Fax: (336) 716-7554.
Survivorship research based in integrated healthcare delivery systems†
The Cancer Research Network
Version of Record online: 21 APR 2008
Copyright © 2008 American Cancer Society
Supplement: Cancer Survivorship: Embracing the Future
Volume 112, Issue Supplement 11, pages 2617–2626, 1 June 2008
How to Cite
Geiger, A. M., Buist, D. S. M., Greene, S. M., Altschuler, A. and Field, T. S. (2008), Survivorship research based in integrated healthcare delivery systems. Cancer, 112: 2617–2626. doi: 10.1002/cncr.23447
This supplement was sponsored by the American Cancer Society's Behavioral Research Center and the National Cancer Institute's Office of Cancer Survivorship.
- Issue online: 19 MAY 2008
- Version of Record online: 21 APR 2008
- Manuscript Accepted: 20 FEB 2008
- Manuscript Revised: 2 JAN 2008
- Manuscript Received: 5 DEC 2007
- National Cancer Institute. Grant Number: U19 CA79689
- integrated delivery of healthcare;
Integrated healthcare delivery systems present unique opportunities for cancer survivorship research. The National Cancer Institute funds the Cancer Research Network (CRN) to leverage these capabilities for all types of cancer research, including survivorship.
The authors gathered information from a recent CRN funding application, Survivorship Interest Group materials, the CRN website, and published articles. CRN studies were selected to illustrate diverse topics and a variety of data-collection approaches.
The 14 systems that participate in the CRN provide care for approximately 10.8 million individuals of all ages and racial/ethnic backgrounds, for whom approximately 38,000 new cancer diagnoses were made in 2005. CRN systems have the ability to use existing data and collect new data on patients, providers, and organizations through well established research centers staffed by independent scientists. Of the 45 funded and 2 pending CRN grant applications as of November 30, 2007, 21 include aspects related to cancer survivorship. These studies have examined clinical trial participation, patterns of care, age and racial/ethnic disparities, diffusion of clinical trial findings, treatment outcomes, surveillance, and end-of-life and palliative care. Breast, colorectal, lung, ovarian, and prostate cancers have been the focus of these studies. Results of these studies have been published widely in leading journals.
Completed and ongoing CRN survivorship studies provide a strong foundation for future studies. Scientists from all institutional affiliations are welcome to approach the CRN with ideas and are encouraged to allow ample time to establish collaborative relationships and design rigorous studies. Cancer 2008. © 2008 American Cancer Society.
In 1997, the National Cancer Institute (NCI) announced a competition for funding “to encourage the expansion of collaborative cancer research among healthcare provider organizations.”1 A key reason for this Request for Applications (RFA) was the recognition that the only readily available electronic data on cancer care came from the linkage of Surveillance, Epidemiology, and End Results (SEER) data and Medicare data, which excluded individuals aged <65 years unless they were disabled, had limited data for individuals enrolled in managed care plans, did not permit direct contact with patients or providers, and lacked data on orally administered treatments. This RFA also foreshadowed later efforts to involve community-based healthcare organizations in research, primarily to support the translation and dissemination of findings into the delivery system.2 Member institutions of the HMO Research Network3 collaboratively submitted a successful application, resulting in the 1999 formation of the Cancer Research Network (CRN). The CRN has reapplied successfully for funding twice and, under the current funding announcement,4 is funded through 2012.
The original RFA defined the desired healthcare provider organizations as oriented to community care with access to large, stable, and diverse patient populations and able to take advantage of existing integrated databases that can provide patient-level information.1 Systems involved in the CRN meet these criteria, in large part because they are integrated healthcare delivery systems. Specifically, CRN systems incorporate providers, facilities, and insurance components into systems that facilitate coordinated delivery of a full range of services with explicit quality and cost management. These systems also emphasize preventive care and chronic disease management to improve long-term health outcomes.
