Correspondence
Quality of Health Care Delivered to Adults in the United States
N Engl J Med 2003; 349:1866-1868November 6, 2003
- Article
To the Editor:
In the study by McGlynn et al. (June 26 issue),1 participants received only about half of recommended care. It may be possible to double efficiency by working harder and smarter, as Steinberg suggests in his accompanying editorial,2 but this is unlikely. Physicians would need 7.4 hours per working day to fully satisfy all the recommendations of the U.S. Preventive Services Task Force.3 This is the time to be allotted to prevention only. Now add the time needed for other tasks, such as care of the sick, continuing medical education, and administration. Completing all these tasks according to recommendation would probably require working more than 15 to 20 hours per working day. It follows that eight hours per working day would allow physicians to satisfy only half of the current recommendations, as suggested by the findings of McGlynn et al.
Recommendations actually compete for limited resources, but only exceptionally do they include measures of feasibility in terms of work force, time, or money available. The cost of implementing the current recommendations probably exceeds real-world budgets. Therefore, the current recommendations need readjustment: they need to account for what can be done and what should have priority.
3 ReferencesMichael Spiegel, M.D.
Christoph Pechlaner, M.D.
University Hospital, A-6020 Innsbruck, Austria
christoph.pechlaner@uibk. ac. at 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 2003;348:2635-2645
Full Text | Web of Science | Medline2
Steinberg EP . Improving the quality of care -- can we practice what we preach? N Engl J Med 2003;348:2681-2683
Full Text | Web of Science | Medline3
Yarnall KS ,Pollak KI ,Ostbye T ,Krause KM ,Michener JL . Primary care: is there enough time for prevention? Am J Public Health 2003;93:635-641
CrossRef | Web of Science | Medline
To the Editor:
We concur with McGlynn et al. that the findings of their study are reason for concern, but we challenge the authors to assess more directly the degree of underreporting involved. With only 22 of 439 indicators receiving verbal confirmation, the extent to which the documented care reflected actual care is unknown. The discrepant findings between the self-reported and documented rates of influenza vaccination (85 percent vs. 15 percent) suggest that underreporting for some indicators may have far exceeded the 10 percentage points estimated by the authors. It would have been helpful to compare all 22 verbally confirmed indicators with data obtained through chart abstraction. In addition, how many of the other 417 indicators (e.g., alcohol counseling) might have been vulnerable to underreporting because of limitations inherent in the standard medical record?
Our concern is that poor performance may have been overestimated with use of the authors' methods. Overestimation is especially likely in domains that required specific provider notations in the chart. The potential effects of this study mean that estimates of substandard performance must be as accurate as possible and threats to validity rigorously explored by the best methods available.
J. Richard Goss, M.D., M.P.H.
Joann G. Elmore, M.D., M.P.H.
Daniel S. Lessler, M.D., M.H.A.
University of Washington, Seattle, WA 98104
rgoss@u.washington. edu To the Editor:
McGlynn et al. acknowledge that the results of their study were biased because they obtained data on only 33.5 percent of the study population. If the actual prevalence of optimal care among nonparticipants was 90 percent, the quality score for the entire study population would have been 77 percent; if the prevalence was 10 percent, the score would have been 27 percent. Adjustments for the differences between the participants and nonparticipants with respect to age, sex, race, education, income, health care utilization, and health status may not have been valid because these factors were not established to be associated with quality scores. Experience shows that participants are not representative of nonparticipants because of differences in the prevalence of the risk factors for the outcome,1 even when the response rate is as high as 82.5 percent.2
Finally, although the study subjects were selected from 12 metropolitan areas, the authors generalized the results to the entire U.S. population. Several studies have shown significant urban–rural differences in the quality of health care.3
(The views expressed in this letter do not represent the views or policies of the Centers for Medicare and Medicaid Services or those of the U.S. government.)
3 ReferencesKazim Sheikh, M.D.
Department of Health and Human Services, Kansas City, MO 641061
Sheikh K . Predicting risk among non-respondents in prospective studies. Eur J Epidemiol 1986;2:39-43
CrossRef | Medline2
Sheikh K ,Mattingly S . Investigating non-response bias in mail surveys. J Epidemiol Community Health 1981;35:293-296
CrossRef | Web of Science | Medline3
Sheikh K ,Bullock C . Urban-rural differences in the quality of care for Medicare patients with acute myocardial infarction. Arch Intern Med 2001;161:737-743
CrossRef | Web of Science | Medline
To the Editor:
In his excellent editorial on improving the quality of health care delivery, Steinberg suggests four actions that are likely to have the greatest chance of success. To these four, I would suggest adding two: reducing unnecessary documentation and coding burdens and making care fulfillment easier.
Quality care depends on having relevant information at the point of care and is made easier and better by systems integration of point-of-care tools. But quality care also requires dedicated time and focus. Physicians often spend more time on the form and formatting of progress notes than on actual patient care. Documentation-based reimbursement, inappropriate coding demands, inefficient referral systems, and paper-based formularies all detract from the time that should be spent with patients.
Rather than layering quality mandates on top of these process frictions, a more successful strategy would be first to reduce these regulatory burdens and to implement information-technology applications that make the delivery of care easier. This incremental approach would create the time and space necessary for enhancing the quality of medical care as well as improving physicians' compliance.
