Improving Accrual of Older Persons to Cancer Treatment Trials: A Randomized Trial Comparing an Educational Intervention With Standard Inform
http://www.100md.com
《临床肿瘤学》
Cancer and Leukemia Group B (CALGB) Elderly Committee, Chicago, IL
CALGB Statistical Center
Duke University Medical Center, Durham, NC
Wake Forest University School of Medicine, Winston-Salem, NC
Dana-Farber Cancer Institute, Boston, MA, Georgetown University Medical Center, Washington, DC
Vermont Cancer Center, Burlington, VT
ABSTRACT
PURPOSE: To design and test a geriatric educational intervention to improve accrual of cancer patients age 65 years and older to cooperative group–sponsored treatment trials.
METHODS: Main member institutions of the Cancer and Leukemia Group B (CALGB) and its affiliates were randomly assigned to receive standard information (n = 73) or educational intervention (n = 53). Standard information included CALGB Web site access and periodic notification about existing trials. The geriatric educational intervention included standard information plus: (1) an educational seminar; (2) educational materials; (3) a list of available protocols for use on charts; (4) a monthly e-mail and mail reminders for 1 year; and (5) a case discussion seminar. The main outcome was percentage of accrual of older persons to phase II and III treatment protocols after study initiation compared with baseline.
RESULTS: There were 3,032 patients entered onto trials in the baseline year, and 2,160 and 1,239 during the 2 years postintervention, respectively. Overall percentage of accrual of older patients was 37% at baseline, and 33% and 31% during the first and second years after intervention. There was no improvement in accrual in the intervention versus control arm: 36% v 32% in the first year and 31% v 31% in the second year.
CONCLUSION: Accrual of older patients was not increased by this intervention. Reasons for lack of effect include low intervention intensity, high baseline accrual rates, and closure of several high-accruing protocols during the study. More intense and multifaceted approaches will be needed to change physician (and patient) behavior and to increase accrual of older persons to clinical trials.
INTRODUCTION
Persons age 65 or older (hereinafter referred to as "older") comprise approximately 14% of the US population, but account for 63% of the cases of cancer in this country.1-3 Moreover, the projected "graying of America," coupled with advances in cancer screening and treatment, will result in a large increase in the absolute number of older cancer patients over the coming decades. Thus, there is a demographic imperative driving the need to improve cancer care to older populations.
Clinical trials are credited for a large proportion of the improvements in cancer therapy. The group at highest risk for cancer, older patients, however, is least likely to be represented in trials.3-5 For example, the mean age of patients having the five cancers with the highest mortality rates that entered National Cancer Institute (NCI)-sponsored treatment trials was significantly lower (48.7 years for men and 54.7 years for women) than that of incident cancer cases registered by Surveillance, Epidemiology, and End Results (SEER; 69.1 years for men and 64.6 years for women).4 In Southwest Oncology Group (SWOG) treatment trials, only 25% of patients enrolled in trials were age 65 years and older.3
Reasons for under-representation of older individuals in clinical trials are multidimensional, and include comorbid illnesses; age bias; physician, patient, and family lack of knowledge about toxicity and benefit of treatment; eligibility criteria excluding older patients from available trials; lack of financial, logistic, and social support; patient or family members’ preference; and excessive time required for enrollment.3-7 Kemeny et al8 found that age itself was a major factor in whether a clinical trial was offered to breast cancer patients, even after controlling for number of comorbidities, physical functioning, and disease stage. However, when offered participation in clinical trials, older women were equally as likely as younger women to participate. In the study of Kemeny et al, 42% of physicians stated that greater education about cancer treatment toxicity in the older person would improve accrual.
Here, we present results of a phase III randomized trial testing the hypothesis that a multifaceted educational intervention directed to physicians and other members of the medical and research team at Cancer and Leukemia Group B (CALGB) institutions would improve the accrual of older persons to cancer treatment trials.
METHODS
Eligibility
The CALGB is a National Cancer Institute–funded cooperative group that conducts studies in adults with cancer. All main member institutions in the CALGB and their affiliates with an average accrual to treatment protocols of 5 or more patients per year over the 3 years before June 2001 were included. The study was exempt from institutional review board review under 45 CFR 46.101(b).1
Pilot Educational Intervention
Registration and random assignment. Main member networks were randomly assigned as a unit to receive standard information or educational intervention. There are three types of institutional memberships within CALGB: main, at-large, and affiliate members. Main members are academic institutions that demonstrate scientific leadership and whose network registers 50 patients annually on CALGB studies. At-large members are academic or community hospitals whose network registers at least 30 patients annually on CALGB studies. Affiliate members participate in group activities as part of a network of sites that are coordinated by a main or at-large member. After random assignment, institutions randomly assigned to receive educational intervention were notified by the study chair or a member of the study team.
Standard information consisted of: 1) periodic notification of all existing CALGB trials per the CALGB Central Office; and 2) CALGB Web site access.
Educational intervention included standard information plus: 1) an educational symposium; 2) geriatric oncology educational materials; 3) monthly mailings and e-mails for 1 year; 4) lists of available protocols for use on patient charts; and 5) a case discussion seminar. Components of educational intervention are listed in Table 1.
Educational symposium. Two or three key members of the research team from each institution were invited and funded to attend the educational symposium. We encouraged designation of one physician and one or more additional member of the oncology research team, which could include a research nurse, clinical research associate (CRA), psychologist, or social worker. The symposium was held in June 2001, in conjunction with the CALGB group meetings in Ottawa, Ontario, Canada.
Geriatric oncology educational materials. Educational materials, including outlines of each lecture and a bibliography, were provided to symposium participants. Lecture slide sets were compiled, and a videotape of the symposium was made; these were available in February 2002. The videotape and slidesets were sent to principal investigators and lead CRAs. All education materials were sent, on request, to individuals within institutions randomly assigned to intervention.