Integrated healthcare delivery systems possess several unique capabilities relevant to cancer survivorship research. The original and 2 subsequent announcements that led to CRN funding included specific objectives related to survivorship research, in part because cancer survivorship research has taken on an increased prominence related to extended survival resulting from treatment advances, concerns about the quality of cancer care, expanding awareness of long-term consequences of cancer and cancer treatment for health and well being, and increased patient advocacy. For example, the Institute of Medicine report5 on cancer survivors' needs underscored the need for in-depth examination of and attention to the medical, functional, and psychosocial needs of patients as they make the transition from active treatment into survivorship and noted that research partnerships across systems of care are needed to address these needs. To illustrate how the CRN has contributed and will continue to contribute to understanding and improving cancer survivorship, this article describes completed and ongoing research within the CRN and highlights how this research leverages the unique features of integrated healthcare delivery systems.
MATERIALS AND METHODS
The CRN has been described previously,6 and this report provides updated information. Since it was established in 1999, the research theme of the CRN has been to improve the effectiveness of cancer prevention and treatment through research that identifies the system, provider, treatment, and patient factors that affect outcomes. The CRN includes 13 full participant systems and 1 affiliate system from across the United States. The affiliate status enables systems to pursue full status by participating in the development of new projects and infrastructure committees and activities; the funding for this work is provided by the system rather than through NCI funding. Group Health in western Washington state has been the lead site since the inception of the CRN. The additional 9 original systems all continue to participate as full members; their full and abbreviated names and locations are Fallon Health Care System (Fallon), central Massachusetts; Harvard Pilgrim Health Care (HPHC), eastern Massachusetts; HealthPartners Research Foundation (HPRF), Minnesota; Henry Ford Health System/Health Alliance Plan (HFHS), Michigan; and Kaiser Permanente (Northern California [KPNC], Southern California [KPSC], Colorado [KPCO], Hawaii [KPH], and Oregon [KPNW]). Subsequently joining the CRN after participating as affiliate systems were Kaiser Permanente Georgia (KPG); Lovelace Sandia Health System (LSHS), New Mexico; and the Marshfield Clinic Research Foundation (MCRF), Wisconsin. Geisinger Health System (Pennsylvania) is currently an affiliate system with an expectation of becoming fully involved in several years. The CRN is funded through a cooperative agreement; thus, NCI program staff also participate actively.
To illustrate the overall infrastructure that has been established to facilitate cancer research, we provide a brief description of the CRN governance structure (Fig. 1). This infrastructure supports collaborative cancer survivorship research across multiple disciplines and CRN systems and other institutions, and it serves as a foundation for future studies. In conjunction with the Principal Investigator (Edward H. Wagner, MD, MPH), a Steering Committee with representation from each system provides scientific and administrative leadership across the CRN. This leadership is informed by 2 advisory committees. The Academic Liaison Committee includes scientists from academia who provide scientific advice. Each site has an Organizational Advisory Committee consisting of health plan, clinical, and research leaders who identify research opportunities arising from medical care and assist in study implementation. Four cores provide infrastructure. The Evaluation Core conducts an annual evaluation of CRN-wide and study-specific processes and productivity.7 The key activity of the Investigator Development Core is the CRN Scholar Program, which provides formal training and mentoring of junior CRN scientists. The Administrative Core includes 3 committees. A Communications Committee facilitates information exchange across systems and studies and with researchers, advocates, and others outside the CRN. A New Proposals Committee ensures that new proposals avoid overlap, are scientifically sound, and involve appropriate collaborators. A Publications Committee encourages publications and presentations, advises studies on writing strategies and authorship issues, and verifies that submitted publications appropriately describe the CRN. The Scientific and Data Resources Core includes expertise in measurement, economics, and data and maintains the CRN's Virtual Data Warehouse (VDW).8
The VDW increases study efficiency by facilitating access to data elements commonly used in research studies. The VDW is virtual in the sense that there is no single location where data are combined and stored. Instead, each site maintains control of their own data, and aggregation of data across sites occurs for specific studies. The VDW consists of a set of programs at each CRN system that translate electronic data from unique local databases into a common format across the CRN along with data quality checks and documentation. Data currently available in the VDW include enrollment, demographics, tumor registry, outpatient pharmacy, and outpatient and inpatient use. Laboratory, pathology, and radiology procedures and death data will be incorporated as part of the new CRN funding cycle. The translation of local data into a common format generally is supported through the CRN infrastructure, whereas specific data aggregation activities are charged to the relevant study.
