Genetic Services for Familial Cancer Patients: A Follow-Up Survey of National Cancer Institute Cancer Centers
http://www.100md.com
《临床肿瘤学》
the Cancer Prevention Program, Division of Public Health Sciences, Fred Hutchinson Cancer Research Center
Departments of Epidemiology and Medicine, University of Washington, Seattle, WA
ABSTRACT
PURPOSE: Anecdotal reports suggest that the volume of services offered to individuals concerned with hereditary cancer risk has increased substantially in recent years. As a follow-up to our 1993 survey, we sought to determine how the scope and volume of genetic services has changed between 1993 and 2002.
METHODS: We surveyed the 61 National Cancer Institute–designated cancer centers in operation in 2002 using an updated version of the questionnaire from 1993. Analysis included frequencies and summary statistics.
RESULTS: The majority of cancer centers responding (46 of 56 centers; 82.1%) provided some genetic services for evaluation of familial cancer, which is a higher proportion than in 1993 (50%; P < .01). Almost all centers (42 of 46 centers; 91.3%) provided services not only to cancer patients and their families, but also to individuals concerned with risk, which is a change (P = .01) from 1993, when 64.7% of centers offered such services. In addition, increases have been found for most other measures of services rendered for familial genetic services.
CONCLUSION: As public awareness of cancer susceptibility genes has grown markedly in recent years, the demand has also grown for genetic services to assess familial cancer risk. Major deleterious genetic mutations are rare, and much of the current research in genetic variation focuses on higher prevalence variants that carry lower risks. This may suggest that testing for mutations will move from genetics clinics to primary care and specialty practices. Thus, it is unclear whether the scope and volume of cancer center genetics services will continue to grow as rapidly as they have over the last decade.
INTRODUCTION
The role of heredity in disease is well established, if not entirely well understood. Cancer is a genetic disease because it is the genetic material that is the target of the disease process. However, a proportion of cancers are genetic also because predisposition is inherited in the form of an abnormal copy of a specific gene. A variety of these genes are identified and, in some cases, understanding the genetic disease has shed light on the related nonfamilial disease.
The identification of abnormal alleles of genes associated with cancer susceptibility has greatly expanded the role of clinical genetic testing for disease susceptibility in adults. Services to identify cancer susceptibility are of interest to those families where an abnormal allele is known to segregate, to individuals with a strong family history of unknown etiology, and to individuals who are concerned about their own genetic risk. Anecdotal reports suggest that the volume of services offered to individuals with or concerned about hereditary cancer risk has increased substantially in recent years. The increase in services is a consequence both of growing awareness of familial factors of diseases and of the greater number of genes that have been cloned and sequenced. With the increasing volume and scope of services, providers, patients, and families face concerns about cost, confidentiality, and accessibility.
In 1993, to gather information about the scope and content of cancer genetic services provided, we surveyed cancer centers receiving support from the National Cancer Institute (NCI).1 As the primary providers of cancer prevention, treatment, and research, these centers may provide the best information on the supply and demand for genetic services associated with cancer susceptibility genes in the United States. The purpose of this follow-up study was to determine how the scope and volume of genetic services provided at NCI-designated cancer centers has changed between 1993 and 2002.
METHODS
Study Population
On August 20, 2001, a list of all NCI-designated cancer centers was obtained from the NCI Web site (http://www.nci.nih.gov/cancercenters/centerslist.html). At that time, there were 60 centers listed (41 comprehensive cancer centers, 11 clinical cancer centers, and eight cancer centers). In July 2002, the University of California Davis Cancer Center was added to the list (as a clinical cancer center), bringing the total to 61. Today, there exist 61 NCI-designated cancer centers, although a few centers have been reclassified according to type so that there are now 39 comprehensive cancer centers, 14 clinical cancer centers, and eight cancer centers (same NCI centers list Web site link as mentioned earlier; accessed August 22, 2003). NCI cancer center support grant funding in dollars for each NCI-designated cancer center as of 2002 was obtained from the FY 2002 NCI FactBook, which is also accessible from the NCI Web site, and through consultation with the Freedom of Information/Privacy Act Coordinator.
Survey Procedures
In December 2002, a survey was sent to the directors of the 61 centers. In April of 2003, follow-up letters along with an additional copy of the survey were sent to the directors who had not yet responded, requesting that they reply by May 1, 2003. At the end of May 2003, follow-up calls to the remaining nonresponders were made.
Survey Content
The survey consisted of a total of 20 open-ended and fixed-response questions that covered the following four areas: services offered, protocol, clinic structure, and sources of revenue (Appendix). The category concerning services included questions on administration of health history questionnaires, collection of family history data, and types and location of genetic services provided. Protocol issues were addressed with questions of clinic policy regarding criteria for determining who is offered genetic testing, record keeping, and notification. The clinic structure section included questions on professional staff and patient populations served. Questions on revenue sources included sources of nonreimbursement-based financial support, sources of reimbursement for patient services, and charges for particular genetic services.
Analysis
Analysis of the data consisted of generating simple frequencies that were categoric in nature and summary statistics (including means, variances, ranges, and distributions) for those responses that were continuous. To compare the responders and nonresponders with respect to NCI cancer center support grant funding in dollars, the Wilcoxon rank sum nonparametric test was performed because the sizes of the two groups were small and the underlying distributions of the two populations could not be assumed to be normally distributed.
