|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
Journal of Clinical Oncology, Vol 26, No 22 (August 1), 2008: pp. 3735-3742 © 2008 American Society of Clinical Oncology. DOI: 10.1200/JCO.2007.13.2555 Variations in Quality of Care for Men With Early-Stage Prostate Cancer
From the Departments of Urology and Epidemiology, College of Physicians and Surgeons, Columbia University, New York, NY; Departments of Urology and Radiation Oncology, University of Michigan, Ann Arbor, MI; Departments of Urology and Health Services, David Geffen School of Medicine, School of Public Health, and Jonsson Comprehensive Cancer Center, University of California, Los Angeles, CA; Department of Epidemiology and Biostatistics, Arnold School of Public Health, University of South Carolina, Columbia, SC; and American College of Surgeons Commission on Cancer, Chicago, IL Corresponding author: John T. Wei, MD, MS, Department of Urology, University of Michigan, 1500 E Medical Center Dr, Women's Trailer Rm 1013, Ann Arbor, MI 48109-0759; e-mail: jtwei{at}umich.edu
Purpose The commencement of quality-improvement initiatives such as Pay for Performance and the Physician Consortium for Performance Improvement has underscored calls to evaluate the quality of cancer care on a patient level for nationally representative samples. Methods We sampled early-stage prostate cancer cases diagnosed in 2000 through 2001 from the American College of Surgeons National Cancer Data Base and explicitly reviewed medical records from 2,775 men (weighted total = 55,160 cases) treated with radical prostatectomy or external-beam radiation therapy. We determined compliance with 29 quality-of-care disease-specific structure and process indicators developed by RAND, stratified by race, geographic region, and hospital type.
Results Overall compliance exceeded 70% for structural and pretherapy disease assessment indicators but was lower for documentation of pretreatment functioning (46.4% to 78.4%), surgical pathology (37.1% to 86.3%), radiation technique (62.6% to 88.3%), and follow-up (55%). Geographic variations were observed as higher compliance in the South Atlantic division than the New England division for having at least one board-certified urologist (odds ratio [OR], 9.2; 95% CI, 1.9 to 45.0), at least one board-certified radiation oncologist (OR, 3.3; 95% CI, 1.2 to 9.0), use of Gleason grading (OR, 4.1; 95% CI, 1.2 to 13.8), and administering total radiation dose Conclusion The significant and unwarranted variations observed for these quality indicators by census division and hospital type illustrate the inconsistencies in prostate cancer care and represent potential targets for quality improvement. The lack of racial disparities suggests equity in care once a patient initiates treatment.
To close gaps in the quality of cancer care delivered in the United States, the Institute of Medicine called for improvements in cancer data systems, refined and expanded analyses of existing cancer data, and the establishment of a national infrastructure for cancer-specific quality assessment and improvement initiatives.1 Attainment of these goals will require the development and implementation of innovative quality-measurement programs that are population based, focused on patients and systems of care, comprehensive in content, and guided by the principle of shared accountability.2,3 In response, third-party payers, who hold a financial stake in quality, have commenced their own pay-for-performance programs.4,5 Despite these efforts, national assessments of the quality of cancer care have been slow to emerge and limited in scope.6 Given the substantial public health and economic burden attributable to prostate cancer,7,8 measuring its quality of care is necessary before monitoring and improving it.1 Recognizing this priority, investigators at RAND developed a comprehensive set of quality indicators for early-stage prostate cancer care.9 These evidence-based indicators are unique in assessing a broad spectrum of care and have demonstrated measurement feasibility from both written and electronic medical records.10,11 More importantly, these indicators measure mutable processes of care that represent potential targets for quality improvement, facilitate evaluation of accountable practitioners, and can be aggregated to yield quality measures at the systems level. Variations in the use of invasive therapies for localized prostate cancer by black/white race, geography, and health care setting have been well documented, are largely unexplained by differences in disease biology, and may thus reflect inequities in the delivery of care.12,13 To address this hypothesis, we used the cancer registry infrastructure available through the American College of Surgeons (ACoS) Commission on Cancer to measure the quality of prostate cancer care nationally. We were particularly interested in identifying quality-of-care disparities by race, geographic region, and hospital type.
