Primary care physicians often misinterpret cancer screening statistics, study indicates

Primary care physicians often misinterpret statistics about the benefits of cancer screening, according to a new study.

Researchers used an Internet survey to conduct a parallel-group, randomized trial of a national sample of U.S. primary care physicians. The physicians were selected from a Harris Interactive research panel, and the trial was randomized only for order effect. The study's objective was to determine whether primary care physicians understand which screening statistics indicate whether cancer screening saves lives. Physicians were given scenarios about two hypothetical screening tests; in the first scenario, the test was described as improving five-year survival rates and increasing early detection, while in the second, it was described as decreasing cancer mortality and incidence. The study's main outcome measures were physicians' general knowledge of screening statistics and their recommendations and perception of screening benefit in each of the scenarios. The results appeared in the March 6 Annals of Internal Medicine.

The study surveyed 297 primary care physicians in 2010 who practiced in both inpatient and outpatient settings and 115 physicians in 2011 who practiced in outpatient settings only. Most physicians (77%) were men, and most (81%) were age 40 or older. The researchers found that physicians were more likely (69% vs. 23%; P<0.001) to recommend a test based on irrelevant evidence (a five-year survival rate that increased from 68% to 99%) than one based on relevant evidence (a cancer mortality reduction from 2 to 1.6 in 1,000 persons). Most physicians did not see a difference between the irrelevant and relevant statistics, with 76% and 81%, respectively, stating that each type proved screening saves lives (P=0.39). Forty-seven percent of physicians erroneously responded that finding more cancer cases in screened versus unscreened populations proves that screening saves lives.

The study authors acknowledged that their results were based on hypothetical scenarios rather than actual clinical practice, and that the scenarios did not include potential harms of screening. However, they concluded that primary care physicians' knowledge of the benefit of cancer screening is lacking. “To better understand the true contribution of specific tests, physicians need to be made aware that in the context of screening, survival and early detection rates are biased metrics and that only decreased mortality in a randomized trial is proof that screening has a benefit,” the authors wrote.