|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
© 2000 American Society for Clinical Oncology Factors That Predict the Referral of Breast Cancer Patients Onto Clinical Trials by Their Surgeons and Medical OncologistsFrom the Case Western Reserve University, Cleveland, OH. Address reprint requests to Laura Siminoff, PhD, School of Medicine WG-37, Case Western Reserve University, 10900 Euclid Ave, Cleveland, OH 44106-4961; email LAS5{at}po.cwru.edu
PURPOSE: To improve understanding of physicians’ reluctance to refer patients to clinical trials. METHODS: This study was conducted in a large metropolitan region from 1993 to 1995 using a two-staged population-based sampling strategy. A total of 147 physicians discussed 245 patient cases and their own knowledge, attitudes, and practices toward clinical trials. RESULTS: Ninety-three patients (38.0%) were offered a trial, and 49 (52.7%) of them agreed to participate. Forty-five patients (18.4%) actually received their adjuvant therapy on trial. Older patients and those with a poorer prognosis were less likely to be referred. Patients who delayed their decision were more than three times as likely to participate in a trial and more than eight times as likely to participate when they were reported to be actively involved in making the decision. Generally, physicians in university settings and who had formal support from a cooperative group were more likely to refer patients to trials. More specifically, surgeons referred more patients to trials when they felt comfortable explaining trials or believed that treatment should not stray from protocol. Oncologists were less likely to make referrals if they perceived the paperwork to be onerous or entry requirements to be too stringent. Surgeons’ participation in recommending adjuvant therapy to patients resulted in more trial referrals unless they treated their patients with tamoxifen. CONCLUSION: (1) Physicians still need to overcome attitudinal and practical barriers to trial participation, (2) more support for physicians is needed, (3) surgeons may play a pivotal role in the recruitment of patients to adjuvant therapy trials, and (4) garnering patient enthusiasm for trial participation and involving them in the choice of adjuvant therapy may be key components to increasing trial enrollment.
FOR YEARS, THE SLOW pace of recruitment to clinical trials has frustrated the medical community. More recently, patients have become aware of the slow pace of medical research and how it hampers the provision of state-of-the-art treatment. Patient advocacy groups, most notably those working on behalf of breast cancer patients and those with acquired immune deficiency syndrome, have been lobbying for the treatment development process to be hastened. These efforts and resultant congressional pressures have resulted in the "fast track" process, whereby at least some pharmaceuticals receive expedited review and approval.1 Despite these changes, the completion of a clinical trial is still a slow and sometimes frustrating endeavor. It has been estimated that no more than 2% to 3% of all breast cancer patients actually participate in a clinical trial during the course of their therapy.2 In truth, despite the clamor by the public to more quickly develop new and better drugs, the first stumbling block to the timely accrual to and completion of clinical trials is the reluctance of physicians to offer patients a chance to participate in a trial. The second major barrier is the low acceptance rate of patients when a trial is offered. From both a scientific and ethical viewpoint, it is extremely desirable that every patient who meets the criteria of a clinical trial has an equal chance of participating in that trial. Scientifically, widespread recruitment minimizes the biases in a trial’s sample. Although a tacit assumption of no bias due to subject refusal is usually made, selection bias has been demonstrated, and this bias compromises the generalizability of trial results to patient populations.3,4 Studies examining the effects of volunteer bias on experimental outcomes5,6 and behavioral interventions7 have found interaction effects between the methods of subject selection and selected outcome variables. The nonrandom event of volunteering for a study, determined in part by situational variables and in part by specific personal attributes, increases the probability of subsequently inflating the effects of treatment compared with what can be expected to occur within the general patient population.8 The resulting accumulated biases inevitably limit study generalizability. In addition, the lack of timely and broad-based accrual to clinical trials poses policy and service delivery problems. When trials are not completed in a timely fashion, treatment decisions must be based on incomplete data. Rapid diffusion of treatment information into the larger health care community is hampered. Slow completion of phase III trials is especially detrimental because an overly long research period creates a tendency for drugs to be used off-label as patients pressure physicians for drugs for which efficacy and safety have not yet been determined. The body of research that addresses the issues surrounding recruitment of patients to clinical trials is small but consistent on at least one point: the central role of the physician in the treatment decision-making process. Studies have reported a wide range of patient recruitment acceptance rates, from 15% to 40% for cancer treatment trials.9-11 However, patients will rarely enter trials unless physicians actively recommend the trials to them. Several studies have suggested that physician treatment preferences account for the majority of trial-eligible patients being entered or not entered onto study protocols.11-13 A better understanding of physicians’ reluctance to refer and recruit patients to trials was a major objective of the study reported here. Two issues are involved: (1) why physicians are generally reluctant to participate in clinical trials at all and (2) why participating physicians refer only a small percentage of their patients to trials. This study examines actual physician referral practices to clinical trials.