Mailings and e-mail reminders. Monthly mailings and e-mail reminders were sent by the study chair to all members of institutions randomly assigned to intervention during the year after the educational symposium. E-mails were sent when addresses were available and functional. Each mailing and e-mail contained a brief educational narrative about cancer in older patients and a reminder to consider clinical trials for older patients. In order to prompt consideration of clinical trials, checklist reminders of available protocols were updated monthly and provided, via e-mail, to CALGB members within institutions randomly assigned to receive educational intervention. Research nurses and/or CRAs received checklists via e-mail and mail and were asked to place the checklists on charts of all older cancer patients.
Case discussion seminar. One or two physicians from each institution were invited to a case discussion seminar held in conjunction with the CALGB group meeting in Washington, DC, in June 2002. At the seminar, five cases illustrating common issues in geriatric oncology were discussed by a panel of experts and the audience. Panelists, chosen to represent different issues within geriatric oncology, included a geriatric psychiatrist (Cornelia Cremens, MD), a medical oncologist (Stuart Lichtman, MD), a surgical oncologist (Margaret Kemeny, MD), and a patient advocate (Edie Fitts).
Evaluation. Efficacy of the intervention was assessed by testing for differences in percentage of accrual of older cancer patients to treatment trials between the standard information and educational intervention arms during the year after the educational symposium and during the year after the case discussion seminar, compared with baseline accrual. Accrual was assessed for CALGB phase II and III treatment trials for all tumor types, and limited to those protocols with no age specification other than age older than 18 years. Phase I protocols with limited access and bone marrow transplantation protocols were excluded because older patients were generally excluded. There were four trials excluded due to age restrictions (132 patient accruals); 10 phase I trials (173 patient accruals); and 9 transplantation trials (132 patient accruals).
Data collection was managed by the CALGB Statistical Center (Durham, NC). Data quality was ensured by careful review of data by CALGB Statistical Center staff and by the study chairperson. Statistical analyses were performed by CALGB statisticians.
Statistical considerations. Thirty-five main member/at-large institutions and 126 institutions in total (main/at-large members plus affiliates) were eligible to participate in this study. The primary end points of the study were percentage of accrual of older patients during the first and second years after the educational seminar compared with baseline accrual. The 1999 accrual to CALGB treatment trials was used as the baseline accrual. Although the unit of random assignment was the main member, the unit of observation for analysis was the individual institutions. Analyses that took into account the intracluster correlation among institution attached to the same main member were necessary when testing arm differences (ie, random effects linear models were used). However, for simplicity of presentation, proportions of total accrual of older patients are presented by arm and by time period. Institutions randomly assigned to the treatment arm that did not send participants to the seminar were analyzed on the basis of intent to treat.
RESULTS
Educational Intervention
There were 53 institutions (comprising 19 main member/at-large networks) randomly assigned to intervention; 72 institutions (comprising 16 main member/at-large networks) served as control.
One hundred three participants from 41 institutions registered to attend the symposium; 101 persons (74 registered; 36 were physicians) from 46 institutions signed in. Approximately 800 mailings and 580 e-mails were sent each month. Forty investigators registered to attend the case discussion seminar; 29 participants (19 registered and 10 not previously registered) signed in at the seminar. During the case discussion, issues raised included the following: more knowledge is needed regarding use of the mental status exam and clinical trial eligibility; depression may mimic decreased cognition, and a brief screening instrument for depression may be helpful in determining who is competent and will understand the informed consent process; many older persons, especially men with prostate cancer, have coexisting illnesses (ie, heart disease) that make them ineligible for available clinical trials, and broadening the eligibility criteria to include patients with stable or controlled illnesses would increase accrual of older persons.
The intervention did not improve accrual of older cancer patients to clinical trials (Table 2). Before the intervention, the percentage of older patients registered to treatment studies was 40% among the intervention institutions compared with 36% in controls (P = .40). During the first and second years postintervention, percentages of older patients in clinical trials were 36% and 31% in the intervention arm and 32% and 31% in the control arm, respectively. There was no significant difference between control and intervention arms in either of the two follow-up evaluations.
Analyzing data with inclusion of all open trials or all CALGB institutions (even those with fewer than five accruals per year for the 3 years before the study) did not change the results. Analysis looking at accrual to trials that were open for the entire 3-year duration of data collection did not change results (data not shown).
Accrual of Older Cancer Patients to CALGB Clinical Trials
During the baseline year and the 2 years following the main intervention of the study, there were 6,431 patient accruals. This included 3,032, 2,160, and 1,239 patients for baseline, years 1 and 2, respectively. The overall accrual of older patients to treatment trials was 37%, 33%, and 31% for baseline, year-1, and year-2 accrual, respectively. For institutions with large accrual rates (> 50 accruals per year), the trend in percentage of accrual of older patients was similar.
The accrual of older persons to trials for treatment of existing disease and to adjuvant/preventive trials was similar at baseline: 37% and 38%, respectively. Over time, however, there was an apparent decline in accrual of older persons to adjuvant trials, whereas percentage of accrual of older patients to treatment trials remained constant (Fig 1).
In looking at accrual across types of cancer, several observations are interesting (Fig 2). First, accrual of older patients to breast cancer trials, whether adjuvant or metastatic, was low: at baseline, 20% in treatment trials and 9% in adjuvant trials. Second, although the percentage of accrual of older patients was higher among those with tumors that were common in older persons than in those with tumors more common in younger persons, older patients were still under-represented. For instance, 20% or less of accruals to breast cancer trials were older, yet nearly half of new cases and two thirds of deaths from breast cancer are in women older than age 65 years.9 Third, there was a decrease over time of accrual of older patients to leukemia, lymphoma, and genitourinary adjuvant treatment trials. Lastly, there was an increase in accrual of older patients to melanoma treatment trials.
DISCUSSION
This is one of the first trials conducted within a large national cooperative group to test an intervention to increase the accrual of older patients to cancer treatment trials. Unfortunately, this multifaceted, brief, directed intervention was not sufficient to change behavior. Accrual to trials was not different between intervention and control, and decreased in both arms of the trial.
