How Are New Cancer Treatments Tested?

We'd like to believe scientists have rigorously studied all tests and treatments and proven beyond a doubt that they are safe and effective. Unfortunately, medical research is complex and drawing definitive conclusions from study results is not always straightforward.

To be deemed reliable, treatment testing results, and any claims about the effectiveness of a treatment, must pass the fair test rule. The fair test rule means researchers must obtain reliable information about treatment effects by reducing biases and taking into consideration the extent to which chance may affect study results. Biases are distortions. For example, a researcher may expect a certain outcome and unintentionally influence the study to produce that outcome.

Improving Testing Results

Progress in science generally occurs in small, incremental steps, not in earth-shattering breakthroughs. One study is rarely reliable for fair treatment comparisons. Researchers should assess all studies alongside evidence from other, similar studies to evaluate if current results support or dispute earlier findings. When you read new medical announcements, if something new sounds too good to be true, it probably is.

One way to eliminate distortions is to compare like to like. Comparing similar groups helps ensure the only real difference in results is due to the treatment in question. Larger studies also reduce the possibility that chance may skew results.

Systematic reviews and meta-analyses evaluate the methods, findings, and conclusions of multiple studies on a given topic. This helps reduce unintended bias and weeds out less reliable study results. For example, according to H. Gilbert Welch, MD, author of Should I Be Tested for Cancer?, in 2001, the Cochrane Collaboration (which he describes as a sort of Consumer Reports for medicine) did a systematic review of all the mammography trials to date and found that many of them to be substandard.

Welch says medical research is not objective and distortions in data can occur at any stage in the research. Furthermore, researchers are naturally interested in gaining attention and, with that, funding for additional research. Medical research tends to favor new tests and treatments, so scientists may exaggerate anticipated benefits and minimize potential risks. You've seen how patients, medical professionals, and patient advocacy groups generally embrace new treatment alternatives. However, a new treatment, or more of a certain treatment, is not necessarily better. There can also be unexpected negative effects.

By understanding what might influence study results, you can ask your oncologist educated questions so you feel comfortable about accepting the cancer treatment she recommends.

 

Sources:

Evans, Imogen, Thornton, Hazel, Chalmers, Iain, and Glasziou, Paul. "Testing Treatment, 2nd edition." Web. 2011.  (http://www.testingtreatments.org/wp-content/uploads/2011/10/TT-interactive-optimised.pdf)

The James Lind Library. "Should I Be Tested for Cancer?" Web.
http://oldjll.sustainabilityforhealth.org/index.html

Welch, Gilbert H. Should You be tested for cancer? Maybe not and here's why. Berkeley and Los Angeles, California, University of California Press, 2004.