Understanding Medical Studies So You Can Get the Prostate Cancer Treatment That’s Right for You
by Gerald Chodak, MD
As you learn about the options for treating your prostate cancer, you may turn to reviewing medical studies. They will tell you the proportion of men who are alive, have a stable prostate-specific antigen (PSA), or did not develop widespread cancer several years after being treated. Because they are all published in medical journals, you probably will assume that they must be well done and the results must be reliable. Unfortunately, that is far from the truth. The fact is, a great many of the studies that are published are not well done. That means the results are unreliable. Without understanding what makes a study good or bad and how results are reported, you run the risk of not getting the treatment that is right for you.
Medical studies are done in different ways. The most common types used for prostate cancer are the following:
- A prospective, randomized, controlled study
- A prospective cohort study
- A case control, retrospective study
- An epidemiological study
Randomly assigning people to their treatment reduces the chance that the results will be biased or incorrect.
Prospective, Randomized, Controlled
This type of study is by far the best because it gives the most reliable information. To understand the reason, you must first know what each word means.
- Prospective means the study was carefully designed before anyone was treated.
- Randomized means that at least two treatments are being compared to each other, and neither the participant nor the doctor can choose what treatment to receive. A computer will make this decision. Randomly assigning people to their treatment reduces the chance that the results will be biased or incorrect.
- Controlled means only people with specific characteristics are allowed to enter the study, and everyone is supposed to be managed in the same way. They all get the same tests before and after they are treated, and the treatments are standardized. For example, if surgery is being compared to external radiation, every person in the radiation group gets the same amount of radiation, and it is delivered using the same technique.
When done properly, a prospective, randomized, controlled study is the only reliable way to prove that one treatment is as good, better, or worse than another treatment. All new drugs must be tested in this way to gain approval from the Food and Drug Administration. Unfortunately, prospective, randomized, controlled studies are very expensive to do, and they can take many years to complete. Without these studies, doctors are unable to determine which therapy is best.
A cohort study also is prospective, but it differs from a randomized study in that all the individuals get to choose their treatment. For example, if a doctor wants to study the effect of a radical prostatectomy, any man healthy enough to undergo an operation can decide to participate. Specific information about each person is recorded, and then they are followed to see what happens to them. The results are compared with other treatments published in medical journals.
Cohort studies are easier to do than prospective, randomized, controlled studies, which is partly the reason why more of these studies are done. The problem with comparing cohort studies is that the results can be very biased, leading to incorrect conclusions. Using these results to compare different treatments may suggest that they are similar, but only a prospective, randomized study can prove if that is true.
Retrospective, Case Control Study
This study design collects information from medical records about men who had their treatment sometime in the past. The results are then analyzed and compared to other studies. The value of retrospective studies is that they are relatively easy to do, are not very costly, and can give immediate information without having to wait many years to get results. Unfortunately, retrospective studies often lead to incorrect conclusions for several reasons:
- There is no way to be sure that the information entered into the chart is correct. Errors are common, and there is no way to correct them.
- Because the study was not prospective, everyone may not have been treated in exactly the same way. For example, in a retrospective study of men having a radical prostatectomy, some of them also may have received radiation or hormone therapy. Combining them with men only having surgery will distort the results.
- Some men having the treatment may be excluded because of missing information. This creates a selection bias, which can make the results appear better than if all the men treated were included.
- Allowing men to select their treatment also can create a bias, making the results appear better than if a randomized study was done.
- When the results of two retrospective studies are compared, the characteristics of the participants rarely are the same, which can lead to misleading conclusions. It would be like trying to compare apples to oranges to decide which fruit is sweeter.
Because of these potential weaknesses, the results of case control, retrospective studies cannot be used to make reliable conclusions. This does not mean that the results are definitely wrong, but there is no way to be sure that they are correct. For that reason, you should use caution when told about the results of a retrospective study.
This also is a retrospective look at what happened to a group of individuals treated in the past. Information is collected from the medical records and then fed into a computer. It searches for anything that separates individuals into those who did well and those who did poorly. For example, consider a group of men who were all treated in the past by external radiation. When the results are analyzed, the study may find a higher survival rate in men who took vitamin C compared to those who did not. The study would then conclude that vitamin C improves survival of men having external radiation.
Is this conclusion valid? Does it mean that men having radiation should be advised to take vitamin C to improve their survival? The answer is no, it is not a valid conclusion or recommendation. Epidemiological studies can suggest vitamin C may be beneficial, but only a prospective, randomized, controlled trial can prove if that is true. There can be many other explanations why some men did better than others that have nothing to do with taking this vitamin.
Epidemiological studies are much easier to do than prospective, randomized, controlled studies, and they are often mentioned in the news. The problem is that the results are made to appear reliable, and the weakness of the study design is rarely reported. You should be very cautious when making decisions based on epidemiological studies.
The Bottom Line
The main message here is that some results reported to you come from well-designed studies, and other results come from poorly designed studies. The better the study design, the more you can trust the results.
♦ ♦ ♦ ♦ ♦
Dr. Gerald Chodak performed the first laparoscopic lymph node dissection for prostate cancer in Sweden, Finland, Norway, and Japan. He helped form the first chapter of Us TOO, an international support group for men with prostate cancer and their significant others. Currently, he produces a weekly video for Medscape.com called “Controversies in Urology” and maintains a prostate educational website at ProstateVideos.com.
Excerpted with permission from Winning the Battle Against Prostate Cancer, by Gerald Chodak, MD, copyright © 2011 by Gerald Chodak. All rights reserved. demoshealth.com
This article was originally published in Coping® with Cancer magazine, July/August 2012.