National Cancer Data Base Infrastructure The National Cancer Data Base Infrastructure (NCDB) was established in 1989 as a joint project of the ACoS and the American Cancer Society. It is a hospital-based information resource that provides data on cancer diagnosis, management, and outcomes for approved institutions, with a primary goal of providing regional and national benchmarks against which participating hospitals can compare their care patterns and outcomes.14 Data from the NCDB have been compared with the population-based Surveillance, Epidemiology, and End Results registry and shown to be similar in terms of patient and disease characteristics.15 In 2000 to 2001, the NCDB collected 70% to 72% of all incident prostate cancer cases in the United States. NCDB data are collected using a standardized, electronic data abstraction format without patient names or other identifying information, maintaining compliance with the Health Insurance Portability and Accountability Act of 1996.15
Sampling Strategy A de-identified file of 117,953 men with early-stage prostate cancer diagnosed during 2000 and 2001 was extracted from the NCDB. A 5% stratified random sample of cases consisting of equal-sized strata based on race (two levels), census division (nine levels), and facility type (three levels) was developed. To limit potential clustering effect at the facility level, a maximum of 30 cases was selected from individual facilities. Selected cases were submitted to participating facilities for medical chart abstraction. Primary treatment cohort was determined using treatment variables collected during the explicit chart abstraction and cross-referenced with existing NCDB treatment variables, such as cancer-specific surgery, surgical approach, radiation modality, total radiation dose, and the availability of conformal radiation facilities. Among the 5,655 cases requested from hospitals, 5,230 (92.5%) were received. Responders and nonresponders were similar at the individual case level by race (P = .59) and clinical stage (P = .69); however, nonresponders were more likely to be at the extremes of age (P = .01) and differed in their types of insurance (P = .02). At the facility level, nonresponders were more likely to come from teaching/research centers (P = .01) and hospitals located in the East North Central and West South Central census divisions (P < .01). Given that these indicators were designed for men undergoing radical prostatectomy (RP) or radiation therapy as monotherapy and evidence in favor of active treatment,17 we excluded those who were treated primarily with observation. The final analytic sample consisted of 2,775 cases (weighted n = 55,160), including 1,390 (weighted n = 27,550) treated with RP and 1,385 (weighted n = 27,610) treated with external-beam radiation therapy. Patient-level sampling weights were developed for the final analytic sample to allow for inference back to the total NCDB population.
Indicator Selection and Chart Abstraction
Assignment of quality-indicator compliance was guided by a chart abstraction tool adapted specifically for this study. We have previously developed a chart abstraction tool to collect RAND indicator data and demonstrated inter-rater reliability (Cohen's Compliance with four structure indicators was determined and reported by the cancer registrars at the facility level. To assign compliance with the 25 process indicators, the registrar at each of the participating facilities performed an explicit and structured review of centralized hospital records and office charts of treating physicians. Each registrar was asked to review and validate case information previously reported to the NCDB. Among the study sample, 48.5% had inpatient hospital records, 47.1% had outpatient hospital records, 32.3% had physician office records, 43.6% had outpatient radiation therapy records, and 42.3% had electronic medical records that were abstracted. The median number of sources accessed per case was two, with 1,028 cases (37%) having three or more sources reviewed and only 171 cases (6%) having no records reviewed. Compliance with a process indicator was defined by documentation in the medical record that the relevant care was performed in a chronologically appropriate time frame with respect to the treatment date. Data confirming compliance may have been recorded by the treating physician or a surrogate (including office, nursing, or resident physician staff). Noncompliance was defined as the lack of specific structure qualities or the absence of documentation for the recommended process of care within the appropriate time frame. The absence of source documentation was also deemed noncompliance, thus there were no missing data for determination of indicator compliance. This operational definition of compliance is consistent with the approach adopted by the Centers for Medicare and Medicaid Services in ongoing pay-for-performance initiatives.3,4 Five of nine covariates had complete data, whereas four covariates had missing data, ranging from 0.2% to 8.8% of weighted cases.
Statistical Analysis All analyses were performed using SAS (version 9.1, SAS Institute, Cary, NC) and SUDAAN (Research Triangle Institute, Research Triangle Park, NC). Hypothesis testing was carried out at the 5% level. Institutional review board approval for this study was obtained from the University of Michigan (Ann Arbor, MI), the University of California at Los Angeles (Los Angeles, CA), and Columbia University (New York, NY).
