- Email: [email protected]
Genetics in obstetricians’ offices: a survey study
Genetics in Obstetricians’ Offices: A Survey Study LOUISE WILKINS-HAUG, MD, PhD, LAUREN HILL, LOUIS SCHMIDT, PhD, GERALD B. HOLZMAN, MD, AND JAY SCHULKIN, PhD Objective: To investigate obstetricians’ genetic knowledge base and practice trends. Methods: A questionnaire survey was sent to 1003 ACOG Fellows, 554 (55%) of whom responded. Results from the 446 respondents practicing obstetrics are reported. Results: The majority of obstetricians surveyed (85.6%) reported completing standardized genetic-history forms for prenatal patients, and about half (48%) performed their own invasive diagnostic procedures. Most (87%) had access to genetic counselors. For aneuploidy risks associated with advanced maternal age, up to 69% of respondents provided at least some patient counseling in their offices. Physician knowledge of risk assessment and diagnostic testing in the areas of aneuploidy and neural tube defects was very good; however, for single-gene disorders such as cystic fibrosis, Tay-Sachs disease, and sickle cell disease, correct risk assessment or appropriate test selection presented difficulties for at least half of the respondents. Respondents cited the rapidity of changes in genetic testing as the greatest obstacle to providing genetic information to patients. Conclusion: Obstetricians’ knowledge of inheritance and test selection pertaining to single-gene disorders was more limited than that for aneuploidy and neural tube defects. Comparable deficits were noted in patient-education efforts for single-gene disorders. (Obstet Gynecol 1999;93:642–7. © 1999 by The American College of Obstetricians and Gynecologists.)
Contemporary scientific theory and research, exemplified by the Human Genome Project, have direct and profound implications for the practice of medicine. Complete mapping of the human genome will affect the clinical practice of obstetrics and gynecology. Genetic From the Department of Obstetrics and Gynecology, Brigham and Women’s Hospital, Harvard Medical School, Boston, Massachusetts; The American College of Obstetricians and Gynecologists, Washington, DC; and the Department of Psychology, University of Maryland College Park, College Park, Maryland. Supported by a grant (MCJ 117016) from the United States Department of Health and Human Services Bureau of Maternal and Child Health.
642 0029-7844/99/$20.00 PII S0029-7844(98)00529-8
alterations associated with an increasing number of inherited diseases, both rare and common, are being classified. Biomedical advances are increasing the availability of tests that provide women and their families with genetic information about carrier detection and prenatal diagnosis. For example, scientific advances within the last decade have influenced DNA testing for cystic fibrosis. The 1991 ACOG Committee Opinion on cystic fibrosis carrier screening stated that until the mid-1980s, “families . . . at increased risk for cystic fibrosis could be offered only genetic counseling. . . . ” Today, however, accurate testing for cystic fibrosis is available for carriers and affected individuals. After publication of the 1997 National Institutes of Health Consensus Statement on genetic testing for cystic fibrosis,1 researchers have considered many issues related to the clinical implications of cystic fibrosis testing and diagnosis.2–5 The challenge of practicing obstetrician-gynecologists to keep up with the array of topics related to medical genetics is multifaceted. Wider and more rapid dissemination of genetic information to physicians and patients will be necessary to ensure quality care. We surveyed a sample of ACOG Fellows practicing obstetrics. Our aims were to investigate current practice trends in genetics in obstetricians’ offices, to identify areas of strength and weakness in physicians’ knowledge base, and to find a potential resource for disseminating genetic information to practicing obstetricians.
Materials and Methods The Collaborative Ambulatory Research Network was established to facilitate the assessment of prevailing patterns in obstetric-gynecologic clinical practice and to help develop professional education where needed. The members of the Network, chosen to be demographically representative of ACOG Fellows in age and sex, are obstetrician-gynecologists who voluntarily participate
Obstetrics & Gynecology
in questionnaire studies such as the present investigation. Questionnaires were mailed in September 1997 to all 253 members of the Network and to a control group of 750 ACOG Fellows who are not Network members. We chose a total sample size based on similar studies in which comparable numbers of surveys were mailed and returned. Reliable effects have been found using sample sizes of approximately 1000. Questionnaires were mailed concurrently to the 750 controls to ensure that Network Fellows’ responses were statistically similar to those of ACOG Fellows in general. To maximize the likelihood that the control group would be representative of the entire population of ACOG Fellows, we oversampled from this group. Non-Network Fellows were sampled randomly, by computer, from the subset of ACOG Fellows who had not been selected previously as a part of a Network study control group. These Fellows were precluded from participation in subsequent studies to avoid overburdening individual ACOG Fellows with questionnaire surveys. Before administering the survey, we conducted 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 approximately 100. We evaluated significance at .05, and all effect sizes were above 1.0. Although we received a sufficient number of completed surveys after our September mailing, we sent a second mailing to all nonrespondents in October 1997 to increase our overall response rate. The survey contained primarily multiple-choice questions about physicians’ practice policies and genetics-related knowledge. Obstetricians were asked to rank the usefulness of various sources of genetic information, and an open-ended question was provided for them to state their educational needs. The questionnaire was divided into four sections: 1) physician knowledge and practices pertaining to patient risk factors, test selection, genetic counseling and referrals, and patient education; 2) ethical considerations (eg, confidentiality of testing); 3) physician education; and 4) physician demographics. The first section comprised several subsections delineated by specific genetic diseases and disorders. Data were analyzed using a personal computer– based social sciences statistical package (SPSS 4.1; SPSS Inc., Chicago, IL). Descriptive statistics were computed for measures used in the analyses. Group differences in responses to continuous measures were assessed using analysis of variance. Differences in categoric measures were assessed using x2. All analyses were examined using an alpha of .05 and 95% confidence intervals. Data were analyzed for group differences in question-
VOL. 93, NO. 5, PART 1, MAY 1999
naire item responses, using a two-way analysis of variance with sex (male versus female) and fellowship status (Collaborative Ambulatory Research Network member versus nonmember) as between-subject factors. We hypothesized that other variables may also affect questionnaire responses. Therefore, we assessed differences in responses between Fellows who used standard or structured sets of questions on genetic risks and those who did not, and between those who did their own invasive prenatal genetic procedures and those who did not. Two groups of Fellows were created by median split: those who had completed specialty training 11 or fewer years ago and those who had completed specialty training more than 11 years ago. Each of these variables was entered as a separate factor in the analyses.