Compared with previous reports, the percentage of accrual of older cancer patients to CALGB treatment trials was high: 36% overall. Only two of the trials used in this study were specifically for the elderly. Thirty-eight percent of the accrual of older patients was those age 65 to 69 years; 54% was age 70 to 79 years, and 8% was 80 years of age or older. Reports in the literature find that persons age 65 years and older represent 22% to 32% of trial accruals.3,6,10,11 The odds ratios for enrollment of persons age 65 to 74 years and 75 years or older into therapeutic, nonsurgical clinical trails for breast, colorectal, lung, and prostate cancer were 0.43 (95% CI, 0.42 to 0.44) and 0.15 (95% CI, 0.15-0.16), respectively, compared with persons age 30 to 64 years.12 The high proportion of older patients in cancer treatment trials in CALGB may be explained by the eligibility criteria of available trials during the years of the study. In addition, the high baseline accrual rate of older patients may have been "optimal" or "maximal" and therefore difficult to improve.
Furthermore, the absolute number of accruals, irrespective of age, decreased during the study. The decline in overall accrual probably represents a lack of available trials and/or closure of high-accruing trials. During the study, there were several large clinical trials for common cancers that closed. This included a trial for node-positive breast cancer (CALGB 49906) that closed in January 2002; a trial of hormonal treatment for adjuvant therapy after 5 years of tamoxifen (CALGB 49805) which closed on May 21, 2002; two adjuvant trials for colon cancer (CALGB 9581 and 89804) which closed May 31, 2002, and July 18, 2002, respectively; and CALGB 9720, a leukemia trial specifically for persons older than age 65 years, which closed April 30, 2002. There was also a lack of trials available for node-negative breast cancer and metastatic colon cancer. Since these are common cancers, especially in older individuals, lack of available trials most likely led to a decrease in overall accrual and in the proportion of older persons accrued. Another possible explanation is that the decline in accrual of older persons to clinical trials seen over the course of this trial represents return to baseline level.
To influence clinical trial accrual in a large group, such as the Cancer and Leukemia Group B, it is important to effectively disseminate influential information. There are at least two key components involved. One is the effective dissemination of information, which was not directly assessed in this study. The second is whether the information is sufficient to influence behavior.
The main thrust of our educational intervention was an educational seminar and educational material generated from it. Funds allowed us to invite two to three key personnel from each institution randomly assigned to receive intervention to attend the symposium. Although this is a limited audience, we hoped that we would motivate an advocate within each institution who would disseminate information regarding the importance and safety of offering clinical trials to older patients. We had no way to monitor whether or not this goal was achieved. In order to extend our intervention to include other CALGB members within the intervention arm who were not able to attend the educational seminar, we offered educational materials and sent monthly direct mailings and e-mails. Mailings can be misplaced or lost; their visibility, therefore, is unreliable. E-mails were sent by the study chairperson in order to heighten interest and visibility, but may have been discarded. Videotapes of and slidesets from the educational seminar were sent to key personnel within institutions randomly assigned to receive intervention, but it was not possible to track how often they were used or how effective they were. There were few additional requests for videotapes and slidesets.
Perhaps our intervention reached the wrong target population (eg, medical oncologists as opposed to surgeons). Siminoff et al5 reported that the chance of referral to a clinical trial was substantially greater when physicians, especially surgeons, were more involved in making a decision about a patient’s adjuvant therapy. In particular, if surgeons started tamoxifen treatment themselves, they were significantly less likely to make referral to a clinical trial (P .05). Age was less of an issue in medical oncologists’ referal to a clinical trial. Directing intervention to surgeons, who often have the initial medical contact with cancer patients, may improve accrual.
Another issue is whether the information provided was sufficient to influence change. Our intervention was designed to educate the medical oncologist and the clinical research team about older cancer patients, the effect of comorbidity on cancer treatment, and toxicity of chemotherapy in older patients, and to remind them to consider available clinical trials for their older cancer patients: themes commonly identified in the literature as needing attention.5-8 Studies have shown that the physicians’ perception about age and tolerance of treatment was the greatest impediment to enrolling older women onto trials, and that concerns about comorbidities (16%) and treatment toxicities (14%) are barriers to accrual.7 There was also, however, concern expressed that older patients would have difficulty understanding the requirements of complicated trials leading to compliance problems (16%) and that additional personnel were needed to explain clinical trials to the older patients and their families (25%). We provided protocol lists to be used to remind physicians and staff about available protocols, but the availability of staff to augment awareness was not provided and could not be assessed. More support staff to assist with clinical trial accrual might be another way of improving accrual.
Since the outcome measure was actual accrual of older patients to CALGB treatment trials during the two years following the intervention, we may have missed changes in clinical practice that happened as a result of the intervention. For instance, we measured accrual and not whether a clinical trial was offered to an older cancer patient, an area where Kemeny et al8 found deficiency. Another CALGB study (CALGB 369901), "Observational Cohort Study: Chemotherapy Decisions and Outcomes in Women Age 65 or Older with Operable, Newly Diagnosed Breast Cancer," will provide further information about whether clinical trials are offered to older women, how they make treatment decisions, and how these decisions affect quality of life and satisfaction with treatment. Another possible reason for lack of change in accrual was that the 1- and 2-year follow-up was not long enough; longer follow-up might be necessary to see an actual change in accrual.