Table 1 lists characteristics of the final analytic sample representing a weighted total of 55,160 cases. The average age was 66.4 years, and 85% of the weighted sample were white. The pretreatment PSA level was below 10 in 72.7% of cases, 60% of cases had clinical stage T1 tumors, 80% of biopsies had a Gleason sum of 6 or 7, and 41% of cases had no documented comorbid disease. White men tended to be older than black men (mean age of 66.9 and 64.0 years for white and black patients, respectively; P < .001), have lower baseline PSA levels (mean PSA of 9.0 and 12.7 ng/mL for white and black patients, respectively; P < .001), have fewer comorbidities (mean number of comorbid conditions was 1.0 and 1.3 for white and black patients, respectively; P < .001), and were more likely to have Medicare coverage (white patients, 49%; black patients, 37%; P < .0001). Teaching/research hospitals tended to care for younger men (mean age of patients treated at teaching/research hospitals, 65.2 years; mean age of patients treated at community cancer center, 66.5 years; and mean age of patients treated at Comprehensive Cancer Centers, 67.5 years; P < .001), with lower-stage disease (teaching/research, 65% cT1; Comprehensive Cancer Center, 54% cT1; and Community Cancer Center, 60% cT1; P = .001), and include more Veterans’ Administration (teaching/research, 11%; Comprehensive Cancer Center, 1%; and Community Cancer Center, 1%; P = .001) and managed care coverage (teaching/research, 32%; Comprehensive Cancer Center, 23%; and Community Cancer Center, 22%; P < .001). The Great Lakes region tended to have fewer men (32%) with stage T2 disease as compared with the Southeast region (46%; P = .02).
Compliance rates were consistently high (> 85%) for the structure indicators, whereas greater variation was observed with the process indicators (Table 2). In particular, lower rates of compliance were observed for several pretherapy evaluation indicators, including assessment of pretreatment sexual (46.4%; 95% CI, 43.7% to 49.1%), urinary (78.4%; 95% CI, 76.3% to 80.5%), and bowel function (52.1%; 95% CI, 49.4% to 54.7%). Missing source documents, as a reason for noncompliance, was generally low, except for the follow-up indicator (29.5%) and two radiation therapy indicators (23.3% to 33.3%; Table 2).
Table 3, Table 4, and Appendix Table A1 (online only) present indicator compliance rates stratified by race, hospital type, and United States census division, with reference groups black race, Community Cancer Centers, and New England division, respectively. Though a black/white racial difference in compliance for these quality indicators was not generally demonstrated (All P values > .05 except for use of radiation dose > 70 Gy; Table 3), significant variations were observed by hospital type (Table 4; indicators marked by "  " in first column have P < .05) and census division (Appendix Table A1, indicators marked by "" in first column have P < .05). Comprehensive Cancer Centers and/or teaching/research hospitals had higher compliance rates than Community Cancer Centers on all of the structure indicators and five of the pretherapy assessment indicators, including documentation of clinical stage, family history of prostate cancer, and urinary, sexual, and bowel function (Table 4; P < .05). Community Cancer Centers had a significantly higher compliance rate for one indicator (rectal protection, P < .05). Two structure indicators, board certification of urologists (range, 76.5% to 97.5%) and radiation oncologists (range, 84.8% to 96.2%), and two process indicators, communication with primary care physician (range, 62.7% to 82.1%) and conformal total radiation dose (range, 55.6% to 84.7%), demonstrated the most regional variation (Table A1; P < .05).