Results Data from respondents who returned the questionnaire within approximately 3 months were included in the analysis. Five hundred fifty-four (55%) of the 1003 questionnaires mailed were returned within the allotted time. Response rates were 81% for the Collaborative Ambulatory Research Network group and 47% for the controls. About 20% of the respondents did not practice obstetrics, leaving a total of 446 (44%) eligible questionnaires. Approximately 54% of the respondents were men and 44% were women. Most practiced in obstetricgynecologic partnerships or groups (50%) or in solo practices (18%). Respondents’ average (mean) age was 43 years (median 41 years), and they had completed their specialty training an average (mean) of 13 years ago (median 11 years). Fifty percent of their practices were obstetric, with an average of 21 deliveries per respondent each month. A majority (85.6%) of the obstetrician-gynecologists had their patients answer a standard or structured set of questions about genetic risks. There were no significant differences in their responses of Collaborative Ambulatory Research Network Fellows and controls (Table 1). There were no significant main effects and no interactions for sex and fellowship status, so the responses of men and women as well as of Collaborative Ambulatory Research Network Fellows and controls were collapsed across factors in subsequent analyses. About half of the respondents (48%) reported performing their own invasive prenatal diagnostic procedures. Specifically, 50% did not perform any, 44% performed amniocentesis only, 4% performed amniocentesis and chorionic villus sampling (CVS), and none performed CVS only. Among the physicians who reported having access to genetic counselors (87%), 21%
Wilkins-Haug et al
Obstetric Genetics
643
Table 1. Demographics Characteristic
Network*
Controls
Mean age Male (%) Mean year respondents completed residency Mean no. of deliveries per month per respondent Mean percentage of practice that is obstetrics Practice type (%) Partnership or group Solo practice Multispecialty group HMO, staff model University-based
39 57 1983
41 53 1986
20
22
50
54
49 24 11 6 9
50 17 11 5 11
†
HMO 5 health maintenance organization. * Collaborative Ambulatory Research Network Fellows. † Non–Collaborative Ambulatory Research Network Fellows.
of the counselors were on site and 66% were in the community; 11% were available for telephone consultation only. The rapidity of advances in testing for genetic diseases was rated as the greatest obstacle to providing patients with information about genetic issues (3.0 on a scale of 1 5 “not a significant factor” to 5 5 “highly significant factor”). This was followed by constraints of office time to adequately address women’s genetic concerns (2.88), lack of available patient-oriented literature (2.33), and lack of insurance for genetic counselors (2.23). Obstetrician-gynecologists provided printed information to patients on a limited number of issues (Table 2). Fifty percent of the obstetricians correctly identified the approximate carrier frequency of cystic fibrosis in the North American white population, at roughly one in 30; 44% underestimated the frequency at one in 50 or less. Recently trained physicians were more likely to estimate accurately than those who were trained less recently (x24 40.61, P , .001). Most respondents (83%) Table 2. Ability to Provide Patient Literature Patient literature topic
Percentage of obstetricians able to provide patient literature
MSAFP screen for aneuploidy Amniocentesis MSAFP screen for spina bifida Advanced maternal age Chorionic villus sampling Tay-Sachs disease Cystic fibrosis Hereditary hemoglobinopathies Fragile X syndrome
92 90 88 80 60 33 33 32 21
MSAFP 5 maternal serum alpha-fetoprotein.