Changing behavior of providers, and of the patients and families with which they interact, is notoriously difficult. Interventions to improve cancer screening rates through education have been only modestly successful, and are most effective when they are multilevel and are conducted in concert with feedback on practice and/or simultaneous patient-directed activities13-15 In a meta-analysis of continuing medical education (CME), Davis concluded that formal CME without practice reinforcing or enabling components has relatively little impact on changing physician behavior.15 The most effective approaches included academic detailing, opinion leaders, reminders, patient mediated activities, and multiple interventions, and interactive workshops.15-18 In one large trial of CME, Parochka and Paprockas19 noted that a single CME session was inadequate to change behavior regarding tamoxifen prescription. Barriers to practice change identified by physicians include lack of professional consensus, lack of time for counseling, lack of time and inadequate reimbursement for time spent talking to patients and families, and lack of regulatory structures.19, 20
We included all treatment trials without an upper age limit. Our finding of lower accrual of older patients to adjuvant versus metastatic trials (29% v 36%, P < .0001) is consistent with what is seen in the literature. In the NCI database, older patients accounted for 25% of participants in trials for early stage disease versus 41% for trials in late stage disease. Likewise, of accruals within the National Cancer Institute of Canada–Clinical Trials Group, older patients accounted for 19% in adjuvant trials and 30% in metastatic disease trials.6 These findings suggest an age bias in treating older cancer patients: perhaps a lower risk:benefit ratio was deemed necessary to enter older patients onto adjuvant treatment trials. Other trials have reported that physicians hesitate to offer older patients clinical trial participation because they are concerned that comorbidity will increase toxicity, even though the presence of the comorbidity, in itself, did not make them ineligible for the trial.7 If ageism is a major factor in offering clinical trials to older cancer patients, education about cancer and its treatment will be inadequate to change accrual patterns, and additional intervention to reduce negative attitudes toward older patients will be needed.
In summary, the educational intervention we studied did not improve accrual of older cancer patients to clinical trials within the study time period. Educating physicians in order to change practice is an enormous task. More intense and/or far-reaching approaches than those used in this study will probably be necessary to improve accrual of older cancer patients to cancer clinical trials. Furthermore, changing patterns of enrollment to clinical trials is probably a multidimensional problem. Ageism, among patients, family members, and the medical team, must be overcome in this process. Currently, trials designed specifically for older patients are accruing and should shed light on this area. Future studies with this growing population in mind are imperative to further understand treatment efficacy and toxicity in this age group.
Authors’ Disclosures of Potential Conflicts of Interest
The authors indicated no potential conflicts of interest.
NOTES
Supported in part by grants from the National Cancer Institute (CA31946) and supplemental funding from the National Institute on Aging.
Presented in abstract form in the Proceedings of the American Society of Clinical Oncology, 2004 (abstr 8040); and in poster form at the American Society of Clinical Oncology meeting, June 8, 2004.
The content of this manuscript is solely the responsibility of the authors and does not necessarily represent the official views of the National Cancer Institute.
Authors’ disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
Ershler W, Longo D: Aging and cancer; issues of basic and clinical science. J Natl Cancer Inst 89:1489-1497, 1997
Jemal A, Tiuari R, Murray T, et al: Cancer Statistics, 2004. Cancer 54:8-29, 2004
Hutchins LF, Unger JM, Crowley JJ, et al: Underrepresentation of patients 65 years of age or older in cancer-treatment trials. N Engl J Med 341:2061-2067, 1999
Trimble EL, Carter CL, Cain D, et al: Representation of older patients in cancer treatment trials. Cancer 74:2208-2214, 1994
Siminoff LA, Zhang A, Colabianchi N, et al: Factors that predict the referral of breast cancer patients onto clinical trials by their surgeons and medical onologists. J Clin Oncol 18:1203-1211, 2000
Yee KWL, Pater JL, Pho L, et al: Enrollment of older patients in cancer treatment trials in Canada: Why is age a barrier J Clin Oncol 21:1618-1623, 2003
Kornblith AB, Kemeny M, Peterson BL, et al: Survey of oncologists’ perceptions of barriers to accrual of older patients with breast carcinoma to clinical trials. Cancer 95:989-996, 2002
Kemeny MM, Peterson BL, Kornblith AB, et al: Barriers to clinical trial participation by older women with breast cancer. J Clin Oncol 21:2268-2275, 2003
Yancik R, Ries LA: Aging and cancer in America: Demographic and epidemiologic perspectives. Hematol Oncol Clin North Am 14:17-23, 2000
Monfardini S, Sorio R, Boes GH, et al: Entry and evaluation of elderly patients in European Organization for Research and Treatment of Cancer (EORTC) new-drug development studies. Cancer 76:333-338, 1995
Lewis JH, Kilgore ML, Goldman DP, et al: Participation of patients 65 years of age or older in cancer clinical trials. J Clin Oncol 21:1383-1389, 2003
Murthy VH, Krumholz HM, Gross CP: Participation in cancer clinical trials: Race-, sex-, and age-based disparities. JAMA 291:2720-2726, 2004
Mandelblatt JS, Yabroff KR: Effectiveness of interventions designed to increase mammography use: A meta-analysis of provider-targeted strategies. Cancer Epidemiol Biomarkers Prev 8:759-767, 1999
Vabroff K, Mandelblatt J: Interventions targeted toward patients to increase mammography use. Cancer Epidemiol Biomarkers Prev 8:749-757, 1999
Davis D: Does CME work An analysis of the effect of educational activities on physician performance or health care outcomes. Int J Psychiatry Med 28:21-39, 1998
Davis DA, Thomson MA, Oxman AD, et al: Changing physician performance: A systematic review of the effect of continuing medical education strategies. JAMA 274:700-705, 1995
Freemantle N, Harvey EL, Wolf F, et al: Printed educational materials: Effects on professional practice and health care outcomes. [Cochrane Database System Review]. Oxford, England, Cochrane Library, CD000172, issue 2, 2000
O’Brien TM, Freemantle N, Oxman AD, et al: Continuing education meetings and workshops: Effects on professional practice and health care outcomes. [Cochrane Database System Review]. Oxford, England, Cochrane Library, CD003030, issue 1, 2001
Parochka J, Paprockas K: A continuing medical education lecture and workshop, physician behavior, and barriers to change. J Contin Educ Health Prof 21:110-116, 2001
Davis DA, Taylor-Valsey A: Translating guidelines into practice, A systematic review of theoretic concepts practical experience and research evidence in the adoption of clinical practice guidelines. CMAJ 157:408-416, 1997(Gretchen G. Kimmick, Berc)
CALGB Statistical Center
Duke University Medical Center, Durham, NC
Wake Forest University School of Medicine, Winston-Salem, NC
Dana-Farber Cancer Institute, Boston, MA, Georgetown University Medical Center, Washington, DC
Vermont Cancer Center, Burlington, VT
ABSTRACT
PURPOSE: To design and test a geriatric educational intervention to improve accrual of cancer patients age 65 years and older to cooperative group–sponsored treatment trials.