RESULTS
Response
Of the 61 NCI-designated cancer centers we contacted, 46 centers completed the survey and returned it via mail or fax, five centers returned the survey by mail or fax indicating that they did not offer any genetic services to familial cancer patients, five centers did not complete the survey but contacted the researchers by phone, e-mail, or post mail to reply that they also did not offer any genetic services to familial cancer patients, and five centers did not respond to our request in any way. All 46 centers that completed the survey and reported that they offer genetic services are comprehensive cancer centers or clinical cancer centers; the eight cancer centers (defined as such because they are neither comprehensive nor clinical centers) all responded that they do not offer such services. Responders and nonresponders were compared with respect to the dollar amount of the cancer center support grant to determine whether or not the nonresponders were comprised of a disproportionate number of the larger cancer centers; there were no significant differences.
Genetic Services Provided
The majority of cancer centers responding (46 of 56 centers; 82%) provided some genetic services for evaluation of familial cancer, which is significantly more than the number of cancer centers responding that provided such services in 1993 (50%; P < .01). Almost two thirds (30 of 46 centers; 65%) of those centers providing genetic services administered a health history questionnaire, and more than three fourths (36 of 46; 78%) collected data on family history. Comments on these questions by those who answered no to either administering a health history questionnaire or collecting family history data, such as that this information is only collected in high-risk clinics or only for certain patients or this is done by the physicians, rather than by the cancer center, indicated that the questions were interpreted in various ways by different responders. Thus, the proportion of centers that do administer a health history questionnaire or collect family history data is higher than the percentages just mentioned.
Referral policies for genetic services were uniform among centers. The majority of centers (41 of 46 centers; 89%) did not require a referral by a physician or other medical professional for genetic counseling. A smaller proportion of centers (36 of 46 centers; 78%) had family history criteria to determine whether patients would be offered genetic testing. These criteria were based on a number of sources, primarily from the American Society of Clinical Oncology (26 of 36 centers; 72%), but also from clinic managers (11 of 36 centers; 31%), other organizations with published recommendations (11 of 36 centers; 31%), and other means (generally applied on an individual basis; 19 of 36 centers; 53%). Once results of genetic testing were obtained, almost all centers (39 of 42 centers; 93%) documented the information on the patient's shadow chart in the genetics clinic. Only a minority of centers (13 of 44 centers; 30%) also documented the information in the patient's hospital chart or the hospital computer. Finally, just less than half of the centers (22 of 46 centers; 48%) notified the referring physician as a matter of clinic policy; for the other clinics and for notification to other individuals, patient consent was required. Questions on referral policies were not asked on the 1993 questionnaire.
The majority of the centers rely on physician referrals within the institution (44 of 46 centers; 95.7%) and outside the institution (41 of 46 centers; 89%). Other methods of recruiting patients include health professional education (40 of 46 centers; 87%), public education (38 of 46 centers; 83%), and use of a brochure (37 of 46 centers; 80%). Fewer centers use the news media (19 of 46 centers; 41%) or letters to licensed health care professionals (15 of 46 centers; 33%). A small number of centers recruited through the Internet, word of mouth (10 of 46 centers; 22%), and the tumor registry (seven of 46 centers; 15%).
Almost all centers (42 of 46 centers; 91%) provided genetic services not only to cancer patients and their families, but also to individuals concerned with risk. This is a significant change (P = .01) from 1993, when 65% of centers (11 of 17 centers) offered services to all three categories of patients. Table 1 compares the types of patients seen by cancer centers offering genetic services in 1993 and 2002.
The cancer centers provided these genetic evaluations in a number of different clinical settings. The majority of centers (32 of 45 centers; 71%) reported that a portion of their services was provided in a dedicated familial cancer clinic; 40% (18 of 45 centers) provided genetic services in an oncology or gynecology/oncology clinic; almost a fifth of the centers (eight of 45 centers; 18%) provided services in a general genetics clinic, and approximately a quarter of centers (11 of 45 centers; 24%) replied "other" to this question and indicated that genetic evaluations took place in offices of genetic counselors, consultation or conference rooms, breast and/or colon centers, and within high-risk programs. (One of the 46 centers that reported that they provided genetic services did not indicate a site at which the genetic evaluation took place.) The proportion of centers offering genetic services in clinics devoted to familial cancer has risen from 47% in 1993 to 71% in 2002 (P < .08).
All 46 centers reported that they provided cancer risk assessment and pedigree analysis; all centers but one also provided genetic counseling; and most centers (37 of 46 centers; 80%) also reported that they provided medical evaluations (including physical examinations). Table 2 shows that these numbers, with the exception of medical evaluation, are only slightly higher than those in 1993.
Although all centers performed pedigree analysis, the process of obtaining the pedigree varied among centers. In the majority of centers, genetic counselors obtained the pedigree (42 of 46 centers; 91%). However, in a number of centers, physicians (10 of 46 centers; 22%), nurses (nine of 46 centers; 20%), and others (nine of 46 centers; 20%), such as clinic, research, and study coordinators, genetic counseling graduate students, health educators, nurse practitioners, and medical geneticists, performed the analysis. Most often, the pedigree was obtained by a face-to-face interview (41 of 46 centers; 89%), but questionnaires (30 of 46 centers; 65%) and telephone interviews (24 of 46 centers; 52%) were common as well. Five centers (11%) also noted other means of pedigree collection, including touch-screen kiosks, computerized correctives, review of medical records, and outside medical collection.