644 Wilkins-Haug et al
Obstetric Genetics
would recommend testing to individuals or couples with a positive family history of cystic fibrosis. Ninetyseven percent of all respondents would not recommend testing to women with a negative family history of cystic fibrosis. About half (51%) of obstetricians correctly chose the percentage of carriers of cystic fibrosis in a white population who can be identified by screening for the common DNA mutation, delta F508, at 75%; 28% overestimated and 20% underestimated. Physicians who did their own invasive genetic diagnostic procedures were more likely to identify correctly the sensitivity of the delta F508 screen than those who did not perform their own procedures (x23 8.9, P 5 .03). Seventy percent of obstetricians would refer a patient with a family history of cystic fibrosis (eg, affected sibling) to a genetic counselor or medical geneticist for carrier testing. Seventy-five percent of obstetricians correctly stated that it was best to use electrophoresis to ensure accurate hemoglobin identification for genetic counseling when initially screening at-risk women for heritable hemoglobinopathies. Forty-two percent of obstetricians correctly recommended hemoglobin electrophoresis for partners of black prenatal women known to be sickle cell carriers. Forty percent erroneously recommended a solubility test (eg, Sickledex; Ortho Diagnostics, Raritan, NJ) for the partner. Respondents who did their own invasive genetic diagnostic procedures were more likely to suggest correctly hemoglobin electrophoresis and were less likely to suggest incorrectly a solubility test for the partner than those who did not perform their own procedures (x23 18.76, P , .001). Seventeen percent of the respondents reported that they would refer that same couple to a genetic counselor for further evaluation of the partner. Forty-one percent of obstetricians correctly estimated the carrier frequency of Tay-Sachs disease in Jews of eastern European descent (Ashkenazi) at approximately one in 30; 53% underestimated carrier frequency at one in 60 or less. Physicians who completed residency training more than 11 years ago were more likely to identify the correct carrier frequency of Tay-Sachs in this population (x24 11.30, P 5 .04). Ninety-eight percent of the respondents reported that they would offer carrier testing for Tay-Sachs disease to a couple of Ashkenazi Jewish heritage; 6% would offer testing for Canavan disease, 11.9% would offer testing for cystic fibrosis, and 23.8% would offer testing for Gaucher disease. Only 18.7% of obstetricians correctly responded that the most accurate test for Tay-Sachs disease carrier screening during pregnancy, or for women taking oral contraceptives, is leukocyte hexosamidase A. For the other options presented, 7.7% of respondents indicated serum hexosamidase A testing,
Obstetrics & Gynecology
8.7% said DNA analysis of the alpha subunit of TaySachs gene, and almost 65% said that they did not know. Thirty-eight percent of respondents correctly answered a question pertaining to situations that would not put a woman at increased risk for a child with fragile X syndrome, by choosing “her brother has a son with fragile X.” Other choices available for placing women at risk were chosen erroneously as risk-free by some of the respondents: “Her father’s brother has fragile X syndrome” (chosen by 36.7%), “her brother has fragile X syndrome” (chosen by 2.5%), and “her sister has fragile X syndrome” (chosen by 3.2%). About 41% said that they did not know. For a woman with a family history of fragile X syndrome (affected sibling), most obstetricians (72%) reported that they would refer her to a genetic counselor or medical geneticist for carrier testing. Approximately 55% of all respondents correctly identified DNA studies for fragile X fragment length as the most sensitive test for determination of the carrier state from among two other options. Respondents who used standard or structured sets of questions about genetic risk were more likely to answer this question correctly (x22 8.65, P 5 .01). Almost all respondents (94.4%) correctly answered that a previous pregnancy with anencephaly puts women at a higher than background risk of neural tube defects. A majority also indicated that maternal insulindependent diabetes (68.9%) and first-trimester exposure to valproic acid (71.1%) increased the risk of neural tube defects. Fewer indicated erroneously that maternal alcohol consumption (14.4%) and previous pregnancy with trisomy 18 (16.45%) increased the risk of neural tube defects. Approximately 87% of respondents indicated that maternal age of 40 years or more placed women at a 1% or greater risk for trisomy 21 in the second trimester. Ninety-three percent correctly selected karyotype of amniocytes as the most sensitive test for detecting trisomy 21 from among five options. Approximately 69% of obstetricians reported that they or their office staff provided prenatal counseling about maternal age and the risk of aneuploidy. When obstetricians provided counseling on advanced maternal age and amniocentesis, the overwhelming majority discussed the age-related risk of trisomy 21 (99%), the risks of the procedure (98%), and options in the event of abnormal results (95%). Fewer discussed the age-related risks of other aneuploidies (88%) or the probability of mosaicism, markers, or rearrangements (40%). Thirty-one percent referred all older patients to a genetic specialist, and 24% referred only those who decided to have prenatal testing. Respondents who performed their own procedures were more likely to counsel and less
VOL. 93, NO. 5, PART 1, MAY 1999
likely to refer older prenatal patients than those who did not (x22 73.45, P , .001). Thirty-nine percent of obstetricians reported that they routinely informed women about who would have access to prenatal test results, 23% rarely informed women, and 34% did not inform women at all. Fifty-one percent reported having no policy in place for disclosure of clinically significant genetic studies when requested by family members other than a spouse. In response to a set of options listed on the questionnaire, physicians indicated the services available to their patients either at the physicians’ office or at a referral site when a fetal abnormality was detected by amniocentesis. Ninety-eight percent offered appointments to discuss results, options, and expected outcomes; 69.2% offered separate appointments at a later date for discussion of recurrence risk; 65% offered appointments with a grief or support counselor; and 62.8% made themselves available by telephone 24 hours a day for the affected woman or family. When asked what resources they had used for their patients over the past year, 39% of respondents reported using a local chapter of a parent support group, 24% used hotlines, or teratogen databases, and 13% used a local chapter of the March of Dimes. Physicians ranked a group of options from 1 5 “use most often” to 8 5 “use least often” to indicate how they stay informed about advances in genetic research, practice trends, and ethical issues. The ACOG Educational Bulletins were ranked first, followed by a tie between professional journals and ACOG Committee Opinions, then college postgraduate courses, textbooks, other continuing medical education activities, American College of Medical Genetics Opinions, and finally “other.” When asked about their preferences for ACOG-produced materials on genetic information, they ranked monographs as most useful, followed by courses, videos, compact disks for computers, and finally “other.”