METHODS: Main member institutions of the Cancer and Leukemia Group B (CALGB) and its affiliates were randomly assigned to receive standard information (n = 73) or educational intervention (n = 53). Standard information included CALGB Web site access and periodic notification about existing trials. The geriatric educational intervention included standard information plus: (1) an educational seminar; (2) educational materials; (3) a list of available protocols for use on charts; (4) a monthly e-mail and mail reminders for 1 year; and (5) a case discussion seminar. The main outcome was percentage of accrual of older persons to phase II and III treatment protocols after study initiation compared with baseline.
RESULTS: There were 3,032 patients entered onto trials in the baseline year, and 2,160 and 1,239 during the 2 years postintervention, respectively. Overall percentage of accrual of older patients was 37% at baseline, and 33% and 31% during the first and second years after intervention. There was no improvement in accrual in the intervention versus control arm: 36% v 32% in the first year and 31% v 31% in the second year.
CONCLUSION: Accrual of older patients was not increased by this intervention. Reasons for lack of effect include low intervention intensity, high baseline accrual rates, and closure of several high-accruing protocols during the study. More intense and multifaceted approaches will be needed to change physician (and patient) behavior and to increase accrual of older persons to clinical trials.
INTRODUCTION
Persons age 65 or older (hereinafter referred to as "older") comprise approximately 14% of the US population, but account for 63% of the cases of cancer in this country.1-3 Moreover, the projected "graying of America," coupled with advances in cancer screening and treatment, will result in a large increase in the absolute number of older cancer patients over the coming decades. Thus, there is a demographic imperative driving the need to improve cancer care to older populations.
Clinical trials are credited for a large proportion of the improvements in cancer therapy. The group at highest risk for cancer, older patients, however, is least likely to be represented in trials.3-5 For example, the mean age of patients having the five cancers with the highest mortality rates that entered National Cancer Institute (NCI)-sponsored treatment trials was significantly lower (48.7 years for men and 54.7 years for women) than that of incident cancer cases registered by Surveillance, Epidemiology, and End Results (SEER; 69.1 years for men and 64.6 years for women).4 In Southwest Oncology Group (SWOG) treatment trials, only 25% of patients enrolled in trials were age 65 years and older.3
Reasons for under-representation of older individuals in clinical trials are multidimensional, and include comorbid illnesses; age bias; physician, patient, and family lack of knowledge about toxicity and benefit of treatment; eligibility criteria excluding older patients from available trials; lack of financial, logistic, and social support; patient or family members’ preference; and excessive time required for enrollment.3-7 Kemeny et al8 found that age itself was a major factor in whether a clinical trial was offered to breast cancer patients, even after controlling for number of comorbidities, physical functioning, and disease stage. However, when offered participation in clinical trials, older women were equally as likely as younger women to participate. In the study of Kemeny et al, 42% of physicians stated that greater education about cancer treatment toxicity in the older person would improve accrual.
Here, we present results of a phase III randomized trial testing the hypothesis that a multifaceted educational intervention directed to physicians and other members of the medical and research team at Cancer and Leukemia Group B (CALGB) institutions would improve the accrual of older persons to cancer treatment trials.
METHODS
Eligibility
The CALGB is a National Cancer Institute–funded cooperative group that conducts studies in adults with cancer. All main member institutions in the CALGB and their affiliates with an average accrual to treatment protocols of 5 or more patients per year over the 3 years before June 2001 were included. The study was exempt from institutional review board review under 45 CFR 46.101(b).1
Pilot Educational Intervention
Registration and random assignment. Main member networks were randomly assigned as a unit to receive standard information or educational intervention. There are three types of institutional memberships within CALGB: main, at-large, and affiliate members. Main members are academic institutions that demonstrate scientific leadership and whose network registers 50 patients annually on CALGB studies. At-large members are academic or community hospitals whose network registers at least 30 patients annually on CALGB studies. Affiliate members participate in group activities as part of a network of sites that are coordinated by a main or at-large member. After random assignment, institutions randomly assigned to receive educational intervention were notified by the study chair or a member of the study team.
Standard information consisted of: 1) periodic notification of all existing CALGB trials per the CALGB Central Office; and 2) CALGB Web site access.
Educational intervention included standard information plus: 1) an educational symposium; 2) geriatric oncology educational materials; 3) monthly mailings and e-mails for 1 year; 4) lists of available protocols for use on patient charts; and 5) a case discussion seminar. Components of educational intervention are listed in Table 1.
Educational symposium. Two or three key members of the research team from each institution were invited and funded to attend the educational symposium. We encouraged designation of one physician and one or more additional member of the oncology research team, which could include a research nurse, clinical research associate (CRA), psychologist, or social worker. The symposium was held in June 2001, in conjunction with the CALGB group meetings in Ottawa, Ontario, Canada.
Geriatric oncology educational materials. Educational materials, including outlines of each lecture and a bibliography, were provided to symposium participants. Lecture slide sets were compiled, and a videotape of the symposium was made; these were available in February 2002. The videotape and slidesets were sent to principal investigators and lead CRAs. All education materials were sent, on request, to individuals within institutions randomly assigned to intervention.
