This article is more than 5 years old. Information may no longer be current.
Many prevention, screening guidelines do not adequately explain benefits, risks
Click Here to Manage Email Alerts
More than two-thirds of cancer prevention and screening recommendation statements do not accurately quantify the benefits and harms associated with the screening platforms, according to results published in Journal of the National Cancer Institute.
“Almost all medical interventions have tradeoffs between different types of benefits and harms,” Tanner J. Caverly, MD, MPH, clinical instructor of internal medicine at University of Michigan Medical School and research investigator at the V.A. Ann Arbor Healthcare System Center for Clinical Management Research, said in a press release. “Patients rely on doctors to base screening recommendations on sound judgment as to whether, in their individual context, the benefits outweigh the harms. 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.”
Caverly and colleagues performed a systematic review to determine how U.S. cancer prevention and screening recommendations present the potential benefits and harms associated with the procedures.
Researchers included 32 screening and prevention recommendations issued by the U.S. Preventive Services Task Force, the American Cancer Society, the American College of Physicians, the National Comprehensive Cancer Network, and other U.S. screening guidelines found within the National Guidelines Clearinghouse. These guidelines offered 55 total recommendations for interventional detection or prevention of breast cancer, cervical cancer, colon cancer, lung cancer and prostate cancer.
Types of screening and preventive interventions included in these recommendations included mammography, clinical breast exam, PSA testing, fecal immunochemical testing, colonoscopy, CT colonography, Pap smear, HPV testing and vaccination, and preventive medications.
The researchers reviewed information about benefits and harms located anywhere in the guideline document, including appendices. They gave each guideline a “comparability rating” based on how benefits and harms were presented, with those that presented absolute effect information on benefits and harms deemed “comparable.” Recommendations that were uneven — because a recommendation was made without listing harms or benefits, or because the benefits and harms were presented in a different manner — were deemed “asymmetric.” “Incomplete” ratings indicating recommendation did not quantify harms or benefits or were explained only using relative effects.
The benefit-to-harm comparability ratings associated with the recommendations served as the primary endpoint.
Twenty-eight recommendations (50.9%) mentioned the incidence of the target cancer over a specified time frame. However, only six recommendations (10.9%) presented the incidence proportionally.
Additionally, six recommendations presented the importance of the target cancer in terms of life-years lost due to disease.
Prior systematic reviews informed 37 (67.3%) recommendations. Thirty-five (63.6%) included an explicit rating scheme for the quality of evidence and 40 (72.7%) included ratings for the strength of the recommendation.
Fourteen recommendation statements (25.4%) did not mention any clinically important benefits associated with the test, whereas 16 (29.1%) did not mention any potential harms.
Researchers deemed the majority of recommendations asymmetric (n = 30; 54.5%). Seventeen recommendations (30.9%) received a rating of comparable and eight recommendations (14.5%) received a rating of incomplete.
The researchers acknowledged limitations of their study.
“A ‘comparable’ rating simply means that benefits and harms are presented in a format to allow potential comparison, while incomplete and asymmetric ratings mean that such a comparison is not possible,” Caverly and colleagues wrote. “Thus, our study focuses on just one component of what is necessary for developing a high-quality guideline.”
They further noted that they did not independently assess the quality of the supporting evidence used to make these recommendations.
“By unevenly presenting the statistics, guidelines can unintentionally give a false picture of the benefits and harms,” Angela Fagerlin, PhD, professor and chair of population health sciences at University of Utah, as well as a research scientist with the Salt Lake City V.A., said in a press release. “The relative risk of ‘twice as likely’ sounds significant, but in terms of absolute risk it might represent a very small difference.” – by Cameron Kelsall
Disclosure: This study was funded in part by grants from the Michigan Center for Diabetes Translational Research and the V.A. Health Services Research & Development Center for Innovation. Caverly reports research support from the V.A. Advanced Fellowship in Health Services Research & Development. The other researchers report no relevant financial disclosures.
Perspective
Back to Top
Kathryn Evers, MD, FACR
Guidelines for cancer screening examinations inform clinical judgments regarding the utility of the various available screening tests. The manner in which the data in the guidelines are presented may influence how clinicians advise patients regarding the use of this testing.
The study by Caverly and colleagues has looked at one aspect of the published screening guidelines for breast, prostate, colorectal, cervical and lung cancers: how benefits and harms are presented.
Less than half of the guidelines presented risks and benefits in a comparable fashion, and in less than half of the guidelines absolute risks of both benefits and risks were not presented. How this actually influences clinical decision-making is unknown. However, the inability to easily compare absolute benefits and harms makes it difficult to make an informed choice regarding desirability of testing for any individual patient or group of patients. This is an important first step in making guidelines transparent and unbiased.
Click Here to Manage Email Alerts