|
| HOME | SUBSCRIPTIONS | CURRENT ISSUE | PAST ISSUES | CARDIOSOURCE | SEARCH | HELP | FEEDBACK |
|
|
|
|||||||||
|
|||||||||||
|
J Am Coll Cardiol, 1998; 32:993-999 © 1998 by the American College of Cardiology Foundation |
* Outcomes Research and Assessment Group, Duke Clinical Research Institute, Durham, North Carolina, USA
Division of Cardiology, Department of Medicine, Duke University Medical Center, Durham, North Carolina, USA
Division of Biometry, Department of Community and Family Medicine, Duke University Medical Center, Durham, North Carolina, USA
Manuscript received November 19, 1997; revised manuscript received June 2, 1998, accepted June 12, 1998.
Address for correspondence: Dr. Eric D. Peterson, Box 3236, Duke University Medical Center, Durham, North Carolina 27710
PETER016{at}MC.DUKE.EDU
 |
Abstract |
|---|
 Top Abstract MethodsResultsDiscussionReferences |
|---|
Background. Since 1989, New York (NY) has compiled provider-specific bypass surgery mortality reports. While some have proposed that "provider profiling" has led to lower surgical mortality rates, critics have suggested that such programs lower in-state procedural access (increasing out-of-state transfers) without improving patient outcomes.
Methods. Using national Medicare data, we examined trends in the percentages of NY residents aged 65 years or older receiving out-of-state bypass surgery between 1987 and 1992 (before and after program initiation). We also examined in-state procedure use among elderly myocardial infarction patients during this period. Finally, we compared trends in surgical outcomes in NY Medicare patients with those for the rest of the nation.
Results. Between 1987 and 1992, the percentage of NY residents receiving bypass out-of-state actually declined (from 12.5% to 11.3%, p < 0.01 for trend). An elderly patient’s likelihood for bypass following myocardial infarction in NY increased significantly since the program’s initiation. Between 1987 and 1992, unadjusted 30-day mortality rates following bypass declined by 33% in NY Medicare patients compared with a 19% decline nationwide (p < 0.001). As a result of this improvement, NY had the lowest risk-adjusted bypass mortality rate of any state in 1992.
Conclusions. We found no evidence that NY’s provider profiling limited procedure access in NY’s elderly or increased out-of-state transfers. Despite an increasing preoperative risk profile, procedural outcomes in NY improved significantly faster than the national average.
The impact of NY’s provider-profiling effort has been controversial. Hannan et al. (1) have reported that overall unadjusted and risk-adjusted bypass surgery mortality rates declined by 21% and 41%, respectively, between 1989 and 1992. They concluded that much of this improvement in procedural outcomes might be due to NY’s profiling efforts. In contrast, others suggested that the improved procedural outcomes in NY resulted from a change in patient selection rather than an actual improvement in quality of care (8). Surgeons may have deferred performing bypass on higher-risk patients either by transferring these patients to hospitals outside NY, or by directly refusing to offer the procedure in high-risk situations. Both of these behaviors have been anecdotally reported. Omoigui et al. (9) noted that patient transfers to the Cleveland Clinic from NY hospitals rose by 31% between 1989 and 1993. These NY transfers had generally higher risk profiles than patients transferred from Ohio or other states. Similarly, a New York Times reporter detailed how her father was refused open heart surgery in NY, presumably because of several surgeons’ concern for their own outcomes statistics (10). A third explanation for NY’s declining mortality was that it merely reflected concurrent national improvements in surgical outcomes. For example, Ghali et al. (11) recently reported that Massachusetts experienced declines in bypass surgery mortality similar to those reported in NY despite not having a statewide profiling effort.
