Study finds more than 2/3 of guidelines don't adequately explain risks, benefits
Listen to an interview with Dr. Fagerlin
With three sets of breast cancer screening guidelines giving conflicting sets of recommendations, it's no wonder that patients and physicians are confused. A new study shows that adding to the confusion are the guidelines themselves. More often than not, cancer screening guidelines either leave out important information about the benefits, or harms, of certain recommendations, or presents information in a biased way.
For example, the study found that one set of guidelines on mammography screening cited both the potential benefits (cancer prevention) and harms (overtreatment) in a balanced way, and presented data on the rate of occurrence of both. Another set of guidelines cited the benefits and harms but did not present supporting evidence. A third cited only potential benefits, without mentioning any harms.
"Physicians and patients are going to go to the guidelines to see what is the benefit and harm of being screened so they can make the decision that is right for them," says senior author Angela Fagerlin, Ph.D. She recently joined the University of Utah School of Medicine as chair of the new Department of Population Health Sciences and is also an investigator at the Huntsman Cancer Institute. "When guidelines don't present the information in a clear way, people can't make informed decisions."
The trend is prevalent among screening and prevention guidelines for a number of types of cancers, says the new report. Among the 55 professional guidelines that were examined - including mammography screening for breast cancer, PSA testing for prostate cancer, colonoscopy for colon cancer, and HPV vaccination for cervical cancer - 69 percent failed to present benefits and harms in comparable terms.
Guidelines were from agencies such as the U.S. Preventive Services Task Force, the American Cancer Society and the National Comprehensive Cancer Network. The research was published in the Journal of the National Cancer Institute
"Patients rely on doctors to base screening recommendations on sound judgment as to whether, in their individual context, the benefits outweigh the harms," says first author Tanner J. Caverly, M.D., MPH, clinical lecturer of internal medicine at the University of Michigan Medical School and a research investigator at the VA Ann Arbor Healthcare System Center for Clinical Management Research. "For this judgment, doctors need a clear understanding, in quantitative and comparable terms, about the degree of potential benefit and the degree of potential harm."
The researchers found that 31 percent of guidelines presented absolute effects for both benefits and harms. Meanwhile, 55 percent presented the tradeoffs unevenly, explaining the benefit but not the harm, and 15 percent either didn't quantify the evidence behind the recommendations or presented statistics in a biased manner.
For example, some studies used relative risk reduction, often a larger number, to show benefit – people who were screened were twice as likely to survive. But they used absolute risk, often a much smaller number, to demonstrate harm – 1 percent of people screened had a false diagnosis.
"By unevenly presenting the statistics, guidelines can unintentionally give a false picture of the benefits and harms. The relative risk of "twice as likely" sounds significant, but in terms of absolute risk it might represent a very small difference," says Fagerlin, also a research scientist at the VA Salt Lake City Health Care System.
The researchers suggest that guidelines should quantify and present all important benefit and harm information in a well-designed table or graphic, so that they can be easily compared.
Additional authors: Rodney A. Hayward, Elyse Reamer, Brian J. Zikmund-Fisher, Daniel Connochie, Michele Heisler
Funding: VA Advanced Fellowship in Health Services Research & Development; Michigan Center for Diabetes Translational Research, National Institutes of Health grant P30DK092926; VA HSR&D Center of Innovation
Reference: Journal of the National Cancer Institute, published online Feb. 24, 2016, doi: 10.1093/jnci/djv436