One goal of measuring the quality of care is to document clinical performance across systems and identify opportunities for improvement in the delivery of care. The use of existing administrative data for clinical quality-of-care assessment, particularly processes of care, has been constrained by a lack of detailed patient-level data, unavoidable misclassification, and a tendency to overestimate compliance.19 The present study overcomes these limitations by leveraging the existing data collection infrastructure available through the ACoS to carry out explicit chart reviews that yield detail-rich, patient-level data, which can be aggregated to provide systems-level evaluations.2,20 Using this approach, we not only identified significant gaps in care and variation on a geographic and facility level, but also demonstrated feasibility for a comprehensive, nationally representative, patient-level evaluation of quality. The relatively high levels of structural indicator compliance suggest that the necessary environment to provide high-quality care is generally available. Notably, the observed variations by hospital type and census division may reflect differing capacities among health care systems to bear the economic and social costs necessary to achieve a favorable environment.21,22 Our finding that certain structural aspects of a high-quality environment, including the availability of board-certified urologists and radiation oncologists, psychological counseling, and conformal radiation therapy, are more prevalent at hospitals with established teaching and/or research activities is consistent with other work demonstrating better performance among academic hospitals in the care of Medicare patients.23 Geographic variation in the availability of structural resources is relevant to the literature documenting unwarranted geographic variation in the use of RP.24,25 Geographic asymmetries in the availability of board-certified urologists, radiation oncologists, and conformal external-beam radiation therapy may result in unbalanced utilization patterns for specific local therapies. Collectively, these structural characteristics may create hot spots of utilization for a particular therapy and provide empiric evidence of supply-sensitive care26 as a possible explanation for the variation in the use of surgery for early-stage prostate cancer. In the current era of evidence-based medicine, there is an increasing expectation for clinicians to incorporate pretherapy clinical information and functional status in the setting of shared decision making.20,27 For the pretherapy disease severity assessment and counseling indicators, the observed compliance of greater than or near 80% implies that high-quality care is possible. In contrast, compliance was substantially lower for pretherapy functional assessment and post-treatment follow-up indicators. Because all therapies for localized prostate cancer have significant and negative impact on health-related quality of life, these specific and mutable processes of care represent potential targets for a systems-oriented quality improvement initiative.28 It is particularly noteworthy that we did not detect significant disparities in structure or process indicator compliance between black and white patients. Despite extensive evidence for racial disparity in prostate cancer mortality, the absence of racial differences between black and white patients in compliance suggests that disparities in care may be diminished among men who have access to care.12,13 Our findings must be considered in light of several limitations. It is possible that processes of care may have occurred but were not recorded or the medical record was unavailable for abstraction by the registrars, leading to the possibility of under-reporting of care that was actually delivered. It is also impossible to capture all the subtleties of community care in such a review. For example, patients may have received counseling or follow-up from other clinicians whose records were not available to the registrars. These limitations highlight the enormous challenges that exist in measuring quality at the community level in the absence of any systems-based quality-assessment infrastructure. Despite the limitations of explicit chart reviews, medical record abstraction remains the standard for assessing the quality of care.19 In an ideal world, quality assessments would be actively integrated into the medical record, obviating the need for retrospective reviews. The rise of electronic medical records, particularly within integrated health systems, will provide rich opportunities to progressively assimilate quality assessment into daily practice and allow for more seamless quality reporting (eg, Pay for Performance) using administrative codes. Although associations of the RAND indicators with critical outcomes such as mortality are desirable, their absence does not invalidate the importance of processes of care as quality measures.23 It was the dearth of randomized studies that document structure-outcome and process-outcome associations that prompted RAND to develop these measures. In reality, most quality-of-care measures rely heavily on levels II and III evidence, and few have had the luxury of clinical trials evidence.29,30 As additional clinical evidence and construct validity for these quality measures become available, ongoing initiatives such as Pay for Performance and the Physician Consortium for Performance Improvement for prostate cancer care are likely to gain even greater traction in the clinical arena. These analyses do not account for potential correlation between patients from the same institution. With 2,775 patients treated at 800 distinct hospitals in our sample, the probability of substantial clustering of outcomes by hospital is low, but there is a chance that the variance of some measures is underestimated, leading to an overstatement of the significance of the group comparisons. This cancer-specific evaluation from the ACoS is an important first step in addressing the call for patient-level evaluation of quality on a nationally representative level. Our findings substantiate the relevance of national physician-led organizations to systematic quality-improvement endeavors. The observed variations in compliance with specific structure and process of care indicators illustrate inconsistencies in the quality of care for men with early-stage prostate cancer in the United States and identify clear targets for future quality-improvement activities.
