Advances in genetics and molecular biology have clinical implications for gynecology practice. Many obstetrician-gynecologists address primary care issues throughout their patients' lives, including diseases and disorders of aging. Diseases with known genetic influences include adult-onset disorders as varied as diabetes, hypertension, certain autoimmune conditions, and hereditary colorectal cancer.1–4 Much has been written about BRCA1 and BRCA2 and particular cases of familial breast and ovarian cancers.5,6
Research on physicians' genetic knowledge has highlighted several areas of concern.7–9 The current study is separate from, but similar to, an earlier survey of ACOG Fellows concerning obstetric genetics.10 That study indicated a need for physician and patient education on genetic issues in obstetric care.
In the present study, we surveyed a sample of ACOG Fellows who practice gynecology. As in the earlier study,10 our aims were to explore current practice trends in gynecologists' offices and to identify areas of strength and weakness in physicians' knowledge.
Materials and Methods
Questionnaires were mailed in September 1998 to all 248 members of the Collaborative Ambulatory Research Network and to a control group of 1000 ACOG Fellows not in the Collaborative Ambulatory Research Network. The Collaborative Ambulatory Research Network is a group of ACOG Fellows who voluntarily participate in questionnaire survey studies. We chose a total sample based on similar studies with comparable numbers of surveys. Reliable effects have been found using survey mailings of approximately 1000.10 Questionnaires were sent concurrently to the control group to find out whether Collaborative Ambulatory Research Network Fellows' responses were statistically similar to those of ACOG Fellows in general. To maximize the likelihood that controls would be representative of the entire population of ACOG Fellows, we oversampled from this group. Non-Network Fellows were randomly sampled by computer from the subset of ACOG Fellows not selected for the Collaborative Ambulatory Research Network study control group since January 1996.
Before we administered the survey, we performed standard preliminary power analyses commonly used in survey research. Our analysis indicated that the minimum number of responses needed to ensure significant effect sizes was 100. We evaluated significance at .05, and all effect sizes were above 1.0. Although a sufficient number of completed surveys were returned, we sent a second mailing in November 1998 to all nonrespondents to increase our overall response rate.
The survey contained primarily multiple-choice questions pertaining to physicians' practice procedures, clinical opinions and experiences, and genetics-related knowledge. Additional items addressed professional education issues. Formal training was defined as didactic (classroom) training or clinical training during medical school, residency, or continuing medical education.
Data were analyzed using a personal computer–based software package, SPSS version 8.0 (SPSS, Inc., Chicago, IL). Descriptive statistics computed for measures used in the analyses are reported as mean ± standard error. Two-tailed t tests were used to compare group means for age. Pearson correlation was used to assess the relation between age and knowledge score. Group differences in knowledge score were assessed using analysis of covariance, with training (yes/no), ordering DNA-based tests (yes/no), and sex as categoric variables, and age as the covariate. Differences in categoric measures were assessed using χ2. All analyses were tested for significance using an alpha of .05.
Data from respondents who returned the questionnaire before January 1999 were included in the analysis. Five hundred sixty-four (45%) of the 1248 questionnaires mailed were returned within the allotted time. The response rate for the Network group was 59%, compared with 25% for the non-Network controls. One hundred thirty-four (about 24%) of all respondents said they did not provide any genetic screening services (ie, DNA-based testing or assessment of patient risk factors for heritable diseases or disorders based on family history), leaving 428 completed questionnaires.
Obstetrician-gynecologists who reported that they did not offer genetic screening services were older (48.6 ± 9 years versus 45.2 ± 5 years; P = .018). They also were more likely to be male (male to female ratio 2.68:1) than those who provided genetic screening services (male to female ratio 1.43:1) (P = .003). Our results were obtained from the 428 respondents who provided some form of genetic screening services. Approximately 59% were men and 41% were women. Respondents' median age was 44 years; men's mean age was 48.3 ± 0.7 years and women's mean age was 40.8 ± 0.5 years. The men were significantly older than the women (P < .001). Not all respondents answered every question.
