This Article Abstract Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Save to my personal folders Download to citation manager Rights & Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Goldman, D. P. Articles by McCabe, M. Search for Related Content PubMed PubMed Citation Articles by Goldman, D. P. Articles by McCabe, M. Social Bookmarking                            What's this?

Measuring the Incremental Cost of Clinical Cancer Research

From RAND, Santa Monica, CA; Division of Cancer Control and Population Sciences, Division of Cancer, Treatment, and Diagnosis, and Office of Education and Special Initiatives, National Cancer Institute, Bethesda, MD; Dana-Farber Cancer Institute, Boston, MA; Johnson Comprehensive Cancer Center, University of California, Los Angeles, Los Angeles, CA; and Center for Cost and Financing Studies, Agency for Healthcare Research and Quality, Rockville, MD.

Address reprint requests to Dana P. Goldman, RAND, 1700 Main St, Santa Monica, CA 90407-2138; email: dgoldman{at}rand.org

	ABSTRACT

TOP ABSTRACT INTRODUCTION POLICY CONTEXT DISCUSSION REFERENCES

 
PURPOSE: To summarize evidence on the costs of treating patients in clinical trials and to describe the Cost of Cancer Treatment Study, an ongoing effort to produce generalizable estimates of the incremental costs of government-sponsored cancer trials.

METHODS: A retrospective study of costs will be conducted with 1,500 cancer patients recruited from a randomly selected sample of institutions in the United States. Patients accrued to either phase II or phase III National Cancer Institute–sponsored clinical trials during a 15-month period will be asked to participate in a study of their health care utilization (n = 750). Costs will be measured approximately 1 year after their trial enrollment from a combination of billing records, medical records, and an in-person survey questionnaire. Similar data will be collected for a comparable group of cancer patients not in trials (n = 750) to provide an estimate of the incremental cost.

RESULTS: Evidence suggests insurers limit access to trials because of cost concerns. Public and private efforts are underway to change these policies, but their permanent status is unclear. Previous studies found that treatment costs in clinical trials are similar to costs of standard therapy. However, it is difficult to generalize from these studies because of the unique practice settings, insufficient sample sizes, and the exclusion of potentially important costs.

CONCLUSION: Denials of coverage for treatment in a clinical trial limit patient access to trials and could impede clinical research. Preliminary estimates suggest changes to these policies would not be expensive, but these results are not generalizable. The Cost of Cancer Treatment Study is an ongoing effort to provide generalizable estimates of the incremental treatment cost of phase II and phase III cancer trials. The results should be of great interest to insurers and the research community as they consider permanent ways to finance cancer trials.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION POLICY CONTEXT DISCUSSION REFERENCES

 
TRADITIONALLY, THE cost of conducting cancer clinical trials has been supported by a combination of research sponsors, institutions, and third-party payers. However, health insurers and other payers are increasingly reluctant to reimburse for direct patient care provided as part of a clinical trial.¹ These policies—driven in part by a perception that patients enrolled onto trials incur substantial additional costs—impede efforts to enroll patients onto clinical trials.² Yet there is a lack of generalizable evidence regarding the costs of treating patients in clinical trials.

Given the importance of clinical trials in extending longevity and improving quality of life,^3-7 there is an urgent need for unbiased information on the possible effects on patient care costs of participation in government-sponsored clinical trials. Such data would make any cost-sharing burden explicit and could lead to better mechanisms for financing clinical trials. This article summarizes current knowledge on the incremental costs of treating cancer patients in clinical trials and the methodologic challenges in generating precise and generalizable estimates of costs. We also introduce the Cost of Cancer Treatment Study (CCTS), an ongoing effort to obtain national estimates of the direct care costs of patients who participate in National Cancer Institute (NCI)–sponsored clinical cancer trials.