The majority of centers (29 of 46 centers; 63%) offering genetic services also maintained a molecular diagnostic laboratory that performs tests for familial cancers. As expected, this proportion is greater than that of centers offering molecular diagnosis in 1993 (53%), but the difference is not statistically significant (P = .47). Of those centers with such a laboratory, the BRCA1/2 genes are by far the most uniformly tested (25 of 29 centers; 86%) and are the only specific alleles mentioned by the majority of the centers that maintained a molecular diagnostic laboratory. Table 3 lists the proportions of centers that test for each of the genes and syndromes that were specifically queried. (Questions on cancers and syndromes tested by the centers were not asked in 1993.)
The types of cancer screening services directly offered varied widely across centers. Nine centers (20%) did not administer any screening tests; for those that did, more clinics performed screening tests for breast cancer than for any other test. Ovarian cancer was second in screening frequency, followed closely by colon cancer. Just more than half of the centers performed screening tests for cervical, prostate, endometrial, and gastric cancer.
Service Volume and Staffing
The number of patients seen each year for genetic evaluation of familial cancer varied substantially across centers. Approximately half of the centers (24 of 46 centers; 52%) see 100 to 300 patients a year; over a quarter (13 of 46 centers; 28%) see over 300 patients a year; and 20% (nine of 46 centers) see fewer than 100 patients a year. The overall volume of patients seen for familial genetics services has increased markedly from 1993, when the majority (eight of 13 centers; 62%) of centers offering such services saw fewer than 100 patients a year, and no center saw more than 300 patients a year.
Centers were also queried on the number of visits required for patient evaluation as a measure of functional capacity. For the majority of the centers (41 of 46 centers; 89%), the initial evaluation consists of one visit, with a mean length of 90 minutes, and one or two follow-up visits, with an average of three quarters of an hour per follow-up visit (mean, 48 minutes). These numbers are approximately the same as the 1993 figures, when the mean length of the initial visit was determined to be 82 minutes and the follow-up visit was 47 minutes. In 2002, the amount of time between initial patient contact and clinic visit ranged from 7 to 75 days, with a mean of 27 days.
The number and types of professional staff providing genetic services for familial cancer patients varied greatly among centers. Most centers have at least one genetic counselor (44 of 46 centers; 96%), medical geneticist (28 of 46 centers; 61%), and medical oncologist (28 of 46 centers; 61%). Table 4 shows that this was also the case in 1993. Although a greater proportion of centers in 1993 employed registered nurses and patient educators for genetic services, in 2002, there were more gynecologic oncologists and a significant number of other physicians (most notably gastroenterologists and surgical oncologists), as well as nurse practitioners and physician assistants (Table 4).
Revenue Sources for Genetic Services
Centers obtained revenue for genetic services from a variety sources. All centers had NCI cancer center support grants, although one center has been on an unfunded extension for 2 years. Specific sources of revenue for genetic services not derived from charges for patient care included institutional support (35 of 46 centers; 76%), grants from private and/or institutional foundations (17 of 46 centers; 37%) research grant awards (13 of 46 centers; 28%), and other sources (16 of 46 centers; 35%), including private donations and state funding. Most centers (37 of 46 centers; 80%) also received revenue from charges for patient care.
Of the 35 centers that received institutional support for genetic services, that support provided, on average, half of all the support provided for genetic services, followed by research grant funding, patient charges, and private funding. Of the 11 centers not receiving institutional support for genetic services, two received 100% of their funding from grants, whereas the other nine centers received the majority of their financial support from patient services and private funding.
Compared with funding sources reported in 1993, there has been an increase (although not a statistically significant one) in the proportion of centers receiving institutional support (+23%), support from private and/or institutional foundations (+14%), and support from other sources (+11%). The proportion of centers receiving support from fees for patient services has remained approximately the same since the previous survey.
Fees charged for patient services varied substantially from center to center. Pretest counseling ranged from $100 to $450 (median, $200); the disclosure visit ranged from $45 to $300 (median, $107); and additional sessions ranged from $60 to $268 (median, $113). Four centers responded that they charged a single fee for all types of counseling sessions, with charges ranging from $225 to $400 (median, $360). One center stated that they do not charge for genetic counseling services.
For centers that billed for patient counseling sessions, private health insurance was cited as the primary source of reimbursement (33 of 46 centers; 72%), followed by Medicare (28 of 46 centers; 61%), self-pay (25 of 46 centers; 54%), Medicaid (21 of 46 centers; 46%), military health care/Veterans Affairs (13 of 46 centers; 28%), research protocol (13 of 46 centers; 28%), other, including charity care (four of 46 centers; 9%), and the Indian Health Service (two of 46 centers; 4%).
Although private health insurance seems to be the main source of funding for genetic services, 40 (91%) of 44 centers responded in the affirmative to the question of whether they have had patients who have been unwilling or unable to use private insurance coverage for their medical services. The primary reason (37 of 40 centers; 93%) was that the patient did not wish to have family or personal cancer history included as a pre-existing condition in their medical records. This was followed closely by the reason that the patient did not have coverage for genetic services (34 of 40 centers; 85%). Having no insurance coverage (25 of 40 centers; 63%) and having poor insurance coverage (25 of 40 centers; 63%) were the next most common reasons why patients did not use private insurance coverage for their medical services.