Our study objectives were to use the national Medicare database to assess the impact of NY’s provider profiling program on procedure access and patient outcomes. First, we examined whether provider profiling had increased the percentage of NY’s elderly residents going out-of-state for bypass surgery. Second, we investigated whether the use of bypass surgery following myocardial infarction (MI) had declined in NY’s elderly since the initiation of provider profiling. Finally, we examined whether bypass surgery outcomes were improving more rapidly in NY than in the rest of the nation during this period.
|
Methods |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
Patient populations.
Our overall bypass surgery study population consisted of all Medicare patients age 
65 years who underwent bypass surgery (ICD-9-CM codes 36.10 through 36.19) between 1987 to 1992 in a U.S. hospital. To avoid counting patients twice, patients undergoing multiple bypass procedures during the study period were included only once, defined by the initial procedure. We also excluded Medicare patients who were under age 65 years, those Medicare benefits as part of the Railroad Retirement Board or because of end-stage renal disease.
We also examined trends in the use of bypass surgery during the initial episode of care (index hospitalization linked with between-hospital transfers) in higher risk elderly patient subsets who were admitted to NY hospitals with an acute MI. Patients were considered to have been admitted for an acute MI if their principal diagnosis was ICD-9-CM code 410, or, if they had a secondary diagnosis of MI and a principal diagnosis of a complication of infarction (such as papillary muscle rupture). We also examined bypass surgery use in very high risk MI subgroups including those with advanced age (e.g., 75 to 80 years) and those with other preoperative risk factors such as congestive heart failure, diabetes mellitus and renal insufficiency (2,12,13).
Data analysis. We compared demographic and clinical characteristics for patients undergoing bypass surgery in NY hospitals during the study period with those receiving procedures in the rest of the nation (U.S. non-NY). We also examined trends in the yearly number, rate and baseline characteristics for CABG procedures performed in NY hospitals. The yearly NY CABG rate was calculated by dividing the number of NY CABG procedures performed per year by the overall NY Medicare resident enrollee population (available from Medicare’s 100% Denominator File). Similar baseline comparisons were also made between NY residents receiving surgery in a NY hospital (in-state) and those receiving procedures in a hospital outside NY (out-of-state). The frequencies of coded comorbid diseases (for example, congestive heart failure or renal disease) were determined using ICD-9-CM applied to discharge abstract information (14). We also calculated a composite measure of comorbidity, the Charlson Index, applied to these ICD-9-CM diagnoses (14–16).
We tested the significance of temporal trends in patient characteristics and out-of-state procedure rates using simple linear regression models (for continuous variables) and logistic regression models (for categorical variables) with year as the independent variable. Comparisons of trends in bypass surgery use in NY vs. U.S. non-NY hospital MI patients were made using regression models which included procedure year, hospital location (NY vs. non-NY) and an interaction term (hospital location by year) as independent predictors and tested the significance of the interaction term. Comparison of temporal trends in unadjusted 30-day mortality rates in NY vs. U.S. non-NY hospitals was made in a similar fashion.
We used a logistic regression model to calculate the risk-adjusted yearly decline in the likelihood for 30-day mortality following bypass surgery for individual states. This model included preoperative risk factors (patient age, race, gender, acute MI admission and Charlson comorbidity index), hospital location (each state), procedure year (continuous variable) and the interaction terms (state by procedure year). For each state, the coefficient of (state by year) interaction term, when added to the coefficient for procedure year and exponentiated, represents the yearly odds of mortality for the particular state. This odds ratio was converted to a yearly decline by comparison to 1.0 (e.g., an odds ratio of 0.93 implies a decline of 7%).