Although all authors completed the disclosure declaration, the following author(s) indicated a financial or other interest that is relevant to the subject matter under consideration in this article. Certain relationships marked with a "U" are those for which no compensation was received; those relationships marked with a "C" were compensated. For a detailed description of the disclosure categories, or for more information about ASCO's conflict of interest policy, please refer to the Author Disclosure Declaration and the Disclosures of Potential Conflicts of Interest section in Information for Contributors. Employment or Leadership Position: None Consultant or Advisory Role: John T. Wei, AUA-AMA PCPI Prostate Cancer Quality Committee (U), GTX Inc (C), Sanofi-aventis (C), Amgen (C) Stock Ownership: None Honoraria: None Research Funding: None Expert Testimony: None Other Remuneration: None
Conception and design: Benjamin A. Spencer, David C. Miller, Mark S. Litwin, Andrew K. Stewart, Rodney L. Dunn, John T. Wei Administrative support: Benjamin A. Spencer, Mark S. Litwin, Jamie D. Ritchey, Andrew K. Stewart, E. Greer Gay, John T. Wei Provision of study materials or patients: Mark S. Litwin, Andrew K. Stewart Collection and assembly of data: Benjamin A. Spencer, David C. Miller, Jamie D. Ritchey, Andrew K. Stewart, Howard M. Sandler, John T. Wei Data analysis and interpretation: Benjamin A. Spencer, David C. Miller, Mark S. Litwin, Jamie D. Ritchey, Andrew K. Stewart, Rodney L. Dunn, E. Greer Gay, Howard M. Sandler, John T. Wei Manuscript writing: Benjamin A. Spencer, David C. Miller, Mark S. Litwin, Jamie D. Ritchey, Andrew K. Stewart, Rodney L. Dunn, John T. Wei Final approval of manuscript: Benjamin A. Spencer, David C. Miller, Mark S. Litwin, Andrew K. Stewart, Rodney L. Dunn, E. Greer Gay, Howard M. Sandler, John T. Wei
We thank Sue Slater for scanning all prostate study forms.
Supported by a Department of Defense Prostate Cancer Research Program Physician Research Training Award (Grant No. PC040167 to B.A.S.) and Grant No. NIH-1-T-32 DKO7782 from the National Institute of Diabetes and Digestive and Kidney Diseases (D.C.M.). The American College of Surgeons has supported this work through its Special Study program. Presented in part at the 2006 American Society of Clinical Oncology Prostate Cancer Symposium, February 24-26, 2006, San Francisco, CA; the 34th Annual Spring Meeting of the American College of Surgeons, April 23-26, 2006, Dallas, TX; and the Annual Meeting of the American Urological Association, May 20-25, 2006, Atlanta, GA. J.T.W. had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Authors’ disclosures of potential conflicts of interest and author contributions are found at the end of this article.
1. Hewitt M, Simone JV, eds. Ensuring Quality Cancer Care: National Cancer Policy Board, Institute of Medicine and Commission on Life Sciences, National Research Council. Washington, DC, National Academy Press, 1999 2. Crossing the Quality Chasm: A New Health System for the 21st Century. Committee on Quality of Health Care in America, Insitute of Medicine, Washington, DC. National Academy Press, 2001 3. Schroeder S, Fisher E, Berwick D: Performance Measurement: Accelerating Improvement. Washington, DC, National Academy of Sciences, Committee on Redesigning Health Insurance Performance Measures, Payment, and Performance Improvement Programs 2006 4. Medicare Begins Performance-Based Payments for Physician Groups. Centers for Medicare and Medicaid Services. http://www.cms.hhs.gov.media/press/release.asp?Counter=1341 5. Rosenthal MB, Frank RG, Li Z, et al: Early experience with pay-for-performance: From concept to practice. JAMA 294:1788-1793, 2005 6. Malin JL, Schneider EC, Epstein AM, et al: Results of the National Initiative for Cancer Care Quality: How can we improve the quality of cancer care in the United States? J Clin Oncol 24:626-634, 2006 7. American Cancer Society: Cancer statistics, 2006. Department of Epidemiology and Research Surveillance, American Cancer Society, Atlanta, GA. http://www.cancer.org/docroot/STT/stt_0.asp. Accessed on March 11, 2006 8. Litwin MS, Saigal CS: Urologic Diseases in America. NIH Publication No. 04-5512 ed. Washington DC, U.S. Department of Health and Human Services, Public Health