Quality of Care in U.S. Hospitals
http://www.100md.com
《新英格兰医药杂志》
To the Editor: Williams et al.1 and Jha et al.2 (July 21 issue) report improvement but variation among hospitals in a study mandated by the Joint Commission on Accreditation of Healthcare Organizations of measures of the quality of care for acute myocardial infarction, heart failure, and pneumonia. In an accompanying editorial, Romano3 speculates that some measures with major improvements, such as smoking-cessation counseling and discharge instructions, may have been "gamed," and the evidence for these measures as well as for pneumococcal vaccination is weak, particularly with regard to the elderly.4 Not surprisingly, mortality after acute myocardial infarction was unchanged, and the incidence of pneumonia, smoking-cessation rates, and even a correlation between performance and outcomes were absent from the articles. Resources spent on improving measures of doubtful efficacy come at the expense of interventions that improve the quality of care, and measuring performance or ranking hospitals with the use of these measures is misleading.
Richard A. Robbins, M.D.
Carl T. Hayden Veterans Affairs Medical Center
Phoenix, AZ 85012
richard.robbins2@med.va.gov
Stephen A. Klotz, M.D.
University of Arizona Health Sciences Center
Tucson, AZ 85724
References
Williams SC, Schmaltz SP, Morton DJ, Koss RG, Loeb JM. Quality of care in U.S. hospitals as reflected by standardized measures, 2002-2004. N Engl J Med 2005;353:255-264.
Jha AK, Li Z, Orav EJ, Epstein AM. Care in U.S. hospitals -- the Hospital Quality Alliance Program. N Engl J Med 2005;353:265-274.
Romano PS. Improving the quality of hospital care in America. N Engl J Med 2005;353:302-304.
Jackson LA, Neuzil KM, Yu O, et al. Effectiveness of pneumococcal polysaccharide vaccine in older adults. N Engl J Med 2003;348:1747-1755.
To the Editor: Instituting performance measures changes our health care system, but will such measures improve health care? The doubt comes because, even though Williams and colleagues state that "the process measures were selected because they had a scientific evidence base, established through randomized clinical trials," this is not true for the pneumonia-related measures.
One example is the practice of obtaining blood cultures before starting antibiotic therapy. This has been recommended in an article on practice guidelines1 and an editorial,2 but the supporting reference3 is a study of bacteremia showing that blood cultures had a "limited effect on antibiotic choice." Attempts to satisfy this criterion could lead hospitals to delay treatment with antibiotics; few would agree that this would enhance care.
I strongly support the practice of drawing blood for cultures before starting antibiotics. What I oppose is using expert opinion rather than scientific studies to drive hospitals to expend substantial resources to meet measures that may only make us feel good about ourselves. Is this why, in the words of Williams et al., there is "an apparent contradiction between improvement in the process measures and the lack of improvement on the inpatient-death measure"?
Bradley S. Bender, M.D.
North Florida–South Georgia Veterans Health System
Gainesville, FL 32608
bradley.bender@med.va.gov
References
Bartlett JG, Dowell SF, Mandell LA, File TM Jr, Musher DM, Fine MJ. Practice guidelines for the management of community-acquired pneumonia in adults. Clin Infect Dis 2000;31:347-382.
Mandell LA. Guidelines for community-acquired pneumonia: a tale of two countries. Clin Infect Dis 2000;31:422-425.
Arbo MDJ, Snydman DR. Influence of blood culture results on antibiotic choice in the treatment of bacteremia. Arch Intern Med 1994;154:2641-2645.
To the Editor: Williams et al. report that, over a two-year period, U.S. hospitals improved their performance on 15 out of 18 quality measures for three conditions. They state that "hospitals that began the study with lower baseline rates tended to improve at faster rates than hospitals with higher baseline rates." However, Table 3 of the article indicates that hospitals in the upper quartile at baseline consistently did worse, not better; all but 1 of the 17 indicators worsened, some minimally but some substantially.
There are a number of possible explanations for this. Regression to the mean probably occurred at both ends of the rankings. For measures near 100 percent, improvement is difficult to impossible to achieve. Complacency with good performance and fatigue as the "newness" of a quality-improvement initiative fades or additional initiatives appear could undercut continued high performance.
Although the overall performance trends are salutary, the findings for the top quartile reinforce the need for caution in the use of performance rankings and indicate a need to study the factors underlying the apparent decline in performance of the "top performers."
Barry G. Saver, M.D., M.P.H.
University of Washington
Seattle, WA 98195-4696
saver@u.washington.edu
The authors reply: We agree with Dr. Saver's conclusion that the apparent decline in performance of hospitals that began the study in the top quartile of performance was largely accounted for by regression to the mean. This trend may have been further influenced by the inclusion of hospitals with small sample sizes, since these hospitals tend to have more extreme values and greater variation in their levels of performance than hospitals with larger sample sizes. After accounting for variability among hospitals, we determined that the decline in the top quartile was not statistically significant. We share Dr. Saver's concern that "complacency with good performance and fatigue . . . could undercut continued high performance." As measurement and quality-improvement initiatives move forward, it will be critical to monitor performance continuously and raise the alarm when performance begins to falter.