Finally, risk adjusted mortality rates by state in 1992 were calculated by using a similar regression model after limiting the dataset to include 1992 procedures only (and excluding temporal factors). State-specific coefficients from this model (relative to the comparison state, NY) were then converted to odds ratios. These odds ratios were then multiplied by the NY mortality rate for 1992 to get state-specific, risk adjusted 1992 mortality rates. This method uses the NY rate as a standard to which all other state rates are compared (17).
|
Results |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
|
|
|
|
|
|
As a final method of examining the relationship between provider profiling programs and surgical outcomes, we displayed each state’s yearly decline in bypass mortality (from 1987 to 1992) by their ending (1992) procedural mortality rate (Fig. 4). In Figure 4 the cross-bars represent the national averages for these two measures, thus creating quadrants. If NY’s provider profiling program influenced surgical outcomes, NY’s surgical mortality rate should have declined faster than the majority of states (right half of the graph) and they should have one of the lowest ending mortality rates (bottom half of the graph). As is evident, NY was one of the nation’s most improved bypass performer’s between 1987 and 1992 and had the nation’s lowest surgical mortality rate in 1992. Interestingly, surgical outcomes in northern New England (Maine, Vermont and New Hampshire), another area which has implemented bypass surgery provider profiling programs, also declined faster than the national average and had lower than average mortality rates in 1992.
|
|
Discussion |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
Effects on physician behavior. Before public release, NY’s provider-specific mortality data are risk-adjusted for potential differences in severity of illness using a validated risk-prediction model (5). State officials maintain that the model adequately compensates a surgeon for taking on higher-risk patients and, if anything, overcompensates for the highest-risk cases (6). Despite this contention, many critics of the program have asserted that surgeons’ natural reaction to having their outcomes monitored is to refer or refuse those patients with higher procedural risk (8,18).
Recently, Omoigui et al. (9) published a single-institution review supporting the contention that NY surgeons were transferring their highest risk patients out-of-state. The authors reported that there was an increase in both the number of patients and their disease severity transferred to the Cleveland Clinic from NY for bypass between 1989 and 1993 (n = 482). In contrast to these single-center results, when using a national database we found no evidence for a widespread increase in out-of-state transfers from NY for bypass surgery. In our study, out-of-state bypass procedures may have occurred in one of three general situations: patients living near NY’s state boarder who cross-over to a neighboring state for a procedure; patients who are away from home when the need for bypass surgery occurs (e.g., vacation, dual homesite); and patients who are specifically transferred out-of-state for bypass surgery. While the "basal rates" for the two former situations should be relatively constant, if physicians were transferring a significant number of higher-risk elderly patients out-of-state for bypass, the overall rate should have increased over time. In fact, NY elderly residents were actually less likely to have their procedure performed outside NY after the profiling program was initiated. Additionally, while claims data provide only limited information on disease severity, we found no significant differences in age, gender, rates of acute MI or comorbid illness burden in NY residents receiving procedures out-of-state as compared with NY residents receiving in-state surgery. Given these national observations, it appears likely that the Cleveland Clinic’s experience was an exception resulting from particular local care patterns rather than the general rule.
New York’s provider profiling program may also have affected surgeons’ willingness to operate on higher-risk patients. While it is difficult to monitor changes in procedure access, we examined temporal trends in bypass surgery use among all elderly Medicare enrollees (Table 2) and among higher-risk elderly MI patients in NY (Fig. 2). Had NY surgeons become more conservative in their care, we would have expected to see a decline in bypass surgery use, particularly among these high-risk patients, corresponding to the initiation of the profiling program. Care patterns in the rest of the nation during this period served as a control for nonspecific changes. Consistent with previous studies, we noted that NY physicians used revascularization procedures less frequently in post-MI patients than physicians in other regions of the country (19,20). However, we also found that NY elderly high-risk MI patients were actually more likely to receive bypass surgery since the program’s initiation, mirroring national trends (Fig. 2). Thus, we were unable to detect systemic evidence for decreased access to bypass surgery in NY’s elderly since provider profiling began.
