The Discussion section can be the most challenging section of a peer-reviewed paper to write. When crafting this section, you are cognizant of the need to convey to the reader why your results are important and what they imply for the future of patient care or treatment. However, it can be difficult to strike the right balance of summary, interpretation, and implication. You can find resources that provide suggestions for a basic structure to follow, and, in general, they suggest something like the following:

  • Summarize key results and interpret their meaning
  • Compare your findings to previous works and discuss similarities and differences
  • Enumerate strengths and limitations and the effect they may have had on observed results
  • Provide a conclusion and suggest next steps for future research

You may be able to find more detailed lists or more specific guidance, but the Discussion section is probably the least structured paper section when you consider the often formulaic structure of the Methods and the detached objectivity required for the Results. Even the Introduction needs a certain structure to it to set the stage for what’s to follow.

Learning to write a good Discussion takes practice and experience, and improvement can come from making mistakes. Luckily, I’ve made more than my share of mistakes, and have been fortunate enough to help a number of others through the bumpy process of crafting a Discussion. From my experience and observation, here are the most common pitfalls authors fall into when drafting this section.

Pitfall 1. Only re-hashing what’s already been said in the Results section.

Because you know that you should summarize your main results, there’s a temptation to simply re-state some or most of the values reported in the Results section. This is so pervasive, in fact, that some journals specifically call it out in their author instructions by cautioning: “don’t just re-state the Results.” While you want to highlight the main results, you also need to provide some interpretation and context so that the reader understands why these results are important or useful. Interweaving results from previously published articles is as much an art as a science, and if not done well it can feel clunky and forced. Often authors, after simply re-stating the Results in one paragraph, will in the next paragraph simply re-state results from other studies, leaving it to the reader to make the connections between them. Don’t assume your reader will make the same connections you will. Start by highlighting a main finding, but then explain why it is important and consider its implications in light of what’s been published previously. When researching the previous studies, don’t focus only on their Results sections – read and consider the interpretation those authors offer within their own Discussion sections.

Pitfall 2. Mentioning something NOT in the Results.

Anything you talk about in the Discussion section should have previously appeared in the Results section. Period. Never introduce new data or results in the Discussion. If it’s important, then it should be presented with the rest of the results. This is true for sub-analyses; if they are necessary for the interpretation you offer in the Discussion then they need to first appear in the Results section.

Pitfall 3. Not placing the results in context.

Interpreting the results requires more than just talking about the specific setting, participants, and situation of your individual study or project. Are your results generalizable to other groups or settings? If you observed results that differ from previous studies, can you explain why? What implications do your results have – if any – for how care should be provided (or, what else needs to be investigated before one can make such claims regarding patient care)?

Here is where understanding your intended audience is key. Consider whether you expect readers to be providers (e.g., physicians, nurses), administrators, policy makers, patients, or something else. Context for a physician (i.e., here’s how these results may inform how you diagnose or treat patients) may be very different than it would be for a policy maker (i.e., here’s how these results may inform how reimbursement can encourage a specific behavior).

Pitfall 4. Overstating strengths or understating limitations

It’s good to be proud of the work that you have done, and it is likely an important contribution to the literature. But you need to be realistic and transparent about both the strengths and the limitations. No one is expecting you to have conducted the perfect study or to have written a seminal paper. Allow the work to stand on its own merits with all of its imperfections and weaknesses. Oh, and just skip the part about your study being “the first one to” do anything. That doesn’t impress reviewers, and the real test of whether a paper is effective and useful is whether it’s read and cited years into the future, not its original publication date.