Discussion Our study results suggest several trends in obstetricians’ practice patterns and knowledge base pertaining to genetics. Overall, the respondents performed a majority of genetic care in their offices whether or not they had access to genetic specialists. This was true for both obstetricians who performed their own invasive diagnostic procedures and those who referred women for diagnostic testing. When statistically significant differences in knowledge were found, physicians who did their own procedures were more knowledgeable than other respondents. Physician knowledge was very good regarding risk assessment and testing for neural tube defects and risk
Wilkins-Haug et al
Obstetric Genetics
645
assessment for certain aneuploidies associated with advanced maternal age. Such knowledge is well established in obstetric practices. More recent knowledge of genetic risk factors and newer testing methods were less familiar to practicing obstetricians. Correct assessment of a patient’s risk based on population frequency of the carrier state occurred half of the time or less, regardless of the disease. Correct risk assessment based on family history varied by disease. Knowledge of which tests to order for carrier screening presented the greatest difficulty. Many respondents could not correctly identify ACOG-recommended carrier tests for sickle cell, Tay-Sachs, and fragile X syndrome.6 – 8 Only 6% offered testing for Canavan disease to women of Ashkenazi Jewish descent. As a result of recent research, the American College of Medical Genetics now supports Canavan disease as a disorder with sufficient frequency and severity to warrant carrier screening in this population.9 Obstetricians ideally should have a working knowledge of single-gene disorders equal to their knowledge of aneuploidy and neural tube defects. These difficulties are not unique to obstetricians, however; findings were similar when DNA testing was initiated for dominantly inherited familial adenomatous polyposis in internists’ offices.10 Patient education is an integral part of counseling and informed consent. Patient pamphlets can readily supplement education by physicians or their staff. Only one-third or fewer respondents in our survey were able to offer written educational material for various singlegene disorders. In contrast, most physicians were able to provide women with adequate written material about genetic concerns such as neural tube defects and Down syndrome. The American Society of Human Genetics recently released a statement addressing issues of genetic testing, confidentiality, and obligation to other family members.11 Survey respondents’ policies for disclosure of test results were inconsistent, with few established avenues to address these issues in their practices. Regarding resources to supplement obstetric care, respondents reported using local support groups or, less frequently, other resources such as the March of Dimes. Only a minority of respondents used nationwide teratogen or DNA databases. Increased awareness of available information and services provides one avenue for supplementing office-based counseling and physician education. Affiliation with genetic counseling centers for education and testing programs could complement office-based obstetrics practices. An effort to ensure ongoing genetic education for health care providers has been tied closely to the scientific initiative to map the human genome.12,13 Our survey identified both strengths and weaknesses in
646 Wilkins-Haug et al
Obstetric Genetics
practitioners’ genetic knowledge. The results suggest that obstetricians need more information about who is at risk for a variety of disorders, which tests are available and appropriate, and how to interpret test results. To bring their knowledge about single-gene disorders up to the level of their knowledge regarding the risks of aneuploidy and neural tube defects, obstetricians will need to develop office-based strategies for retrieval and dissemination of information about singlegene disorders. Affiliation with genetic counseling centers could facilitate education and referral. Providing quality educational opportunities for physicians and their patients is a challenge for obstetrics practices, ACOG, and educators in general.
References 1. Genetic testing for cystic fibrosis. National Institutes of Health Consensus Statement 1997 Apr 14 –16;15(4):1–37. 2. Asch DA, Hershey JC, Dekay ML, Pauly MV, Patton JP, Jedrziewski MK, et al. Carrier screening for cystic fibrosis: Costs and clinical outcomes. Med Decis Making 1998;18:202–12. 3. Asch DA, Hershey JC, Pauly MV, Patton JP, Jedrziewski MK, Mennuti MT. Genetic screening for reproductive planning: Methodological and conceptual issues in policy analysis. Am J Public Health 1996;86:684 –90. 4. Asch DA, Patton JP, Hershey JC, Mennuti MT. Reporting the results of cystic fibrosis carrier screening. Am J Obstet Gynecol 1993;168:1– 6. 5. Grody W, Dunkel-Schetter C, Tatsugawa ZH, Fox MA, Fang CY, Cantor RM, et al. PCR-based screening for cystic fibrosis carrier mutations in an ethnically diverse pregnant population. Am J Hum Genet 1997;60:935– 47. 6. American College of Obstetricians and Gynecologists. Genetic screening for hemoglobinopathies. ACOG committee opinion no. 168. Washington, DC: American College of Obstetricians and Gynecologists, 1996. 7. American College of Obstetricians and Gynecologists. Screening for Tay-Sachs disease. ACOG committee opinion no. 162. Washington, DC: American College of Obstetricians and Gynecologists, 1995. 8. American College of Obstetricians and Gynecologists. Fragile X syndrome. ACOG committee opinion no. 161. Washington, DC: American College of Obstetricians and Gynecologists, 1995. 9. American College of Medical Genetics. Position statement on carrier testing for Canavan disease. Bethesda, Maryland: American College of Medical Genetics, 1998. 10. Giardiello FM, Brensinger JD, Petersen GM, Luce MC, Hylind LM, Bacon JA, et al. The use and interpretation of commercial APC gene testing for familial adenomatous polyposis. N Engl J Med 1997;336:823–7. 11. American College of Medical Genetics statement. Professional disclosure of familial genetic information. The American College of Human Genetics Social Issues Subcommittee on Familial Disclosure. Am J Hum Genet 1998;62:474 – 83. 12. 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. 13. Bernhardt BA, Geller G, Doksum T, Larson SM, Roter D, Holtzman NA. Prenatal genetic testing: Content of discussions between
Obstetrics & Gynecology
obstetric providers and pregnant women. Obstet Gynecol 1998;91: 648 –55.
Received August 4, 1998. Received in revised form October 26, 1998. Accepted November 13, 1998.