Drs. Robbins and Klotz and Dr. Bender express concern about the evidence base supporting the use of several measures. Although the accumulated scientific knowledge that supports the measures is dynamic and evolutionary, there is ample support for each of the measures from either clinical trials or expert consensus and clinical practice guidelines produced by specialty societies. Guidelines of the American Thoracic Society and the Infectious Diseases Society of America for the management of community-acquired pneumonia emphasize the importance of obtaining blood cultures before the administration of antibiotics1 and the value of pneumococcal vaccination. Clinical practice guidelines of the American College of Cardiology emphasize the importance of patient education in the treatment of heart failure.2 Data supporting the benefits of smoking-cessation counseling in the inpatient setting are well established.3 Therefore, although many of the measures we examined are supported by data from clinical trials, others are supported by clinical practice guidelines and observational studies and should be part of the practice of caring for patients who have the three conditions examined. Hospitals should not have to choose between obtaining blood cultures before antibiotics are administered and providing antibiotics in a timely manner.
Drs. Robbins and Klotz and Dr. Bender also suggest that the lack of statistically significant improvement on the inpatient mortality measure due to acute myocardial infarction reflects on the usefulness of the process measures. As stated in one of our articles, we do not believe that it is reasonable to expect a direct relationship between process measures that address actions taken at discharge and a measure of inpatient mortality. Furthermore, given the many clinical practices affecting patient outcomes and the imperfection of risk-adjustment methods, the lack of improvement in inpatient mortality for acute myocardial infarction does not detract from the value of these clinical processes. We recognize that no performance measure is perfect in its design or implementation, and we believe that most readers accept the current state of hospital-performance measurement as a good beginning to a very long journey.
Scott C. Williams, Psy.D.
Jerod M. Loeb, Ph.D.
Joint Commission on Accreditation of Healthcare Organizations
Oakbrook Terrace, IL 60181
Ashish K. Jha, M.D., M.P.H.
Arnold M. Epstein, M.D.
Harvard School of Public Health
Boston, MA 02115
References
Metersky ML, Ma A, Bratzler DW, Houck PM. Predicting bacteremia in patients with community-acquired pneumonia. Am J Respir Crit Care Med 2004;169:342-347.
ACC/AHA 2005 guideline update for the diagnosis and management of chronic heart failure in the adult. Bethesda, Md.: American College of Cardiology Foundation, 2005. (Accessed October 6, 2005, at http://www.acc.org/clinical/guidelines/failure/index.pdf.)
Fiore MC, Bailey WC, Cohen SJ, et al. Treating tobacco use and dependence. Clinical practice guideline no. 18. Rockville, Md.: Public Health Service, June 2000.
Richard A. Robbins, M.D.
Carl T. Hayden Veterans Affairs Medical Center
Phoenix, AZ 85012
richard.robbins2@med.va.gov
Stephen A. Klotz, M.D.
University of Arizona Health Sciences Center
Tucson, AZ 85724
References
Williams SC, Schmaltz SP, Morton DJ, Koss RG, Loeb JM. Quality of care in U.S. hospitals as reflected by standardized measures, 2002-2004. N Engl J Med 2005;353:255-264.
Jha AK, Li Z, Orav EJ, Epstein AM. Care in U.S. hospitals -- the Hospital Quality Alliance Program. N Engl J Med 2005;353:265-274.
Romano PS. Improving the quality of hospital care in America. N Engl J Med 2005;353:302-304.
Jackson LA, Neuzil KM, Yu O, et al. Effectiveness of pneumococcal polysaccharide vaccine in older adults. N Engl J Med 2003;348:1747-1755.
To the Editor: Instituting performance measures changes our health care system, but will such measures improve health care? The doubt comes because, even though Williams and colleagues state that "the process measures were selected because they had a scientific evidence base, established through randomized clinical trials," this is not true for the pneumonia-related measures.
One example is the practice of obtaining blood cultures before starting antibiotic therapy. This has been recommended in an article on practice guidelines1 and an editorial,2 but the supporting reference3 is a study of bacteremia showing that blood cultures had a "limited effect on antibiotic choice." Attempts to satisfy this criterion could lead hospitals to delay treatment with antibiotics; few would agree that this would enhance care.
I strongly support the practice of drawing blood for cultures before starting antibiotics. What I oppose is using expert opinion rather than scientific studies to drive hospitals to expend substantial resources to meet measures that may only make us feel good about ourselves. Is this why, in the words of Williams et al., there is "an apparent contradiction between improvement in the process measures and the lack of improvement on the inpatient-death measure"?
Bradley S. Bender, M.D.