DISCUSSION
Public awareness of cancer susceptibility genes, most notably BRCA1/2 and the genes associated with hereditary colorectal cancer, has grown substantially since these genes have been identified and reported in the medical literature. In this follow-up survey of NCI-designated cancer centers, we find that the scope and volume of genetic services provided for familial cancer risk assessment has expanded rapidly since 1993. A substantially higher proportion of centers provided genetic evaluation for cancer risk in 2002 than in 1993. In sharp contrast to 1993, most centers are now providing services for individuals who are simply concerned with their cancer risk but are not cancer patients or close family members of cancer patients. Over half of all NCI-designated cancer centers now have clinics dedicated to familial cancer. The scope of services has increased substantially, although most centers still do not have molecular diagnostic laboratories on site (a notable exception is BRCA1/2 testing). A smaller proportion of centers offered standard screening services, such as mammography and pap smears in 2003 compared with 1994, although the proportion of those offering nonstandard high-tech tests seems to have increased substantially. The volume of patients seen has increased markedly, as evidenced by the fact that the number of centers that provided services to more than 300 patients changed from none in 1993 to 13 in 2002.
The new questions provided in the 2002 survey suggest that traditional physician-directed referrals continue to be the most important source for new patients for genetics clinics that provide cancer services. Responses to the clinic revenue source questions highlighted the sometimes precarious state of financing for genetics clinics. Institutional support is an important source of funds for operations for most clinics. Presumably, this institutional support is provided out of institutional overhead distributed from a departmental source. This contrasts with most other departments in academic centers, which rely on patient revenue to cover the bulk of their clinical operations. Notably, Medicare revenues accounted for a marginal fraction of income for genetic services, again in contrast to most academic clinical specialties that serve adult patients. Although the scope and volume of clinical (and presumably billable) services increased across clinics, the proportion receiving institutional support increased as well, showing that operating expenses are exceeding patient-derived revenues at many clinics. Clinic charges varied widely across clinics, but it is important to note that, for most medical services, charges typically exceed insurer reimbursements by 50% to 100%.
Although referrals were not needed to enter genetics clinics, more than three fourths of the clinics applied their own screening criteria to determine whether testing was appropriate for self-referrals. Even though physician referrals provided the vast majority of new patients, only a minority of clinics routinely notified the referring physician of the test results. We suspect that recognition of the sensitive nature of genetic tests and concerns about interpretation of tests by nonspecialists are driving the relatively low rate of voluntary information sharing among clinical professionals. Nevertheless, studies indicate that primary care physicians are often dissatisfied because of delays or inadequate referral letter content from referring specialists.2,3 Miscommunication between consultants and generalists has been cited as a cause of disruptions in continuity of care, delayed diagnoses, unnecessary testing, and iatrogenic complications.4 As generalists' knowledge of genetic testing grows and the volume of referrals increases, generalists may come to demand that genetics specialty clinics share test information about their patients.
It is certain that new genetic variants associated with cancer risk will be found. As their role in cancer risk and pathogenesis becomes clearer, demand for clinic-based genetic services will undoubtedly grow. Nevertheless, genetic mutations are, by definition, rare, and the number of cancer family syndromes (and associated mutations) characterized is unlikely to increase substantially. This raises the question of whether the scope and volume of cancer genetics services will continue to grow as rapidly as they did from 1993 to 2002. Much of the new research in genetic variation focuses on higher prevalence variants that carry lower risks, notably polymorphisms and haplotypes.5 Pharmacogenomics and studies of gene-environment interactions and their role in cancer risk are two other major growth industries in genetics. It is not yet known whether services provided with the genetic tests that are developed from this research will come under the purview of NCI-designated cancer centers or will bypass them and diffuse into general clinical practice. At present, the clinical utility of higher prevalence, lower penetrance variants, which often are relevant only in the presence of specific exposures,6 is far from clear. It is possible that genetics clinics will serve as the clinical laboratories for these newer tests; however, unlike mutation testing, use of lower risk cancer genetic variants that are found to have clinical utility will rapidly move from genetics clinics to primary care and specialty practices.
Authors' Disclosures of Potential Conflicts of Interest
The authors indicated no potential conflicts of interest.
NOTES
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
Thompson JA, Wiesner GL, Sellers TA, et al: Genetic services for familial cancer patients: A survey of National Cancer Institute cancer centers. J Natl Cancer Inst 87:1446-1455, 1995
Forrest CB, Glade GB, Baker AE, et al: Coordination of specialty referrals and physician satisfaction with referral care. Arch Pediatr Adolesc Med 154:499-506, 2000
Gandhi TK, Sittig DF, Franklin M, et al: Communication breakdown in the outpatient referral process. J Gen Intern Med 15:626-631, 2000
Epstein RM: Communications between primary care physicians and consultants. Arch Fam Med 4:403-409, 1995
National Human Genome Research Institute: International consortium launches genetic variation mapping project. http://www.genome.gov/page.cfmpageID=10005336
Potter JD: At the interfaces of epidemiology, genetics, and genomics. Nat Rev Genet 2:142-147, 2001(Meira Epplein, Ksenia P. )
Departments of Epidemiology and Medicine, University of Washington, Seattle, WA
ABSTRACT
PURPOSE: Anecdotal reports suggest that the volume of services offered to individuals concerned with hereditary cancer risk has increased substantially in recent years. As a follow-up to our 1993 survey, we sought to determine how the scope and volume of genetic services has changed between 1993 and 2002.