	POLICY CONTEXT

TOP ABSTRACT INTRODUCTION POLICY CONTEXT DISCUSSION REFERENCES

 
The issue of coverage for patient care costs in clinical trials is complex and involves at least two important policy questions. The first is, do insurance barriers impede patient recruitment onto clinical trials. Most insurers or plans have policies that exclude coverage for services provided as part of a clinical trial and define them as experimental.^2,8-10 However, since most payers do not track who is enrolled onto clinical trials, these policies are not always enforced.¹¹ Exclusions tend to be invoked only for expensive experimental procedures—eg, stem-cell transplantation—which, even when given outside a research protocol, typically require preauthorization for coverage. Thus, the question is really an empirical one.

Unfortunately, there are few, if any, careful studies quantifying this problem, but a recent study by the General Accounting Office (GAO) does provide some qualitative evidence. After interviewing health insurers, clinical researchers, government officials, and medical directors at health plans, the GAO concluded that there are not widespread limitations on patient access but cautioned that "neither the insurers nor the cancer centers were statistically representative groups, and thus the findings from our interviews cannot be generalized."² In addition, the mere existence of these exclusions probably discourages enrollment.² Other studies also suggest that physicians may automatically exclude patients with certain types of insurance from trials because of reimbursement concerns¹¹—a practice that may explain why participation rates are lower in some managed-care settings.¹²

The second question is, do insurers erect these barriers because of cost concerns? Here the answer is clearer. In the GAO report, medical directors indicated that decisions about trial participation are made on a case-by-case basis, with costs tantamount to scientific merit as the two most important factors in the decision.

As a consequence, public and private efforts are underway to remove insurance barriers to trial enrollment. Public agreements allow less discretion in the choice of trials. For example, NCI has entered into agreements with the United States Department of Defense and the Department of Veterans Affairs to provide their beneficiaries coverage when participating in any NCI-sponsored clinical trial; and Virginia, Illinois, Maryland, and Rhode Island have enacted laws mandating at least partial coverage for participants in federally approved clinical trials. Most importantly, President Clinton signed an executive order in June 2000 mandating that Medicare cover routine costs associated with all clinical trials, with the purpose of encouraging elderly patients, who comprise 65% of all new cases, to enroll onto cancer clinical trials.¹³

Several private initiatives are also underway, although these provide insurers with more discretion. One exception is the agreement signed by the Governor of New Jersey with a coalition of insurance companies to provide an estimated 25,000 cancer patients in that state access to federally approved clinical trials.

Even with these policy changes, however, there is still a need for generalizable estimates of the cost of trial participation. First, many of these arrangements are being conducted as demonstrations (Journal of the National Cancer Institute news articles¹⁴), and cost information must be developed before coverage will become a permanent benefit. Second, and most importantly, no clear consensus has emerged regarding the definition of standard care costs. The Institute of Medicine recently recommended that Medicare reimburse routine care for patients in clinical trials in the same way it reimburses routine care for patients not in clinical trials.¹¹ The Institute of Medicine also recommended that the Health Care Financing Administration pay more than routine costs, at least for selected trials.¹¹ Obtaining valid and precise estimates of costs would help insurers develop simpler and more consistent financing strategies for clinical trials.

ESTIMATES OF THE COST OF TRIAL PARTICIPATION
Three recent studies have investigated the costs of care among cancer patients in single institutions or health plans and provide some useful evidence. Wagner et al¹⁵ found that 61 cancer patients in phase II and phase III cancer trials at the Mayo Clinic had at most 10% higher costs than a set of matched controls not enrolled onto trials over a 5-year period, although the difference was not statistically significant. Fireman et al¹⁶ estimated that 135 patients in NCI-sponsored cancer trials at a large group model health maintenance organization (HMO) (Kaiser Permanente, Northern California) had approximately 10% higher costs over 1 year than 135 matched controls, with most of the difference attributable to chemotherapy administration costs. Finally, Barlow et al (manuscript in preparation) estimated treatment costs over a 2-year period among 77 patients in NCI-sponsored breast and colorectal cancer trials at another large HMO (Group Health Cooperative–Puget Sound). Compared with a general sample of nontrial patients with the same age range, time of diagnosis, and initial cancer stage, trial patients incurred slightly lower treatment costs, although the difference was not statistically significant; however, using data from 26 patients in breast cancer trials and matched controls, trial patients incurred 26% higher costs over a 2-year period. (Barlow et al, manuscript in preparation).

