A Checklist for Communicating Science and Health Research to the Public

Thinkstock / Peshkova

As science and health communicators, our main goal is to share our institutions’ wealth of science and health knowledge. We strive to make the information accessible to a broad range of people — from scientists and health professionals to health educators to patients and the general public. By pooling the experience and advice from experts in our community, we’ve started a list of strategies for communicating science and health research to the public.

As a growing community effort, we welcome your comments, advice, and additions to the list. We hope to hear from you, whether you’re new or an experienced communications professional. Please share your thoughts with us by sending an email to sciencehealthandpublictrust@mail.nih.gov.

Setting the stage

• Explain why the research is important. When choosing a scientific or health study to highlight, consider what’s being contributed to the research field. Does it represent a major advance? Does it change the way people think about a problem?
• Provide perspective for the study. Explain the context of the research and how it fits into what we already know.
• Take care to not overstate the importance or statistical significance of a study, finding, or emerging situation when relaying what’s interesting or exciting about a scientific development.
• Use clear language to describe the science. Use nontechnical words when possible, and define jargon when needed. Use as few words as possible to describe health messages or scientific concepts.
• Use conditional language when appropriate (language that hedges or highlights the potential gaps or unknowns). Researchers often use such language to convey their best educated explanation of the data, while noting what still needs to be addressed.
• Choose quotes from scientific experts that explain the significance of the research in nontechnical terms. If a quote reaches beyond the study’s data itself, place it in more context with a balanced explanation.
• Convey information in a respectful tone that doesn’t stigmatize or assign blame to individuals or groups involved or affected by the study or disease/disorder.
• Carefully review the chosen language. Some words and phrases — such as miracle, hope, breakthrough, game-changer, and paradigm shift — may give false hope. Others — like victim, affliction, or suffer — make assumptions about how people are affected. These are unhelpful and potentially offensive.
• Match images to your text with care. Empower the intended audience by using accurate, informative, and engaging visuals.
• Provide resources for additional education on complex topics. Include links to credible health information webpages.
• Provide information for locating the original study. Consider linking to paper citations on PubMed, which includes abstracts and links where people can access the studies, if available to the public.
• Note funding sources and potential conflicts of interest to help maintain transparency: patent applications, start-up companies, potential royalty payments, or associations/funding from biotech/pharmaceutical companies.

Describing the science

• Identify whether the study involved human or other model systems. If an animal model, clearly identify the type of animal at the top of the story — in the headline, the lede, or in bullet point summaries when applicable.
• Explicitly state whether the study shows an association or causation. An association is a relationship, or correlation. A positive association means as one goes up, so does the other. A negative association means as one goes up, the other goes down. Causation is when an event or variable is shown to cause a specific outcome. Whether a study shows association or causation depends on the study design.
• If a human study, describe the type of study:
  • Observational studies — studies in which researchers do not carry out any interventions. Observational studies can find associations. The causes might not be clear and could only be a coincidence.
  • Randomized controlled trials — controlled experiments designed to test the effects of an intervention or treatment. A well-designed randomized controlled trial is the best way to establish causation.
• For clinical trials:
  • Describe the clinical trial phase (I, II, III, or IV). Even if you don’t explicitly state the phase, be clear about how far along the drug or treatment is in development.
  • Give a breakdown of the study participants’ demographics, e.g. sex, gender, age range, race, and/or ethnicity when applicable.

• Identify and explain the limitations of the study’s endpoint — the outcome that is used to measure the effect of a drug or treatment. Be alert to:
  • Surrogate endpoints — indicators or signs used in place of another to tell more quickly if a treatment works. For example, a study may measure a biomarker previously linked to death rates, rather than directly measure rates of death.
  • Composite endpoints — individual endpoints that are combined together (e.g. hospitalization + death). These may be used to increase the statistical power of a study, but be careful when describing exactly what they show.
• Check the study’s sample size, methodology, and potential limitations. Given the study design, was the sample size large enough to draw firm conclusions? Were essential design elements of the study clearly communicated? Are the results consistent with previous studies? Were the results preliminary, necessitating a larger follow-up study? Ideally, the peer reviewers will ensure these are stated in the paper.
• Were there any confounding factors and how were they addressed?
• Clearly explain risk, or the estimated likelihood of a certain outcome. Put new risks in perspective. Include the values for both absolute risk and relative risk.
  • Absolute risk — the measure of the risk of a certain event happening. This is usually expressed as the number of cases within a specific group in a certain time period.
  • Relative risk/risk ratio — the measure of the risk of a certain event happening in one group compared to the risk of the same event happening in another group. Relative risk numbers can be misleading in certain situations. For example, if one group has a 0.2% chance of getting a disease and the other has a 0.3% chance, group two has 50% greater risk than group one (or group one a 33% lower risk), but the difference is still not very meaningful.
• Discuss both the benefits and drawbacks of any potential treatment, as health care decisions must take many different factors into account, e.g. treatment effectiveness, side effects, and overall risk of the intervention.