Journal of Cancer Education
ISSN: 0885-8195 (Print) 1543-0154 (Online) Journal homepage: https://www.tandfonline.com/loi/hjce20
Improving accrual into cancer clinical trials John F. Foley MD & Charles G. Moertel MD To cite this article: John F. Foley MD & Charles G. Moertel MD (1991) Improving accrual into cancer clinical trials, Journal of Cancer Education, 6:3, 165-173 To link to this article: https://doi.org/10.1080/08858199109528114
Published online: 01 Oct 2009.
Submit your article to this journal
Article views: 4
View related articles
Citing articles: 4 View citing articles
Full Terms & Conditions of access and use can be found at https://www.tandfonline.com/action/journalInformation?journalCode=hjce20
J. Cancer Education. Vol. 6, No. 3. pp. 165-173, 1991 Primed in the U.S.A. Pergamon Press pic
0885-8195/91 $3.00 + .00 © 1991 American Association for Cancer Education
IMPROVING ACCRUAL INTO CANCER CLINICAL TRIALS JOHN F . FOLEY, MD* and CHARLES G. MOERTEL,
MD†
Abstract — It is thought by some that there are inadequate patient entries on high priority national cancer treatment protocols, resulting in intolerable delays in addressing crucial research issues. The North Central Cancer Treatment Group‡ in Rochester, Minnesota generated a questionnaire with possible impediments to patient treatment entry as well as suggested measures for approving accrual. Of 334 questionnaires, 63% were completed and returned by medical oncologists, radiation therapists, surgeons and nurse/data managers. The most frequently listed impediments included patients' advanced age or fragility, inadequate health insurance, and excess travel distance. Protocol concerns included too many tests that were too complex and too costly. Practice concerns included physicians too busy and inadequate help. Suggestions toward improving patient accrual included subsidizing drugs, subsidizing nonessential tests for patient care, and increasing funding for support personnel. Popular suggestions for education efforts included teaching the importance of clinical trials, emphasizing personal benefit to the patient and recognition of physicians and institutions participating. It was felt that protocols should be simplified, be more consistent with practice, and be more flexible. These considerations are relevant not only in the interests of more effective clinical research but also in the interests of appropriate education of residents and oncology fellows involved in clinical trials.
INTRODUCTION
reasons, the conduct of clinical trials is one of the most important responsibilities of the NaThe clinical trial is the ultimate and only mean- tional Cancer Program. To be of greatest service ingful method of establishing the value of new to the public, it is essential that cancer clinical cancer treatment approaches. For new drug ther- trials be conducted with the highest standards of apy it is required that value be proven by clin- quality, involve sufficient patients to achieve ical trial, usually randomized, before such therapy statistical validity, and be completed in a timely can be offered to the public at large. For these fashion. To expeditiously achieve these objectives, a large national clinical trial organization has been created that represents one of the largThis work was supported by grants CA 25224, CA 37404, est single investments of the public funds enCA 47404, CA 35269, CA 35101, CA 37417, CA 35448, trusted to the National Cancer Institute. In spite CA 35272, CA 35113, CA 35103 from the National Canof this, however, concerns have been raised recer Institute, National Institutes of Health. * F r o m the University of Nebraska Medical Center, Omaha, garding what is perceived as inadequate patient Nebraska. entries on high priority national protocols, re† From the Mayo Clinic, Rochester, Minnesota. ‡ sulting in intolerable delays in addressing cruFrom the North Central Cancer Treatment Group and the Mayo Clinic: The Billings Clinic, D. Twito; Quain and cial research issues. Dr. Michael Friedman, Chief Ramstad Clinic, D. Pfeifle; Oncology Associates of Cedar of the Clinical Investigations Branch of the DiRapids, M. Wiesenfeld; The Duluth Clinic CCOP, J. Krook; vision of Cancer Treatment, NCI, has stated, The Fargo Clinic CCOP, G. McCormack; Geisinger Clinic and Medical Center CCOP, R. Goldberg; Ochsner Medical "Examining national trends reveals that only Foundation CCOP, C. Kardinal; Illinois Oncology Research about 1% - 1.5% of available breast cancer, Association CCOP, S. Cullinan; Rapid City Regional On0.5% of rectal cancer, and 1% of colon cancer cology Group, L. Ebbert; Saskatchewan Cancer Foundation, M. Poon; Sioux Falls Community Cancer Consortium CCOP, patients are actually recruited into such studL. Tschetter, St. Cloud Clinic of Internal Medicine, Ltd., ies." 1 He asks, "Why is an $8,000,000 CCOP H. Windschitl; Grand Forks Clinic Ltd., J. Laurie; Nebraska (and a $55,000,000 Cooperative Group) proOncology Group - Creighton University/University of Nebraska Medical Center and Associates, J. Foley, J. Mailliard. gram not capable of generating larger numbers Reprint requests to: John F. Foley, M.D., Section of of patients with specific types of malignancies Oncology/Hematology, University of Nebraska Medical Cento resolve important scientific dilemmas conter, 600 South 42nd Street, Omaha, NE 68198-3330.
165
166
J.F. FOLEY and C G .
fronting oncologists?" More recently, former NCI director, Dr. Vincent DeVita was reported to have stated that " . . .the cooperative group system is broken" and that the slow patient accrual of cooperative groups is a "national disgrace." 2 ' 3 Whereas these statements may be a bit pejorative, they are certainly attention-grabbing. Even if the practicing oncologist is not concerned with these national political issues, he must be concerned that such a small proportion of cancer patients with poor prognosis or fatal disease is given the hope that can be provided through entry into clinical research protocols. The evidence is convincing that although an adequate resource of protocol-eligible patients is available through currently funded clinical trial resources, only a small proportion of this reservoir is actually entered on protocols. In a study of the Eastern Cooperative Oncology Group (ECOG), it was found that among 3,534 cancer patients seen by their members and affiliates during a survey period, only 22% were entered on protocols.4 Among those patients eligible for available protocols, only 53% were actually on a protocol. In a similar survey of NCI funded Community Clinical Oncology Programs (CCOPs) only 34% of 9,508 patients clinically eligible for available protocols were entered on protocol.5 In both the ECOG and CCOP surveys, the major reason for not entering an elibible patient on protocol was a decision made by the physician. It is also probable that when the decision was made by patients not to enter protocol, this decision was in large measure influenced by the attitude of physicians. Dr. Michael Friedman has concluded, "Speculation as to what financial, psychological or other factors influenced the physicians' and patients' decisions not to enter a study is sterile. An insightful analysis of reasons for non-participation would be methologically difficult, but is manifestly necessary. ' ' 1 The current study attempts to meet that need. Perhaps of more importance, we attempted to elicit constructive recommendations for improvement in patient protocol accrual.