METHODS: We surveyed the 61 National Cancer Institute–designated cancer centers in operation in 2002 using an updated version of the questionnaire from 1993. Analysis included frequencies and summary statistics.
RESULTS: The majority of cancer centers responding (46 of 56 centers; 82.1%) provided some genetic services for evaluation of familial cancer, which is a higher proportion than in 1993 (50%; P < .01). Almost all centers (42 of 46 centers; 91.3%) provided services not only to cancer patients and their families, but also to individuals concerned with risk, which is a change (P = .01) from 1993, when 64.7% of centers offered such services. In addition, increases have been found for most other measures of services rendered for familial genetic services.
CONCLUSION: As public awareness of cancer susceptibility genes has grown markedly in recent years, the demand has also grown for genetic services to assess familial cancer risk. Major deleterious genetic mutations are rare, and much of the current research in genetic variation focuses on higher prevalence variants that carry lower risks. This may suggest that testing for mutations will move from genetics clinics to primary care and specialty practices. Thus, it is unclear whether the scope and volume of cancer center genetics services will continue to grow as rapidly as they have over the last decade.
INTRODUCTION
The role of heredity in disease is well established, if not entirely well understood. Cancer is a genetic disease because it is the genetic material that is the target of the disease process. However, a proportion of cancers are genetic also because predisposition is inherited in the form of an abnormal copy of a specific gene. A variety of these genes are identified and, in some cases, understanding the genetic disease has shed light on the related nonfamilial disease.
The identification of abnormal alleles of genes associated with cancer susceptibility has greatly expanded the role of clinical genetic testing for disease susceptibility in adults. Services to identify cancer susceptibility are of interest to those families where an abnormal allele is known to segregate, to individuals with a strong family history of unknown etiology, and to individuals who are concerned about their own genetic risk. Anecdotal reports suggest that the volume of services offered to individuals with or concerned about hereditary cancer risk has increased substantially in recent years. The increase in services is a consequence both of growing awareness of familial factors of diseases and of the greater number of genes that have been cloned and sequenced. With the increasing volume and scope of services, providers, patients, and families face concerns about cost, confidentiality, and accessibility.
In 1993, to gather information about the scope and content of cancer genetic services provided, we surveyed cancer centers receiving support from the National Cancer Institute (NCI).1 As the primary providers of cancer prevention, treatment, and research, these centers may provide the best information on the supply and demand for genetic services associated with cancer susceptibility genes in the United States. The purpose of this follow-up study was to determine how the scope and volume of genetic services provided at NCI-designated cancer centers has changed between 1993 and 2002.
METHODS
Study Population
On August 20, 2001, a list of all NCI-designated cancer centers was obtained from the NCI Web site (http://www.nci.nih.gov/cancercenters/centerslist.html). At that time, there were 60 centers listed (41 comprehensive cancer centers, 11 clinical cancer centers, and eight cancer centers). In July 2002, the University of California Davis Cancer Center was added to the list (as a clinical cancer center), bringing the total to 61. Today, there exist 61 NCI-designated cancer centers, although a few centers have been reclassified according to type so that there are now 39 comprehensive cancer centers, 14 clinical cancer centers, and eight cancer centers (same NCI centers list Web site link as mentioned earlier; accessed August 22, 2003). NCI cancer center support grant funding in dollars for each NCI-designated cancer center as of 2002 was obtained from the FY 2002 NCI FactBook, which is also accessible from the NCI Web site, and through consultation with the Freedom of Information/Privacy Act Coordinator.
Survey Procedures
In December 2002, a survey was sent to the directors of the 61 centers. In April of 2003, follow-up letters along with an additional copy of the survey were sent to the directors who had not yet responded, requesting that they reply by May 1, 2003. At the end of May 2003, follow-up calls to the remaining nonresponders were made.
Survey Content
The survey consisted of a total of 20 open-ended and fixed-response questions that covered the following four areas: services offered, protocol, clinic structure, and sources of revenue (Appendix). The category concerning services included questions on administration of health history questionnaires, collection of family history data, and types and location of genetic services provided. Protocol issues were addressed with questions of clinic policy regarding criteria for determining who is offered genetic testing, record keeping, and notification. The clinic structure section included questions on professional staff and patient populations served. Questions on revenue sources included sources of nonreimbursement-based financial support, sources of reimbursement for patient services, and charges for particular genetic services.
Analysis
Analysis of the data consisted of generating simple frequencies that were categoric in nature and summary statistics (including means, variances, ranges, and distributions) for those responses that were continuous. To compare the responders and nonresponders with respect to NCI cancer center support grant funding in dollars, the Wilcoxon rank sum nonparametric test was performed because the sizes of the two groups were small and the underlying distributions of the two populations could not be assumed to be normally distributed.