The CRN established a Survivorship Interest Group in March 2004 to provide a focal point for scientific activities in this area. The group meets by teleconference at least quarterly and uses an electronic mail list to share and discuss information and ideas. The electronic mail list includes representatives from nearly every CRN system as well as several scientists based in academia. In addition to sharing information about new opportunities in survivorship research and assisting in the development of research ideas, the Interest Group also tracks the status of all CRN research involving survivors as defined by the NCI Office of Cancer Survivorship,9 specifically, individuals from the diagnosis of cancer through the balance of his or her life and family members, friends, and caregivers.
We gathered information from the August 2006 CRN funding application, Survivorship Interest Group materials, the private CRN website, and published articles. To illustrate the survivorship research conducted by the CRN, we selected 10 CRN studies that illustrate a variety of approaches to data collection and a diversity of topics. Survivorship studies conducted outside the CRN but within a participating system are not tracked centrally and, thus, are beyond the scope of this report.
The 13 systems fully involved in the CRN provide care for approximately 10.8 million individuals in the United States (Table 1). The majority of this care is delivered in integrated healthcare delivery systems. Increasingly, however, the CRN systems are diversifying their delivery organizations and practice arrangements, which provides an opportunity to examine variations across different structures (for example, capitated models vs fee-for-service models). The enrolled individuals include people of all ages and substantial groups of African Americans, Asian Americans, and Latinos, particularly in systems located where sizable populations of these groups reside. Over half of the individuals enrolled in 2000 remained enrolled in 2005.
|Total enrollment, in thousands||171||540||805||620||295||462||286||225||3260||487||3190||240||175|
|Members enrolled in integrated system, %||53||80||20||64||65||100||90||100||100||100||100||90||85|
|Retention of 2000 cohort, %|
|Research center established||1996||1983||1969||1989||1987||1987||1988||1991||1961||1964||1975||1990||1959|
|Research affiliation with NCI-designated cancer center||No||Yes||Yes||Yes||No||Yes||No||IP||No||Yes||IP||No||IP|
|Data resources, first y available|
|Electronic medical record||2006||2005||1969||2004||1988||1997||2006||2001||2005||1997||2006||N/A||1994|
|Outpatient visit claims||1987||1992||1969||1990||1988||1992||1995||1995||1994||1987||1993||1999||1991|
|Oncologists within system‡||10||13||8||9||38||6||N/A||6||83||12||81||N/A||N/A|
All systems participating in the CRN have well established research centers, with the earliest established in 1959 and the most recent established in 1996 (Table 1). These centers are akin to academic departments, in that they conduct research in the public domain, rely on external funding to support their studies, and disseminate their results through peer-reviewed publications. In 2005, these centers obtained >$100 million in direct research support, with about half provided by the National Institutes of Health. Independent scientists within the centers include biostatisticians, clinicians, epidemiologists, health services researchers, and social scientists, representing a broad array of topical interests as well as substantial expertise about their healthcare delivery system's patients, providers, and data systems. These scientists lead their own studies and also collaborate actively with scientists from other institutions. To support collaborative cancer research and help meet the CRN's goals, 5 systems have formal affiliations with NCI-designated Cancer Centers located in their communities, and 3 systems are pursuing such affiliations. Other partnerships with academic scientists and institutions have been formed based on mutual research interests. Many of the CRN systems also participate in clinical trials through cooperative groups and the Community Clinical Oncology Program.10
Cancer registries are in place at 11 of the CRN systems. All 13 fully participating systems have electronic databases that contain information on membership and utilization claims for outpatient visits, hospitalization, and pharmacy. These databases feed the VDW8 described above (see Materials and Methods). All but 1 of the systems have implemented an electronic medical record, most using the EpicCare system,11 and 9 systems have information on deaths among all individuals who have ever been members.