Mailings and e-mail reminders. Monthly mailings and e-mail reminders were sent by the study chair to all members of institutions randomly assigned to intervention during the year after the educational symposium. E-mails were sent when addresses were available and functional. Each mailing and e-mail contained a brief educational narrative about cancer in older patients and a reminder to consider clinical trials for older patients. In order to prompt consideration of clinical trials, checklist reminders of available protocols were updated monthly and provided, via e-mail, to CALGB members within institutions randomly assigned to receive educational intervention. Research nurses and/or CRAs received checklists via e-mail and mail and were asked to place the checklists on charts of all older cancer patients.
Case discussion seminar. One or two physicians from each institution were invited to a case discussion seminar held in conjunction with the CALGB group meeting in Washington, DC, in June 2002. At the seminar, five cases illustrating common issues in geriatric oncology were discussed by a panel of experts and the audience. Panelists, chosen to represent different issues within geriatric oncology, included a geriatric psychiatrist (Cornelia Cremens, MD), a medical oncologist (Stuart Lichtman, MD), a surgical oncologist (Margaret Kemeny, MD), and a patient advocate (Edie Fitts).
Evaluation. Efficacy of the intervention was assessed by testing for differences in percentage of accrual of older cancer patients to treatment trials between the standard information and educational intervention arms during the year after the educational symposium and during the year after the case discussion seminar, compared with baseline accrual. Accrual was assessed for CALGB phase II and III treatment trials for all tumor types, and limited to those protocols with no age specification other than age older than 18 years. Phase I protocols with limited access and bone marrow transplantation protocols were excluded because older patients were generally excluded. There were four trials excluded due to age restrictions (132 patient accruals); 10 phase I trials (173 patient accruals); and 9 transplantation trials (132 patient accruals).
Data collection was managed by the CALGB Statistical Center (Durham, NC). Data quality was ensured by careful review of data by CALGB Statistical Center staff and by the study chairperson. Statistical analyses were performed by CALGB statisticians.
Statistical considerations. Thirty-five main member/at-large institutions and 126 institutions in total (main/at-large members plus affiliates) were eligible to participate in this study. The primary end points of the study were percentage of accrual of older patients during the first and second years after the educational seminar compared with baseline accrual. The 1999 accrual to CALGB treatment trials was used as the baseline accrual. Although the unit of random assignment was the main member, the unit of observation for analysis was the individual institutions. Analyses that took into account the intracluster correlation among institution attached to the same main member were necessary when testing arm differences (ie, random effects linear models were used). However, for simplicity of presentation, proportions of total accrual of older patients are presented by arm and by time period. Institutions randomly assigned to the treatment arm that did not send participants to the seminar were analyzed on the basis of intent to treat.
RESULTS
Educational Intervention
There were 53 institutions (comprising 19 main member/at-large networks) randomly assigned to intervention; 72 institutions (comprising 16 main member/at-large networks) served as control.
One hundred three participants from 41 institutions registered to attend the symposium; 101 persons (74 registered; 36 were physicians) from 46 institutions signed in. Approximately 800 mailings and 580 e-mails were sent each month. Forty investigators registered to attend the case discussion seminar; 29 participants (19 registered and 10 not previously registered) signed in at the seminar. During the case discussion, issues raised included the following: more knowledge is needed regarding use of the mental status exam and clinical trial eligibility; depression may mimic decreased cognition, and a brief screening instrument for depression may be helpful in determining who is competent and will understand the informed consent process; many older persons, especially men with prostate cancer, have coexisting illnesses (ie, heart disease) that make them ineligible for available clinical trials, and broadening the eligibility criteria to include patients with stable or controlled illnesses would increase accrual of older persons.
The intervention did not improve accrual of older cancer patients to clinical trials (Table 2). Before the intervention, the percentage of older patients registered to treatment studies was 40% among the intervention institutions compared with 36% in controls (P = .40). During the first and second years postintervention, percentages of older patients in clinical trials were 36% and 31% in the intervention arm and 32% and 31% in the control arm, respectively. There was no significant difference between control and intervention arms in either of the two follow-up evaluations.
Analyzing data with inclusion of all open trials or all CALGB institutions (even those with fewer than five accruals per year for the 3 years before the study) did not change the results. Analysis looking at accrual to trials that were open for the entire 3-year duration of data collection did not change results (data not shown).
Accrual of Older Cancer Patients to CALGB Clinical Trials
During the baseline year and the 2 years following the main intervention of the study, there were 6,431 patient accruals. This included 3,032, 2,160, and 1,239 patients for baseline, years 1 and 2, respectively. The overall accrual of older patients to treatment trials was 37%, 33%, and 31% for baseline, year-1, and year-2 accrual, respectively. For institutions with large accrual rates (> 50 accruals per year), the trend in percentage of accrual of older patients was similar.
The accrual of older persons to trials for treatment of existing disease and to adjuvant/preventive trials was similar at baseline: 37% and 38%, respectively. Over time, however, there was an apparent decline in accrual of older persons to adjuvant trials, whereas percentage of accrual of older patients to treatment trials remained constant (Fig 1).
In looking at accrual across types of cancer, several observations are interesting (Fig 2). First, accrual of older patients to breast cancer trials, whether adjuvant or metastatic, was low: at baseline, 20% in treatment trials and 9% in adjuvant trials. Second, although the percentage of accrual of older patients was higher among those with tumors that were common in older persons than in those with tumors more common in younger persons, older patients were still under-represented. For instance, 20% or less of accruals to breast cancer trials were older, yet nearly half of new cases and two thirds of deaths from breast cancer are in women older than age 65 years.9 Third, there was a decrease over time of accrual of older patients to leukemia, lymphoma, and genitourinary adjuvant treatment trials. Lastly, there was an increase in accrual of older patients to melanoma treatment trials.
DISCUSSION
This is one of the first trials conducted within a large national cooperative group to test an intervention to increase the accrual of older patients to cancer treatment trials. Unfortunately, this multifaceted, brief, directed intervention was not sufficient to change behavior. Accrual to trials was not different between intervention and control, and decreased in both arms of the trial.
