Cancer Studies: A Primer

CURE, Winter 2011, Volume 10, Issue 4

When trying to understand cancer studies, consider these questions.

Ever wonder why one headline says that something causes cancer and the next one says it doesn’t? Most of the time it’s because a news report is based on the results of one study, but scientific conclusions are drawn from looking at a whole collection of work. Otherwise, it’s like trying to learn about the Super Bowl from a story about a single play in the game. Each hike of the ball is important, and some of them lead to touchdowns, but what matters is the score at the end of the game.

When trying to understand cancer studies, ask these questions:

What kind of study is it? Some are better than others at establishing links between a cause and an effect. The weakest type of epidemiological study is a cross-sectional study, which only provides a snapshot in time. A case-control study is better because it compares sick people to healthy ones who have been selected to be otherwise similar in other characteristics. A prospective cohort study is stronger still because it begins before people get cancer. The strongest type of study is a randomized trial, in which people are randomly assigned some type of medical intervention. A systematic review or meta-analysis pools information from several randomized trials and analyzes it to create the broadest possible view of the data.

How big was the study? Bigger is better when it comes to research studies, particularly when we are looking for small differences, but bigger studies cost more to carry out.

How was the analysis conducted? Proper statistical tools should be used. Results should be adjusted for imbalances in the study populations, and there are many other caveats in the field of biostatistics that can af fect a study’s validity and interpretability.

Is there an alternate explanation for the results? Researchers try to account for known causes of cancer when they conduct studies, but they can’t limit or adjust for everything.

What are the numbers? News stories often report how much a given exposure increases or decreases risk. But consider: How big was the risk to begin with? (And this will vary, depending on age, family history and other factors.) Think of it this way: If something doubles your risk of getting struck by lightning, it’s still not very likely to affect you. But if it doubles your risk of having a car crash, it would be a much bigger threat to your health.

Who paid for the study? If an industry that had a stake in the findings funded the research, it doesn’t mean the results are wrong. Nonetheless, it is a factor to consider when trying to put the results in context.

How does it fit with previous research? Maybe the study found an increased risk of cancer, but most of the ones before it have not. If so, see if this research was stronger (or weaker) scientifically than previous research. For more information on understanding all medical news, visit healthnewsreview.org.