In 2005, there were >38,000 new cancer diagnoses among the 9 CRN systems that were full members at that time and had cancer registries; these 9 systems provide care for approximately 9.5 million individuals. The cancer diagnoses included approximately 6500 women with in situ and invasive breast cancers, 6250 prostate cancers, 3700 colorectal cancers, 875 pancreatic cancers, 350 esophageal cancers, and 400 hematologic malignancies. The majority of care for these cases is provided by >250 adult and pediatric surgical, gynecologic, medical, and radiation oncologists who work within these delivery systems. More detailed demographic and other data about individuals diagnosed with cancer are available only to scientists collaborating with the CRN to plan or conduct studies; please see the end of this report for information about initiating a collaboration with the CRN.
Overview of Survivorship Studies
By November 30, 2007, 45 studies had been funded under the auspices of the CRN, and 2 proposals were pending. Survivorship issues are addressed in 17 (53%) of the funded studies and in both of the pending proposals. We note that several survivorship studies are conducted within individual CRN member systems and are not considered CRN studies per se, so they are not tracked in a centralized manner. To illustrate the capacity of the CRN and, by extension, the capacity of integrated healthcare delivery systems for survivorship research, we have elected to feature 10 studies that address a variety of topics and that use different approaches to data collection. These studies are outlined in Table 2 and are described below.
|Descriptive title and selected references||Leader and site||Key methodological details|
|Disenrollment among cancer patients: Field 200412||Field TS (Fallon)||Electronic data from 132,580 adults enrolled in 5 systems who were newly diagnosed with invasive cancer from 1993 to 1998 and were followed for disenrollment or death through 1999|
|Racial and ethnic disparities in survival after cancer: Field 200413, Doubeni 200714||Field TS (Fallon)||Electronic data from women with invasive breast cancer (2276 African Americans, 18,879 whites) and men and women with invasive colorectal cancer (1479 African Americans, 985 Hispanics, 909 Asians/Pacific Islanders, 10,585 whites), all diagnosed from 1993 to 1998 and followed for death through 2003|
|Diffusion of adjuvant aromatase inhibitor treatment for breast cancer: Aiello 200815||Aiello E (GH)||Electronic data from the CRN Virtual Data Warehouse from 13,245 women enrolled in 7 systems who were newly diagnosed with invasive breast cancer from 1996 to 2003 and followed for adjuvant hormone therapy use through 2004|
|Efficacy of contralateral prophylactic mastectomy: Herrinton 2005,16 Geiger 2003,17 Rolnick 200418||Fletcher SW (HPHC)||Electronic data and computer-assisted medical record abstraction among 1072 women with unilateral breast cancer who had their contralateral breast prophylactically removed from 1979 to 1999, and were followed through 2000 for contralateral cancer and breast cancer death|
|Breast cancer treatment effectiveness in older women: Enger 2006,19 Geiger 2007,20 Ulcickas Yood 2008,21 Lash 2007,22 Thwin 2007,25 Lash 200726||Silliman RA (BU)||Electronic data and computer-assisted medical record abstraction among 1859 women aged ≥65 y and diagnosed with early invasive breast cancer at 1 of 6 systems from 1990 to 1994, and followed for recurrence and death through 2004|
|End-of-life and palliative care for ovarian cancer: Rolnick 200727||Neslund-Dudas C (HFHS)||Electronic data and medical record abstraction of utilization during the last 6 mo of life among 421 women who died of ovarian cancer while members of 1 of 3 systems from 1995 to 2000|
|Outcomes of contralateral prophylactic mastectomy: Geiger 2006,28 Nekhlyudov 2005,29 Greene 2006,30 West 2005,31 Altschuler 2008,32 Rolnick 200733||Geiger AM (KPSC)||Mailed satisfaction and quality-of-life survey of 519 women with unilateral breast cancer who had their contralateral breast prophylactically removed from 1979 to 1999|
|Organizational and physician barriers to clinical trials participation: Somkin 200534||Somkin CP (KPNC)||Assessment of 10 systems through semistructured qualitative interviews with key informants and a mailed survey of 221 oncologists|