Several unpublished studies presented at the Thirty-Sixth Annual Meeting of the American Society of Clinical Oncology (May 20-23, 2000) reached similar conclusions. Quirk et al¹⁷ found that 77 patients in phase II and phase III cancer trials at Memorial Sloan-Kettering Cancer Center had lower mean treatment costs (hospital costs plus physician charges) than 75 matched controls not enrolled onto a clinical trial, although the difference was not statistically significant. Stinson et al¹⁸ found that 35 patients in phase II clinical trials at five cancer centers belonging to the American Association of Cancer Institutes had lower mean treatment costs than 35 matched controls, although the difference was not statistically significant.

These studies provide important evidence about the costs of care associated with trials. Nevertheless, as Brown¹⁹ has argued previously, more study is warranted. There are a number of technical reasons for this conclusion. First, existing studies have had sample sizes that were insufficient to detect cost differences that may be important for policy purposes, mainly because of the limited number of available trial patients at any single institution or health plan. Second, cases and controls matched at a single institution may differ in unobserved but important ways that affect treatment costs due to self-selection into trials. Third, these studies excluded some potentially important dimensions of treatment. For instance, each published study excluded treatment provided by clinicians outside the delivery system in which the respective study was conducted^15,16 (Barlow et al, manuscript in preparation), and one study excluded the costs of medications.¹⁵

More important, however, is that treatment patterns differ across institutions, and—with the exception of the study by Stinson et al¹⁸—each of these studies was conducted within a single institution or health system. In all the previous studies, trial patients and matched controls in each study received care at major research institutions or specialized HMOs, where the costs of treatment may be different from costs in the community settings where many patients receive care. This makes the results difficult to generalize.

THE CCTS
Precise and generalizable estimates of the effects of trial participation on patient care costs could help policy makers refine mechanisms for financing clinical trials, with the goal of facilitating timely clinical research. Given the inherent limitations of previous studies, the most likely way to obtain such estimates is via a large, nationally representative sample of institutions involved in cancer trial research.

The CCTS is a 3-year study designed to provide such estimates. Approximately 1,500 cancer patients will be recruited from a broad cross-section of trials and institutions nationwide. Ultimately, the CCTS will yield a precise answer to the question "How much more expensive is it to treat a patient on an NCI-sponsored clinical trial?" This answer will allow policy makers, insurers, and others to estimate the additional treatment cost, if any, of providing blanket access to these trials.

The CCTS will use a retrospective cohort design. Patients accrued to NCI-sponsored clinical trials in 1998 will be asked to participate in a study of their health care utilization approximately 1 year after their trial enrollment. Costs will be measured for all services used by this sample. Similar data will be collected for a comparable group of cancer patients not receiving care in a research study who will serve as the controls for the CCTS. Efforts will be made to estimate the cost of care for a wide spectrum of services from all of a patient’s providers using a combination of billing records, medical records, and an in-person survey questionnaire. Further details about the CCTS, and how it is designed to address previous limitations, are presented below.

METHODOLOGIC CHALLENGES
A number of conceptual and methodologic issues must be resolved in order to produce cost estimates in a timely and cost-effective way. They include the sample size necessary to assess policy-relevant differences in treatment costs, the characteristics of the appropriate control patients, the appropriate follow-up period, and the types of medical care that should be included in cost estimates.

Sample Size and Design After consulting with policy makers and insurance industry leaders, members of the CCTS group determined that to be useful to policy makers, a national cost study should have sufficient sample size to detect a 10% difference in costs. Power calculations then dictated that the study needed 750 patients on trials and 750 matched controls in order to have an 80% chance of detecting such a difference.²⁰ In addition, cross-institution samples would be necessary for the study to be generalizable, given the likelihood that practice patterns and treatment costs vary across institutions. However, it was clear at the outset that the CCTS would not be feasible if it were necessary to approach hundreds of different providers and institutional review boards and collect records and data from huge numbers of sites. Otherwise, the costs of obtaining permissions from a multitude of study investigators and institutional review boards, and the effort needed to collect and record data, would make the study prohibitively expensive.