Address reprint requests to:
Jay Schulkin, PhD Department of Research American College of Obstetricians and Gynecologists 409 12th Street, SW Washington, DC 20024-2188
WRITING
Copyright © 1999 by The American College of Obstetricians and Gynecologists. Published by Elsevier Science Inc.
GUIDE AVAILABLE
To help prospective authors, especially those who are beginning in medical journal writing, to produce better papers, this journal has developed “A Guide to Writing for Obstetrics & Gynecology.” Copies of the second edition of this booklet are available, free of charge, from: Editorial Office, Obstetrics & Gynecology, 10921 Wilshire Boulevard, Suite 403, Los Angeles, CA 90024-3908. Requests may also be submitted by FAX: (310) 208-2838.
VOL. 93, NO. 5, PART 1, MAY 1999
Wilkins-Haug et al
Obstetric Genetics
647
Contemporary scientific theory and research, exemplified by the Human Genome Project, have direct and profound implications for the practice of medicine. Complete mapping of the human genome will affect the clinical practice of obstetrics and gynecology. Genetic From the Department of Obstetrics and Gynecology, Brigham and Women’s Hospital, Harvard Medical School, Boston, Massachusetts; The American College of Obstetricians and Gynecologists, Washington, DC; and the Department of Psychology, University of Maryland College Park, College Park, Maryland. Supported by a grant (MCJ 117016) from the United States Department of Health and Human Services Bureau of Maternal and Child Health.
642 0029-7844/99/$20.00 PII S0029-7844(98)00529-8
alterations associated with an increasing number of inherited diseases, both rare and common, are being classified. Biomedical advances are increasing the availability of tests that provide women and their families with genetic information about carrier detection and prenatal diagnosis. For example, scientific advances within the last decade have influenced DNA testing for cystic fibrosis. The 1991 ACOG Committee Opinion on cystic fibrosis carrier screening stated that until the mid-1980s, “families . . . at increased risk for cystic fibrosis could be offered only genetic counseling. . . . ” Today, however, accurate testing for cystic fibrosis is available for carriers and affected individuals. After publication of the 1997 National Institutes of Health Consensus Statement on genetic testing for cystic fibrosis,1 researchers have considered many issues related to the clinical implications of cystic fibrosis testing and diagnosis.2–5 The challenge of practicing obstetrician-gynecologists to keep up with the array of topics related to medical genetics is multifaceted. Wider and more rapid dissemination of genetic information to physicians and patients will be necessary to ensure quality care. We surveyed a sample of ACOG Fellows practicing obstetrics. Our aims were to investigate current practice trends in genetics in obstetricians’ offices, to identify areas of strength and weakness in physicians’ knowledge base, and to find a potential resource for disseminating genetic information to practicing obstetricians.
Materials and Methods The Collaborative Ambulatory Research Network was established to facilitate the assessment of prevailing patterns in obstetric-gynecologic clinical practice and to help develop professional education where needed. The members of the Network, chosen to be demographically representative of ACOG Fellows in age and sex, are obstetrician-gynecologists who voluntarily participate
Obstetrics & Gynecology
in questionnaire studies such as the present investigation. Questionnaires were mailed in September 1997 to all 253 members of the Network and to a control group of 750 ACOG Fellows who are not Network members. We chose a total sample size based on similar studies in which comparable numbers of surveys were mailed and returned. Reliable effects have been found using sample sizes of approximately 1000. Questionnaires were mailed concurrently to the 750 controls to ensure that Network Fellows’ responses were statistically similar to those of ACOG Fellows in general. To maximize the likelihood that the control group would be representative of the entire population of ACOG Fellows, we oversampled from this group. Non-Network Fellows were sampled randomly, by computer, from the subset of ACOG Fellows who had not been selected previously as a part of a Network study control group. These Fellows were precluded from participation in subsequent studies to avoid overburdening individual ACOG Fellows with questionnaire surveys. Before administering the survey, we conducted 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 approximately 100. We evaluated significance at .05, and all effect sizes were above 1.0. Although we received a sufficient number of completed surveys after our September mailing, we sent a second mailing to all nonrespondents in October 1997 to increase our overall response rate. The survey contained primarily multiple-choice questions about physicians’ practice policies and genetics-related knowledge. Obstetricians were asked to rank the usefulness of various sources of genetic information, and an open-ended question was provided for them to state their educational needs. The questionnaire was divided into four sections: 1) physician knowledge and practices pertaining to patient risk factors, test selection, genetic counseling and referrals, and patient education; 2) ethical considerations (eg, confidentiality of testing); 3) physician education; and 4) physician demographics. The first section comprised several subsections delineated by specific genetic diseases and disorders. Data were analyzed using a personal computer– based social sciences statistical package (SPSS 4.1; SPSS Inc., Chicago, IL). Descriptive statistics were computed for measures used in the analyses. Group differences in responses to continuous measures were assessed using analysis of variance. Differences in categoric measures were assessed using x2. All analyses were examined using an alpha of .05 and 95% confidence intervals. Data were analyzed for group differences in question-
VOL. 93, NO. 5, PART 1, MAY 1999
naire item responses, using a two-way analysis of variance with sex (male versus female) and fellowship status (Collaborative Ambulatory Research Network member versus nonmember) as between-subject factors. We hypothesized that other variables may also affect questionnaire responses. Therefore, we assessed differences in responses between Fellows who used standard or structured sets of questions on genetic risks and those who did not, and between those who did their own invasive prenatal genetic procedures and those who did not. Two groups of Fellows were created by median split: those who had completed specialty training 11 or fewer years ago and those who had completed specialty training more than 11 years ago. Each of these variables was entered as a separate factor in the analyses.