Compared with previous reports, the percentage of accrual of older cancer patients to CALGB treatment trials was high: 36% overall. Only two of the trials used in this study were specifically for the elderly. Thirty-eight percent of the accrual of older patients was those age 65 to 69 years; 54% was age 70 to 79 years, and 8% was 80 years of age or older. Reports in the literature find that persons age 65 years and older represent 22% to 32% of trial accruals.3,6,10,11 The odds ratios for enrollment of persons age 65 to 74 years and 75 years or older into therapeutic, nonsurgical clinical trails for breast, colorectal, lung, and prostate cancer were 0.43 (95% CI, 0.42 to 0.44) and 0.15 (95% CI, 0.15-0.16), respectively, compared with persons age 30 to 64 years.12 The high proportion of older patients in cancer treatment trials in CALGB may be explained by the eligibility criteria of available trials during the years of the study. In addition, the high baseline accrual rate of older patients may have been "optimal" or "maximal" and therefore difficult to improve.
Furthermore, the absolute number of accruals, irrespective of age, decreased during the study. The decline in overall accrual probably represents a lack of available trials and/or closure of high-accruing trials. During the study, there were several large clinical trials for common cancers that closed. This included a trial for node-positive breast cancer (CALGB 49906) that closed in January 2002; a trial of hormonal treatment for adjuvant therapy after 5 years of tamoxifen (CALGB 49805) which closed on May 21, 2002; two adjuvant trials for colon cancer (CALGB 9581 and 89804) which closed May 31, 2002, and July 18, 2002, respectively; and CALGB 9720, a leukemia trial specifically for persons older than age 65 years, which closed April 30, 2002. There was also a lack of trials available for node-negative breast cancer and metastatic colon cancer. Since these are common cancers, especially in older individuals, lack of available trials most likely led to a decrease in overall accrual and in the proportion of older persons accrued. Another possible explanation is that the decline in accrual of older persons to clinical trials seen over the course of this trial represents return to baseline level.
To influence clinical trial accrual in a large group, such as the Cancer and Leukemia Group B, it is important to effectively disseminate influential information. There are at least two key components involved. One is the effective dissemination of information, which was not directly assessed in this study. The second is whether the information is sufficient to influence behavior.
The main thrust of our educational intervention was an educational seminar and educational material generated from it. Funds allowed us to invite two to three key personnel from each institution randomly assigned to receive intervention to attend the symposium. Although this is a limited audience, we hoped that we would motivate an advocate within each institution who would disseminate information regarding the importance and safety of offering clinical trials to older patients. We had no way to monitor whether or not this goal was achieved. In order to extend our intervention to include other CALGB members within the intervention arm who were not able to attend the educational seminar, we offered educational materials and sent monthly direct mailings and e-mails. Mailings can be misplaced or lost; their visibility, therefore, is unreliable. E-mails were sent by the study chairperson in order to heighten interest and visibility, but may have been discarded. Videotapes of and slidesets from the educational seminar were sent to key personnel within institutions randomly assigned to receive intervention, but it was not possible to track how often they were used or how effective they were. There were few additional requests for videotapes and slidesets.
Perhaps our intervention reached the wrong target population (eg, medical oncologists as opposed to surgeons). Siminoff et al5 reported that the chance of referral to a clinical trial was substantially greater when physicians, especially surgeons, were more involved in making a decision about a patient’s adjuvant therapy. In particular, if surgeons started tamoxifen treatment themselves, they were significantly less likely to make referral to a clinical trial (P .05). Age was less of an issue in medical oncologists’ referal to a clinical trial. Directing intervention to surgeons, who often have the initial medical contact with cancer patients, may improve accrual.
Another issue is whether the information provided was sufficient to influence change. Our intervention was designed to educate the medical oncologist and the clinical research team about older cancer patients, the effect of comorbidity on cancer treatment, and toxicity of chemotherapy in older patients, and to remind them to consider available clinical trials for their older cancer patients: themes commonly identified in the literature as needing attention.5-8 Studies have shown that the physicians’ perception about age and tolerance of treatment was the greatest impediment to enrolling older women onto trials, and that concerns about comorbidities (16%) and treatment toxicities (14%) are barriers to accrual.7 There was also, however, concern expressed that older patients would have difficulty understanding the requirements of complicated trials leading to compliance problems (16%) and that additional personnel were needed to explain clinical trials to the older patients and their families (25%). We provided protocol lists to be used to remind physicians and staff about available protocols, but the availability of staff to augment awareness was not provided and could not be assessed. More support staff to assist with clinical trial accrual might be another way of improving accrual.
Since the outcome measure was actual accrual of older patients to CALGB treatment trials during the two years following the intervention, we may have missed changes in clinical practice that happened as a result of the intervention. For instance, we measured accrual and not whether a clinical trial was offered to an older cancer patient, an area where Kemeny et al8 found deficiency. Another CALGB study (CALGB 369901), "Observational Cohort Study: Chemotherapy Decisions and Outcomes in Women Age 65 or Older with Operable, Newly Diagnosed Breast Cancer," will provide further information about whether clinical trials are offered to older women, how they make treatment decisions, and how these decisions affect quality of life and satisfaction with treatment. Another possible reason for lack of change in accrual was that the 1- and 2-year follow-up was not long enough; longer follow-up might be necessary to see an actual change in accrual.