Seventy-eight percent of the respondents said that they ask for patients' family histories at initial visits, and 47% do so at annual examinations. Approximately 13% order DNA-based genetic tests and 89% of these take family histories at initial visits. Respondents were allowed to choose more than one answer, and many indicated that they review family histories more than once. Gynecologists who take family histories to assess the risk of adult-onset disorders or to perform preconception screening use that information to screen for increased risk of breast cancer (95.3%); ovarian cancer (94.4%); other cancers (86.6%); adult-onset disorders such as diabetes, heart disease, or hypertension (88.3%); and preconception reproductive risks for population-specific disorders (77.8%). Respondents were allowed to choose more than one option. When a woman identified a first-degree relative with breast or ovarian cancer, colon cancer, or certain other adult-onset disorders, most gynecologists reported that they monitor earlier or more frequently for signs and symptoms of disease in each case. The second most frequently endorsed option was referral to an internist.
When asked whether their practices currently send patient samples or order DNA-based genetic tests for a selection of diseases, approximately 30% of gynecologists did so for ovarian cancer, 24% for breast cancer, 15% for colon cancer, 6% for Huntington disease, 1% for Alzheimer disease, and 60% for preconception screening. When asked who was most likely to suggest DNA-based genetic tests in their gynecologic practices, most responded that they would be the most likely to do so (40%), followed by patients (26%) or other professionals (7%). Approximately 17% responded “not applicable” and less than 4% responded “don't know.” Gynecologists who order DNA-based tests were more likely than those who do not to suggest genetic testing (67.9% versus 36.3%; P = .005).
Slightly more than 57% of respondents have had patients request DNA-based genetic screening for breast, ovarian, or colon cancers. Those who order DNA-based tests were more likely than those who do not to report having had patients make such requests (83.9% versus 54.2%; P < .001). Only 8% of respondents have had healthy patients request prophylactic nonsurgical treatment (eg, Tamoxifen; tamoxifen citrate, Zeneca Pharmaceuticals, Wilmington, DE) Labs) after genetic screens indicated that “cancer genes” were present. Nineteen percent have had healthy patients request prophylactic surgery after a positive result for some form of cancer. Among the group that had patients with positive tests, those who order DNA-based tests were more likely to have patients request prophylactic surgery (P < .01).
The gynecologists were presented with a set of hypothetical options pertaining to prophylactic surgery and were asked under which conditions they might recommend prophylactic mastectomy or oophorectomy for a patient who is a known carrier of BRCA1 or BRCA2 but is currently healthy. Very few respondents (3.8%) would make such a recommendation solely on the basis of a positive test result. Twenty-seven percent would if the patients also had family histories of cancer. Slightly more than 44% would respond to a patient's request for surgery by recommending surgery. About 13% said they would be unlikely to recommend surgery, and less than 2% would not recommend surgery at all. Others responded “don't know” (8.6%), “not applicable” (11%), and “other” (17.3%) to this question.
Gynecologists were asked to choose the best answer from a list of options about their current prescribing practices for hormone replacement therapy or oral contraceptives when patients have particular positive genetic test results. Many respondents indicated that they have not yet had to consider these decisions in practice, by choosing the “not applicable” option. Results are presented in Table 1.
Almost 90% of the respondents answered correctly that available genetic tests for gynecologic cancers are most informative when used with family history. The next most common answer, chosen by 8% of respondents, was that the tests are not very informative. The next set of questions sampled gynecologists' genetic knowledge. For each question, the percentage of respondents who answered correctly, incorrectly, or said “don't know” are presented in Table 2.
We created a summary score based on correct answers to the genetic knowledge questions. Answers to questions that were dependent on each other were counted as one response. Respondents received 1 to 13 points out of a possible 14-point total. Age was negatively correlated with knowledge score (r = −.165, P < .01). Respondents who had formal training scored higher than those who did not (8.9 ± 0.2 versus 8.1 ± 0.2; P < .001). Training had no effect on knowledge scores among physicians who order DNA-based tests. Those who order DNA-based tests scored similarly to formally trained physicians, but significantly higher than those with no formal training who do not order DNA-based tests (8.7 ± 0.3 versus 8.0 ± 0.2; P < .05). Respondents' sex had no effect.