Mortality outcomes. The most important measure of any quality improvement effort is its effect on patient outcomes. Hannan et al. (1) previously reported that overall unadjusted and risk-adjusted in-hospital mortality rates in NY have declined significantly between 1989 and 1992. Our study confirmed that overall unadjusted bypass mortality rates in NY Medicare patients declined significantly faster than the rest of the nation from 1987 to 1992, 33% vs. 19% (p < 0.01 for trend comparisons). Furthermore, this decline in mortality was most profound in NY between 1989 to 1992, after the initiation of the profiling efforts and during a period when national mortality averages declined only slightly (Fig. 3).
We also found that risk-adjusted surgical mortality in NY improved faster than almost all individual states between 1987 to 1992 (Fig. 4). As a result, NY had the lowest risk-adjusted mortality rate in the country in 1992. It is also interesting to note that the other area of the country which initiated a provider profiling program during similar time periods, northern New England, experienced significantly larger improvements in their surgical outcomes than other areas of the country (Fig. 4) (21,22).
These results contrast with those recently reported by Ghali et al. (11). They found that Massachusetts (a state without provider profiling) experienced significant improvements in surgical outcomes from 1990 to 1994. While our study confirmed that bypass mortality improved in Massachusetts (a 25% decline in unadjusted Medicare procedural mortality between 1987 to 1992), this decline was significantly less than that experienced by either NY or northern New England. In part these differences in study results may be due to differences in the years of study and patient inclusion criteria (our study included all bypass surgery cases, but was limited to the elderly, while theirs included all age patients, but was limited to certain diagnosis-related groups). Finally, even if bypass outcomes in Massachusetts improved to a similar degree as in NY and northern New England, this may represent a "spill-over" of the effect of the quality improvement programs (from NY and northern New England to Massachusetts) resulting from its geographic location directly between two areas with statewide profiling programs.
If provider profiling has led to improved outcomes, the mechanism by which this program works remains unclear. While public disclosure of outcomes scorecards could have affected patients’ or physicians’ surgical referral decisions, NY State officials have not found evidence for mass patients migration from high- to low-mortality hospitals since the public release of provider profiles (5). Additionally, our study demonstrated that northern New England (which does not publicly disclose its provider’s outcome rating) achieved similar improvements in surgical outcomes to those in NY (21,22). Alternatively, provider profiling outcomes data may have been used internally by hospitals and physician to motivate internal, individualized, quality improvement initiatives. For example, Chassin et al. (5) have reported that certain NY hospitals restricted operating privileges for surgeons with poor outcomes while other institutions have improved their management of unstable patients. Malenka and O’Connor (22) similarly reported that outcomes feedback has promoted provider-initiated quality improvement programs in northern New England.
Limitations. While Medicare is currently the only data source available for examining national care trends, these data are limited in terms of the extent and accuracy of clinical information provided (23). This limitation prevented us from comparing fully risk-adjusted outcomes results. It is notable, however, that baseline characteristics for NY and U.S. non-NY bypass patients were strikingly similar, making significant differences in disease severity less likely. New York outcomes were also consistently better than nationwide outcomes after adjusting for available risk markers. Our results were limited to patients aged 65 years or older and may not reflect care patterns or outcomes in younger patients. However, we would have expected that if NY’s profiling efforts had adversely affected access to care, the higher-risk elderly would have experienced the largest decline in in-state procedures. Our results only reflect care through the end of 1992 and NY surgeons may have altered their practice patterns since then. Finally, we examined general trends in care and outcomes over time and cannot rule out isolated incidences of altered clinical behavior resulting from the profiling effort.