MOERTEL
vestigators of each of the 15 North Central Cancer Treatment Group (NCCTG) member institutions (all medical oncologists), the chairman of each modality committee (medical oncology, radiation oncology, surgical oncology and pathology), the NCCTG chairman and cochairman (both medical oncologists), and Dr. Robert Wittes, Associate Director, Cancer Therapy Evaluation Program, Division of Cancer Treatment, NCI. The objective of this meeting was to prepare a compilation of possible reasons why patients were not entered on protocol and of suggested measures for improving protocol entry. Based on the contributions from participants in this initial meeting, a questionnaire was composed listing 17 possible impediments to patient entry. Respondents were asked to score these items on a 0 to 4 basis with 0 representing no impediment to patient entry and 4 a major and very frequently encountered impediment. The questionnaire also listed 18 suggested measures to improving patient accrual, again with a 0 to 4 scoring system with 0 indicating the opinion that the measure would not improve accrual and 4 indicating that the measure was highly likely to improve accrual. The questionnaires were then distributed to each individual listed on the NCCTG membership roster, and results were tabulated for all members who were directly or indirectly responsible for patient protocol entry. Respondents were classified into four groups: medical oncologists and/or hematologists, general surgeons and surgical subspecialists, radiation oncologists, and oncology nurses and/or data managers. Respondents were given the option of signing questionnaires or submitting them anonymously. Results were analyzed by the mean score accorded each item and also by the percentage of respondents who scored each item as one of major importance, i.e., score 3 or 4. Since there were no substantive differences in results by either means of analysis, we have elected to report the percentages of respondents who scored each item as 3 or 4. RESULTS
METHOD Our first step in addressing the purposes stated above was a meeting involving the principal in-
Of 334 questionnaires sent to appropriate group members, 209 (63%) were completed and returned. According to specialty, question-
Improving accrual into cancer clinical trials
naires were returned by 75% of 91 oncologists/ hématologiste, by 75% of 80 nurse/data managers, 70% of 40 radiation oncologists, and 43% of 123 surgeons. It would seem reasonable to assume that the responses are representative of the first three disciplines and probably were contributed by almost all members in these disciplines who are active NCCTG participants rather than just pro forma listings on our roster. It is clear, however, that the respondents composed only a minority of listed surgeons. This may be explained, at least in part, by the fact that a number of uncompleted questionnaires were returned by surgeons with accompanying notes stating that they did not feel they could properly respond since they relied on their medical oncology colleagues for handling of protocol entries. In the main, returned questionnaires were completely filled out. Only 4.8% of a total 7,315 potential answers were left blank, and most of these were on forms returned by surgeons. It would, therefore, appear that the results of this survey are reasonably representative of opinions of those who are more directly involved in the patient protocol entry process (medical oncologists/hematologists/radiation oncologists and nurse/data managers), and less representative of those who are more peripherally involved in this process (surgeons). Figure 1 graphically presents the proportion of respondents who considered the factors described as major impediments to patient protocol entry. These factors have been grouped according to concerns specifically related to the nature of the research protocol, specifically related to the patient, and specifically related to the physician and his practice. It is of interest that although 78% of the respondents considered one or more of the factors listed to be major detriments to patient protocol entry, there was little uniformity as to which of these individual factors were dominant concerns. This was true for the respondents considered as a whole or according to specialty interest. Indeed, there was not a single factor that was considered a major impediment by even a simple majority of all respondents. The most frequently listed concerns were those directly related tö the nature of specific patients
167
(Figure 1A). Leading these was advanced age or perceived fragility of the patient, considered a major impediment to protocol entry by 42% of all respondents and by 53% of medical oncologists/hematologists. Following this were the very practical problems of inadequate health care coverage and geographic location of the patient's residence, which precluded easy and affordable return visits for follow-up treatments or observation. Although refusal of the patient to be an experimental subject or to be randomized were considered important factors by some, it is of interest that three of four respondents did not regard these factors as major impediments to protocol entry. The most frequently mentioned major concerns regarding specific protocols pertained to practical issues, ie, overly complex protocol procedures, excessive testing requirements, excessive cost, and an excess personal burden placed on the patient (Figure IB). It is thought provoking that only 7% of medical oncologists/hematologists regarded excess cost as an entry impediment, whereas 50% of nurse/data managers considered this a major factor. Conversely, 29% of medical oncologists/hematologists considered physician time requirements a major impediment, but only 10% of nurse/data managers agreed. Another interesting contrast, and possibly a lesson for our group, is that inadequate support personnel and problems with interdisciplinary coordination were important considerations for surgeons (32% and 27%) but less troublesome to medical oncologists/hematologists (18% and 10%) (Figure 1C). Perhaps correcting popular misconceptions, the least frequent concerns were excessive toxicity of protocol treatment, scientifically uninteresting protocols, or that the referring physician would object to research protocol entry. There was considerably greater unanimity of respondents regarding recommendations for improving patient protocol entry (Figure 2). We have classified these into measures for improvement of financial support of research, public relations and education efforts, changes in protocol design and conduct, and changes that the research units could make in their practice patterns. The most strongly recommended measures
168
J.F. FOIEY and C G . MOERTEL
Too »any testa
Too old op too fragile
sax o complex
| Xnadoquats health car« coverage 331%
io auch pereonal burden on patient Excessive travel distance from patlent'a residence
25X o costly
quires too aucn physician time 21X Inefusal to be experlnental eubiect «uch toxlcity
J24X
|Unllkely to coaply with protocol requirements Scientifically uninteresting jlBX
50
100
50
loo
(b)
(a) Physician too busy
I Inadequate support personnel J22X |Problem »lth Interdisciplinary coordination 116%
concern that referring phyalclan Mill rebel
50
100
(C)
Figure 1. Major impediments to patient protocol entry: (A) Patient concerns; (B) Protocol concerns; (C) Practice concerns (black, score 4; shaded, score 3).