Results Data from respondents who returned the questionnaire within approximately 3 months were included in the analysis. Five hundred fifty-four (55%) of the 1003 questionnaires mailed were returned within the allotted time. Response rates were 81% for the Collaborative Ambulatory Research Network group and 47% for the controls. About 20% of the respondents did not practice obstetrics, leaving a total of 446 (44%) eligible questionnaires. Approximately 54% of the respondents were men and 44% were women. Most practiced in obstetricgynecologic partnerships or groups (50%) or in solo practices (18%). Respondents’ average (mean) age was 43 years (median 41 years), and they had completed their specialty training an average (mean) of 13 years ago (median 11 years). Fifty percent of their practices were obstetric, with an average of 21 deliveries per respondent each month. A majority (85.6%) of the obstetrician-gynecologists had their patients answer a standard or structured set of questions about genetic risks. There were no significant differences in their responses of Collaborative Ambulatory Research Network Fellows and controls (Table 1). There were no significant main effects and no interactions for sex and fellowship status, so the responses of men and women as well as of Collaborative Ambulatory Research Network Fellows and controls were collapsed across factors in subsequent analyses. About half of the respondents (48%) reported performing their own invasive prenatal diagnostic procedures. Specifically, 50% did not perform any, 44% performed amniocentesis only, 4% performed amniocentesis and chorionic villus sampling (CVS), and none performed CVS only. Among the physicians who reported having access to genetic counselors (87%), 21%
Wilkins-Haug et al
Obstetric Genetics
643
Table 1. Demographics Characteristic
Network*
Controls
Mean age Male (%) Mean year respondents completed residency Mean no. of deliveries per month per respondent Mean percentage of practice that is obstetrics Practice type (%) Partnership or group Solo practice Multispecialty group HMO, staff model University-based
39 57 1983
41 53 1986
20
22
50
54
49 24 11 6 9
50 17 11 5 11
†
HMO 5 health maintenance organization. * Collaborative Ambulatory Research Network Fellows. † Non–Collaborative Ambulatory Research Network Fellows.
of the counselors were on site and 66% were in the community; 11% were available for telephone consultation only. The rapidity of advances in testing for genetic diseases was rated as the greatest obstacle to providing patients with information about genetic issues (3.0 on a scale of 1 5 “not a significant factor” to 5 5 “highly significant factor”). This was followed by constraints of office time to adequately address women’s genetic concerns (2.88), lack of available patient-oriented literature (2.33), and lack of insurance for genetic counselors (2.23). Obstetrician-gynecologists provided printed information to patients on a limited number of issues (Table 2). Fifty percent of the obstetricians correctly identified the approximate carrier frequency of cystic fibrosis in the North American white population, at roughly one in 30; 44% underestimated the frequency at one in 50 or less. Recently trained physicians were more likely to estimate accurately than those who were trained less recently (x24 40.61, P , .001). Most respondents (83%) Table 2. Ability to Provide Patient Literature Patient literature topic
Percentage of obstetricians able to provide patient literature
MSAFP screen for aneuploidy Amniocentesis MSAFP screen for spina bifida Advanced maternal age Chorionic villus sampling Tay-Sachs disease Cystic fibrosis Hereditary hemoglobinopathies Fragile X syndrome
92 90 88 80 60 33 33 32 21
MSAFP 5 maternal serum alpha-fetoprotein.
644 Wilkins-Haug et al
Obstetric Genetics
would recommend testing to individuals or couples with a positive family history of cystic fibrosis. Ninetyseven percent of all respondents would not recommend testing to women with a negative family history of cystic fibrosis. About half (51%) of obstetricians correctly chose the percentage of carriers of cystic fibrosis in a white population who can be identified by screening for the common DNA mutation, delta F508, at 75%; 28% overestimated and 20% underestimated. Physicians who did their own invasive genetic diagnostic procedures were more likely to identify correctly the sensitivity of the delta F508 screen than those who did not perform their own procedures (x23 8.9, P 5 .03). Seventy percent of obstetricians would refer a patient with a family history of cystic fibrosis (eg, affected sibling) to a genetic counselor or medical geneticist for carrier testing. Seventy-five percent of obstetricians correctly stated that it was best to use electrophoresis to ensure accurate hemoglobin identification for genetic counseling when initially screening at-risk women for heritable hemoglobinopathies. Forty-two percent of obstetricians correctly recommended hemoglobin electrophoresis for partners of black prenatal women known to be sickle cell carriers. Forty percent erroneously recommended a solubility test (eg, Sickledex; Ortho Diagnostics, Raritan, NJ) for the partner. Respondents who did their own invasive genetic diagnostic procedures were more likely to suggest correctly hemoglobin electrophoresis and were less likely to suggest incorrectly a solubility test for the partner than those who did not perform their own procedures (x23 18.76, P , .001). Seventeen percent of the respondents reported that they would refer that same couple to a genetic counselor for further evaluation of the partner. Forty-one percent of obstetricians correctly estimated the