Study Sample This study was conducted in a large metropolitan region in Pennsylvania from 1993 to 1995 using a two-staged population-based sampling strategy. The first stage identified all physicians who provided primary surgical care to breast cancer patients in the sampling region. A list of 272 physicians who possibly provided primary surgical care to breast cancer patients was compiled from data provided by county medical associations. The majority was general and oncology surgeons, but a small number of plastic surgeons and obstetrician-gynecologists also provided such treatment. A telephone screen yielded 198 surgeons who reported providing surgical treatment to breast cancer patients. Of this sample, 48 refused to participate in the study beyond the screening process. The final sample of 150 surgeons represents a study acceptance rate of 75.8%, of whom 107 had eligible patients and participated in the study. The 150 eligible surgeons were mailed a letter explaining the nature of the study and asking for their participation. A telephone call followed to schedule a time to conduct chart reviews in their offices and a 1-hour interview with the surgeon. The in-person interview centered on discussing adjuvant therapy. We selected up to four of the surgeon’s most recent breast cancer patients for discussion. Patients chosen for the interview protocol had been technically eligible to participate in at least one of the phase III breast cancer treatment trials open for recruitment in the region when the patient’s surgical care was provided. A trained research nurse performed the record reviews to identify and select cases. Basic medical, disease, and demographic characteristics of the patients were abstracted from the medical records. Cases that were referred to a medical oncologist for consideration of adjuvant therapy were followed up directly with the medical oncologist. Fifty-five oncologists were eligible to participate in the study and 40 agreed to participate in the study, for a response rate of 72.73%. The final practitioner sample was composed of 147 physicians (107 surgeons and 40 medical oncologists). Written informed consent was obtained from all participants and was approved by the sponsoring institution’s institutional review board. Physicians were told that we were studying how treatment decisions are made for newly diagnosed breast cancer patients.
Elicitation of Case Histories The first half of the interview reviewed patient case histories to elicit decision criteria. An initial overall review of the individual case helped the physician "recreate" the case and aided subject recall.14,15 This initial review asked physicians to recount what the patient looked like, who accompanied her, and what the first impression of the patient was. Interviewers were trained to conduct these interviews during a 6-month training period, and interviewers covered a series of scripted questions. Probing of responses was done using structured probes. Elicited data included the following: (1) a general history of the patient’s breast cancer illness; (2) an examination of the treatment options the physician considered, including knowledge of whether the patient was eligible for a clinical trial and if one was offered; (3) whether there was a discussion of more than one treatment option with the patient; (4) a visual analog scale to rate the strength of the physicians’ treatment recommendations; (5) review of the physician’s rationale for his or her treatment recommendations; (6) items considered (treatment benefits and side effects); (7) decision-making style; (8) final treatment choice; (9) his or her impression of the patient’s reaction to the various treatment recommendations (using a five-point rating scale); (10) the physician’s impression of the character and/or emotional status of the patient and ability to understand the information being conveyed (using a five-point rating scale); and (11) the physician’s previous treatment relationship with the patient. The second part of the interview focused on the physician’s attitudes, knowledge, and general practices concerning clinical trials. Physicians were queried about their attitudes toward trial participation, including such issues as perceived barriers to patient participation (eg, patient resistance to participating) and their own participation (eg, too much paperwork). We also collected demographic information about the physicians, including age, sex, ethnicity, hospital and cooperative trial group affiliations, training, and practice settings.
