|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
Originally published as JCO Early Release 10.1200/JCO.2005.03.6111 on November 21 2005 © 2005 American Society of Clinical Oncology.
Using Practice Guidelines to Assess Cancer Care QualityThe Centers for Medicare and Medicaid Services, Washington, DC Healthcare provided to patients in the United States neither parallels long-established best practices nor takes rapid advantage of important biomedical advances. The widely cited Community Quality Index study estimated the global gap between what is optimal and what actually occurs in practice.1 The investigators studied 439 separate processes of care covering both prevention and treatment of acute and chronic conditions, and found that only about half of the care patients received constituted best practices. This finding was relatively consistent across all measures. In this issue, Harlan et al2 conduct a similar analysis of cancer care, using data from the National Cancer Institute's Patterns of Care study. They report on whether the initial treatment of patients with a broad range of cancer diagnoses follows widely accepted practice guidelines. In the study by Harlan et al, about two thirds of cancer patients receive care that parallels practice guidelines. The study by Harlan et al2 faced some methodologic challenges that are worth noting. The treatments provided to patients from 1995 to 1999 were compared with practice guidelines from a different era (2002); some practice standards must have changed during that time interval. In addition, it is always difficult to identify treatments provided to patients through chart review, and it is challenging to identify patients who refused treatment. I suspect that both of these limitations led Harlan et al to underestimate the percentage of patients who received guideline-based treatment. The study was also limited explicitly, in that it focused only on initial treatment of newly diagnosed patients. To capture more fully the quality of cancer care, all treatment courses should be scrutinized. A sizable share of cancer treatment is not initial treatment, but rather subsequent treatment (such as treatments after recurrence and second-line treatments). These minor issues do not reduce the importance of this study, which characterizes the shortfall between what could be provided to cancer patients in optimal circumstances and what occurs in practice. The sizeable gap the authors describe is made more important because of cancer's high prevalence, and can reasonably be posited to be responsible for a large number of avoidable cancer deaths. Physicians, patients, and payers have all taken stock of studies such as this one, and have come to a common set of conclusions. The quality of care is too low as a result of care system fragmentation and a poorly designed payment system. For instance, patients' medical records cannot be exchanged easily between physicians or care locations, leading to redundant care, elevated costs, and errors. Doctors and hospitals are paid for delivering services (ie, fee for service)—not for the quality of the care they provide or the outcomes their patients experience. However, this common consensus on the problems is not mirrored by consensus on the correct solutions. The one exception is that providers, payers, and patients all agree that we need to find ways of assessing care quality routinely. Some working models exist. Health plans have been reporting on their quality through the Health Plan Employer Data and Information Set measures for years. Hospitals and nursing homes more recently started reporting on specific processes of care in a uniform fashion. (To review the quality measures for individual US hospitals, refer to www.hospitalcompare.hhs.gov). Recent articles in the New England Journal of Medicine suggest that these quality reporting activities may be improving quality. One study found that hospitals that reported on the quality of care they delivered as part of the Joint Commission on Hospital Accreditation certification showed improvement across multiple measures over time.3 Another study examined Medicare managed care plans, and found that across multiple Health Plan Employer Data and Information Set measures, there were steady increases in performance over time, with concomitant reductions in treatment disparities.4 Interestingly, this latter study was published with two other articles that examined the same time period in the broader healthcare system (where quality measurement is not widespread). In these latter studies, there were neither meaningful improvements in quality nor reduction in treatment disparities.5,6 The differences in findings across these studies are consistent with the hypothesis that quality measurement and reporting improves quality. The study by Harlan et al2 suggests one way that a quality measurement and reporting system in cancer could be developed: track adherence to practice guidelines. This approach would be different from what has been proposed more generally in ambulatory care, which is to use individual quality measures. For instance, one accepted quality indicator is "Do patients with a heart attack receive a beta-blocker?" An analog to this in cancer might be "Are patients with node-positive colon cancer or stage 2/3 rectal cancer offered adjuvant chemotherapy?" (This is a measure incorporated into the American Society of Clinical Oncology's groundbreaking Quality Oncology Practice Initiative, which includes many other practice level measures as well.7) These two measures have similarities and also important differences. Regarding the similarities, both the beta-blocker and the adjuvant chemotherapy measures address important questions in their respective diseases, and both are supported by high-quality evidence, which demonstrates that the two interventions yield roughly similar reductions in risk of death. But the differences are important. Myocardial infarctions are frequent, and beta-blockers are a mainstay of therapy, making beta-blocker use itself relevant to the public health. Fewer than 5% of all patients diagnosed with cancer have node-positive colon cancer or stage 2/3 rectal cancer, so the adjuvant chemotherapy measure addresses only a small fraction of cancer care (and only addresses initial treatment of this subgroup). In addition, unlike beta-blockers, the recommended adjuvant therapy in colorectal cancer changes frequently