Conclusions. Since the enactment of NY State’s bypass surgery provider profiling, there has been an increase in the number and rate of bypass surgery procedures performed in NY (including a higher percentage of older patients, more acute MI patients and more comorbidity). We found no empirical evidence that NY residents were being forced to seek out-of-state bypass surgery, or that access to procedures had declined in high-risk elderly NY residents since the program’s initiation. Furthermore, mortality following bypass surgery has declined significantly faster in NY as compared with the rest of the nation. These data support NY’s provider profiling program as a potential means of improving patient outcomes while maintaining access to care.
|
Footnotes |
|---|
|
References |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  J. P. Drozda Jr, E. P. Hagan, M. J. Mirro, E. D. Peterson, and J. S. Wright ACCF 2008 Health Policy Statement on Principles for Public Reporting of Physician Performance Data: A Report of the American College of Cardiology Foundation Writing Committee to Develop Principles for Public Reporting of Physician Performance Data J. Am. Coll. Cardiol., May 20, 2008; 51(20): 1993 - 2001. [Full Text] [PDF]
|
|
|
|
 |
|
 C. H. Fung, Y.-W. Lim, S. Mattke, C. Damberg, and P. G. Shekelle Systematic Review: The Evidence That Publishing Patient Care Performance Data Improves Quality of Care Ann Intern Med, January 15, 2008; 148(2): 111 - 123. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. J. Epstein Do cardiac surgery report cards reduce mortality? Assessing the evidence. Med Care Res Rev, August 1, 2006; 63(4): 403 - 426. [Abstract] [PDF]
|
|
|
|
|
|
Z. G. Turi The Big Chill: The Deleterious Effects of Public Reporting on Access to Health Care for the Sickest Patients J. Am. Coll. Cardiol., June 7, 2005; 45(11): 1766 - 1768. [Full Text] [PDF]
|
|
|
|
 |
|
 M N Marshall and P S Romano Impact of reporting hospital performance Qual. Saf. Health Care, April 1, 2005; 14(2): 77 - 78. [Full Text] [PDF]
|
|
|
|
 |
|
 R. M. Werner and D. A. Asch The Unintended Consequences of Publicly Reporting Quality Information JAMA, March 9, 2005; 293(10): 1239 - 1244. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. M. Shahian, E. H. Blackstone, F. H. Edwards, F. L. Grover, G. L. Grunkemeier, D. C. Naftel, S. A.M. Nashef, W. C. Nugent, and E. D. Peterson Cardiac Surgery Risk Models: A Position Article Ann. Thorac. Surg., November 1, 2004; 78(5): 1868 - 1877. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. E. Herzlinger and R. Parsa-Parsi Consumer-Driven Health Care: Lessons From Switzerland JAMA, September 8, 2004; 292(10): 1213 - 1220. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. D. Wirtschafter and R. J. Powers Organizing Regional Perinatal Quality Improvement: Global Considerations and Local Implementation NeoReviews, February 1, 2004; 5(2): e50 - 59. [Full Text] [PDF]
|
|
|
|
 |
|
 A. M. Epstein, T. H. Lee, and M. B. Hamel Paying Physicians for High-Quality Care N. Engl. J. Med., January 22, 2004; 350(4): 406 - 410. [Full Text] [PDF]
|
|
|
|
 |
|
 E. L. Hannan, C. Wu, T. J. Ryan, E. Bennett, A. T. Culliford, J. P. Gold, A. Hartman, O. W. Isom, R. H. Jones, B. McNeil, et al. Do Hospitals and Surgeons With Higher Coronary Artery Bypass Graft Surgery Volumes Still Have Lower Risk-Adjusted Mortality Rates? Circulation, August 19, 2003; 108(7): 795 - 801. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. N. Marshall, P. G. Shekelle, H. T.O. Davies, and P. C. Smith Public Reporting On Quality In The United States And The United Kingdom Health Aff., May 1, 2003; 22(3): 134 - 148. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. B. Mukamel and W. D. Spector Quality Report Cards and Nursing Home Quality Gerontologist, April 1, 2003; 43(90002): 58 - 66. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Schneeweiss, A. Manstetten, M. Wildner, O. Sangha, M. Liebetrau, and A. Paeger Costs of Measuring Outcomes of Acute Hospital Care in a Longitudinal Outcomes Measurement System American Journal of Medical Quality, January 1, 2003; 18(1): 3 - 9. [Abstract] [PDF]
