform future research to explore new ideas.

6. References

Any articles that the researchers cited can be found here.

Knowing the different parts of a research paper make it easier to find specific information. For instance, if we wanted to see how the control and experimental groups were divided, we would look at the methods section. Or, if we wanted to see a table of figures that showed measurements taken after the study, we would look at the results section.

When looking for findings that could promise a better quality of life for anyone dealing with a disease, we want to focus on clinically relevant breakthroughs in medical science. In order to do that, we need to dig in deep to the articles that announce these discoveries. It may be challenging at first, but reading primary research articles and understanding their implications is a skill that becomes easier with practice. In learning more about how research is conducted and reported, we can better recognize promising findings out of a sea of data.