RESULTS
Response
Of the 61 NCI-designated cancer centers we contacted, 46 centers completed the survey and returned it via mail or fax, five centers returned the survey by mail or fax indicating that they did not offer any genetic services to familial cancer patients, five centers did not complete the survey but contacted the researchers by phone, e-mail, or post mail to reply that they also did not offer any genetic services to familial cancer patients, and five centers did not respond to our request in any way. All 46 centers that completed the survey and reported that they offer genetic services are comprehensive cancer centers or clinical cancer centers; the eight cancer centers (defined as such because they are neither comprehensive nor clinical centers) all responded that they do not offer such services. Responders and nonresponders were compared with respect to the dollar amount of the cancer center support grant to determine whether or not the nonresponders were comprised of a disproportionate number of the larger cancer centers; there were no significant differences.
Genetic Services Provided
The majority of cancer centers responding (46 of 56 centers; 82%) provided some genetic services for evaluation of familial cancer, which is significantly more than the number of cancer centers responding that provided such services in 1993 (50%; P < .01). Almost two thirds (30 of 46 centers; 65%) of those centers providing genetic services administered a health history questionnaire, and more than three fourths (36 of 46; 78%) collected data on family history. Comments on these questions by those who answered no to either administering a health history questionnaire or collecting family history data, such as that this information is only collected in high-risk clinics or only for certain patients or this is done by the physicians, rather than by the cancer center, indicated that the questions were interpreted in various ways by different responders. Thus, the proportion of centers that do administer a health history questionnaire or collect family history data is higher than the percentages just mentioned.
Referral policies for genetic services were uniform among centers. The majority of centers (41 of 46 centers; 89%) did not require a referral by a physician or other medical professional for genetic counseling. A smaller proportion of centers (36 of 46 centers; 78%) had family history criteria to determine whether patients would be offered genetic testing. These criteria were based on a number of sources, primarily from the American Society of Clinical Oncology (26 of 36 centers; 72%), but also from clinic managers (11 of 36 centers; 31%), other organizations with published recommendations (11 of 36 centers; 31%), and other means (generally applied on an individual basis; 19 of 36 centers; 53%). Once results of genetic testing were obtained, almost all centers (39 of 42 centers; 93%) documented the information on the patient's shadow chart in the genetics clinic. Only a minority of centers (13 of 44 centers; 30%) also documented the information in the patient's hospital chart or the hospital computer. Finally, just less than half of the centers (22 of 46 centers; 48%) notified the referring physician as a matter of clinic policy; for the other clinics and for notification to other individuals, patient consent was required. Questions on referral policies were not asked on the 1993 questionnaire.
The majority of the centers rely on physician referrals within the institution (44 of 46 centers; 95.7%) and outside the institution (41 of 46 centers; 89%). Other methods of recruiting patients include health professional education (40 of 46 centers; 87%), public education (38 of 46 centers; 83%), and use of a brochure (37 of 46 centers; 80%). Fewer centers use the news media (19 of 46 centers; 41%) or letters to licensed health care professionals (15 of 46 centers; 33%). A small number of centers recruited through the Internet, word of mouth (10 of 46 centers; 22%), and the tumor registry (seven of 46 centers; 15%).
Almost all centers (42 of 46 centers; 91%) provided genetic services not only to cancer patients and their families, but also to individuals concerned with risk. This is a significant change (P = .01) from 1993, when 65% of centers (11 of 17 centers) offered services to all three categories of patients. Table 1 compares the types of patients seen by cancer centers offering genetic services in 1993 and 2002.
The cancer centers provided these genetic evaluations in a number of different clinical settings. The majority of centers (32 of 45 centers; 71%) reported that a portion of their services was provided in a dedicated familial cancer clinic; 40% (18 of 45 centers) provided genetic services in an oncology or gynecology/oncology clinic; almost a fifth of the centers (eight of 45 centers; 18%) provided services in a general genetics clinic, and approximately a quarter of centers (11 of 45 centers; 24%) replied "other" to this question and indicated that genetic evaluations took place in offices of genetic counselors, consultation or conference rooms, breast and/or colon centers, and within high-risk programs. (One of the 46 centers that reported that they provided genetic services did not indicate a site at which the genetic evaluation took place.) The proportion of centers offering genetic services in clinics devoted to familial cancer has risen from 47% in 1993 to 71% in 2002 (P < .08).
All 46 centers reported that they provided cancer risk assessment and pedigree analysis; all centers but one also provided genetic counseling; and most centers (37 of 46 centers; 80%) also reported that they provided medical evaluations (including physical examinations). Table 2 shows that these numbers, with the exception of medical evaluation, are only slightly higher than those in 1993.
Although all centers performed pedigree analysis, the process of obtaining the pedigree varied among centers. In the majority of centers, genetic counselors obtained the pedigree (42 of 46 centers; 91%). However, in a number of centers, physicians (10 of 46 centers; 22%), nurses (nine of 46 centers; 20%), and others (nine of 46 centers; 20%), such as clinic, research, and study coordinators, genetic counseling graduate students, health educators, nurse practitioners, and medical geneticists, performed the analysis. Most often, the pedigree was obtained by a face-to-face interview (41 of 46 centers; 89%), but questionnaires (30 of 46 centers; 65%) and telephone interviews (24 of 46 centers; 52%) were common as well. Five centers (11%) also noted other means of pedigree collection, including touch-screen kiosks, computerized correctives, review of medical records, and outside medical collection.