Changing behavior of providers, and of the patients and families with which they interact, is notoriously difficult. Interventions to improve cancer screening rates through education have been only modestly successful, and are most effective when they are multilevel and are conducted in concert with feedback on practice and/or simultaneous patient-directed activities13-15 In a meta-analysis of continuing medical education (CME), Davis concluded that formal CME without practice reinforcing or enabling components has relatively little impact on changing physician behavior.15 The most effective approaches included academic detailing, opinion leaders, reminders, patient mediated activities, and multiple interventions, and interactive workshops.15-18 In one large trial of CME, Parochka and Paprockas19 noted that a single CME session was inadequate to change behavior regarding tamoxifen prescription. Barriers to practice change identified by physicians include lack of professional consensus, lack of time for counseling, lack of time and inadequate reimbursement for time spent talking to patients and families, and lack of regulatory structures.19, 20
We included all treatment trials without an upper age limit. Our finding of lower accrual of older patients to adjuvant versus metastatic trials (29% v 36%, P < .0001) is consistent with what is seen in the literature. In the NCI database, older patients accounted for 25% of participants in trials for early stage disease versus 41% for trials in late stage disease. Likewise, of accruals within the National Cancer Institute of Canada–Clinical Trials Group, older patients accounted for 19% in adjuvant trials and 30% in metastatic disease trials.6 These findings suggest an age bias in treating older cancer patients: perhaps a lower risk:benefit ratio was deemed necessary to enter older patients onto adjuvant treatment trials. Other trials have reported that physicians hesitate to offer older patients clinical trial participation because they are concerned that comorbidity will increase toxicity, even though the presence of the comorbidity, in itself, did not make them ineligible for the trial.7 If ageism is a major factor in offering clinical trials to older cancer patients, education about cancer and its treatment will be inadequate to change accrual patterns, and additional intervention to reduce negative attitudes toward older patients will be needed.
In summary, the educational intervention we studied did not improve accrual of older cancer patients to clinical trials within the study time period. Educating physicians in order to change practice is an enormous task. More intense and/or far-reaching approaches than those used in this study will probably be necessary to improve accrual of older cancer patients to cancer clinical trials. Furthermore, changing patterns of enrollment to clinical trials is probably a multidimensional problem. Ageism, among patients, family members, and the medical team, must be overcome in this process. Currently, trials designed specifically for older patients are accruing and should shed light on this area. Future studies with this growing population in mind are imperative to further understand treatment efficacy and toxicity in this age group.
Authors’ Disclosures of Potential Conflicts of Interest
The authors indicated no potential conflicts of interest.
NOTES
Supported in part by grants from the National Cancer Institute (CA31946) and supplemental funding from the National Institute on Aging.
Presented in abstract form in the Proceedings of the American Society of Clinical Oncology, 2004 (abstr 8040); and in poster form at the American Society of Clinical Oncology meeting, June 8, 2004.
The content of this manuscript is solely the responsibility of the authors and does not necessarily represent the official views of the National Cancer Institute.
Authors’ disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
Ershler W, Longo D: Aging and cancer; issues of basic and clinical science. J Natl Cancer Inst 89:1489-1497, 1997
Jemal A, Tiuari R, Murray T, et al: Cancer Statistics, 2004. Cancer 54:8-29, 2004
Hutchins LF, Unger JM, Crowley JJ, et al: Underrepresentation of patients 65 years of age or older in cancer-treatment trials. N Engl J Med 341:2061-2067, 1999
Trimble EL, Carter CL, Cain D, et al: Representation of older patients in cancer treatment trials. Cancer 74:2208-2214, 1994
Siminoff LA, Zhang A, Colabianchi N, et al: Factors that predict the referral of breast cancer patients onto clinical trials by their surgeons and medical onologists. J Clin Oncol 18:1203-1211, 2000
Yee KWL, Pater JL, Pho L, et al: Enrollment of older patients in cancer treatment trials in Canada: Why is age a barrier J Clin Oncol 21:1618-1623, 2003
Kornblith AB, Kemeny M, Peterson BL, et al: Survey of oncologists’ perceptions of barriers to accrual of older patients with breast carcinoma to clinical trials. Cancer 95:989-996, 2002
Kemeny MM, Peterson BL, Kornblith AB, et al: Barriers to clinical trial participation by older women with breast cancer. J Clin Oncol 21:2268-2275, 2003
Yancik R, Ries LA: Aging and cancer in America: Demographic and epidemiologic perspectives. Hematol Oncol Clin North Am 14:17-23, 2000
Monfardini S, Sorio R, Boes GH, et al: Entry and evaluation of elderly patients in European Organization for Research and Treatment of Cancer (EORTC) new-drug development studies. Cancer 76:333-338, 1995
Lewis JH, Kilgore ML, Goldman DP, et al: Participation of patients 65 years of age or older in cancer clinical trials. J Clin Oncol 21:1383-1389, 2003
Murthy VH, Krumholz HM, Gross CP: Participation in cancer clinical trials: Race-, sex-, and age-based disparities. JAMA 291:2720-2726, 2004
Mandelblatt JS, Yabroff KR: Effectiveness of interventions designed to increase mammography use: A meta-analysis of provider-targeted strategies. Cancer Epidemiol Biomarkers Prev 8:759-767, 1999
Vabroff K, Mandelblatt J: Interventions targeted toward patients to increase mammography use. Cancer Epidemiol Biomarkers Prev 8:749-757, 1999
Davis D: Does CME work An analysis of the effect of educational activities on physician performance or health care outcomes. Int J Psychiatry Med 28:21-39, 1998
Davis DA, Thomson MA, Oxman AD, et al: Changing physician performance: A systematic review of the effect of continuing medical education strategies. JAMA 274:700-705, 1995
Freemantle N, Harvey EL, Wolf F, et al: Printed educational materials: Effects on professional practice and health care outcomes. [Cochrane Database System Review]. Oxford, England, Cochrane Library, CD000172, issue 2, 2000
O’Brien TM, Freemantle N, Oxman AD, et al: Continuing education meetings and workshops: Effects on professional practice and health care outcomes. [Cochrane Database System Review]. Oxford, England, Cochrane Library, CD003030, issue 1, 2001
Parochka J, Paprockas K: A continuing medical education lecture and workshop, physician behavior, and barriers to change. J Contin Educ Health Prof 21:110-116, 2001
Davis DA, Taylor-Valsey A: Translating guidelines into practice, A systematic review of theoretic concepts practical experience and research evidence in the adoption of clinical practice guidelines. CMAJ 157:408-416, 1997(Gretchen G. Kimmick, Berc)