Sixty-five percent of the respondents reported that they had received no formal training (either didactic or clinical) in DNA-based genetic testing in gynecologic practice. About 20% reported training in continuing medical education courses; fewer reported training during medical school, internship, or residency. Almost 71% of those over 55 years old, 65.6% of those aged 41–55, and 62.4% of those aged 40 or younger had not received formal training. Among those who had received formal training, younger respondents were significantly more likely to have had formal training during internship or residency, whereas those 55 years and older were more likely to have had training in continuing medical education courses (P < .05).
Physicians' knowledge of issues relevant to DNA-based genetic testing was mixed. Gynecologists appeared more knowledgeable about breast and ovarian cancers than they were about colon cancer and other adult-onset disorders. We found that gynecologists recognize that test results are most informative when used with family history. Obtaining structured reviews of patients' family histories in the gynecology office could provide opportunities for preventive care and patient education.
Our study suggests that formal training improves knowledge of genetics and issues pertaining to DNA-based testing. However, 65% of the respondents reported that they had received no formal training in DNA-based genetic testing in gynecologic practice. Only 25% of the non-Network group responded to the survey. It is likely that those who are least knowledgeable and least likely to have had training were among the nonrespondents.
Incorporating genetics into gynecology practice is a formidable challenge, and limited training is a concern. This and other studies7–10 indicate that further education is needed for most physicians. It will also be important for physicians to form collaborative approaches to patient care across disciplines, which will benefit physicians and their patients. Coordinated-care strategies will make more manageable the amount of genetic knowledge the practicing physician will need to have and will increase the amount and type of resources available to physicians and patients. Nonphysician health care providers will likely emerge in complementary roles in patient education and testing programs. Increased education of allied health professionals and closer affiliation with regional genetic services that are already available could advance medical genetics in gynecologic practice.
1. Savill J. Role of molecular cell biology in understanding disease. BMJ 1997;314:203–6.
2. Luft FC. Molecular genetics of human hypertension. J Hypertens 1998;16:1871–8.
3. Shaw JT, Lovelock PK, Kesting JB, Cardinal J, Duffy D, Wainwright B, et al. Novel susceptibility gene for late-onset NIDDM is localized to human chromosome 12q. Diabetes 1998;47:1793–6.
4. Kinzler KW, Vogelstien B. Lessons from hereditary colorectal cancer. Cell 1996;87:159–70.
5. Thorlacius S, Struewing JP, Hartge P, Olafsdottir GH, Sigvaldason H, Tryggvadottir L, et al. Population-based study of risk of breast cancer in carriers of BRCA2 mutation. Lancet 1998;352:1337–9.
6. Tengs TO, Winer EP, Paddock S, Aguilar-Chavez O, Berry DA. Testing for the BRCA1 and BRCA2 breast-ovarian cancer susceptibility genes: A decision analysis. Med Decis Making 1998;18:365–75.
7. Collins FS, Bochm K. Avoiding casualties in the genetic revolution: The urgent need to educate physicians about genetics. Acad Med 1999;74:48–9.
8. Touchette N, Holtzman NA, Davis JG, Feetham S. Toward the 21st century: Incorporating genetics into primary health care. Plainview, New York: Cold Spring Harbor Laboratory Press, 1997.
9. Hunter A. Wright P, Cappelli M, Kasaboski A, Surh L. Physician knowledge and attitudes towards molecular genetic (DNA) testing of their patients. Clin Genet 1998;53:447–55.
10. Wilkins-Haug L, Hill L, Schmidt L, Holzman GB, Schulkin J. Genetics in obstetricians' offices: A survey study. Obstet Gynecol 1999;93:642–7.