The existence of timely national data on accrual from the cooperative groups—which are responsible for most of the clinical trial accrual on NCI-sponsored protocols—provided a convenient sampling frame for a clustered, multistage design. (Details are provided elsewhere.²⁰) However, using data supplied by NCI’s Cancer Therapy Evaluation Program, we sampled 35 NCI-sponsored phase III treatment trials active in 1998. A list of all institutions affiliated with these 35 trials was then compiled, and a final list of 55 study sites was then sampled randomly. The study sites include a heterogenous mix of providers in the cancer research community, ie, 14 NCI-designated cancer centers, 12 community clinical oncology programs, and 29 other institutions, such as academic medical centers. These institutions also affiliate with community clinics and hospitals that accrue patients under the auspices of the core institution (in some cases, the affiliates are owned or operated by the core institution, but not always). Patients in the sampled phase III trials, as well as an additional sample of phase II patients, will be selected from these institutions for recruitment into the CCTS.

Identifying Controls One critical issue in assessing the incremental costs of trial participation is to identify an appropriate comparison population. Most fundamentally, to find controls for patients in a particular trial, CCTS will select patients who met the protocol eligibility criteria for that trial during the same time period during which the case enrolled but who did not receive cancer therapy as part of any treatment trial. Patients enrolled onto the control arm of a clinical trial are considered "cases" in the CCTS.

Ideally, the CCTS would identify controls through a systematic review of medical records to determine which patients met the relevant the protocol entry criteria.¹⁴ However, such a strategy is prohibitively expensive given the number of patients and institutions involved. Instead, the CCTS will use a variety of sources at each institution. Patient logs will be consulted in some cases to identify people who were approached for trial enrollment but who did not enroll. The number of such patients will be limited, however, so the majority will be identified through tumor registries (or other administrative data) followed by a brief medical record screen. An expert panel of oncologists has been convened to identify the screener criteria for each sampled trial.

Selection Bias Trial participation is voluntary, so patients who are eligible for a trial but do not enroll may be systematically different from those who do enroll. These differences may affect treatment costs if, for instance, patients with a preference for aggressive treatment proactively seek out trials or are more likely to enroll onto a trial if offered the opportunity. Within the CCTS, we will attempt to minimize this potential source of bias in several ways. First, we will use chart reviews to identify any differences in underlying disease severity. Any residual severity differences between cases and controls will then be controlled for using multi-variate analysis. Second, the patient survey will assess characteristics that might affect the type of care patients might seek out or agree to if offered, such as patients’ preferences for care, their access to trials (eg, whether they had ever sought out a trial, or whether a trial had ever been offered to them), and their health locus of control.

Third, the CCTS will try to identify control patients who were prevented from enrolling onto the particular trial for external reasons. For instance, the CCTS will seek controls by examining logs of patients approached for each sampled trial at a particular institution but whose insurance refused to authorize trial participation. In addition, the CCTS will identify patients who were actually ineligible for a particular trial but whose clinical characteristics differed from the protocol entry criteria in ways that were likely to have a minor effect on the course of treatment. In the CCTS, a panel of oncologists is reviewing the protocol eligibility criteria of each sampled trial to identify criteria that may be unlikely to affect patients’ course of treatment.

Fourth, the CCTS will seek some controls for particular trials at institutions where that trial was not taking place. Matching patients and controls within the same institution may be difficult, because such patients may be hard to find: most patients at an institution who are clinically eligible for an open trial at that institution may be enrolled. Indeed, Wagner et al¹⁵ report having to drop more than half of their available trial patients because no matched control could be identified at the Mayo Clinic. Thus, the CCTS will draw some control patients from other institutions where the trial is or was not being conducted. These patients, who were never approached about entering a trial, will not be self-selected as cases.

Measuring Costs Data on treatment costs could be collected prospectively or retrospectively. In principle, a prospective design could substantially improve data quality relative to retrospective data collection, which is why such a design is preferred by studies such as the Medical Expenditure Panel Study.²¹ However, prospective cost studies are expensive to implement when patients are accruing slowly and enrollment is taking place at hundreds of institutions. (This also partly explains why administration of clinical trials is so expensive.) Because of this expense—as well as the delay in getting results—the CCTS chose to follow other studies and use a retrospective design to assess costs^15,16,21,22 (Barlow et al, manuscript in preparation).