Statistical Analysis
A total of 245 patients were enrolled onto this study, and complete data were obtained for 244. One hundred seventy patients (69.4%) were observed for follow-up by a medical oncologist after surgical management. A total of 14.7% of patients received no adjuvant therapy of any kind, 44.5% received chemotherapy, and 40.8% received tamoxifen only. Although all patients could have been participants in a phase III clinical trial, only 93 patients (38.1%) were offered a trial, with 49 (52.7%) of them agreeing to participate. Forty-five patients (48.4% of those who were offered a trial and 18.4% of all patients included in this study) eventually received their adjuvant therapy as part of a trial. Table 1 summarizes physicians’ sociodemographic and professional characteristics. The average age of surgeons was 47.8 years; 84.1% were white, 1.9% were black; and 94.4% were male. Nearly half (40.2%) of the surgeons practiced alone, 49.5% practiced in a private group, and 9.3% practiced in a university setting. The majority of surgeons (54.2%) received specialty training at a university. On average, surgeons were affiliated with more than two hospitals and less than one cooperative group. They occasionally referred patients to clinical trial. Oncologists’ average age was 45.2 years. All oncologists were white and 85% of them were male; 12.5% practiced in a university setting, while the majority (75%) practiced in a private group. Eighty percent of oncologists received their specialty training at a university. On average, oncologists were affiliated with a little over three hospitals and more than two cooperative groups and reported that they frequently referred patients to a clinical trial.
Table 2 summarizes the patient sociodemographic and disease profile. Among surgical patients, the average age was 56.2 years. Most (91.4%) were white, 4.5% were black, 66.8% were married, and 59.8% had children. About half (48.8%) lived in urban or suburban areas and 39.8% were employed. A total of 82.4% of patients had health insurance of some type; Medicare or Medicaid covered 31.5% of patients. Patients received primary surgical management in the form of a modified radical mastectomy (52.9%) or lumpectomy followed by radiation (44.3%). Most (82.8%) had a unifocal tumor, and about one third (31.6%) had a family history of breast cancer.
The effects of physicians’ demographic and professional characteristics on referral to trials are presented in Table 3. Surgeons who practiced in a university setting or who received support from a cooperative group were significantly more likely to refer their patients to a clinical trial than those who practiced alone or in private groups (P  .01) or did not receive support from a cooperative group (P .01). Surgeons who were white, had a smaller percentage of all of their breast patients who were breast cancer patients, or had fewer affiliations with cooperative groups were also significantly more likely to make a trial referral than those who were non-white (P .05), had a larger percentage of breast cancer patients (P < .05), or who had more affiliations with cooperative groups (P .05). Oncologists who practiced in a university setting or who treated fewer breast cancer patients per year were significantly more likely to make a referral to a trial than oncologists who practiced alone or in private groups (P .05). Those who had fewer hospital affiliations or a larger percentage of breast cancer patients among their patients were marginally more likely to refer their patients to a clinical trial than oncologists who had more hospital affiliations (P .10) or a smaller percentage of breast cancer patients (P .10).
The effects of patient characteristics on physicians’ trial referrals were examined. The only factor that predicted a significant chance of referral to a clinical trial by a surgeon was the patient’s age. The older the patient, the less likely she was to be referred (P .05). For the patients who were seen by an oncologist, those who had a family history of breast cancer or Medicare or Medicaid coverage had a marginally lower chance of referral to a trial by their oncologist than those without a family history of breast cancer (P .10) or without Medicare and Medicaid coverage (P .10).
Table 4 presents the effects of the patient-physician interaction as they considered adjuvant therapy together. For all patients, the chance of a trial referral was substantially greater when physicians, especially surgeons, were more involved in making a decision about patient’s adjuvant therapy (P
We also examined physician’s knowledge, referral patterns, and decision-making processes when making or not making trial referrals. As Table 5 indicates, surgeons and oncologists were significantly more likely to refer their patients to a clinical trial when they knew which trial the patient was eligible for (P .01) or when patients were more involved in decision making (P .01). Moreover, surgeons were significantly more likely to make a trial referral if they reported referring patients to trials regularly (P .01), but were marginally less likely to make the referral if they had to refer patients indirectly through an oncologist (P .10). Oncologists were more likely to make a trial referral if they were more involved in making decisions about the trial (P .05) or if the patient had also been considered by their surgeon for a trial (P .10).
