Breast Cancer Risk Assesment in the Primary Care Setting: A Pilot Teleconference / E-mail Program Targeting Gynecologists, Nurse Practitioners and Physicians Assistants
Sally Bogert, NP; Michelle Finklestein, MD, Elena Cyrus, BA; Jennifer Motley, MD; Neil Love, MD

Publication of the Breast Cancer Prevention Trial (NSABP P-1) created a clinical and perhaps medical-legal imprimatur for primary care clinicians to identify women at increased risk for breast cancer and counsel them regarding options. The P-1 study used the highly predictive Gail model to identify women at increased risk, and although this model is widely available for healthcare clinicians, it is not commonly used. In addition, once a high-risk patient is identified, many primary care clinicians refer these women to specialists rather than counsel them and offer options such as chemoprevention or referral to the NSABP's STAR trial (Study of Tamoxifen and Raloxifene).

As a result, it is a public healthcare challenge to educate primary care providers regarding the calculation of a woman's breast cancer risk and appropriate management. Gynecologists, nurse practitioners (NP) and physician assistants (PA) were targeted for a pilot educational program to develop a method to facilitate the incorporation of this strategy into clinical practice.

Primary care clinicians are inundated with educational and promotional materials recommending a variety of preventive strategies that require time to implement. To increase the likelihood of breast cancer risk assessment receiving a high priority for the clinician's time, we assessed how this strategy could meet other existing clinical needs in women's health care. To do this, workshops using interactive keypad polling were conducted at five continuing medical education programs for gynecologists. The P-1 data and Gail model risk assessment tools were reviewed, and clinicians were queried about the potential role these might play in clinical practice. Following the meetings, e-mail correspondence took place with these physicians and others who were recruited to the project. Over 150 physicians provided input on an ongoing basis as did a panel of breast cancer researchers who were consulted on this project. Our goal was to determine how Gail model risk assessment and counseling of high-risk women could be effectively incorporated into primary women's health care, meeting both the needs of patients and busy clinicians. As we developed ideas, we transmitted these to clinicians by e-mail, who evaluated the concepts in their practice and reported back to us.

The conclusions of this initiative were as follows:

1. The most useful aspect of Gail model assessment was in counseling regarding the use of estrogen replacement therapy (ERT) in normal risk menopausal women.

Clinicians uniformly agreed that the greatest concern for women considering ERT was a potential increased risk of breast cancer, and related discussions were perceived by clinicians as challenging and time-consuming. The Gail assessment was seen as very valuable in reassuring women that their baseline risk of breast cancer was lower than the patients (and clinicians) had believed, and this assisted clinicians in educating women about the risks and benefits of ERT.

Clinicians preferred a hand-held calculator compared to the computer disk or slide rule.

2. Women at increased risk of breast cancer based solely on their age were not deemed appropriate candidates for Gail assessment or referral to the STAR trial.

While the NSABP accepted these women into P-1 based on an increased five-year risk, clinicians generally believed that breast cancer was not a key preventive issue for these women (mainly over age 65), particularly compared to other issues such as heart disease and osteoporosis. While clinicians were concerned about medical-legal implications of not identifying and informing high-risk women about their options, it was widely believed that for a woman to be considered "high-risk," she must have had a prior breast biopsy or a first-degree relative with breast cancer.

A strategy supported by clinicians that encompassed these principals was to offer and encourage Gail model assessment and counseling to women age 35-60 with either a prior breast biopsy or a first-degree relative with breast cancer.

Methods

To test the validity of the above conclusions, a series of educational teleconferences was conducted with 344 primary care clinicians, mainly nurse practitioners. The teleconferences were moderated by a nurse practitioner (Sally Bogert, NP) and a physician (Neil Love, MD, or Michelle Finkelstein, MD). The objectives of the program are outlined in Figure 1.

Prior to the teleconference, the participants received a teaching packet of materials including a 16-page teleconference guide containing the agenda and several supportive articles, a hand-held Gail model risk calculator, patient aid materials, a post-teleconference questionnaire, and additional articles on the P-1 study, risk assessment and breast cancer treatment. During the teleconference, audio sound bites from interviews with national breast cancer research and gynecology leaders were played, and interactive touchpad polling was used to allow teleconference participants to answer survey questions using the keypad on their telephone. A question and answer session was also conducted. Use of the hand-held Gail model risk calculator was taught and practiced during case presentations. After the teleconference, participants completed and returned a detailed follow-up questionnaire and received a $50 honorarium for their participation.