North Florida–South Georgia Veterans Health System
Gainesville, FL 32608
bradley.bender@med.va.gov
References
Bartlett JG, Dowell SF, Mandell LA, File TM Jr, Musher DM, Fine MJ. Practice guidelines for the management of community-acquired pneumonia in adults. Clin Infect Dis 2000;31:347-382.
Mandell LA. Guidelines for community-acquired pneumonia: a tale of two countries. Clin Infect Dis 2000;31:422-425.
Arbo MDJ, Snydman DR. Influence of blood culture results on antibiotic choice in the treatment of bacteremia. Arch Intern Med 1994;154:2641-2645.
To the Editor: Williams et al. report that, over a two-year period, U.S. hospitals improved their performance on 15 out of 18 quality measures for three conditions. They state that "hospitals that began the study with lower baseline rates tended to improve at faster rates than hospitals with higher baseline rates." However, Table 3 of the article indicates that hospitals in the upper quartile at baseline consistently did worse, not better; all but 1 of the 17 indicators worsened, some minimally but some substantially.
There are a number of possible explanations for this. Regression to the mean probably occurred at both ends of the rankings. For measures near 100 percent, improvement is difficult to impossible to achieve. Complacency with good performance and fatigue as the "newness" of a quality-improvement initiative fades or additional initiatives appear could undercut continued high performance.
Although the overall performance trends are salutary, the findings for the top quartile reinforce the need for caution in the use of performance rankings and indicate a need to study the factors underlying the apparent decline in performance of the "top performers."
Barry G. Saver, M.D., M.P.H.
University of Washington
Seattle, WA 98195-4696
saver@u.washington.edu
The authors reply: We agree with Dr. Saver's conclusion that the apparent decline in performance of hospitals that began the study in the top quartile of performance was largely accounted for by regression to the mean. This trend may have been further influenced by the inclusion of hospitals with small sample sizes, since these hospitals tend to have more extreme values and greater variation in their levels of performance than hospitals with larger sample sizes. After accounting for variability among hospitals, we determined that the decline in the top quartile was not statistically significant. We share Dr. Saver's concern that "complacency with good performance and fatigue . . . could undercut continued high performance." As measurement and quality-improvement initiatives move forward, it will be critical to monitor performance continuously and raise the alarm when performance begins to falter.
Drs. Robbins and Klotz and Dr. Bender express concern about the evidence base supporting the use of several measures. Although the accumulated scientific knowledge that supports the measures is dynamic and evolutionary, there is ample support for each of the measures from either clinical trials or expert consensus and clinical practice guidelines produced by specialty societies. Guidelines of the American Thoracic Society and the Infectious Diseases Society of America for the management of community-acquired pneumonia emphasize the importance of obtaining blood cultures before the administration of antibiotics1 and the value of pneumococcal vaccination. Clinical practice guidelines of the American College of Cardiology emphasize the importance of patient education in the treatment of heart failure.2 Data supporting the benefits of smoking-cessation counseling in the inpatient setting are well established.3 Therefore, although many of the measures we examined are supported by data from clinical trials, others are supported by clinical practice guidelines and observational studies and should be part of the practice of caring for patients who have the three conditions examined. Hospitals should not have to choose between obtaining blood cultures before antibiotics are administered and providing antibiotics in a timely manner.
Drs. Robbins and Klotz and Dr. Bender also suggest that the lack of statistically significant improvement on the inpatient mortality measure due to acute myocardial infarction reflects on the usefulness of the process measures. As stated in one of our articles, we do not believe that it is reasonable to expect a direct relationship between process measures that address actions taken at discharge and a measure of inpatient mortality. Furthermore, given the many clinical practices affecting patient outcomes and the imperfection of risk-adjustment methods, the lack of improvement in inpatient mortality for acute myocardial infarction does not detract from the value of these clinical processes. We recognize that no performance measure is perfect in its design or implementation, and we believe that most readers accept the current state of hospital-performance measurement as a good beginning to a very long journey.
Scott C. Williams, Psy.D.
Jerod M. Loeb, Ph.D.
Joint Commission on Accreditation of Healthcare Organizations
Oakbrook Terrace, IL 60181
Ashish K. Jha, M.D., M.P.H.
Arnold M. Epstein, M.D.
Harvard School of Public Health
Boston, MA 02115
References
Metersky ML, Ma A, Bratzler DW, Houck PM. Predicting bacteremia in patients with community-acquired pneumonia. Am J Respir Crit Care Med 2004;169:342-347.
ACC/AHA 2005 guideline update for the diagnosis and management of chronic heart failure in the adult. Bethesda, Md.: American College of Cardiology Foundation, 2005. (Accessed October 6, 2005, at http://www.acc.org/clinical/guidelines/failure/index.pdf.)
Fiore MC, Bailey WC, Cohen SJ, et al. Treating tobacco use and dependence. Clinical practice guideline no. 18. Rockville, Md.: Public Health Service, June 2000.