Physicians and nurse practitioners who participated in the 2012 DocStyles survey included those who practiced in the USA, actively saw patients, and had practiced for at least 3 years. The physician sample was randomly selected from the Epocrates Honors Panel^®, which included 275,000 medical practitioners, to match the American Medical Association’s (AMA) Masterfile^® proportions for age, gender, and region. The identities of panel members were verified by checking each physician’s first name, last name, date of birth, medical school, and graduation date against the AMA Masterfile at the time of panel registration. The nurse practitioner sample was drawn from Epocrates’ Allied Health Panel^® of over one million health professionals, including 78,668 nurse practitioners; the identities of nurse practitioners were not verified.

Quota sampling involves deliberately setting the proportions of selected participant characteristics within a sample [15] and was incorporated into the 2012 DocStyles survey methods to ensure adequate representation of all provider groups surveyed. Sampling quotas were set at 1,000 primary care physicians (internists and family practitioners), 250 nurse practitioners, and 250 gynecologists. Email invitations to participate in the survey were sent to 2,175 primary care physicians, 456 nurse practitioners, and 489 gynecologists. Per sampling quotas, completed surveys were accepted from 1,001 primary care physicians, 252 nurse practitioners, and 250 gynecologists. In addition, providers who did not treat adult female patients and those who worked primarily in inpatient care settings were excluded from the analyses, which narrowed the sample to 892 primary care physicians, 240 nurse practitioners, and 248 gynecologists, (N = 1,380). To protect confidentiality, no individual identifiers were included in the dataset provided to investigators, and the analyses reported here were exempted from CDC Institutional Review Board approval.

Participants were asked to indicate where gynecologic cancer education materials were available in their offices. Responses included (1) patient waiting area, (2) examination room, (3) other location in my office, and (4) this material is not available in my office. Multiple responses were accepted unless “this material is not available in my office” was selected.

Participants who reported using gynecologic cancer education materials in their offices were asked what organization produced the materials. Responses included (1) American Cancer Society (ACS), (2) American Congress of Obstetricians and Gynecologists (ACOG), (3) Centers for Disease Control and Prevention (CDC), (4) Facts for Women: Reproductive Cancer Awareness campaign, (5) Inside Knowledge: Get the Facts About Gynecologic Cancer campaign, (6) National Cancer Institute, (7) Society of Gynecologic Oncologists, (8) WebMD, (9) other, and (10) not sure. Multiple responses were accepted unless “not sure” was selected.

Providers were also provided a list of symptoms highlighted in CDC gynecologic cancer education materials (Table 1) and asked which cancers they associated with each symptom. Responses included (1) cervical cancer, (2) ovarian cancer, (3) uterine cancer, (4) vaginal cancer, (5) vulvar cancer, and (6) none of these. Multiple responses were accepted unless “none of these” was selected.

Within each provider group, descriptive statistics were used to examine and summarize participant characteristics and use of gynecologic cancer education materials. Next, Pearson chi-square and t tests were performed to investigate the associations between provider characteristics (gender, race, practice size, geographic region of practice location, and number of Pap tests performed monthly) and use of CDC gynecologic cancer education materials.

Pearson chi-square tests were used to compare the recognition of the symptoms highlighted in CDC gynecologic cancer education materials between providers who used CDC gynecologic cancer education materials and providers who did not (providers who used non-CDC gynecologic education materials and those who used no gynecologic cancer education materials).

Most participants were male (60.1 %), white (77.8 %), and practiced in settings with 2–9 other practitioners (54.1 %) (results not shown). Participants were relatively evenly distributed by geographic region: Northeastern states (27.0 %), Southern states (22.5 %), Western states (23.6 %), and Midwestern states (27.0 %). The mean number of Pap tests performed during a typical month varied by specialty: primary care physicians (13.7 tests), nurse practitioners (24.6 tests), and gynecologists (119.1 tests).

Among gynecologists, reported in-office use of gynecologic cancer education materials was almost universal (96.0%), and the majority of primary care physicians (73.4 %) and nurse practitioners (71.1 %) also reported the use of such materials (results not shown). More than half of the participants across provider groups reported that gynecologic cancer education materials were available in their patient waiting rooms (55.6%) and examination rooms (54.4%). The most common sources of gynecologic cancer education materials were ACOG (33.7 %), ACS (32.0 %), and CDC (18.1 %). Less than a quarter of providers (19.4 %) reported using materials developed by CDC or the Inside Knowledge campaign. In the remaining analyses, participants were classified as using materials produced by CDC if they reported using materials produced by CDC or the Inside Knowledge campaign.

Primary care physicians practicing in Northeastern states were less likely to report using CDC materials (11.0 %) compared with those practicing in other regions (18.8–19.6 %) (χ²=8.62, p = 0.035). Among primary care physicians, the use of CDC materials was associated with higher Pap test volume: mean monthly Pap tests was 18.8 tests among those who used CDC materials and 12.7 tests among those who did not use CDC materials (t = 3.79, p <0.001).White primary care physicians and nurse practitioners were less likely to report use of CDC materials than practitioners of other races: 15.1 % of white primary care physicians compared with 21.8 % of primary care physicians of other races (χ²=5.66, p = 0.017), and 24.2 % of white nurse practitioners compared with 52.9 % of nurse practitioners of other races (χ²=6.73, p = 0.009). No significant predictors of using CDC gynecologic cancer education materials were detected among gynecologists.

Recognition of gynecologic cancer symptoms was generally higher among gynecologists than among primary care physicians and nurse practitioners (Table 1). Across provider groups, low recognition (<25 %) of at least one symptom of the cancers studied was found with the exception of uterine cancer—the majority in all provider groups recognized all of the uterine cancer symptoms studied. Symptom recognition was consistently higher among providers who reported using CDC gynecologic cancer education materials compared with other providers. When symptom recognition was analyzed in aggregate (all provider groups combined), providers who reported use of CDC materials were significantly more likely to recognize 11 of the 21 gynecologic cancer symptoms studied than providers who did not use CDC materials.