were those that would require efforts and expenditures on a national level. These involved relieving patients or their third-party carriers of the protocol-related financial burdens of expensive marketed drugs or tests done purely for research purposes (Figure 2A). They also involved efforts through national media to improve the public image of clinical trials in general, of the
benefits of clinical trial participation for individual patients, and of the physician and/or institutions participating in clinical trials (Figure 2B). There were no divisions among the various specialty interest in their strong support for these measures. Available on-scene educational materials for patients and families regarding clinical trials as well as regarding specific protocols
169
Improving accrual into cancer clinical trials
National/Local media: - Importance of clinical trials
V//////A'™ National/Local Redis: - Personal benefit to patients in clinical trials
V//////A* National/Local media: - Recognition of physicians or institutions participating in clinical trials
Patient education video tapes regarding clinical trials Subsidize research testing not essential to patient care tlont education brochures regarding V//////////A'"
ent education brochures regarding
[Incraaae funding for physician participation tlaa
individual protocols
clinical trials
tlent education vldao tapes regarding Individual protocols
ÀA1X
3
50
100 0
100
50
(b)
(a) Simplify protocol procedures and requirements
Y/////////A™ Hake protocol more consistent with clinical practice
Score-3
simollfy forais Prearrange availability of protocol required .•ateríais before patient Is seen
Y////////A™ Increase flexibility In protocol conduct
1Schedule potential patients for protocol entry [at earlier and lesa hectic tines of day
53X Y/////////A* Reduce protocol required testing ¡SOX
50
100
(c)
0
50
100
(d)
Figure 2. Recommendations for improving patient protocol accrual: (A) National financial support; (B) National/local public and patient educational efforts; (C) Changes in protocol and clinical research procedures; (D) Changes in practice of physician investigators (black, score 4; shaded, score 3).
were endorsed by most, and particularly by responding surgeons. There was also uniform enthusiasm for modifying protocols to make them more expeditious and less cumbersome in con-
duct (Figure 2C). With regard to practice patterns, most physician respondents did not feel that scheduling of new patients influenced protocol entry, but 52% of nurse/data managers felt
170
J.F. FOIÍY and C G .
that scheduling at an earlier and less hectic time of day could be advantageous (Figure 2D). Sixtyfour percent-of nurse/data managers and 55% of all respondents endorsed the importance of prearranging availability of necessary materials for protocol entry (eg, operative and pathology reports) when the patient was seen in initial consultation. DISCUSSION Certainly, any opinion survey produces data that are, by definition, purely subjective and therefore to be interpreted with appropriate caution. However, the joint decision making process by physician and eligible patient regarding enrollment in a clinical research protocol is also highly subjective in nature. With all this, a body opinion from those devoted to the objectives of the National Cancer Program probably represents the most rational basis for decision regarding constructive change. The results of this study must be interpreted in light of the nature of the participants, the demography of the regions they represent, and the atmosphere of the time at which the study was conducted. At the time of this survey the NCCTG had 15 member institutions. In cooperation with the Mayo Clinic, it is currently accruing patients to cancer treatment protocols at the rate of approximately 1,500 per year, with a near equal number of patients entered on cancer control protocols. Fourteen members are primarily engaged in the private practice of medicine and only one represents a university setting. Eleven members are located in the north central region of the United States and only one member is located in a large urban area. These characteristics are probably atypical of most clinical cancer trial groups, but in some respects they may be more typical of overall cancer care delivery in this country. The relatively large protocol case entry in relationship to membership may also be atypical. In the main, our membership institutions represent tertiary care facilities in their respective regions; and, for most, the nature of oncology practice is perhaps not as highly competitive as it is in the coastal or sun-belt regions of this country. The long established tradition of interdisciplinary practice among our members may
MOERTEL
also not be nationally representative. The dominance of medical oncologists/hematologists among our respondents may perhaps be criticized, although certainly for most national clinical trial groups this specialty area takes principal responsibility for patient protocol entry. We included oncology nurse/data managers among our respondents without apology, because if they are not primarily instrumental in patient protocol entry, they are at least active collaborators and certainly can be regarded as reliable, if not dispassionate, observers of this phenomenon. Our study was conducted at a time when the major funding source for clinical cancer research, the National Cancer Institute, had expressed serious concerns regarding the relatively small en-. try of patients on nationally approved cancer treatment protocols when viewed in the light of the large available patient reservoir. There had been at least the implication, and perhaps a commitment, of additional funding for measures to effectively improve case accrual. With or without the funding carrot, our group supported the NCI initiative and this study was undertaken as a first step in a positive and constructive initiative to improve clinical trial patient accrual. Perhaps not surprisingly, the results of this study place much of the burden on the shoulders of the National Cancer Institute. Any committed participants in cancer clinical trials, particularly if they derive from the private practice sector, must hold the conviction that the most ideal medical care offered to the poor prognosis cancer patient today is offered in the setting of a well-designed clinical trial. A loud, strong, and clear recommendation of this study is that the American public be educated through the public media regarding the overall importance and ethical nature of clinical trials and the benefit the cancer patient will receive by treatment in such a setting in the hands of physicians who care enough about the cancer patient, both today and tomorrow, to make this extra effort. The public has many misconceptions about treatment on clinical trial protocols, and these may be reinforced by physicians not participating in clinical trials or by promoters of "alternative" therapies. Certainly this is true with regard to randomized trials, particularly if those trials involve no treatment or placebo control arms. It
Improving accrual into cancer clinical trials
is vitally important that the public be assured that the highest quality cancer care is provided by entry into a clinical trial, which will offer not only the best in available treatment but also the hope of treatment improvement both for themselves and for other patients in the future. They should also be assured that randomization is not evidence of scientific depersonalization, but rather is employed only when there is no acceptable evidence of superiority of any of the management programs to which they may be assigned. They should also understand that they will be fully informed if there is a possibility that they will be assigned to no added treatment or to placebo treatment and that this will only be done if there is no acceptable evidence that additional treatment will be of benefit for their .problem, e.g, the surgical adjuvant setting. Public recognition for the efforts and accomplishments of the physician is an impotant inducement to his participation in clinical trials. For the physician in private community practice, this will be the only reward he will receive, since the extra time he devotes to clinical trial participation usually involves a financial sacrifice for him or his institution. An important secondary benefit of such recognition through public media is that patients may be attracted to preferentially seek their cancer care from physicians and institutions who they know have the high motivation and recognized competence necessary for participation in nationally approved clinical cancer research programs. By thus making the community physician who participates in clinical trials more competitive, the reservoir or patients available for clinical trials may in turn be increased. Utilizing national and regional media, the Office of Cancer Communications of the National Cancer Institute has already initiated an aggressive and innovative program for public education in cancer clinical trials. We very much hope that the results of this study will encourage them to stay on this course. The results of this survey have also emphatically indicated that a major impediment to clinical trial accrual is the frequently complex and cumbersome nature of protocols and procedures. Begg6 documented that in institutional trials in Stage II breast cancer, there were 23 exclusion criteria per study. The Community Clini-
171
cal Oncology Program logs demonstrated that among patients with a relevant protocol, 44% were ineligible. Goodwin et al7 reported that although 31% of adult cancer patients were over 70 years old, only 7% of these patients were enrolled in Southwest Oncology Group protocols in New Mexico. Certainly some of the requirements of protocols are essential to their scientific purposes. On the other hand, it must be admitted that data is frequently demanded that have no direct relation to protocol objectives and which will probably never be analyzed for any purpose. The practicing physician may rightfully regard the gathering of such data on extraordinarily detailed study forms as inconsequential busy work and a waste of his time and the time of his staff. Superfluous testing is sometimes more the rule than the exception in clinical research protocols, eg, weekly blood counts for a patient who has already been treated for months on the same regimen without experiencing hazardous hématologie toxicity, chest x-rays and chemistry profiles done every month without clinical indication, meticulous restaging procedures without evidence that these contribute to clinical outcome. Frequency of follow-up is often far in excess of what conscientious clinical practice would dictate. Is it really necessary, for example, to bring an asymptomatic surgical adjuvant patient back at three-month intervals in order to fine-tune the time of recurrence on a protocol for which survival is the endpoint of primary interest? It would seem appropriate for each clinical trial cooperative group to carefully scrutinize protocols and periodically review procedures to ensure that their demands are rational, represent prudent utilization of the time of the physician and data manager, and are properly conservative of health care costs. There is little doubt that the closer our protocol procedures conform to high-quality clinical practice, the more conducive they will be to protocol accrual and to physician and patient compliance. Note in Figure IB, Protocol Concerns, 24% felt that the protocols were too costly. Note in Figure 2A, that 75% felt that there should be subsidization of expensive marketed drugs and 71% felt there should be subsidization for research testing not essential to patient care. We believe this apparent contradiction relates to the fact
172
J.F. FOLEY and C.G.