Data on health care use and costs will come from a combination of patient report and administrative records. The design requires patients to report about past health care use and to identify health care providers from whom administrative records can be obtained (with patients’ consent). Most of the literature on recall bias supports this design. The major problem with self-reported utilization data is the net omission of medical events,²³ a phenomenon experimental psychologists associate with the exponential decay of memory.^24-26 Fortunately, the omissions tend to be less salient—and costly—events, with the recall of inpatient episodes better than that of outpatient care.²⁷ Patients also sometimes include utilization outside of the recall window—a process known as telescoping.^28,29 Overall, the literature supports retrospective assessment of health care use over a 6-month period, at most; however, patients may be able to identify their medical providers over a longer period, and administrative records can be collected from these providers for periods covering more than the previous 6 months.

Estimating Costs in Multiple Settings Patients reach clinical trials in a number of ways. In many cases, patients may seek out or be referred to a clinical trial being conducted at the institution where they are already receiving care. However, some patients may change providers in order to participate in a clinical trial, for instance, from a community provider to an academic medical center (or from a community provider that does not do clinical research to one that does). If practice patterns and/or health care costs differ across different types of institutions, as seems plausible, treatment costs for these patients will change. The CCTS will measure costs for the subgroups shown in the matrix in Table 1.

As a consequence, the CCTS will be able to provide generalizable cost estimates in both tertiary and community settings, as well as for patients who leave community settings to enroll onto a clinical trial.

	DISCUSSION

TOP ABSTRACT INTRODUCTION POLICY CONTEXT DISCUSSION REFERENCES

 
Evidence suggests that, at least in some circumstances, health plans are reluctant to pay for care delivered as part of a clinical trial. Such denials of coverage limit patient access to trials and could impede clinical research. Public and private efforts are underway to change these policies, but their permanent status is unclear given cost concerns.

Preliminary estimates of the cost differences are available, but these come from studies at selected institutions. The CCTS is an effort to measure the incremental treatment cost of cancer trials and to address limitations of previous studies.

Successful implementation is predicated on the participation and cooperation of the cancer community. If successful, the study will provide generalizable results that should be of great use to insurers, the cancer research community, and policy makers as they consider ways to finance clinical trial research.

	ACKNOWLEDGMENTS

 
Principal funding for the Cost of Cancer Treatment Study is being provided by the National Cancer Institute. Additional funding comes from the Office of the Director, National Institutes of Health, and the National Science Foundation as part of its support for the White House’s Office of Science and Technology Policy.

We are indebted to Martin Brown, Susan Hubbard, Elise Kreiss, Joseph Lipscomb, Michael Montello, and Jeanette Sian at the National Cancer Institute; John Crowley, Charles Coltman, Marjorie Godfrey, Jeanne-Marie Smith, and Dana Sparks at the Southwestern Oncology Group; Robert Coomis, Rick Magnan, and Tom Malone at the Eastern Cooperative Oncology Group; John Kellner, Robert Park, and Edward Trimble at the Gynecologic Oncology Group; Walter Cronin, Joan Goldberg, and Norman Wolmark at the National Surgical Adjuvant Breast and Bowel Project; Sharon Elcombe, Janis Gjervik, Sonja Hamilton, Michael O’Connell, and Michael Pfenning at the North Central Cancer Treatment Group; Walter Curran and Tim McKeough at the Radiation Therapy Oncology Group; and Michael Moloney, Karen Sartell, and Richard Schilsky at the Cancer and Leukemia Group B. We thank Deborah Rivera for her assistance in preparing the manuscript.

	NOTES

 
The opinions expressed herein are solely those of the authors.

