To learn how to read a medical journal article, read the introduction first, then the abstract, methods and results, and finish by checking the authors, funding and conflict of interest disclosures. You do not need to follow every technical detail.
This reading order suits patients checking a claim they met online or in the news. For the wider habit of judging sources, see how to find reliable health information online.
What to have ready
- The full paper, not only a summary. The steps below depend on the introduction, methods, results and disclosure statements, which a news story or abstract page leaves out.
- A pen or notes file. Writing down what the paper says, step by step, keeps your own reading separate from what the authors conclude.
- A realistic goal. The aim is to understand the question, what was done, what was found and who stands to gain. You are not trying to repeat the experiment.
- Time to read in sequence. The order below is deliberate, so avoid jumping straight to the conclusion.
The parts of a research paper
Primary research papers are usually divided into five sections, according to Duke University’s guide for non-scientists. Knowing what each part does tells you what to look for in it.
- Abstract. A dense summary of the entire paper.
- Introduction. The big question and the background to it.
- Methods. What the authors actually did.
- Results. What they found, largely in figures and tables.
- Discussion. What the authors think the findings mean, sometimes labelled conclusions or interpretations.
The order depends on the journal, and some journals also publish extra material online called Supplementary Online Information. If a paper seems to leave out detail, look for those files.
Reading order for a medical study
Step 1: Check what kind of paper it is and who wrote it
A primary research article is a peer-reviewed report of new research on a specific question, so confirm that is what you hold. A review of other studies or an opinion piece follows different rules. Duke’s guide also notes that reputable biomedical journals will be indexed by PubMed, which gives you a quick test of the journal. If you need to look a paper up, how to search pubmed for medical information explains the process.
Then note the authors and their institutional affiliations before reading further. This takes less than a minute. It gives you context for the funding and disclosure checks in the last step.
Step 2: Read the introduction before the abstract
Duke’s guide advises beginning with the introduction, not the abstract. The introduction should identify the big question and give the background. The University of Wisconsin-Milwaukee library guide suggests summarizing that background in five sentences or less. Try it in your own words. If you cannot say what question the researchers set out to answer, reread before moving on, because every later step depends on it.
Step 3: Use the abstract as a check, not a verdict
The abstract is the dense first paragraph at the very beginning of a paper, and it contains a succinct summary of the entire paper. Once you have read the introduction, the abstract becomes a useful test. Does it match the question you wrote down? Does it promise more than the introduction set up?
Treat it as a summary, not as evidence. Reading only the abstract of a study is how many headlines end up overstating what a paper found. Abstracts compress the methods and results into a few lines, so the detail that qualifies a claim is usually missing.
Step 4: Read the methods for who was studied and how
The methods section tells you what the researchers did and to whom. Duke’s guide suggests drawing a diagram for each experiment, showing exactly what the authors did. For a patient reader, a short list works as well. Note who took part, what was done to or measured in them, and what was compared with what.
You do not need to understand the methods in enough detail to replicate the experiment. A basic explanation of the design is enough for a non-expert. The useful question is whether the people studied resemble you or the person you are researching for. If they clearly do not, the findings may not transfer.
Step 5: Read the results, then the p-values
Duke’s guide advises you not to decide what the results mean yet, only to write down what they are. The majority of results are summarized in the figures and tables, so spend time there instead of relying on the paragraph that describes them.
Statistical wording needs particular care. Duke’s guide tells readers to pause any time the words “significant” or “non-significant” are used. In plain language, a p-value is the statistical number that papers testing a null hypothesis use to back those words. Not all research tests a null hypothesis, but when a paper does, identify it. Then check which result each p-value belongs to. A p-value is evidence about one specific comparison in the paper, not a verdict on the whole study.
Step 6: Read the discussion against your notes
Now compare the authors’ interpretation with the list of results you wrote down. Look for places where the discussion claims more than the tables showed. Also check whether the paper’s own limitations are acknowledged. Because section order varies by journal, the discussion may carry a different label, such as conclusions or interpretations.
Step 7: Find the funding and conflict of interest disclosures
Finish with the check many readers skip. The University of Wisconsin-Milwaukee guide tells readers to pay attention to the author and journal affiliations or funding sources for the research. Funding and affiliations can affect how results are framed, so this step shapes how firmly you hold the conclusions from the earlier ones.
Journal policies show what a disclosure is meant to contain. One example is the conflict of interest policy of Insight Surgery. It defines a conflict of interest as financial, personal or professional interests that may influence research. It lists financial relationships such as employment, consulting, honoraria, stock ownership or patents. It also asks authors to include a statement in their manuscript saying whether or not they have anything to disclose.
That policy treats perceived conflicts as equally important as actual ones, and reviewers must disclose theirs too. Other journals set their own rules, so read the policy page of the journal you are reading. A statement saying no conflicts exist is still information. A paper with no statement at all is a prompt to look harder.
After the paper itself, the same questions apply to the website or news story that brought you to it. For that, read how to tell if a health website is reliable.
Checkpoints before you trust a finding
| Reading step | What to confirm | Move on when |
|---|---|---|
| Paper type and authors | A peer-reviewed report of new research, in a journal indexed by PubMed, with authors and affiliations noted | You know who wrote it and where it was published |
| Introduction | The big question and background, summarized in your own words | You can state the question briefly |
| Abstract | The summary matches the question you identified | Nothing in it is a surprise relative to the introduction |
| Methods | Who was studied and what was done, in outline only | You can sketch the design without replicating it |
| Results | Findings listed from tables and figures before interpretation; each use of “significant” or “non-significant” checked | Your list of results is written down |
| Funding and disclosures | Affiliations, funding sources and the conflict of interest statement | You know who paid and who may benefit |
The checkpoints draw on Duke University’s reading guide and the University of Wisconsin-Milwaukee library guide.
If a step stays unclear after a second reading, stop treating the paper as settled and look for other studies or a clinician’s view. A paper that has not passed these checks should not drive a treatment decision on its own.
Frequently asked questions
What does a p-value mean?
It is the statistical number that papers testing a null hypothesis report alongside words such as significant or non-significant. Read it as evidence about one comparison, and check which result it belongs to.
Where do I find who funded a study?
Look at the author and journal affiliations, and search the full text for the funding and conflict of interest statements. Journals such as Insight Surgery require authors to include a disclosure statement in the manuscript.
Do I have to understand every detail of the methods?
No. A basic grasp of what was done is enough for a non-expert, and you do not need to be able to replicate the experiment.
What if a paper seems to be missing information?
Some journals publish additional files called Supplementary Online Information. Check whether the paper has them before deciding the detail is absent.
Is the journal worth trusting?
Reputable biomedical journals are indexed by PubMed, so a quick lookup there is a sensible first test. It does not replace reading the paper’s methods, results and disclosures.