MOERTEL
that at that time most of the patients have been able to participate in the protocols and get reimbursement. However, the third-party carriers and Medicare are becoming much more difficult to deal with when it relates to expensive marketed drugs and are less willing to fund some of these, particularly when it is stated that they are in an experimental program. In addition, most everyone in protocol studies will agree that more studies and imaging procedures are done in these patients than in patients who are off study. This again reflects the concerns regarding third-party reimbursement. Because of this problem, we have developed a subcommittee that continually looks over testing requirements in the various protocols and recommends that we drop those that seem to be not necessary.
make treatment of a patient on protocol more economical than treatment with a comparable regimen off protocol. The requirement of accounting for research testing not essential for patient care would encourage greater discrimination by investigators before incorporating such testing as a protocol requirement. Both measures would undoubtedly make support of treatment on research protocols more attractive to the health insurance industry. These advantages would, of course, have to be balanced against the total cost to the national clinical cancer trials program. Our survey does make a strong case that if additional investment of public funds is contemplated to increase accrual on national protocols, defraying research-related costs to the patient could be a very effective measure.
The loudest and strongest recommendations expressed by this survey are those regarding the costs that clinical trial participation impose on the patient or the patient's third-party carrier. Additional funding is urged to defer charges for expensive marketed drugs and for testing required for research purposes but not essential to best patient care. Such patient care funding is almost nonexistent in today's clinical trial grants and contracts. The argument has been made that if any such funding is allowed, even for purely research related patient charges, third parties may then be encouraged to refuse payment for all care charges accrued by patients on research protocols, even those required for best patient care. Certainly there is mounting evidence of reluctance of insurance carriers to reimburse charges for cancer research treatment. This has been motivated, at least in part, by the massive expenses involved in treatment with some biologics or demanded by certain commercial research treatment enterprises. On the other hand, a very persuasive argument can be made that third-party carriers have no desire to deny their subscribers afflicted with cancer the hope that accompanies treatment on a research protocol, provided the carriers can be assured that the cost is comparable to that charged to them when the patient is treated in a nonresearch setting. It is clear that government buying power would make drugs available at considerably lower cost than could be obtained in the open market, and defraying the expense of these agents would
There is an apparent contradiction in the data in Figure 1C that only 22% felt there was inadequate support personnel, whereas in Figure 2A, 61% of the respondents felt that there should be increased funding for support personnel. The addition of cancer control to our research activities is beginning to markedly increase the data managers' workload, with no commensurate increase in funding thus far. Thus, the data from Figure 2A indicate our ongoing needs. There are two obvious alternatives for enhancing accrual on important national clinical cancer research trials. The first is to recruit new physicians not currently active in this effort. Whereas this might add numbers, genuine concern must be raised regarding the possible adverse effects on quality of research data. New physicians would have no track record of quality performance, and they would be unaccustomed to research discipline. The fact that they had not previously affiliated with the clinical trials program, in the face of ample opportunity, would raise questions regarding their motivation to produce high quality data over a sustained period of time. The very large trials for which these added numbers would be needed are large because they are looking for small differences. Even a modest rate of poor quality entries on such protocols (ie, patients ineligible, cancelled, lost to follow-up, or treated in gross violation of protocol) could completely obscure a positive result that would have been evident if the protocol had been conducted with high-quality stan-
Improving accrual into cancer clinical trials
dards. Considering the importance of these trials to the cancer patient and the large public investment involved, this would seem to be an intolerable risk. Efforts to improve protocol accrual within existent resources would appear far more attractive. Well-trained and experienced quality control personnel, methods, and logistics would already be in place. Quality of performance would be established by record and approved by peers. Both the ECOG and CCOP studies had shown that the reservoir of eligible patients exists within already funded institutions to increase protocol accrual by two- or three-fold. If our study has validity, it indicates the strong likelihood that protocol accrual can be substantially increased by providing public education on both the national and local level regarding the importance and benefits of clinical trial participation, by public recognition of the efforts and accomplishments of the physicians and institutions participating in clinical trials, by reducing nonessential complexities of clinical trial protocols, and by easing the financial burden of clinical
173
trial participation for the patient or his thirdparty carrier. Perhaps wisdom will dictate that these measures should be given serious consideration before risking the possible major damage that might be done to important national protocols by recruitment of untried participants who have not previously evidenced a devotion to cancer clinical trial programs. REFERENCES 1. Friedman MA: Patient accrual to clinical trials. Cancer Treatment Rep 71:557-558, 1987. 2. Cancer Letter 13, (35), 1987. 3. Cancer Utter 14, (10), 1988. 4. Begg CB, Zelen M, Carbone PP et al: Cooperative groups and community hospitals. Cancer 52:1760-1767, 1983. 5. Hunter CP, Frelick RW, Feldman AR et al: Selection factors in clinical trials: Results from the Community Clinical Oncology Program Physician's Patient Log. Cancer Treatment Rep 71:559-566, 1987. 6. Begg CB: Selection of patients for clinical trials. Semin Oncol 15:434-440, 1988. 7. Goodwin JS, Hunt WC, Humble CG, Key CR, Samet JM: Cancer treatment protocols. Who gets chosen? Arch Intern Med 148:2258-2260, 1988.
