SEATTLE – Women at average risk are often screened for ovarian cancer, even though national guidelines recommend against this practice, based on a study reported at the annual meeting of the North American Primary Care Research Group.
In a cross-sectional survey of 1,088 primary care physicians presented with examples cases, 72% said they would almost never offer low-risk women transvaginal ultrasound or the cancer antigen (CA)-125 blood test, and 35% said they would almost never do so for medium-risk women.
Of concern, 30% of physicians overall said they would almost always offer or order the tests for such women.
"Screening is actually not recommended for either of these groups, suggesting that physicians are either not accurately assessing the risk or that they are choosing to offer or order the test despite recommendations to the contrary," commented lead investigator Dr. Laura-Mae Baldwin.
Besides risk assessment, a mistaken belief that the tests are effective for screening in average-risk women – endorsed by a third of the physicians – was the next strongest predictor of nonadherence to guidelines.
In fact, transvaginal ultrasound and the CA-125 test have high false-positive rates and low positive predictive values when used for ovarian cancer screening, according to Dr. Baldwin, who is a professor of family medicine and director of the Family Medicine Research Section at the University of Washington, Seattle.
"There is no professional organization or government agency that currently recommends routine ovarian cancer screening," she noted. "The U.S. Preventive Services Task Force (USPSTF) actually gives it a D grade, meaning that the harms exceed the benefits, and they recommend against the test."
"But as we know, physicians have demonstrated enthusiasm for some cancer screening tests that don’t have clear proven benefit, and we thought it might be possible that women may be exposed to the potential harms of ovarian cancer screening despite these recommendations," she said.
The investigators sent vignette-based surveys to a cross-sectional sample of U.S. primary care physicians selected from American Medical Association Physician Masterfiles.
The vignettes portrayed women at low, medium, and high risk for ovarian cancer. Respondents were asked, for the patient described, how often they would offer or order each of a list of tests for cancer screening, and were given response options of almost always, sometimes, and almost never.
"We excluded the physicians who had vignettes of women at high risk for ovarian cancer because there are some recommendations that suggest that it may be appropriate to screen in those women, so we wanted really to look at a group that had a more average risk of ovarian cancer," Dr. Baldwin explained.
Results were based on 1,088 respondents who represented a weighted sample of 106,001 physicians nationally. Some 42% were general internists, 41% were family physicians, and 17% were obstetrician-gynecologists.
Nearly three-fourths of the physicians worked in group practices. When asked about their sources of information on cancer screening, the most common were the American Cancer Society (66% listed it among their top three), the USPSTF (53%), the National Institutes of Health and National Cancer Institute (33%), and the American College of Obstetricians and Gynecologists (31%).
In a finding that Dr. Baldwin described as highly surprising, 33% of physicians believed that transvaginal ultrasound, the CA-125 test, or both were effective for ovarian cancer screening in average-risk women.
For low-risk patients, 72% of physicians were adherent to screening recommendations (defined as almost never offering or ordering either screening test); for medium-risk patients, 35% were adherent.
On the flip side, 6% of physicians said they would almost always offer or order a screening test for a low-risk patient, and 24% said they would almost always do so for a medium-risk patient.
In multivariate analyses, physicians were less likely to be adherent to recommendations against screening if the patient was at medium risk or requested testing, or they had been in practice for at least 10 years, Dr. Baldwin reported.
On the other hand, physicians were more likely to be adherent if they high-listed USPSTF as an information source, did not have any personal or family experience with cancer, were involved with clinical teaching, or were in a group practice.
However, when belief about the effectiveness of the screening tests was added to the other physician factors, physicians believing them to be effective were less likely to adhere to recommendations than their peers believing them to be ineffective.
Also, three of the other factors – years in practice, high-listing the USPSTF, and group practice – were no longer significantly related to adherence. "So basically, those associations were mediated by the degree to which those groups believed in the effectiveness of these tests," she explained.