carrier frequency of Tay-Sachs disease in Jews of eastern European descent (Ashkenazi) at approximately one in 30; 53% underestimated carrier frequency at one in 60 or less. Physicians who completed residency training more than 11 years ago were more likely to identify the correct carrier frequency of Tay-Sachs in this population (x24 11.30, P 5 .04). Ninety-eight percent of the respondents reported that they would offer carrier testing for Tay-Sachs disease to a couple of Ashkenazi Jewish heritage; 6% would offer testing for Canavan disease, 11.9% would offer testing for cystic fibrosis, and 23.8% would offer testing for Gaucher disease. Only 18.7% of obstetricians correctly responded that the most accurate test for Tay-Sachs disease carrier screening during pregnancy, or for women taking oral contraceptives, is leukocyte hexosamidase A. For the other options presented, 7.7% of respondents indicated serum hexosamidase A testing,
Obstetrics & Gynecology
8.7% said DNA analysis of the alpha subunit of TaySachs gene, and almost 65% said that they did not know. Thirty-eight percent of respondents correctly answered a question pertaining to situations that would not put a woman at increased risk for a child with fragile X syndrome, by choosing “her brother has a son with fragile X.” Other choices available for placing women at risk were chosen erroneously as risk-free by some of the respondents: “Her father’s brother has fragile X syndrome” (chosen by 36.7%), “her brother has fragile X syndrome” (chosen by 2.5%), and “her sister has fragile X syndrome” (chosen by 3.2%). About 41% said that they did not know. For a woman with a family history of fragile X syndrome (affected sibling), most obstetricians (72%) reported that they would refer her to a genetic counselor or medical geneticist for carrier testing. Approximately 55% of all respondents correctly identified DNA studies for fragile X fragment length as the most sensitive test for determination of the carrier state from among two other options. Respondents who used standard or structured sets of questions about genetic risk were more likely to answer this question correctly (x22 8.65, P 5 .01). Almost all respondents (94.4%) correctly answered that a previous pregnancy with anencephaly puts women at a higher than background risk of neural tube defects. A majority also indicated that maternal insulindependent diabetes (68.9%) and first-trimester exposure to valproic acid (71.1%) increased the risk of neural tube defects. Fewer indicated erroneously that maternal alcohol consumption (14.4%) and previous pregnancy with trisomy 18 (16.45%) increased the risk of neural tube defects. Approximately 87% of respondents indicated that maternal age of 40 years or more placed women at a 1% or greater risk for trisomy 21 in the second trimester. Ninety-three percent correctly selected karyotype of amniocytes as the most sensitive test for detecting trisomy 21 from among five options. Approximately 69% of obstetricians reported that they or their office staff provided prenatal counseling about maternal age and the risk of aneuploidy. When obstetricians provided counseling on advanced maternal age and amniocentesis, the overwhelming majority discussed the age-related risk of trisomy 21 (99%), the risks of the procedure (98%), and options in the event of abnormal results (95%). Fewer discussed the age-related risks of other aneuploidies (88%) or the probability of mosaicism, markers, or rearrangements (40%). Thirty-one percent referred all older patients to a genetic specialist, and 24% referred only those who decided to have prenatal testing. Respondents who performed their own procedures were more likely to counsel and less
VOL. 93, NO. 5, PART 1, MAY 1999
likely to refer older prenatal patients than those who did not (x22 73.45, P , .001). Thirty-nine percent of obstetricians reported that they routinely informed women about who would have access to prenatal test results, 23% rarely informed women, and 34% did not inform women at all. Fifty-one percent reported having no policy in place for disclosure of clinically significant genetic studies when requested by family members other than a spouse. In response to a set of options listed on the questionnaire, physicians indicated the services available to their patients either at the physicians’ office or at a referral site when a fetal abnormality was detected by amniocentesis. Ninety-eight percent offered appointments to discuss results, options, and expected outcomes; 69.2% offered separate appointments at a later date for discussion of recurrence risk; 65% offered appointments with a grief or support counselor; and 62.8% made themselves available by telephone 24 hours a day for the affected woman or family. When asked what resources they had used for their patients over the past year, 39% of respondents reported using a local chapter of a parent support group, 24% used hotlines, or teratogen databases, and 13% used a local chapter of the March of Dimes. Physicians ranked a group of options from 1 5 “use most often” to 8 5 “use least often” to indicate how they stay informed about advances in genetic research, practice trends, and ethical issues. The ACOG Educational Bulletins were ranked first, followed by a tie between professional journals and ACOG Committee Opinions, then college postgraduate courses, textbooks, other continuing medical education activities, American College of Medical Genetics Opinions, and finally “other.” When asked about their preferences for ACOG-produced materials on genetic information, they ranked monographs as most useful, followed by courses, videos, compact disks for computers, and finally “other.”