ISSN: 0885-8195 (Print) 1543-0154 (Online) Journal homepage: https://www.tandfonline.com/loi/hjce20
Improving accrual into cancer clinical trials John F. Foley MD & Charles G. Moertel MD To cite this article: John F. Foley MD & Charles G. Moertel MD (1991) Improving accrual into cancer clinical trials, Journal of Cancer Education, 6:3, 165-173 To link to this article: https://doi.org/10.1080/08858199109528114
Published online: 01 Oct 2009.
Submit your article to this journal
Article views: 4
View related articles
Citing articles: 4 View citing articles
Full Terms & Conditions of access and use can be found at https://www.tandfonline.com/action/journalInformation?journalCode=hjce20
J. Cancer Education. Vol. 6, No. 3. pp. 165-173, 1991 Primed in the U.S.A. Pergamon Press pic
0885-8195/91 $3.00 + .00 © 1991 American Association for Cancer Education
IMPROVING ACCRUAL INTO CANCER CLINICAL TRIALS JOHN F . FOLEY, MD* and CHARLES G. MOERTEL,
MD†
Abstract — It is thought by some that there are inadequate patient entries on high priority national cancer treatment protocols, resulting in intolerable delays in addressing crucial research issues. The North Central Cancer Treatment Group‡ in Rochester, Minnesota generated a questionnaire with possible impediments to patient treatment entry as well as suggested measures for approving accrual. Of 334 questionnaires, 63% were completed and returned by medical oncologists, radiation therapists, surgeons and nurse/data managers. The most frequently listed impediments included patients' advanced age or fragility, inadequate health insurance, and excess travel distance. Protocol concerns included too many tests that were too complex and too costly. Practice concerns included physicians too busy and inadequate help. Suggestions toward improving patient accrual included subsidizing drugs, subsidizing nonessential tests for patient care, and increasing funding for support personnel. Popular suggestions for education efforts included teaching the importance of clinical trials, emphasizing personal benefit to the patient and recognition of physicians and institutions participating. It was felt that protocols should be simplified, be more consistent with practice, and be more flexible. These considerations are relevant not only in the interests of more effective clinical research but also in the interests of appropriate education of residents and oncology fellows involved in clinical trials.
INTRODUCTION
reasons, the conduct of clinical trials is one of the most important responsibilities of the NaThe clinical trial is the ultimate and only mean- tional Cancer Program. To be of greatest service ingful method of establishing the value of new to the public, it is essential that cancer clinical cancer treatment approaches. For new drug ther- trials be conducted with the highest standards of apy it is required that value be proven by clin- quality, involve sufficient patients to achieve ical trial, usually randomized, before such therapy statistical validity, and be completed in a timely can be offered to the public at large. For these fashion. To expeditiously achieve these objectives, a large national clinical trial organization has been created that represents one of the largThis work was supported by grants CA 25224, CA 37404, est single investments of the public funds enCA 47404, CA 35269, CA 35101, CA 37417, CA 35448, trusted to the National Cancer Institute. In spite CA 35272, CA 35113, CA 35103 from the National Canof this, however, concerns have been raised recer Institute, National Institutes of Health. * F r o m the University of Nebraska Medical Center, Omaha, garding what is perceived as inadequate patient Nebraska. entries on high priority national protocols, re† From the Mayo Clinic, Rochester, Minnesota. ‡ sulting in intolerable delays in addressing cruFrom the North Central Cancer Treatment Group and the Mayo Clinic: The Billings Clinic, D. Twito; Quain and cial research issues. Dr. Michael Friedman, Chief Ramstad Clinic, D. Pfeifle; Oncology Associates of Cedar of the Clinical Investigations Branch of the DiRapids, M. Wiesenfeld; The Duluth Clinic CCOP, J. Krook; vision of Cancer Treatment, NCI, has stated, The Fargo Clinic CCOP, G. McCormack; Geisinger Clinic and Medical Center CCOP, R. Goldberg; Ochsner Medical "Examining national trends reveals that only Foundation CCOP, C. Kardinal; Illinois Oncology Research about 1% - 1.5% of available breast cancer, Association CCOP, S. Cullinan; Rapid City Regional On0.5% of rectal cancer, and 1% of colon cancer cology Group, L. Ebbert; Saskatchewan Cancer Foundation, M. Poon; Sioux Falls Community Cancer Consortium CCOP, patients are actually recruited into such studL. Tschetter, St. Cloud Clinic of Internal Medicine, Ltd., ies." 1 He asks, "Why is an $8,000,000 CCOP H. Windschitl; Grand Forks Clinic Ltd., J. Laurie; Nebraska (and a $55,000,000 Cooperative Group) proOncology Group - Creighton University/University of Nebraska Medical Center and Associates, J. Foley, J. Mailliard. gram not capable of generating larger numbers Reprint requests to: John F. Foley, M.D., Section of of patients with specific types of malignancies Oncology/Hematology, University of Nebraska Medical Cento resolve important scientific dilemmas conter, 600 South 42nd Street, Omaha, NE 68198-3330.
165
166
J.F. FOLEY and C G .
fronting oncologists?" More recently, former NCI director, Dr. Vincent DeVita was reported to have stated that " . . .the cooperative group system is broken" and that the slow patient accrual of cooperative groups is a "national disgrace." 2 ' 3 Whereas these statements may be a bit pejorative, they are certainly attention-grabbing. Even if the practicing oncologist is not concerned with these national political issues, he must be concerned that such a small proportion of cancer patients with poor prognosis or fatal disease is given the hope that can be provided through entry into clinical research protocols. The evidence is convincing that although an adequate resource of protocol-eligible patients is available through currently funded clinical trial resources, only a small proportion of this reservoir is actually entered on protocols. In a study of the Eastern Cooperative Oncology Group (ECOG), it was found that among 3,534 cancer patients seen by their members and affiliates during a survey period, only 22% were entered on protocols.4 Among those patients eligible for available protocols, only 53% were actually on a protocol. In a similar survey of NCI funded Community Clinical Oncology Programs (CCOPs) only 34% of 9,508 patients clinically eligible for available protocols were entered on protocol.5 In both the ECOG and CCOP surveys, the major reason for not entering an elibible patient on protocol was a decision made by the physician. It is also probable that when the decision was made by patients not to enter protocol, this decision was in large measure influenced by the attitude of physicians. Dr. Michael Friedman has concluded, "Speculation as to what financial, psychological or other factors influenced the physicians' and patients' decisions not to enter a study is sterile. An insightful analysis of reasons for non-participation would be methologically difficult, but is manifestly necessary. ' ' 1 The current study attempts to meet that need. Perhaps of more importance, we attempted to elicit constructive recommendations for improvement in patient protocol accrual.