Peter Basch, M.D.
MedStar Health, Washington, DC 20003
peter.basch@medstar. net Author/Editor Response
Spiegel and Pechlaner highlight the challenges of providing all needed services under the current organization of the U.S. health care system. We agree that future guidelines should set priorities for what is done during individual physician–patient encounters. However, broad improvements in quality may require substantial reorganization of the health care system, including infrastructure support, care provided by teams of health professionals, and group visits.
Goss and colleagues raise issues related to the effect of medical-record documentation on quality. We specifically included in our survey questions about the areas in which we had the greatest concern about medical-record documentation, to help ensure an accurate evaluation of those areas. Most of the 22 indicators whose scoring was augmented by respondents' reports were not as disparate with respect to scores from each source of data as the example we selected for our article. Previous studies have found both overreporting and underreporting in the medical record, with potentially offsetting effects in aggregate scores.1,2 Finally, our expert panel deliberately selected indicators for which the failure to document care was itself a quality problem.
Although we acknowledge in our article that nonresponse rates are an important limitation in our study, we disagree with Dr. Sheikh that the nonresponse factors for which we adjusted (patients' demographic characteristics and utilization patterns) are not related to quality. Furthermore, we are unaware of any evidence that the providers used by nonrespondents in research projects are consistently and systematically different from the providers used by participants in such studies.
Combined information from respondents and medical records allowed us to produce a comprehensive, national estimate of quality consistent with that in previous studies.3 Research on quality is challenging because we lack a national health-information infrastructure that would allow us to estimate quality performance at all levels of the health care system without obtaining written permission — an approach that is increasingly difficult to use. We support substantial investments to enhance the routine availability of such data while protecting patients' privacy, so that progress in improving the U.S. health care system is not held up while we debate whether or not research findings showing that there is a chasm between what works and what we do are sufficient evidence for action.
3 ReferencesElizabeth A. McGlynn, Ph.D.
RAND Health, Santa Monica, CA 90401
mcglynn@rand.org Steven M. Asch, M.D., M.P.H.
Veterans Affairs Greater Los Angeles Health System, Los Angeles, CA 90073Eve A. Kerr, M.D., M.P.H.
Veterans Affairs Center for Practice Management and Outcomes Research, Ann Arbor, MI 481051
Peabody JW ,Luck J ,Glassman P ,Dresselhaus TR ,Lee M . Comparison of vignettes, standardized patients, and chart abstraction: a prospective validation of 3 methods for measuring quality. JAMA 2000;283:1715-1722
CrossRef | Web of Science | Medline2
Dresselhaus TR ,Luck J ,Peabody JW . The ethical problem of false positives: a prospective evaluation of physician reporting in the medical record. J Med Ethics 2002;28:291-294
CrossRef | Medline3
Institute of Medicine. Crossing the quality chasm: a new health system for the 21st century. Washington, D.C.: National Academy Press, 2001.
Author/Editor Response
Drs. Spiegel and Pechlaner and Dr. Basch point out that limitations in the time physicians have available to care for patients probably contribute to noncompliance with recommended care. Without question, physicians today spend substantial time on administrative tasks that are frustrating and that make poor use of their time. Efforts to reduce this administrative burden and to develop more efficient ways to deal with it should be strongly encouraged. One example of such an effort is the Institute for Healthcare Improvement's Idealized Design of Clinical Office Practice initiative, which has already achieved some documented success.1 Other strategies for improving office efficiency that have been found to be successful in some settings should be considered for broader implementation as well.2 Patients would suffer unnecessarily, however, if quality-improvement efforts were deferred until administrative hassles had been eliminated.
Drs. Spiegel and Pechlaner rightly point out that compliance with different recommended clinical practices have varying effects on patient outcomes. Quality-improvement efforts should therefore be prioritized according to the clinical benefits of improved compliance with particular practice recommendations as well as the time and other resources required to increase this compliance. It is likely, however, that compliance with many recommended clinical practices can be increased without an incremental expenditure of physicians' time, particularly if information technology is used to support their practice.
2 ReferencesEarl P. Steinberg, M.D., M.P.P.
Resolution Health, San Jose, CA 95126
esteinberg@resolutionhealth.com 1
Idealized design of clinical office practices. Boston: Institute for Healthcare Improvement, 2003. (Accessed October 16, 2003, at http://www.ihi.org/idealized/idcop/index.asp.)
2
Office practices: case for improvement. Boston: Institute for Healthcare Improvement, 2003. (Accessed October 16, 2003, at http://www.qualityhealthcare.org/QHC/Topics/OfficePractices/.)
- Citing Articles (3)
Citing Articles
1
Frederick L. Greene, Sharon Gilkerson, Paige Tedder, Kathy Smith. (2009) The role of the hospital registry in achieving outcome benchmarks in cancer care. Journal of Surgical Oncology 99:8, 497-499
CrossRef2
Robert E. White. (2008) Health Information Technology Will Shift the Medical Care Paradigm. Journal of General Internal Medicine 23:4, 495-499
CrossRef3
Blumenthal, David, . (2004) Doctors and Drug Companies. New England Journal of Medicine 351:18, 1885-1890
Full Text