A number of attitudinal and experiential variables were significantly associated with physicians’ referrals of patients to clinical trials (Table 6). Surgeons and oncologists were more likely to refer patients to a trial if they felt comfortable explaining trials to patients (P .01 for surgeons and P .10 for oncologists), but were significantly less likely to make a trial referral if they perceived a lack of patient interest (P .01 for surgeons and P .05 for oncologists). The chance of a trial referral was significantly greater among surgeons who felt that local trials keep patients in the community (P .10), that patients are observed closely when they participate in a trial (P .05), and that treatment regimens should not stray from protocols (P .05). Surgeons who did not like the specific trials available were significantly less likely to refer their patients to a trial (P .05).
Oncologists, on the other hand, were less likely to make a trial referral if they considered the paperwork associated with a clinical trial to be too time-consuming (P .01) or believed that trial entry requirements were too stringent and resulted in too many of their patients being ineligible (P .05). Nevertheless, oncologists who considered the trials extra work (P .01) or expensive to participate in (P .05) or who found explaining medical uncertainty difficult (P .01) were more likely to refer patients to trials. This finding could indicate that physicians who were active trial participants were more aware of the burdens and benefits of trial participation than those who were not involved in clinical trials. Indeed, most of these oncologists had a history of referring patients into trials and indicated that these negative aspects of trial work had little impact on their enthusiasm about clinical trials.
The significant and marginally significant predictors examined above were selected and entered into two final models to predict surgeons’ and oncologists’ decisions to refer patients to clinical trials. The two models are summarized in Table 7. Surgeons who reported that they referred patients to clinical trials on a regular basis were more likely to have referred the patient to a trial by a ratio of 2.5:1 (P
Six prominent predictors of oncologists’ referrals to trials were identified and together they explained 73% of the variance (Table 7). According to our current model, oncologists who practiced in a university setting were more likely to make a trial referral than those who practiced alone or in private groups in the community at a ratio of 56.2:1 (P .05). Factors related to the interactional features of the decision-making process assumed great importance. The chance of a trial referral was 24.2 times higher among patients who were more involved with making decisions about clinical trials (P .01) and 6.9 times higher among patients whose oncologists were more involved with making decisions about clinical trials (P .01). The patients whose surgeons were more involved with decisions about adjuvant therapy had a significantly greater chance (2.5) of being referred by their oncologists to a trial (P .01). Knowing which trial the patient was specifically eligible for also increased the chance of a trial referral among oncologists by 5.3 times (P .05). Nevertheless, oncologists who considered the paperwork too time-consuming were significantly less likely to refer patients to a trial (P .05).
Adequate participation in randomized controlled trials is one of the most important requirements for the development of new and more effective treatments for cancer. This study has demonstrated, once again, that accrual to clinical trials is unacceptably low. This study sought to understand how trial-eligible breast cancer patients are or are not referred to phase III clinical trials by following the treatment trajectory of a series of newly diagnosed breast cancer patients. Patient treatment histories were traced from the surgeons who provided primary surgical care through consultation with a medical oncologist. Despite all patients being trial-eligible, only 38.1% were actually asked to participate in a trial and only 18.4% eventually received their adjuvant therapy as part of a trial. A number of factors played a role in the low accrual rate. These factors were related to both the patients’ and the physicians’ individual characteristics. However, only two patient characteristics were found to play any significant role in trial referral, whereas a number of physician characteristics were important. This study found, as have others,16-18 that older patients were less likely to be referred to a trial, as were patients whose prognosis was deemed to be poor. As anticipated, physicians whose practices were based in university settings and had formal support from a cooperative group were more likely to refer to trials. Furthermore, attitudes toward trials varied somewhat between the two types of specialists. Surgeons who felt more comfortable explaining trials and those who were white were more apt to refer patients to trials as were those who stated that treatment should not stray far from formal protocols. Oncologists were less likely to refer patients to a trial if they perceived the paperwork to be onerous and trial entry requirements too stringent. Nevertheless, certain specific burdens, such as additional expenses, were associated with a greater likelihood to participate. This may be a function of greater knowledge of what trial participation entails for the oncologist and a strong belief that trials are, nonetheless, a priority. Greater knowledge and interest in trials on the part of surgeons had a surprisingly strong effect on trial referral, especially specific knowledge about which trials were open for patient recruitment. In addition, surgeons’ participation in recommending adjuvant therapy to patients also resulted in more trial referrals, although when surgeons treated their patients with tamoxifen, the patients were less likely to be referred to trial. Although few patient medical or demographic characteristics played a strong role in trial referral, patients’ interactions with their physicians did. The chance of a referral to a trial was 24 times higher among patients who were reported to be actively involved in making decisions about adjuvant therapy treatment and participation in a clinical trial. Furthermore, patients who delayed making an adjuvant therapy decision were also more likely to participate in a trial. Recruitment to trials suffers because of a variety of factors, most of which have nothing to do with the disease characteristics of patients. Physicians still need to overcome attitudinal and practical barriers to trial participation, such as a belief that patients are not interested in trials and a lack of comfort explaining trials. More support to physicians is clearly needed in terms of completion of paperwork and making trial entry easier. This study indicates that surgeons may play a more pivotal role in the recruitment of patients to adjuvant therapy trials than previously thought. Lastly, the role of patients in the process should not be overlooked. Garnering patient enthusiasm for trial participation and involving them in the choice of adjuvant therapy may be a key component to increasing trial enrollment.