|Lung and colorectal cancer outcomes (CanCORS): Ayanian 200435||Fletcher RH (HPHC)||Rapid, electronic data-based identification of newly diagnosed lung and colorectal cancer patients at five systems, with data collected through medical record abstraction and patient and physician surveys|
|Clinical and pathologic predictors of recurrence after ductal carcinoma in situ: Nekhlyudov 200636||Habel L (KPNC)||Electronic data, medical record abstraction and tissue specimens for a cohort of women diagnosed with ductal carcinoma in situ and treated with breast-conserving surgery at 1 of 3 systems from 1990 to 2001, and followed for recurrence through 2005|
Survivorship Studies Relying on Electronic Data
To examine the ability of the CRN to follow cancer survivors over time, 1 of the first CRN survivorship studies examined the extent to which individuals diagnosed with cancer remain enrolled in their healthcare delivery system. Relying exclusively on electronic data, the study indicated that 96% of adults diagnosed with any invasive cancer remained enrolled at 1 year after cancer diagnosis (excluding deaths), and 84% remained enrolled at 5 years.12 The reasons for disenrollment were not gathered in the study; however, anecdotally, we believe that the most common reasons are loss of coverage because of changes in employment of the patient or family member who subscribes, a geographic move out of the plan's coverage area, or an employer's decision to no longer to offer the plan. The study demonstrated the CRN's capacity for secondary data analysis and established that the CRN could follow a large proportion of survivors for at least 5 years after diagnosis.
The same electronic data were used as a foundation for a study that delved into the issues of racial differences in cancer survival in settings in which all individuals are insured and have access to necessary care. The study suggested that, even in this setting, African Americans are more likely to die from breast and colorectal cancer than whites.13, 14 This work served as a foundation for a proposed program project grant to explore disparities in more detail.
Electronic data also have been used to examine the diffusion of adjuvant aromatase inhibitor treatment for breast cancer. That study revealed striking increases in aromatase inhibitor use and decreases in tamoxifen use after 2001, suggesting that early results from clinical trials changed medical practices in 7 CRN systems before national breast cancer treatment guidelines were modified based on trial results.15 The study made use of the CRN VDW,8 which is described above (see Materials and Methods), and approached data analysis in a disseminated manner, such that only aggregated data were centralized, thus reducing administrative demands related to Institutional Review Board (IRB) and Health Insurance Portability and Accessibility Act requirements. The study provides a model for using the CRN in rapid turnaround studies of the dissemination of research results.
Studies of Survivorship Combining Electronic Data With Medical Record Abstraction
Combining the use of electronic data with medical record abstraction is a common approach in the CRN. One of the original CRN studies used this approach to examine the efficacy of contralateral prophylactic mastectomy (the preventive removal of the contralateral breast in women with unilateral breast cancer). The results indicated that contralateral prophylactic mastectomy significantly reduced the risk of contralateral breast cancer as well as breast cancer and overall death.16 The study was among the first to look at these outcomes, particularly in the community rather than in academic medical settings. In addition, the study designed and implemented a computerized system to facilitate medical record abstraction,17 which was an important data-collection innovation for the CRN. The study also included methodological work demonstrating the sensitivity and specificity of utilization claims data for identifying incident breast cancer cases relative to the gold standard of cancer registries.18 The study illustrated the combined use of electronic data and medical record abstraction that is common in the CRN and established approaches that have been adopted and improved on in later studies.
