Vignette-based study of ovarian cancer screening: do U.S. physicians report adhering to evidence-based recommendations?

Abstract

Background: No professional society or group recommends routine ovarian cancer screening, yet physicians' enthusiasm for several cancer screening tests before benefit has been proven suggests that some women may be exposed to potential harms.

Objective: To provide nationally representative estimates of physicians' reported nonadherence to recommendations against ovarian cancer screening.

Design: Cross-sectional survey of physicians offering women's primary care. The 12-page questionnaire contained a woman's annual examination vignette and questions about offers or orders for transvaginal ultrasonography (TVU) and cancer antigen 125 (CA-125).

Setting: United States.

Participants: 3200 physicians randomly sampled equally from the 2008 American Medical Association Physician Masterfile lists of family physicians, general internists, and obstetrician-gynecologists; 61.7% responded. After exclusions, 1088 respondents were included; their responses were weighted to represent the specialty distribution of practicing U.S. physicians nationally.

Measurements: Reported nonadherence to screening recommendations (defined as sometimes or almost always ordering screening TVU or CA-125 or both).

Results: Twenty-eight percent (95% CI, 24.5% to 32.9%) of physicians reported nonadherence to screening recommendations for women at low risk for ovarian cancer; 65.4% (CI, 61.1% to 69.4%) did so for women at medium risk for ovarian cancer. Six percent (CI, 4.4% to 8.9%) reported routinely ordering or offering ovarian cancer screening for low-risk women, as did 24.0% (CI, 20.5% to 28.0%) for medium-risk women (P ≤ 0.001). Thirty-three percent believed TVU or CA-125 was an effective screening test. In adjusted analysis, actual and physician-perceived patient risk, patient request for ovarian cancer screening, and physician belief that TVU or CA-125 was an effective screening test were the strongest predictors of physician-reported nonadherence to published recommendations.

Limitation: The results are limited by their reliance on survey methods; there may be respondent-nonrespondent bias.

Conclusion: One in 3 physicians believed that ovarian cancer screening was effective, despite evidence to the contrary. Substantial proportions of physicians reported routinely offering or ordering ovarian cancer screening, thereby exposing women to the documented risks of these tests.

Primary funding source: Centers for Disease Control and Prevention and the National Cancer Institute.