Understanding Scientific Research
If you are interested in reading scholarly or scientific articles about prostate cancer and treatments, two good places to start are Google Scholar and PubMed. These databases feature articles from peer-reviewed journals.
There are a number of different types of articles that can be found on these databases. Review articles tend to summarize what is currently known about a disease with focus on causes, diagnosis or treatment. Original research articles report on the outcome of experiments designed to answer a specific research question. Case reports describe the clinical experience of a single patient or a very small group of patients that often have complex disease characteristics.
When looking at scholarly articles, it is important to keep the following in mind:
- Randomized controlled trials (RCTs) are the gold standard of clinical investigations. The word “randomized” refers to the fact that the patients do not choose which treatment they receive, and neither do their doctors which is important for reducing biases in the results. The word “controlled” means that there is a control group of patients who receive a mock treatment or placebo, which is important for distinguishing whether the outcome is actually a result of the treatment under investigation. If the study does not present the data for a control group, there is no way of knowing whether their outcome is positive or negative. Ideally, the RCT should be “double-blind” which means that neither the patient nor the investigators are aware of which treatment the patient is receiving. This is especially pertinent to the testing of new pharmaceuticals, where the neither party should be aware of whether the drug or the placebo is given.
- Statistical significance is important when drawing conclusions and you don’t need to be a statistician to understand what you are reading. In many cases, investigators will establish a P-value prior to performing an experiment with a low P-value being a greater significance. In clinical research the P-value is often 0.05, which means that if P<0.05, the investigators can be 95% confident that their findings did not occur by chance. In scientific papers, a statistically significant result will often be denoted with an asterisk or other small symbol, or conversely the P-value will be reported. A statistically insignificant result should be viewed merely as a trend, but keep in mind that the result could have occurred by chance.
- The number of individuals that were investigated in order to gather data and generate conclusions is known as the sample size and is often represented with the letter “n”. The n value should be reported in all reputable articles. Another thing to keep in mind is whether the sample tested by the investigators is representative of the population that the conclusions apply to. For example, if the investigators are testing a new drug to be given to men with advanced prostate cancer and claim that it is more effective than the existing drugs, they should test their new drug with a population of men with advanced prostate cancer and use controls that are similar to the individuals in the experimental group in every respect except for the treatment they receive. These variables will often be reported in a table near the beginning of the article. If there are statistically significant differences between experimental and control groups, the results should be viewed cautiously.
- The outcome measures should be appropriate for the conclusions that the investigators draw. There are a number of different measures of how effective a treatment is. For example, if the investigators are interested in cancer control, an appropriate measure could be prostate cancer recurrence or PSA recurrence. If the investigator is interested in quality of life outcomes between different treatments, health related quality of life scores would be appropriate. Consider which outcome measures are being used and evaluate whether or not they are appropriate for the conclusions being drawn. Also, the outcome measures should be reported for a sufficient length of time following treatment.
- Keep in mind that correlation is not the same as causation and just because two variables appear to be related does not mean that there is a cause and effect relationship. For example, if a paper reports that one treatment is associated with better outcomes than another treatment it might not actually be a result of the treatment, but some other factor. For example, if Treatment A is found to have a better cure rate than Treatment B, it might be because patients who select treatment B have more advanced prostate cancer and tend to relapse more.
- The date of publication is also an important consideration. Cancer treatments are constantly changing, so what might have been true five or ten years ago might have been disproven by the time you read the article. Also keep an eye out for updates or comments that have subsequently been published.
Remember, if something you read doesn’t make sense you can always ask your doctor about it. In addition, keep in mind that even scholarly sources can have biases and if you read something that seems too good to be true, it probably is. Finally, the experience of the groups of people being reported might not apply to your own situation depending on the stage and grade of your cancer as well as other factors that your doctors will take into consideration.