Discussion Our study results suggest several trends in obstetricians’ practice patterns and knowledge base pertaining to genetics. Overall, the respondents performed a majority of genetic care in their offices whether or not they had access to genetic specialists. This was true for both obstetricians who performed their own invasive diagnostic procedures and those who referred women for diagnostic testing. When statistically significant differences in knowledge were found, physicians who did their own procedures were more knowledgeable than other respondents. Physician knowledge was very good regarding risk assessment and testing for neural tube defects and risk
Wilkins-Haug et al
Obstetric Genetics
645
assessment for certain aneuploidies associated with advanced maternal age. Such knowledge is well established in obstetric practices. More recent knowledge of genetic risk factors and newer testing methods were less familiar to practicing obstetricians. Correct assessment of a patient’s risk based on population frequency of the carrier state occurred half of the time or less, regardless of the disease. Correct risk assessment based on family history varied by disease. Knowledge of which tests to order for carrier screening presented the greatest difficulty. Many respondents could not correctly identify ACOG-recommended carrier tests for sickle cell, Tay-Sachs, and fragile X syndrome.6 – 8 Only 6% offered testing for Canavan disease to women of Ashkenazi Jewish descent. As a result of recent research, the American College of Medical Genetics now supports Canavan disease as a disorder with sufficient frequency and severity to warrant carrier screening in this population.9 Obstetricians ideally should have a working knowledge of single-gene disorders equal to their knowledge of aneuploidy and neural tube defects. These difficulties are not unique to obstetricians, however; findings were similar when DNA testing was initiated for dominantly inherited familial adenomatous polyposis in internists’ offices.10 Patient education is an integral part of counseling and informed consent. Patient pamphlets can readily supplement education by physicians or their staff. Only one-third or fewer respondents in our survey were able to offer written educational material for various singlegene disorders. In contrast, most physicians were able to provide women with adequate written material about genetic concerns such as neural tube defects and Down syndrome. The American Society of Human Genetics recently released a statement addressing issues of genetic testing, confidentiality, and obligation to other family members.11 Survey respondents’ policies for disclosure of test results were inconsistent, with few established avenues to address these issues in their practices. Regarding resources to supplement obstetric care, respondents reported using local support groups or, less frequently, other resources such as the March of Dimes. Only a minority of respondents used nationwide teratogen or DNA databases. Increased awareness of available information and services provides one avenue for supplementing office-based counseling and physician education. Affiliation with genetic counseling centers for education and testing programs could complement office-based obstetrics practices. An effort to ensure ongoing genetic education for health care providers has been tied closely to the scientific initiative to map the human genome.12,13 Our survey identified both strengths and weaknesses in
646 Wilkins-Haug et al
Obstetric Genetics
practitioners’ genetic knowledge. The results suggest that obstetricians need more information about who is at risk for a variety of disorders, which tests are available and appropriate, and how to interpret test results. To bring their knowledge about single-gene disorders up to the level of their knowledge regarding the risks of aneuploidy and neural tube defects, obstetricians will need to develop office-based strategies for retrieval and dissemination of information about singlegene disorders. Affiliation with genetic counseling centers could facilitate education and referral. Providing quality educational opportunities for physicians and their patients is a challenge for obstetrics practices, ACOG, and educators in general.
References 1. Genetic testing for cystic fibrosis. National Institutes of Health Consensus Statement 1997 Apr 14 –16;15(4):1–37. 2. Asch DA, Hershey JC, Dekay ML, Pauly MV, Patton JP, Jedrziewski MK, et al. Carrier screening for cystic fibrosis: Costs and clinical outcomes. Med Decis Making 1998;18:202–12. 3. Asch DA, Hershey JC, Pauly MV, Patton JP, Jedrziewski MK, Mennuti MT. Genetic screening for reproductive planning: Methodological and conceptual issues in policy analysis. Am J Public Health 1996;86:684 –90. 4. Asch DA, Patton JP, Hershey JC, Mennuti MT. Reporting the results of cystic fibrosis carrier screening. Am J Obstet Gynecol 1993;168:1– 6. 5. Grody W, Dunkel-Schetter C, Tatsugawa ZH, Fox MA, Fang CY, Cantor RM, et al. PCR-based screening for cystic fibrosis carrier mutations in an ethnically diverse pregnant population. Am J Hum Genet 1997;60:935– 47. 6. American College of Obstetricians and Gynecologists. Genetic screening for hemoglobinopathies. ACOG committee opinion no. 168. Washington, DC: American College of Obstetricians and Gynecologists, 1996. 7. American College of Obstetricians and Gynecologists. Screening for Tay-Sachs disease. ACOG committee opinion no. 162. Washington, DC: American College of Obstetricians and Gynecologists, 1995. 8. American College of Obstetricians and Gynecologists. Fragile X syndrome. ACOG committee opinion no. 161. Washington, DC: American College of Obstetricians and Gynecologists, 1995. 9. American College of Medical Genetics. Position statement on carrier testing for Canavan disease. Bethesda, Maryland: American College of Medical Genetics, 1998. 10. Giardiello FM, Brensinger JD, Petersen GM, Luce MC, Hylind LM, Bacon JA, et al. The use and interpretation of commercial APC gene testing for familial adenomatous polyposis. N Engl J Med 1997;336:823–7. 11. American College of Medical Genetics statement. Professional disclosure of familial genetic information. The American College of Human Genetics Social Issues Subcommittee on Familial Disclosure. Am J Hum Genet 1998;62:474 – 83. 12. 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. 13. Bernhardt BA, Geller G, Doksum T, Larson SM, Roter D, Holtzman NA. Prenatal genetic testing: Content of discussions between
Obstetrics & Gynecology
obstetric providers and pregnant women. Obstet Gynecol 1998;91: 648 –55.
Received August 4, 1998. Received in revised form October 26, 1998. Accepted November 13, 1998.
Address reprint requests to:
Jay Schulkin, PhD Department of Research American College of Obstetricians and Gynecologists 409 12th Street, SW Washington, DC 20024-2188
WRITING
Copyright © 1999 by The American College of Obstetricians and Gynecologists. Published by Elsevier Science Inc.
GUIDE AVAILABLE
To help prospective authors, especially those who are beginning in medical journal writing, to produce better papers, this journal has developed “A Guide to Writing for Obstetrics & Gynecology.” Copies of the second edition of this booklet are available, free of charge, from: Editorial Office, Obstetrics & Gynecology, 10921 Wilshire Boulevard, Suite 403, Los Angeles, CA 90024-3908. Requests may also be submitted by FAX: (310) 208-2838.
VOL. 93, NO. 5, PART 1, MAY 1999
Wilkins-Haug et al
Obstetric Genetics
647