The majority of centers (29 of 46 centers; 63%) offering genetic services also maintained a molecular diagnostic laboratory that performs tests for familial cancers. As expected, this proportion is greater than that of centers offering molecular diagnosis in 1993 (53%), but the difference is not statistically significant (P = .47). Of those centers with such a laboratory, the BRCA1/2 genes are by far the most uniformly tested (25 of 29 centers; 86%) and are the only specific alleles mentioned by the majority of the centers that maintained a molecular diagnostic laboratory. Table 3 lists the proportions of centers that test for each of the genes and syndromes that were specifically queried. (Questions on cancers and syndromes tested by the centers were not asked in 1993.)
The types of cancer screening services directly offered varied widely across centers. Nine centers (20%) did not administer any screening tests; for those that did, more clinics performed screening tests for breast cancer than for any other test. Ovarian cancer was second in screening frequency, followed closely by colon cancer. Just more than half of the centers performed screening tests for cervical, prostate, endometrial, and gastric cancer.
Service Volume and Staffing
The number of patients seen each year for genetic evaluation of familial cancer varied substantially across centers. Approximately half of the centers (24 of 46 centers; 52%) see 100 to 300 patients a year; over a quarter (13 of 46 centers; 28%) see over 300 patients a year; and 20% (nine of 46 centers) see fewer than 100 patients a year. The overall volume of patients seen for familial genetics services has increased markedly from 1993, when the majority (eight of 13 centers; 62%) of centers offering such services saw fewer than 100 patients a year, and no center saw more than 300 patients a year.
Centers were also queried on the number of visits required for patient evaluation as a measure of functional capacity. For the majority of the centers (41 of 46 centers; 89%), the initial evaluation consists of one visit, with a mean length of 90 minutes, and one or two follow-up visits, with an average of three quarters of an hour per follow-up visit (mean, 48 minutes). These numbers are approximately the same as the 1993 figures, when the mean length of the initial visit was determined to be 82 minutes and the follow-up visit was 47 minutes. In 2002, the amount of time between initial patient contact and clinic visit ranged from 7 to 75 days, with a mean of 27 days.
The number and types of professional staff providing genetic services for familial cancer patients varied greatly among centers. Most centers have at least one genetic counselor (44 of 46 centers; 96%), medical geneticist (28 of 46 centers; 61%), and medical oncologist (28 of 46 centers; 61%). Table 4 shows that this was also the case in 1993. Although a greater proportion of centers in 1993 employed registered nurses and patient educators for genetic services, in 2002, there were more gynecologic oncologists and a significant number of other physicians (most notably gastroenterologists and surgical oncologists), as well as nurse practitioners and physician assistants (Table 4).
Revenue Sources for Genetic Services
Centers obtained revenue for genetic services from a variety sources. All centers had NCI cancer center support grants, although one center has been on an unfunded extension for 2 years. Specific sources of revenue for genetic services not derived from charges for patient care included institutional support (35 of 46 centers; 76%), grants from private and/or institutional foundations (17 of 46 centers; 37%) research grant awards (13 of 46 centers; 28%), and other sources (16 of 46 centers; 35%), including private donations and state funding. Most centers (37 of 46 centers; 80%) also received revenue from charges for patient care.
Of the 35 centers that received institutional support for genetic services, that support provided, on average, half of all the support provided for genetic services, followed by research grant funding, patient charges, and private funding. Of the 11 centers not receiving institutional support for genetic services, two received 100% of their funding from grants, whereas the other nine centers received the majority of their financial support from patient services and private funding.
Compared with funding sources reported in 1993, there has been an increase (although not a statistically significant one) in the proportion of centers receiving institutional support (+23%), support from private and/or institutional foundations (+14%), and support from other sources (+11%). The proportion of centers receiving support from fees for patient services has remained approximately the same since the previous survey.
Fees charged for patient services varied substantially from center to center. Pretest counseling ranged from $100 to $450 (median, $200); the disclosure visit ranged from $45 to $300 (median, $107); and additional sessions ranged from $60 to $268 (median, $113). Four centers responded that they charged a single fee for all types of counseling sessions, with charges ranging from $225 to $400 (median, $360). One center stated that they do not charge for genetic counseling services.
For centers that billed for patient counseling sessions, private health insurance was cited as the primary source of reimbursement (33 of 46 centers; 72%), followed by Medicare (28 of 46 centers; 61%), self-pay (25 of 46 centers; 54%), Medicaid (21 of 46 centers; 46%), military health care/Veterans Affairs (13 of 46 centers; 28%), research protocol (13 of 46 centers; 28%), other, including charity care (four of 46 centers; 9%), and the Indian Health Service (two of 46 centers; 4%).
Although private health insurance seems to be the main source of funding for genetic services, 40 (91%) of 44 centers responded in the affirmative to the question of whether they have had patients who have been unwilling or unable to use private insurance coverage for their medical services. The primary reason (37 of 40 centers; 93%) was that the patient did not wish to have family or personal cancer history included as a pre-existing condition in their medical records. This was followed closely by the reason that the patient did not have coverage for genetic services (34 of 40 centers; 85%). Having no insurance coverage (25 of 40 centers; 63%) and having poor insurance coverage (25 of 40 centers; 63%) were the next most common reasons why patients did not use private insurance coverage for their medical services.