Supported by grant no. CA49641 from the National Cancer Institute, Bethesda, MD.
1. Morris L: FDA Modernization Act: Implications for oncology. Oncology 12:139-141, 1998[Medline] 2. Lippman ME, Chabner BA: Overview of proceedings of the NIH consensus development conference on adjuvant chemotherapy and endocrine therapy for breast cancer. NCI Monogr 1, 1986 (editorial)
3.
Antman K, Amato D, Wood W, et al: Selection bias in clinical trials. J Clin Oncol 3:1142-1147, 1985 4. Green SB: Randomized clinical trials: Design and analysis. Semin Oncol 8:417-423, 1981[Medline] 5. Robinson D, Woemer MG, Pollak S, et al: Subject selection biases in clinical trials: Data from a multicenter schizophrenia treatment study. J Clin Psychopharmacol 2:170-176, 1992 6. Lennox EL, Stiller CA, Morris PH: Nephroblastoma: Treatment during 1970-3 and the effect on survival of inclusion in the first MRC trial. J 2:567-569, 1979 7. Arnett B, Rikli R: Effects of method of subject selection (volunteer vs random) and treatment variable on motor performance. Res Q Exerc Sport 52:433-440, 1981[Medline] 8. Remington DR, Taylor HL, Buskirk ER: Methods for assessing volunteer bias and its application to a cardiovascular disease prevention programme involving physical activity. Epidemiol Community Health 32:250-255, 1978 9. Morrow GR, Hickok JT, Burish TG: Behavioral aspects of clinical trials: An integrated framework from behavior theory. Cancer Suppl 74:2676-2682, 1994 10. Taylor KM, Margolese RG, Soskolne CL: Physicians’ reasons for not entering eligible patients in a randomized clinical trial of surgery for breast cancer. N Engl J Med 310:1363-1367, 1984[Abstract] 11. Siminoff LA, Fetting JH, Abeloff MD: Doctor-patient communication about breast cancer adjuvant therapy. J Clin Oncol 7:1192-1200, 1989[Abstract] 12. Lee JY, Breaux SR: Accrual of radiotherapy patients to clinical trials. Cancer 52:1014-1016, 1983[Medline] 13. McCusker J, Wax A, Bennett JM: Cancer patient accessions into clinical trials. Am J Clin Oncol 5:227-236, 1982[Medline] 14. Cicorel AV: Method & Measurement in Sociology. New York, NY,Free Press, 1964 15. Shaffer G, Saunders V, Owens W: Additional evidence for the accuracy of biographical data: Long-term retest and observer ratings. Personnel Psychol 39:791-809, 1986 16. Simon MS, Brown DR, Du W, et al: Accrual to breast cancer clinical trials at a university-affiliated hospital in metropolitan Detroit. Am J Clin Oncol 22:42-46, 1999[Medline]
17.
Benson AB, Pregler JP, Bean JA, et al: Oncologists’ reluctance to accrue patients onto clinical trials: An Illinois Cancer Center study. J Clin Oncol 9:2067-2075, 1991 18. Silliman RA, Guadagnoli E, Weitberg AB, et al: Age as a predictor of diagnostic and initial treatment intensity in newly diagnosed breast cancer patients. J Gerontol 44:646-750, 1989 Submitted June 28, 1999; accepted November 17, 1999.
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2000 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
TOP
ABSTRACT
INTRODUCTION