The major items included:

• Historical perspectives on breast cancer and review of the P-1 study
  
  
• Pertinent breast cancer risk factors and development of the Gail model
  
  
• Use of the hand-held Gail model risk calculator
  
  
• Clinical relevance of five-year and lifetime risks, as well as relative risk (see Figure 2)
  
  
• Counseling menopausal patients considering ERT regarding breast cancer risk
  
  
• Management options for high-risk women

Following the teleconference, the participants mailed in the post-teleconference questionnaire. Four months later, a shorter questionnaire was e-mailed to the participants to assess their current practices with regards to breast cancer risk assessment and management. Other than the pre-teleconference packets, most of the correspondence was conducted via e-mail.

Nine teleconferences were held with the number of participants per teleconference ranging from 16 to 97. Of the 344 participants, 83% were nurse practitioners and 17% were physician assistants. The majority practiced in family medicine or women's health care and provided care to 20 to 60 women in their practice on a weekly basis. When asked, "To what extent is breast health and the risk of breast cancer a part of your daily practice?" (on a scale of 1-10: 1 = not at all, 10 = a major issue), the mean response was 6.8.

Teleconference Questionnaire

During the teleconference, keypad polling was conducted to assess the participants' use of the Gail model and the educational value of the teleconference. Responses revealed that prior to the teleconference, only 17% of participants had ever used the Gail model and less than 1% used it routinely. Figure 3 illustrates the clinicians' intentions to use the Gail model following the teleconference. As illustrated, following the teleconference 98.2% had identified a use for the Gail model in their practice, and 56.6% planned to use it routinely on all women over age 35.

With regard to the educational value of the teleconference, the participants were asked "How useful and educational has this teleconference been to you?" (rating it 1-9: 1 = not useful, 9 = very useful). The mean response was 8.3.

Post-Teleconference Questionnaire

Following the teleconference, each participant mailed in a 27-item questionnaire. Participants were asked to rate their familiarity with the results of the P-1 study prior to the teleconference. On a scale of 1 to 10 (1 = not at all, 10 = extremely), the mean response was 3.5. When asked to rate their belief in the validity and applicability of the P-1 study to their practice following the teleconference, using the same scale, the mean response was 8.9. Participants were also asked to rate their interest in using the Gail model as a result of the teleconference. The results can be seen in Figure 4.

Four Month Follow-Up Questionnaire

Four months following the teleconferences, a questionnaire was e-mailed to participants, and 53 responded. The respondents saw an average of 26 female patients over the age of 35 on a weekly basis. Of the respondents, only 18% had ever used the Gail model prior to the teleconference. Four months later, 80% of the respondents had used the Gail model. Prior to the teleconference, respondents calculated a patient's breast cancer risk on an average of two patients per week. Four months following the teleconference, the mean increased to six patients per week.

Figure 5 illustrates the subsets of patients in which the teleconference had the greatest impact in changing the clinicians' practice standard to include the Gail model risk assessment.

For practical purposes, the most frequent use of the Gail model tool was in menopause counseling for ERT, but the majority of participants also used the tool for assessment of women with major risk factors. Not only did the teleconference impact the clinicians' use of the Gail model for risk assessment, but it also appeared to impact the clinicians' comfort with and understanding of chemoprevention. Participants were asked how they would manage a high-risk patient eligible to take tamoxifen for prevention who had a positive risk/benefit ratio. Following the teleconference, there was a fourfold increase in the number that reported they would prescribe tamoxifen.

Historically, primary care clinicians have identified women at increased risk for breast cancer based on history alone, without specific numeric risk estimates. Usually women identified as being very high-risk are referred to a breast cancer specialist for counseling and management, rather than being managed by the primary care clinician. However, with the accuracy and accessibility of the Gail model for breast cancer risk assessment, primary care clinicians are in an ideal setting to assess and identify women at high risk. With the FDA approval of tamoxifen for chemoprevention and the opportunity to participate in clinical research, primary care clinicians can now offer intervention options other than just close follow-up. In addition, the Gail model offers primary care clinicians a valuable tool for a challenging clinical scenario that is a daily part of primary women's health care, namely counseling on the option of ERT in menopausal patients.

The educational packet and one-hour teleconference proved successful in affecting this paradigm shift in clinical practice, as evidenced by the results of self-reported clinical behavior. Prior to the teleconferences, only 17.2% had ever used the Gail model at all, and the mean use by participants was only two patients per week. Following the teleconferences, 98.2% identified a use for the Gail model in their practice. While the greatest practical impact of the Gail model risk assessment was seen in normal-risk women receiving menopause counseling, there was also a major effect in terms of identifying and counseling high-risk women. Not only were clinicians more likely to identify these patients, but they were also more likely to prescribe tamoxifen or refer for trial accrual when indicated.

Top of Page