MOERTEL
vestigators of each of the 15 North Central Cancer Treatment Group (NCCTG) member institutions (all medical oncologists), the chairman of each modality committee (medical oncology, radiation oncology, surgical oncology and pathology), the NCCTG chairman and cochairman (both medical oncologists), and Dr. Robert Wittes, Associate Director, Cancer Therapy Evaluation Program, Division of Cancer Treatment, NCI. The objective of this meeting was to prepare a compilation of possible reasons why patients were not entered on protocol and of suggested measures for improving protocol entry. Based on the contributions from participants in this initial meeting, a questionnaire was composed listing 17 possible impediments to patient entry. Respondents were asked to score these items on a 0 to 4 basis with 0 representing no impediment to patient entry and 4 a major and very frequently encountered impediment. The questionnaire also listed 18 suggested measures to improving patient accrual, again with a 0 to 4 scoring system with 0 indicating the opinion that the measure would not improve accrual and 4 indicating that the measure was highly likely to improve accrual. The questionnaires were then distributed to each individual listed on the NCCTG membership roster, and results were tabulated for all members who were directly or indirectly responsible for patient protocol entry. Respondents were classified into four groups: medical oncologists and/or hematologists, general surgeons and surgical subspecialists, radiation oncologists, and oncology nurses and/or data managers. Respondents were given the option of signing questionnaires or submitting them anonymously. Results were analyzed by the mean score accorded each item and also by the percentage of respondents who scored each item as one of major importance, i.e., score 3 or 4. Since there were no substantive differences in results by either means of analysis, we have elected to report the percentages of respondents who scored each item as 3 or 4. RESULTS
METHOD Our first step in addressing the purposes stated above was a meeting involving the principal in-
Of 334 questionnaires sent to appropriate group members, 209 (63%) were completed and returned. According to specialty, question-
Improving accrual into cancer clinical trials
naires were returned by 75% of 91 oncologists/ hématologiste, by 75% of 80 nurse/data managers, 70% of 40 radiation oncologists, and 43% of 123 surgeons. It would seem reasonable to assume that the responses are representative of the first three disciplines and probably were contributed by almost all members in these disciplines who are active NCCTG participants rather than just pro forma listings on our roster. It is clear, however, that the respondents composed only a minority of listed surgeons. This may be explained, at least in part, by the fact that a number of uncompleted questionnaires were returned by surgeons with accompanying notes stating that they did not feel they could properly respond since they relied on their medical oncology colleagues for handling of protocol entries. In the main, returned questionnaires were completely filled out. Only 4.8% of a total 7,315 potential answers were left blank, and most of these were on forms returned by surgeons. It would, therefore, appear that the results of this survey are reasonably representative of opinions of those who are more directly involved in the patient protocol entry process (medical oncologists/hematologists/radiation oncologists and nurse/data managers), and less representative of those who are more peripherally involved in this process (surgeons). Figure 1 graphically presents the proportion of respondents who considered the factors described as major impediments to patient protocol entry. These factors have been grouped according to concerns specifically related to the nature of the research protocol, specifically related to the patient, and specifically related to the physician and his practice. It is of interest that although 78% of the respondents considered one or more of the factors listed to be major detriments to patient protocol entry, there was little uniformity as to which of these individual factors were dominant concerns. This was true for the respondents considered as a whole or according to specialty interest. Indeed, there was not a single factor that was considered a major impediment by even a simple majority of all respondents. The most frequently listed concerns were those directly related tö the nature of specific patients
167
(Figure 1A). Leading these was advanced age or perceived fragility of the patient, considered a major impediment to protocol entry by 42% of all respondents and by 53% of medical oncologists/hematologists. Following this were the very practical problems of inadequate health care coverage and geographic location of the patient's residence, which precluded easy and affordable return visits for follow-up treatments or observation. Although refusal of the patient to be an experimental subject or to be randomized were considered important factors by some, it is of interest that three of four respondents did not regard these factors as major impediments to protocol entry. The most frequently mentioned major concerns regarding specific protocols pertained to practical issues, ie, overly complex protocol procedures, excessive testing requirements, excessive cost, and an excess personal burden placed on the patient (Figure IB). It is thought provoking that only 7% of medical oncologists/hematologists regarded excess cost as an entry impediment, whereas 50% of nurse/data managers considered this a major factor. Conversely, 29% of medical oncologists/hematologists considered physician time requirements a major impediment, but only 10% of nurse/data managers agreed. Another interesting contrast, and possibly a lesson for our group, is that inadequate support personnel and problems with interdisciplinary coordination were important considerations for surgeons (32% and 27%) but less troublesome to medical oncologists/hematologists (18% and 10%) (Figure 1C). Perhaps correcting popular misconceptions, the least frequent concerns were excessive toxicity of protocol treatment, scientifically uninteresting protocols, or that the referring physician would object to research protocol entry. There was considerably greater unanimity of respondents regarding recommendations for improving patient protocol entry (Figure 2). We have classified these into measures for improvement of financial support of research, public relations and education efforts, changes in protocol design and conduct, and changes that the research units could make in their practice patterns. The most strongly recommended measures
168
J.F. FOIEY and C G . MOERTEL
Too »any testa
Too old op too fragile
sax o complex
| Xnadoquats health car« coverage 331%
io auch pereonal burden on patient Excessive travel distance from patlent'a residence
25X o costly
quires too aucn physician time 21X Inefusal to be experlnental eubiect «uch toxlcity
J24X
|Unllkely to coaply with protocol requirements Scientifically uninteresting jlBX
50
100
50
loo
(b)
(a) Physician too busy
I Inadequate support personnel J22X |Problem »lth Interdisciplinary coordination 116%
concern that referring phyalclan Mill rebel
50
100
(C)
Figure 1. Major impediments to patient protocol entry: (A) Patient concerns; (B) Protocol concerns; (C) Practice concerns (black, score 4; shaded, score 3).