Building on the approaches developed in the contralateral prophylactic mastectomy study and in other CRN studies, a scientist based in academia led a team that used electronic and abstracted medical record data to examine the patterns of care and outcomes of treatment among older women with early-stage, invasive breast cancer. Analyses are ongoing, but the results published thus far indicate that older women are at risk of not receiving the full complement of effective therapies,19 and failure to receive these therapies results in an increased risk of recurrences, second primary breast cancers, and death from breast cancer.20, 21 In addition, surveillance mammograms within 5 years of the original breast cancer diagnosis appear to reduce the risk of breast cancer mortality in older women.22 Results from this study have been the subject of 2 editorials.23, 24 The study also continues to advance CRN methods by further refining the computer-assisted medical record abstraction process to assess interrater and intrarater reliability25 and by examining the use of reliability data to make probabilistic corrections to data during analysis.26 The results of this study address important issues that are difficult to examine in clinical trials because of the lack of inclusion of older women and for which SEER-Medicare data alone are insufficient.
The ability to integrate electronic data with medical record abstraction also has been applied to understanding end-of-life care. A study assessing analgesic drug therapies indicated that only 54% of women who were dying of ovarian cancer received high-intensity medication, suggesting room for improvement in end-of-life pain management in this population.27
Studies of Survivorship Combining Electronic Data, Medical Record Abstraction, and Surveys of Patients, Providers, and/or Organizations
The contralateral prophylactic mastectomy work was extended in a second study that demonstrated the use of mailed surveys in CRN survivorship research. The study demonstrated that women who had breast cancer and who had undergone contralateral prophylactic mastectomy reported high levels of satisfaction and current contentment with quality of life years after their treatment, in proportions virtually identical to those among women who had breast cancer but had not undergone prophylactic mastectomy.28 The study also revealed that the majority of women undergoing the procedure took an active role in the decision-making process and that these women reported higher levels of satisfaction than women in less active roles.29 Methodological work in this study included an assessment of the impact of multicenter IRB requirements on the study protocol30 and documentation of the validity of race and ethnicity classifications gathered from medical records compared with self-report.31 The study also used qualitative methods to develop a survey and analyze results.32, 33 This study illustrates the CRN's capability for studies of long-term, patient-oriented outcomes and has contributed to the development of strategies for IRB submissions.
Assessing organizations and studying physicians also are distinctive capabilities of the CRN. An original charge of the CRN was to increase clinical trials participation, so a study used survey and qualitative interview methods to examine the barriers to oncologist participation in oncology clinical trials.34 The oncologist survey was designed with a tracking system separate from the surveys; thus, responses were anonymous, but reminders could be sent out. The response rate for oncologists was 90%. Telephone interviews were conducted with 18 oncology and administrative leaders across the CRN. The results suggested an overall favorable attitude toward trials but identified challenges related to the often stringent eligibility criteria for enrolling patients and the need for staff to support patient enrollment and trial-related documentation.
CRN participation in the ongoing Cancer Care Outcomes Research and Surveillance (CanCORS) Consortium35 continues a long-standing interest in using observational study designs to understand care patterns and outcomes better, including quality of life. CanCORS makes use of electronic data to rapidly identify newly diagnosed patients, then collects data by using established processes for medical record abstraction and patient and physician surveys. This study also demonstrates the CRN's ability to collaborate across the country and with multiple types of delivery systems.
Studies of Survivorship Incorporating Pathology Specimens
Another ongoing CRN study seeks to understand the clinical and pathologic predictors of recurrence after ductal carcinoma in situ.36 This study relies on electronic data and medical record abstraction but also has arranged for the acquisition of archived tissue slides and specimens. These materials will be submitted for expert pathology review and will be used in biomarker studies. Assembling a large, multiethnic cohort of women, along with tissue specimens and data on long-term outcomes, demonstrates another aspect of the unique capabilities of the CRN.