	REFERENCES

TOP ABSTRACT INTRODUCTION POLICY CONTEXT DISCUSSION REFERENCES

 
1. Fleming ID: Barriers to clinical trials: Part I. Reimbursement problems. Cancer 74: 2662-2665, 1994 (suppl 9)[Medline]

2. US General Accounting Office: NIH Clinical Trials: Various Factors Affect Patient Participation. Washington, DC, US General Accounting Office, 1999, Publication No. GAO/HEHS-99-1821

3. National Institutes of Health Consensus Conference: Treatment of early-stage breast cancer. JAMA 265:391-395, 1991

4. National Institutes of Health Consensus Conference: Adjuvant therapy for patients with colon and rectal cancer. JAMA 264:1444-1450, 1990

5. Fisher B, Redmond C, Poisson R, et al: Eight-year results of a randomized clinical trial comparing total mastectomy and lumpectomy with or without irradiation in the treatment of breast cancer. N Engl J Med 320: 822-828, 1989[Abstract]

6. Fisher B, Brown A, Mamounas E, et al: Effect of preoperative chemotherapy on local-regional disease in women with operable breast cancer: Findings from National Surgical Adjuvant Breast and Bowel Project B-18. J Clin Oncol 15: 2483-2493, 1997[Abstract/Free Full Text]

7. Perez EA, Hesketh P, Sandbach J, et al: Comparison of single-dose oral granisetron versus intravenous ondansetron in the prevention of nausea and vomiting induced by moderately emetogenic chemotherapy: A multicenter, double-blind, randomized parallel study. J Clin Oncol 16: 754-760, 1998[Abstract]

8. Schain WS: Barriers to clinical trials: Part II: Knowledge and attitudes of potential participants. Cancer 74: 2666-2671, 1994 (suppl 9)[Medline]

9. Mansour EG: Barriers to clinical trials: Part III: Knowledge and attitudes of health care providers. Cancer 74: 2672-2675, 1994 (suppl 9)[Medline]

10. Morrow GR, Hickok JT, Burish TG: Behavioral aspects of clinical trials: An integrated framework from behavior theory. Cancer 74: 2676-2682, 1994 (suppl 9)[Medline]

11. Aaron HJ, Gelband H (eds): Extending Medicare Reimbursement in Clinical Trials. Washington, DC, Institute of Medicine, National Academy Press, 2000

12. Kirman C: Clinical research in the managed care setting. Drug Benefit Trends 8: 17-26, 1996

13. Hutchins LF, Unger JM, Crowley JJ, et al: Underrepresentation of patients 65 years of age or older in cancer treatment trials. N Engl J Med 341: 2061-2067, 1999[Abstract/Free Full Text]

14. Reynolds T: Cost studies show clinical trials, standard therapy may be equal. J Natl Cancer Inst 92: 1116-1118, 2000[Free Full Text]

15. Wagner JL, Alberts SR, Sloan JA, et al: Incremental costs of enrolling cancer patients in clinical trials: A population-based study. J Natl Cancer Inst 91: 847-853, 1999 (published erratum appears in 92:164-165, 2000)[Abstract/Free Full Text]

16. Fireman B, Fehrenbacher L, Gruskin EP, et al: Cost of care for patients in cancer trials. J Natl Cancer Inst 92: 136-142, 2000[Abstract/Free Full Text]

17. Quirk J, Schrag D, Radzyner M, et al: Clinical trial costs are similar to and may be less than standard care and inpatient (INPT) charges at an academic medical center (AMC) are similar to major, minor, and non-teaching hospitals. Proc Am Soc Clin Oncol 18: 433a, 2000 (abstr 1696)

18. Stinson TJ, Bennett CL, Vogel V, et al: A multi-center study of the costs of enrolling cancer patients in phase II clinical trials. Proc Am Soc Clin Oncol 18: 432a, 2000 (abstr 1695)

19. Brown ML: Cancer patient care in clinical trials sponsored by the National Cancer Institute: What does it cost? J Natl Cancer Inst 91: 818-819, 1999[Free Full Text]

20. Goldman DP, Adams JL, Berry SH, et al: The Cost of Cancer Treatment Study’s Design and Methods. Santa Monica, CA, RAND, 2000 (RAND Publication No. MR-1169-NCI/NIH/OSTP)