were those that would require efforts and expenditures on a national level. These involved relieving patients or their third-party carriers of the protocol-related financial burdens of expensive marketed drugs or tests done purely for research purposes (Figure 2A). They also involved efforts through national media to improve the public image of clinical trials in general, of the
benefits of clinical trial participation for individual patients, and of the physician and/or institutions participating in clinical trials (Figure 2B). There were no divisions among the various specialty interest in their strong support for these measures. Available on-scene educational materials for patients and families regarding clinical trials as well as regarding specific protocols
169
Improving accrual into cancer clinical trials
National/Local media: - Importance of clinical trials
V//////A'™ National/Local Redis: - Personal benefit to patients in clinical trials
V//////A* National/Local media: - Recognition of physicians or institutions participating in clinical trials
Patient education video tapes regarding clinical trials Subsidize research testing not essential to patient care tlont education brochures regarding V//////////A'"
ent education brochures regarding
[Incraaae funding for physician participation tlaa
individual protocols
clinical trials
tlent education vldao tapes regarding Individual protocols
ÀA1X
3
50
100 0
100
50
(b)
(a) Simplify protocol procedures and requirements
Y/////////A™ Hake protocol more consistent with clinical practice
Score-3
simollfy forais Prearrange availability of protocol required .•ateríais before patient Is seen
Y////////A™ Increase flexibility In protocol conduct
1Schedule potential patients for protocol entry [at earlier and lesa hectic tines of day
53X Y/////////A* Reduce protocol required testing ¡SOX
50
100
(c)
0
50
100
(d)
Figure 2. Recommendations for improving patient protocol accrual: (A) National financial support; (B) National/local public and patient educational efforts; (C) Changes in protocol and clinical research procedures; (D) Changes in practice of physician investigators (black, score 4; shaded, score 3).
were endorsed by most, and particularly by responding surgeons. There was also uniform enthusiasm for modifying protocols to make them more expeditious and less cumbersome in con-
duct (Figure 2C). With regard to practice patterns, most physician respondents did not feel that scheduling of new patients influenced protocol entry, but 52% of nurse/data managers felt
170
J.F. FOIÍY and C G .
that scheduling at an earlier and less hectic time of day could be advantageous (Figure 2D). Sixtyfour percent-of nurse/data managers and 55% of all respondents endorsed the importance of prearranging availability of necessary materials for protocol entry (eg, operative and pathology reports) when the patient was seen in initial consultation. DISCUSSION Certainly, any opinion survey produces data that are, by definition, purely subjective and therefore to be interpreted with appropriate caution. However, the joint decision making process by physician and eligible patient regarding enrollment in a clinical research protocol is also highly subjective in nature. With all this, a body opinion from those devoted to the objectives of the National Cancer Program probably represents the most rational basis for decision regarding constructive change. The results of this study must be interpreted in light of the nature of the participants, the demography of the regions they represent, and the atmosphere of the time at which the study was conducted. At the time of this survey the NCCTG had 15 member institutions. In cooperation with the Mayo Clinic, it is currently accruing patients to cancer treatment protocols at the rate of approximately 1,500 per year, with a near equal number of patients entered on cancer control protocols. Fourteen members are primarily engaged in the private practice of medicine and only one represents a university setting. Eleven members are located in the north central region of the United States and only one member is located in a large urban area. These characteristics are probably atypical of most clinical cancer trial groups, but in some respects they may be more typical of overall cancer care delivery in this country. The relatively large protocol case entry in relationship to membership may also be atypical. In the main, our membership institutions represent tertiary care facilities in their respective regions; and, for most, the nature of oncology practice is perhaps not as highly competitive as it is in the coastal or sun-belt regions of this country. The long established tradition of interdisciplinary practice among our members may
MOERTEL
also not be nationally representative. The dominance of medical oncologists/hematologists among our respondents may perhaps be criticized, although certainly for most national clinical trial groups this specialty area takes principal responsibility for patient protocol entry. We included oncology nurse/data managers among our respondents without apology, because if they are not primarily instrumental in patient protocol entry, they are at least active collaborators and certainly can be regarded as reliable, if not dispassionate, observers of this phenomenon. Our study was conducted at a time when the major funding source for clinical cancer research, the National Cancer Institute, had expressed serious concerns regarding the relatively small en-. try of patients on nationally approved cancer treatment protocols when viewed in the light of the large available patient reservoir. There had been at least the implication, and perhaps a commitment, of additional funding for measures to effectively improve case accrual. With or without the funding carrot, our group supported the NCI initiative and this study was undertaken as a first step in a positive and constructive initiative to improve clinical trial patient accrual. Perhaps not surprisingly, the results of this study place much of the burden on the shoulders of the National Cancer Institute. Any committed participants in cancer clinical trials, particularly if they derive from the private practice sector, must hold the conviction that the most ideal medical care offered to the poor prognosis cancer patient today is offered in the setting of a well-designed clinical trial. A loud, strong, and clear recommendation of this study is that the American public be educated through the public media regarding the overall importance and ethical nature of clinical trials and the benefit the cancer patient will receive by treatment in such a setting in the hands of physicians who care enough about the cancer patient, both today and tomorrow, to make this extra effort. The public has many misconceptions about treatment on clinical trial protocols, and these may be reinforced by physicians not participating in clinical trials or by promoters of "alternative" therapies. Certainly this is true with regard to randomized trials, particularly if those trials involve no treatment or placebo control arms. It
Improving accrual into cancer clinical trials
is vitally important that the public be assured that the highest quality cancer care is provided by entry into a clinical trial, which will offer not only the best in available treatment but also the hope of treatment improvement both for themselves and for other patients in the future. They should also be assured that randomization is not evidence of scientific depersonalization, but rather is employed only when there is no acceptable evidence of superiority of any of the management programs to which they may be assigned. They should also understand that they will be fully informed if there is a possibility that they will be assigned to no added treatment or to placebo treatment and that this will only be done if there is no acceptable evidence that additional treatment will be of benefit for their .problem, e.g, the surgical adjuvant setting. Public recognition for the efforts and accomplishments of the physician is an impotant inducement to his participation in clinical trials. For the physician in private community practice, this will be the only reward he will receive, since the extra time he devotes to clinical trial participation usually involves a financial sacrifice for him or his institution. An important secondary benefit of such recognition through public media is that patients may be attracted to preferentially seek their cancer care from physicians and institutions who they know have the high motivation and recognized competence necessary for participation in nationally approved clinical cancer research programs. By thus making the community physician who participates in clinical trials more competitive, the reservoir or patients available for clinical trials may in turn be increased. Utilizing national and regional media, the Office of Cancer Communications of the National Cancer Institute has already initiated an aggressive and innovative program for public education in cancer clinical trials. We very much hope that the results of this study will encourage them to stay on this course. The results of this survey have also emphatically indicated that a major impediment to clinical trial accrual is the frequently complex and cumbersome nature of protocols and procedures. Begg6 documented that in institutional trials in Stage II breast cancer, there were 23 exclusion criteria per study. The Community Clini-
171
cal Oncology Program logs demonstrated that among patients with a relevant protocol, 44% were ineligible. Goodwin et al7 reported that although 31% of adult cancer patients were over 70 years old, only 7% of these patients were enrolled in Southwest Oncology Group protocols in New Mexico. Certainly some of the requirements of protocols are essential to their scientific purposes. On the other hand, it must be admitted that data is frequently demanded that have no direct relation to protocol objectives and which will probably never be analyzed for any purpose. The practicing physician may rightfully regard the gathering of such data on extraordinarily detailed study forms as inconsequential busy work and a waste of his time and the time of his staff. Superfluous testing is sometimes more the rule than the exception in clinical research protocols, eg, weekly blood counts for a patient who has already been treated for months on the same regimen without experiencing hazardous hématologie toxicity, chest x-rays and chemistry profiles done every month without clinical indication, meticulous restaging procedures without evidence that these contribute to clinical outcome. Frequency of follow-up is often far in excess of what conscientious clinical practice would dictate. Is it really necessary, for example, to bring an asymptomatic surgical adjuvant patient back at three-month intervals in order to fine-tune the time of recurrence on a protocol for which survival is the endpoint of primary interest? It would seem appropriate for each clinical trial cooperative group to carefully scrutinize protocols and periodically review procedures to ensure that their demands are rational, represent prudent utilization of the time of the physician and data manager, and are properly conservative of health care costs. There is little doubt that the closer our protocol procedures conform to high-quality clinical practice, the more conducive they will be to protocol accrual and to physician and patient compliance. Note in Figure IB, Protocol Concerns, 24% felt that the protocols were too costly. Note in Figure 2A, that 75% felt that there should be subsidization of expensive marketed drugs and 71% felt there should be subsidization for research testing not essential to patient care. We believe this apparent contradiction relates to the fact
172
J.F. FOLEY and C.G.