Survivorship research has been a focus of the CRN since its inception and leverages many capabilities of integrated healthcare delivery systems. Perhaps foremost among these capabilities is the use of a large population base with >38,000 cancer diagnoses per year and an ability to follow survivors over time.12 The population base includes substantial minority populations that have been incorporated in studies, allowing investigation of disparities.13, 14 Also key is the availability of cancer registries and many other types of electronic data that can be analyzed independent of other information12–15, 18 or used in conjunction with other types of data.16, 19–22, 25–29, 31, 35, 36 The CRN has experience with many data-collection modalities, including medical record abstraction,16, 19–22, 25–27, 35, 36 patient and physician surveys,28, 29, 31–33, 35 tissue specimen acquisition,36 and organizational assessment.37 The CRN also has played a role in developing methodologies for cancer survivorship research.17, 18, 25, 30, 31, 38
To date, the CRN survivorship research portfolio includes studies of breast, colorectal, lung, ovarian, and prostate cancers as well as ductal carcinoma in situ.12–20, 22, 25–31, 34–36 Studies have examined patterns of care and how age and racial and ethnic disparities, decision-making, and diffusion of research results impact patterns of care.13–15, 19, 29, 35 Studies also have focused on treatment outcomes, including effectiveness in preventing recurrences and death, the occurrence of unanticipated late effects, and quality of life.16, 20, 28, 36 Other topics to date include approaches to surveillance for recurrences and for end-of-life and palliative care.22, 27 Results of these studies have been published widely in leading journals, such as Cancer, the Journal of Clinical Oncology, and the Journal of the National Cancer Institute.
None of the survivorship work of the CRN would be possible without the long-standing research centers within the CRN systems and the scientists and staff at these research centers, some of whom had well established survivorship research programs before the advent of the CRN. We note that many of these systems maintain their own programs of survivorship research separate from the CRN. These programs are not tracked centrally; consequently, summarizing them was beyond the scope of this report.
Completed and current CRN studies provide a solid foundation for future work. The CRN is an excellent laboratory for studying survivorship care, particularly the issues relating to transitions in and continuity of care outlined in the 2006 Institute of Medicine report.5 The CRN also could be used to develop and test or replicate interventions to improve survivorship care, symptom management, and other outcomes. Studies of many other adult cancers are possible, including less common cancers, and there is potential to use the CRN to study pediatric, adolescent, and young adult cancers. Although there have been several studies of patterns of care, treatment outcomes, and late effects, substantial opportunity for further study of these issues exists within the CRN. Many economic issues also remain open for examination within the CRN. The advent of the electronic medical record and the ever-expanding use of the Internet present a new set of options for survivorship research and interventions at the system, provider, family, and patient levels. The CRN Survivorship Interest Group exists to encourage and support the pursuit of these and any other ideas that will improve the lives of cancer survivors.
A frequent criticism of CRN research is that the results are not generalizable outside integrated healthcare delivery systems. From the perspective of the composition of patient populations, the CRN member systems are highly representative of the communities in which they provide care; therefore, this generally is not a major issue. Instead, concern often arises from the belief that integrated delivery systems are uncommon and include features that are not found in other settings, thus raising the question of whether findings from these environments will apply to other settings. Although it is true that integrated delivery systems are not present in every community, approximately 1 in 4 United States residents is enrolled in a health maintenance organization,39 and CRN findings likely generalize to this large proportion of the population. There also are settings that share features with integrated delivery settings, such as the Veteran's Administration Health System and primary care networks organized around academic medical centers. In addition, integrated delivery settings may provide a “best case” scenario in the translation and dissemination of research results from academic settings to the delivery system. If care improvements cannot be achieved in integrated delivery systems like those involved in the CRN, then it seems highly unlikely that improvements would be possible in other settings. Considering issues of generalizability and balancing them with other conceptual and methodological strengths and limitations is important for any study. We suggest that this is no more or less true within the CRN than in other settings.