because of scientific advances. Thus, in cancer, the number of individual quality measures that would be needed to characterize a meaningful fraction of cancer care, including treatments after disease recurrence or progression, would be very large indeed, and would require frequent updating. By using guidelines as the quality benchmark in their study, Harlan et al2 overcame the first of these obstacles: they were able to assess initial treatments across a broad range of cancers. In fact, there were guidelines relevant to the initial treatment of all of the cancers included in the Surveillance, Epidemiology, and End Results patterns of care study, suggesting that a sizable fraction of cancer care could be evaluated. Buttressing this point, the National Comprehensive Cancer Network reports that their guidelines alone cover the majority of all clinical scenarios encountered in cancer, and together with practice guidelines from other sources, such as Cancer Care Ontario and the American Society of Clinical Oncology, coverage should be near complete. Addressing the second concern, the guidelines emanating from these groups are updated with sufficient frequency to keep them abreast of scientific progress. However, using guideline adherence to judge cancer quality is an approach that should be pursued cautiously. The analyses by Harlan et al2 suggest that adherence to guidelines is probably a valid indicator of quality in many situations, given that rates of adherence track well with demographic variables that are usually associated with care quality, such as patient race, education, and marital status. But guideline adherence in the study by Harlan et al varied greatly across cancers, raising some concerns about the guidelines themselves. For instance, adherence to guidelines was extraordinarily high for testicular cancer, suggesting that the guidelines are up to date, the recommendations are based on high-quality evidence, and the vast majority of treating physicians concur with them. In testicular cancer, this finding is not surprising, given the success rate of optimal treatment. In contrast, the rate of adherence to guidelines for ovarian cancer, cervical cancer, and rectal cancer were very low. It is not reasonable to assume that physicians simply are less careful in their treatment decisions for these patients, or that patients with these diseases are more reluctant to accept recommended therapies. Rather, these findings should provoke questions about the guidelines. Are the guidelines out of sync with community standards, and if so, why? Is the evidence supporting the guidelines of insufficient quality, or are variations in community standards not properly appreciated? Alternatively, are there health system barriers that prevent the recommendations from being followed? The low rates of adherence to guidelines in rectal cancer treatment in this study are an example. Are they due to low rates of physician acceptance or knowledge about the recommendations, low rates of patient acceptance, or low rates of availability of radiation facilities themselves? Knowing the answer to these questions should help us learn what should be done to improve rectal cancer outcomes. In other words, comparisons between actual care and practice guidelines might yield interpretable results about care quality. But those comparisons must also be a vehicle for identifying where there is clinical consensus, identifying where evidence is lacking, and identifying systemic barriers to high-quality care. Assessing adherence to guidelines could lead to quality measurement where there is consensus, and research where there is not. Such a process could be sufficiently flexible to allow the rapid changes in practice standards, and allow for more rapid identification of practical clinical questions that need more evaluation. There is a central purpose to such an effort. For many of the cancers Harlan et al2 examined, there was a pattern of poor-quality care that needs to be understood and addressed. As in many other studies, the poor, the underinsured, and racial minorities were the worst off.8 Studies have suggested that a way to achieve improvement, and also to narrow some of the gaps in quality, is through routine quality measurement. The study by Harlan et al 2 suggests evaluating adherence to practice guidelines is an approach with some promise, with the caveat that the quality of the guidelines themselves must always be considered. Nonadherence may tell us about quality shortfalls, but also important information about the extent of consensus on best practices, the availability of resources to treat cancer patients, and areas where guidelines might need improvement through research. Author's Disclosures of Potential Conflicts of Interest The author indicated no potential conflicts of interest. Acknowledgment The author is Senior Adviser to the Administrator of the Centers for Medicare and Medicaid Services (CMS). The views expressed in this editorial are his, and do not necessarily reflect the views of the CMS. REFERENCES
1. 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 2. Harlan L, Greene AL, Limin X, et al: Insurance status and the use of guideline therapy in the treatment of selected cancers. J Clin Oncol 23:9079-9088, 2005 3. Williams SC, Schmaltz SP, Morton DJ, et al: Quality of care in U.S. hospitals as reflected by standardized measures, 2002-2004. N Engl J Med 353:255-264, 2005 4. Trivedi AN, Zaslavsky AM, Schneider EC, et al: Trends in the quality of care and racial disparities in Medicare managed care. N Engl J Med 353:692-700, 2005 5. Jha AK, Fisher ES, Li Z, et al: Racial trends in the use of major procedures among the elderly. N Engl J Med 353:683-691, 2005 6. Vaccarino V, Rathore SS, Wenger NK, et al: Sex and racial differences in the management of acute myocardial infarction, 1994 through 2002. N Engl J Med 353:671-682, 2005 7. Neuss MN, Desch CE, McNiff KK, et al: A process for measuring the quality of cancer care: The Quality Oncology Practice Initiative. J Clin Oncol 23:6233-6239, 2005 8. Smedley BD, Stith AY, Nelson AR (eds; Committee on Understanding and Eliminating Racial and Ethic Dispartirs inHealth Care): Unequal treatment: Confronting racial and ethnic disparities in health care. Washington, DC, National Academy Press, 2002 Related Article
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2005 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