DISCUSSION
Public awareness of cancer susceptibility genes, most notably BRCA1/2 and the genes associated with hereditary colorectal cancer, has grown substantially since these genes have been identified and reported in the medical literature. In this follow-up survey of NCI-designated cancer centers, we find that the scope and volume of genetic services provided for familial cancer risk assessment has expanded rapidly since 1993. A substantially higher proportion of centers provided genetic evaluation for cancer risk in 2002 than in 1993. In sharp contrast to 1993, most centers are now providing services for individuals who are simply concerned with their cancer risk but are not cancer patients or close family members of cancer patients. Over half of all NCI-designated cancer centers now have clinics dedicated to familial cancer. The scope of services has increased substantially, although most centers still do not have molecular diagnostic laboratories on site (a notable exception is BRCA1/2 testing). A smaller proportion of centers offered standard screening services, such as mammography and pap smears in 2003 compared with 1994, although the proportion of those offering nonstandard high-tech tests seems to have increased substantially. The volume of patients seen has increased markedly, as evidenced by the fact that the number of centers that provided services to more than 300 patients changed from none in 1993 to 13 in 2002.
The new questions provided in the 2002 survey suggest that traditional physician-directed referrals continue to be the most important source for new patients for genetics clinics that provide cancer services. Responses to the clinic revenue source questions highlighted the sometimes precarious state of financing for genetics clinics. Institutional support is an important source of funds for operations for most clinics. Presumably, this institutional support is provided out of institutional overhead distributed from a departmental source. This contrasts with most other departments in academic centers, which rely on patient revenue to cover the bulk of their clinical operations. Notably, Medicare revenues accounted for a marginal fraction of income for genetic services, again in contrast to most academic clinical specialties that serve adult patients. Although the scope and volume of clinical (and presumably billable) services increased across clinics, the proportion receiving institutional support increased as well, showing that operating expenses are exceeding patient-derived revenues at many clinics. Clinic charges varied widely across clinics, but it is important to note that, for most medical services, charges typically exceed insurer reimbursements by 50% to 100%.
Although referrals were not needed to enter genetics clinics, more than three fourths of the clinics applied their own screening criteria to determine whether testing was appropriate for self-referrals. Even though physician referrals provided the vast majority of new patients, only a minority of clinics routinely notified the referring physician of the test results. We suspect that recognition of the sensitive nature of genetic tests and concerns about interpretation of tests by nonspecialists are driving the relatively low rate of voluntary information sharing among clinical professionals. Nevertheless, studies indicate that primary care physicians are often dissatisfied because of delays or inadequate referral letter content from referring specialists.2,3 Miscommunication between consultants and generalists has been cited as a cause of disruptions in continuity of care, delayed diagnoses, unnecessary testing, and iatrogenic complications.4 As generalists' knowledge of genetic testing grows and the volume of referrals increases, generalists may come to demand that genetics specialty clinics share test information about their patients.
It is certain that new genetic variants associated with cancer risk will be found. As their role in cancer risk and pathogenesis becomes clearer, demand for clinic-based genetic services will undoubtedly grow. Nevertheless, genetic mutations are, by definition, rare, and the number of cancer family syndromes (and associated mutations) characterized is unlikely to increase substantially. This raises the question of whether the scope and volume of cancer genetics services will continue to grow as rapidly as they did from 1993 to 2002. Much of the new research in genetic variation focuses on higher prevalence variants that carry lower risks, notably polymorphisms and haplotypes.5 Pharmacogenomics and studies of gene-environment interactions and their role in cancer risk are two other major growth industries in genetics. It is not yet known whether services provided with the genetic tests that are developed from this research will come under the purview of NCI-designated cancer centers or will bypass them and diffuse into general clinical practice. At present, the clinical utility of higher prevalence, lower penetrance variants, which often are relevant only in the presence of specific exposures,6 is far from clear. It is possible that genetics clinics will serve as the clinical laboratories for these newer tests; however, unlike mutation testing, use of lower risk cancer genetic variants that are found to have clinical utility will rapidly move from genetics clinics to primary care and specialty practices.
Authors' Disclosures of Potential Conflicts of Interest
The authors indicated no potential conflicts of interest.
NOTES
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
Thompson JA, Wiesner GL, Sellers TA, et al: Genetic services for familial cancer patients: A survey of National Cancer Institute cancer centers. J Natl Cancer Inst 87:1446-1455, 1995
Forrest CB, Glade GB, Baker AE, et al: Coordination of specialty referrals and physician satisfaction with referral care. Arch Pediatr Adolesc Med 154:499-506, 2000
Gandhi TK, Sittig DF, Franklin M, et al: Communication breakdown in the outpatient referral process. J Gen Intern Med 15:626-631, 2000
Epstein RM: Communications between primary care physicians and consultants. Arch Fam Med 4:403-409, 1995
National Human Genome Research Institute: International consortium launches genetic variation mapping project. http://www.genome.gov/page.cfmpageID=10005336
Potter JD: At the interfaces of epidemiology, genetics, and genomics. Nat Rev Genet 2:142-147, 2001(Meira Epplein, Ksenia P. )