21. Cohen S: Sample Design of the 1996 Medical Expenditure Panel Household Survey Component. Rockville, MD, Agency for Health Care Policy and Research, 1997 (MEPS Methodology Report No. 2, AHCPR Publication No. 97-0027)

22. Bozzette SA, Berry SH, Duan N, et al: The care of HIV-infected adults in the United States: HIV Cost and Services Utilization Study Consortium. N Engl J Med 339: 1897-1904, 1998[Abstract/Free Full Text]

23. Means B, Mingay DJ, Nigam A, et al: A cognitive approach to enhancing health survey reports of medical visits, in Gruneberg MM, et al (ed): Practical Aspects of Memory: Current Research and Issues—Vol 1. Memory in Everyday Life. Chichester, United Kingdom, Wiley, 1998, pp 537-542

24. Balamuth E: Health Interview Responses Compared to Medical Records. Washington, DC, National Center for Health Statistics, 1965, series 2, no 7

25. Jobe JB, White AA, Kelley CL, et al: Recall strategies and memory for health-care visits. Milbank Q 68: 171-189, 1990[Medline]

26. Sudman S, Bradburn NM: Effects of time and memory factors on response surveys. JASA 68: 805-815, 1973

27. Cohen S, Burt V: Data frequency effect in the National Medical Care Expenditure Survey. J Econ Soc Meas 13: 125-151, 1985[Medline]

28. Roberts RO, Bergstralh EJ, Schmidt L, et al: Comparison of self-reported and medical record health care utilization measures. J Clin Epidemiol 49: 989-995, 1996[Medline]

29. Neter J, Waksberg J: A study of response errors in expenditures data from household interviews. JASA 59: 18-55, 1964

Submitted January 6, 2000; accepted July 25, 2000.

CiteULike    Complore    Connotea    Del.icio.us    Digg    Facebook    Reddit    Technorati    Twitter    What's this?

This article has been cited by other articles:

				E. D. Paskett, D. Schrag, A. Kornblith, E. B. Lamont, J. C. Weeks, J. R. Marshall, C. Shapiro, and J. Holland Cancer and leukemia group B cancer control and health outcomes committee: origins and accomplishments. Clin. Cancer Res., June 1, 2006; 12(11): 3601s - 3605s. [Abstract] [Full Text] [PDF]

				E. J. Sherman, D. M. Rubin, E. Venkatraman, G. K. Schwartz, V. A. Miller, M. H. Radzyner, H. S. Ruchlin, D. Spriggs, and D. G. Pfister Using Patients As Their Own Controls for Cost Evaluation of Phase I Clinical Trials J. Clin. Oncol., April 1, 2004; 22(7): 1308 - 1314. [Abstract] [Full Text] [PDF]

				J. R. Wright and M. N. Levine Researching the Cost of Research J. Clin. Oncol., November 15, 2003; 21(22): 4081 - 4082. [Full Text] [PDF]

				D. P. Goldman, S. H. Berry, M. S. McCabe, M. L. Kilgore, A. L. Potosky, M. L. Schoenbaum, M. Schonlau, J. C. Weeks, R. Kaplan, and J. J. Escarce Incremental Treatment Costs in National Cancer Institute-Sponsored Clinical Trials JAMA, June 11, 2003; 289(22): 2970 - 2977. [Abstract] [Full Text] [PDF]

				M. L. Pearson, P. A. Ganz, K. McGuigan, J. R. Malin, J. Adams, and K. L. Kahn The Case Identification Challenge in Measuring Quality of Cancer Care J. Clin. Oncol., November 1, 2002; 20(21): 4353 - 4360. [Abstract] [Full Text] [PDF]

				C. L. Bennett, J. R. Adams, K. S. Knox, A. M. Kelahan, S. M. Silver, and J. S. Bailes Clinical Trials: Are They a Good Buy? J. Clin. Oncol., December 1, 2001; 19(23): 4330 - 4339. [Abstract] [Full Text]

This Article Abstract Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Save to my personal folders Download to citation manager Rights & Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Goldman, D. P. Articles by McCabe, M. Search for Related Content PubMed PubMed Citation Articles by Goldman, D. P. Articles by McCabe, M. Social Bookmarking                            What's this?