Survivorship research is a major focus of the CRN and has benefited from the capability of the CRN to use multiple approaches to gather data on patients, providers, and organizations. Completed and ongoing research provides a strong base for future studies. All CRN research in this area has been highly collaborative across institutions and disciplines and among junior, middle career, and senior career scientists. Future studies are likely to be highly successful if they build on the unique capabilities and prior survivorship research of the CRN and honor its strong values promoting collaboration and the use of studies to develop research capacities and scientists and staff across the CRN systems. Scientists from all institutional affiliations are welcome to approach the CRN with ideas and are encouraged to allow ample time to establish collaborative relationships and design rigorous studies. Initial steps in this process include assessing whether an idea is well suited to the CRN, forming a partnership with at least 1 CRN-based scientist, and submitting a concept proposal to the CRN New Proposals Committee. Scientists who are interested in pursuing a collaboration with the CRN are encouraged to consult the following webpage to identify the current contact to initiate this process: http://www.crn.cancer.gov/about/work.html [Accessed March 26, 2008].
We thank all of the scientists and staff of the Cancer Research Network (CRN) who have conducted survivorship research and acknowledge in particular the leadership of the Site Principal Investigators of the CRN: Terry Field, DSc, MPH (Fallon Health Care System); Edward H. Wagner, MD, MPH. (Group Health); Suzanne W. Fletcher, MD, MSc (Harvard Pilgrim Health Care); Sharon J. Rolnick, PhD, MPH, MA (HealthPartners Research Foundation); Christine C. Johnson, PhD (Henry Ford Health System/Health Alliance Plan): Debra P. Ritzwoller, PhD (Kaiser Permanente Colorado); Douglas Roblin, PhD (Kaiser Permanente Georgia); Tom Vogt, MD, MPH (Kaiser Permanente Hawaii); Larry H. Kushi, ScD (Kaiser Permanente Northern California); Mark Hornbrook, PhD (Kaiser Permanente Northwest); Virginia P. Quinn, PhD (Kaiser Permanente Southern California); Margaret Gunter, PhD (Lovelace Sandia Health System); and Robert Greenlee, PhD, MPH (Marshfield Clinical Research Foundation). We also thank the members of the CRN Survivorship Interest Group, particularly Mark Hornbrook, PhD (Kaiser Permanente Northwest) and Carolyn Gotay (University of Hawaii). Finally, we thank our many colleagues at the National Cancer Institute, particularly our project officer, Martin Brown, PhD.
- 1National Institutes of Health. RFA CA-97-017: Cancer Research Network Across Health Care Systems, 1997. Bethesda, Md: National Institutes of Health. Available at: http://grants.nih.gov/grants/guide/rfa-files/RFA-CA-97-017.html. Accessed July 11, 2007.
- 2National Institutes of Health. NIH Roadmap for Medical Research: Re-engineering the Clinical Research Enterprise, 2007. Bethesda, Md: National Institutes of Health. Available at: http://nihroadmap.nih.gov/clinicalresearchtheme/. Accessed July 11, 2007.
- 3HMO Research Network. About Our Organization, 2007. Available at: http://www.hmoresearchnetwork.org/about.htm. Accessed July 11, 2007.
- 4National Institutes of Health. RFA-CA-06-505: Cancer Research Network (U19), 2006. Bethesda, Md: National Institutes of Health. Available at: http://grants.nih.gov/grants/guide/rfa-files/RFA-CA-06-505.html. Accessed July 11, 2007.
- 5From Cancer Patient to Cancer Survivor: Lost in Transition. Washington, DC: National Academies Press; 2005., , .
- 9National Cancer Institute. About Cancer Survivorship Research: Survivorship Definitions, 2007. Bethesda, Md: National Cancer Institute. Available at: http://dccps.nci.nih.gov/ocs/definitions.html. Accessed July 11, 2007.
- 11Epic Systems Corporation. EpicCare website. Available at: http://www.epicsystems.com/. Accessed October 23, 2007.
- 36Radiologic evaluations and invasive procedures following breast-conserving surgery in a large cohort of women with ductal carcinoma in situ. J Gen Intern Med. 2006; 21: 119., , , et al.
- 38National Center for Health Statistics. Health, United States, 2006. Hyattsville, Md: National Center for Health Statistics; 2006.