MOERTEL
that at that time most of the patients have been able to participate in the protocols and get reimbursement. However, the third-party carriers and Medicare are becoming much more difficult to deal with when it relates to expensive marketed drugs and are less willing to fund some of these, particularly when it is stated that they are in an experimental program. In addition, most everyone in protocol studies will agree that more studies and imaging procedures are done in these patients than in patients who are off study. This again reflects the concerns regarding third-party reimbursement. Because of this problem, we have developed a subcommittee that continually looks over testing requirements in the various protocols and recommends that we drop those that seem to be not necessary.
make treatment of a patient on protocol more economical than treatment with a comparable regimen off protocol. The requirement of accounting for research testing not essential for patient care would encourage greater discrimination by investigators before incorporating such testing as a protocol requirement. Both measures would undoubtedly make support of treatment on research protocols more attractive to the health insurance industry. These advantages would, of course, have to be balanced against the total cost to the national clinical cancer trials program. Our survey does make a strong case that if additional investment of public funds is contemplated to increase accrual on national protocols, defraying research-related costs to the patient could be a very effective measure.
The loudest and strongest recommendations expressed by this survey are those regarding the costs that clinical trial participation impose on the patient or the patient's third-party carrier. Additional funding is urged to defer charges for expensive marketed drugs and for testing required for research purposes but not essential to best patient care. Such patient care funding is almost nonexistent in today's clinical trial grants and contracts. The argument has been made that if any such funding is allowed, even for purely research related patient charges, third parties may then be encouraged to refuse payment for all care charges accrued by patients on research protocols, even those required for best patient care. Certainly there is mounting evidence of reluctance of insurance carriers to reimburse charges for cancer research treatment. This has been motivated, at least in part, by the massive expenses involved in treatment with some biologics or demanded by certain commercial research treatment enterprises. On the other hand, a very persuasive argument can be made that third-party carriers have no desire to deny their subscribers afflicted with cancer the hope that accompanies treatment on a research protocol, provided the carriers can be assured that the cost is comparable to that charged to them when the patient is treated in a nonresearch setting. It is clear that government buying power would make drugs available at considerably lower cost than could be obtained in the open market, and defraying the expense of these agents would
There is an apparent contradiction in the data in Figure 1C that only 22% felt there was inadequate support personnel, whereas in Figure 2A, 61% of the respondents felt that there should be increased funding for support personnel. The addition of cancer control to our research activities is beginning to markedly increase the data managers' workload, with no commensurate increase in funding thus far. Thus, the data from Figure 2A indicate our ongoing needs. There are two obvious alternatives for enhancing accrual on important national clinical cancer research trials. The first is to recruit new physicians not currently active in this effort. Whereas this might add numbers, genuine concern must be raised regarding the possible adverse effects on quality of research data. New physicians would have no track record of quality performance, and they would be unaccustomed to research discipline. The fact that they had not previously affiliated with the clinical trials program, in the face of ample opportunity, would raise questions regarding their motivation to produce high quality data over a sustained period of time. The very large trials for which these added numbers would be needed are large because they are looking for small differences. Even a modest rate of poor quality entries on such protocols (ie, patients ineligible, cancelled, lost to follow-up, or treated in gross violation of protocol) could completely obscure a positive result that would have been evident if the protocol had been conducted with high-quality stan-
Improving accrual into cancer clinical trials
dards. Considering the importance of these trials to the cancer patient and the large public investment involved, this would seem to be an intolerable risk. Efforts to improve protocol accrual within existent resources would appear far more attractive. Well-trained and experienced quality control personnel, methods, and logistics would already be in place. Quality of performance would be established by record and approved by peers. Both the ECOG and CCOP studies had shown that the reservoir of eligible patients exists within already funded institutions to increase protocol accrual by two- or three-fold. If our study has validity, it indicates the strong likelihood that protocol accrual can be substantially increased by providing public education on both the national and local level regarding the importance and benefits of clinical trial participation, by public recognition of the efforts and accomplishments of the physicians and institutions participating in clinical trials, by reducing nonessential complexities of clinical trial protocols, and by easing the financial burden of clinical
173
trial participation for the patient or his thirdparty carrier. Perhaps wisdom will dictate that these measures should be given serious consideration before risking the possible major damage that might be done to important national protocols by recruitment of untried participants who have not previously evidenced a devotion to cancer clinical trial programs. REFERENCES 1. Friedman MA: Patient accrual to clinical trials. Cancer Treatment Rep 71:557-558, 1987. 2. Cancer Letter 13, (35), 1987. 3. Cancer Utter 14, (10), 1988. 4. Begg CB, Zelen M, Carbone PP et al: Cooperative groups and community hospitals. Cancer 52:1760-1767, 1983. 5. Hunter CP, Frelick RW, Feldman AR et al: Selection factors in clinical trials: Results from the Community Clinical Oncology Program Physician's Patient Log. Cancer Treatment Rep 71:559-566, 1987. 6. Begg CB: Selection of patients for clinical trials. Semin Oncol 15:434-440, 1988. 7. Goodwin JS, Hunt WC, Humble CG, Key CR, Samet JM: Cancer treatment protocols. Who gets chosen? Arch Intern Med 148:2258-2260, 1988.
