Most prevention and screening guidelines skirt the details on harms, benefits

Guidelines for cancer prevention and screening are “ideally suited” for explaining the benefits and harms of medical interventions, but according to a recent study published by the Journal of the National Cancer Institute, a majority of these statements do an insufficient job of providing such insight.

Tanner J. Caverly, MD, MPH, Veterans Affairs Medical Center in Ann Arbor, Mich., and colleagues studied a total of 55 recommendation statements for preventing or detecting breast, prostate, colon, cervical, or lung cancer from organizations such as the U.S. Preventive Services Task Force, American Cancer Society, American College of Physicians, and National Comprehensive Cancer Network. Overall, 69 percent of those statements either did not fully list benefits and harms or listed them in an “asymmetric manner.”

“If patients and physicians have inaccurate perceptions about the magnitude of benefits and harms from cancer prevention and screening, it may be, at least in part, because the essential information is not readily accessible in important patient-care resources,” the authors wrote.

Caverly et al. suggested that while organizations may be making their statements brief on purpose, they appear to be going too far in that direction.

“Although it is useful for clinicians if guidelines are concise in their recommendations, estimates of the magnitude of both benefits and harms should be clearly delineated, along with information on the reliability and quality of the evidence that provides the basis of those estimates,” the authors wrote. “The provision of such information is essential for allowing clinicians to assess the guideline committee’s conclusions and provides important information for patient consultations.”

The authors also provided examples of “asymmetric” guidelines, saying they were a significant concern. Some recommendation statements, for example, presented the benefits and the harms in different ways, mischaracterizing the trade-off between the two statistics.

In other cases, the number of patients impacted by the target cancer would be presented as just a numerator, leading to statements such as, “In 2013, an estimated 232,340 women in the United States will be diagnosed with breast cancer.” Caverly and colleagues said this isn’t the most helpful way to provide such information.

“These large numbers do have some population health relevance,” the authors wrote. “However, they can lead to a misperception about how likely an individual is to benefit from an intervention when the denominator is large. Using incidence rates is much more informative when proposing individual-level patient interventions (ie, ‘The age-adjusted annual incidence rate of cervical cancer is 6.6 cases per 100,000 women’).”