Service, National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases, U.S. Government Publishing Office, 2004 9. Spencer BA, Steinberg M, Malin J, et al: Quality-of-care indicators for early-stage prostate cancer. J Clin Oncol 21:1928-1936, 2003 10. Miller DC, Litwin MS, Sanda MG, et al: Use of quality indicators to evaluate the care of patients with localized prostate carcinoma. Cancer 97:1428-1435, 2003[CrossRef][Medline] 11. Litwin MS, Steinberg M, Malin J, et al: Prostate cancer patient outcomes and choice of providers: Development of an infrastructure for quality assessment. Santa Monica, CA, RAND Corporation, 2000 12. Fowler JE Jr, Terrell F: Survival in blacks and whites after treatment for localized prostate cancer. J Urol 156:133-136, 1996[CrossRef][Medline] 13. Johnstone PA, Kane CJ, Sun L, et al: Effect of race on biochemical disease-free outcome in patients with prostate cancer treated with definitive radiation therapy in an equal-access health care system: Radiation oncology report of the Department of Defense Center for Prostate Disease Research. Radiology 225:420-426, 2002 14. American College of Surgeons National Cancer Data Base: What is the NCDB? http://www.facs.org/cancer/ncdb/ncdbabout.html). Accessed May 1, 2006 15. Mettlin CJ, Menck HR, Winchester DP, et al: A comparison of breast, colorectal, lung, and prostate cancers reported to the National Cancer Data Base and the Surveillance, Epidemiology, and End Results Program. Cancer 79:2052-2061, 1997[CrossRef][Medline] 16. American Joint Committee on Cancer: AJCC Cancer Staging Manual (ed 6). New York, NY, Springer-Verlag, 2002 17. Bill-Axelson A, Holmberg L, Ruutu M, et al: Radical prostatectomy versus watchful waiting in early prostate cancer. N Engl J Med 352:1977-1984, 2005 18. Spencer BA, McGlynn EA, Steinberg M, e al: Quality of care indicators for early stage prostate cancer. Presented at the Annual Meeting of the Am Urol Assoc, San Antonio, TX, May 21-26, 2005 19. MacLean CH, Louie R, Shekelle PG, et al: Comparison of administrative data and medical records to measure the quality of medical care provided to vulnerable older patients. Med Care 44:141-148, 2006[CrossRef][Medline] 20. Wei JT, Dunn RL, Sandler HM, et al: Comprehensive comparison of health-related quality of life after contemporary therapies for localized prostate cancer. J Clin Oncol 20:557-566, 2002 21. Allison JJ, Kiefe CI, Weissman N, et al: Quality of care at teaching and nonteaching hospitals. JAMA 284:2994-2995, 2000 22. Ayanian JZ, Weissman JS: Teaching hospitals and quality of care: A review of the literature. Milbank Q 80:569-593, 2002[CrossRef][Medline] 23. Jha AK, Li Z, Orav EJ, et al: Care in U.S. hospitals: The Hospital Quality Alliance program. N Engl J Med 353:265-274, 2005 24. Lu-Yao GL, McLerran D, Wasson J, et al: An assessment of radical prostatectomy: Time trends, geographic variation, and outcomes—The Prostate Patient Outcomes Research Team. JAMA 269:2633-2636, 1993 25. Spencer BA, Fung CH, Wang M, et al: Geographic variation across veterans affairs medical centers in the treatment of early stage prostate cancer. J Urol 172:2362-2365, 2004[CrossRef][Medline] 26. Fisher ES, Wennberg JE: Health care quality, geographic variations, and the challenge of supply-sensitive care. Perspect Biol Med 46:69-79, 2003[Medline] 27. Litwin MS, Hays RD, Fink A, et al: Quality-of-life outcomes in men treated for localized prostate cancer. JAMA 273:129-135, 1995 28. Schriger DL, Baraff LJ, Rogers WH, et al: Implementation of clinical guidelines using a computer charting system: Effect on the initial care of health care workers exposed to body fluids. JAMA 278:1585-1590, 1997 29. Asch SM, Kerr EA, Keesey J, et al: Who is at greatest risk for receiving poor-quality health care? N Engl J Med 354:1147-1156, 2006 30. McGlynn EA, Asch SM, Adams J, et al: The quality of health care delivered to adults in the United States. N Engl J Med 348:2635-2645, 2003 Submitted July 2, 2007; accepted February 25, 2008.
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2008 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
TOP
ABSTRACT
INTRODUCTION
70 Gy (OR, 3.1; 95% CI, 1.6 to 6.1). Teaching/research hospitals and Comprehensive Cancer Centers had higher compliance than Community Cancer Centers, whereas racial differences were not observed for any indicator. 
statistic > 0.95).