|
|
|
|
 |
|
 V. A. Ferraris and S. P. Ferraris Risk Stratification and Comorbidity Card. Surg. Adult, January 1, 2003; 2(2003): 187 - 224. [Full Text]
|
|
|
|
|
|
K. E. Flynn, M. A. Smith, and M. K. Davis From Physician to Consumer: The Effectiveness of Strategies to Manage Health Care Utilization Med Care Res Rev, December 1, 2002; 59(4): 455 - 481. [Abstract] [PDF]
|
|
|
|
|
|
R. Galvin and A. Milstein Large Employers' New Strategies in Health Care N. Engl. J. Med., September 19, 2002; 347(12): 939 - 942. [Full Text] [PDF]
|
|
|
|
 |
|
 N. Gupta, P. L. Kotler, and R. A. Dudley Considerations in the Development of Intensive Care Unit Report Cards J Intensive Care Med, September 1, 2002; 17(5): 211 - 217. [Abstract] [PDF]
|
|
|
|
|
|
M. R. Chassin Achieving And Sustaining Improved Quality: Lessons From New York State And Cardiac Surgery Health Aff., July 1, 2002; 21(4): 40 - 51. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. D. Clough, D. Engler, R. Snow, and P. E. Canuto Lack of Relationship Between the Cleveland Health Quality Choice Project and Decreased Inpatient Mortality in Cleveland American Journal of Medical Quality, March 1, 2002; 17(2): 47 - 55. [Abstract] [PDF]
|
|
|
|
|
|
D. M. Shahian, S.-L. Normand, D. F. Torchiana, S. M. Lewis, J. O. Pastore, R. E. Kuntz, and P. I. Dreyer Cardiac surgery report cards: comprehensive review and statistical critique Ann. Thorac. Surg., December 1, 2001; 72(6): 2155 - 2168. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. C. Becher and M. R. Chassin Improving The Quality Of Health Care: Who Will Lead? Health Aff., September 1, 2001; 20(5): 164 - 179. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E C Schneider and T Lieberman Publicly disclosed information about the quality of health care: response of the US public Qual. Saf. Health Care, June 1, 2001; 10(2): 96 - 103. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. R. Hemmelgarn, W. A. Ghali, and H. Quan A case study of hospital closure and centralization of coronary revascularization procedures Can. Med. Assoc. J., May 1, 2001; 164(10): 1431 - 1435. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. C. Becher and M. R. Chassin Improving Quality, Minimizing Error: Making It Happen Health Aff., May 1, 2001; 20(3): 68 - 81. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. J. Harlan Statewide reporting of coronary artery surgery results: A view from California J. Thorac. Cardiovasc. Surg., March 1, 2001; 121(3): 409 - 417. [Full Text] [PDF]
|
|
|
|
|
|
D. J. O'Rourke, D. J. Malenka, E. M. Olmstead, H. B. Quinton, J. H. Sanders Jr, S. J. Lahey, M. Norotsky, R. D. Quinn, Y. R. Baribeau, F. Hernandez Jr, et al. Improved in-hospital mortality in women undergoing coronary artery bypass grafting Ann. Thorac. Surg., February 1, 2001; 71(2): 507 - 511. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. D. Peterson, E. R. DeLong, L. H. Muhlbaier, A. B. Rosen, H. E. Buell, C. I. Kiefe, and T. F. Kresowik Challenges in comparing risk-adjusted bypass surgery mortality results: Results from the Cooperative Cardiovascular Project J. Am. Coll. Cardiol., December 1, 2000; 36(7): 2174 - 2184. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. J.O. Birkmeyer, C. A.S. Marrin, J. R. Morton, B. J. Leavitt, S. J. Lahey, D. C. Charlesworth, F. Hernandez, E. M. Olmstead, and G. T. O'Connor Decreasing mortality for aortic and mitral valve surgery in northern New England Ann. Thorac. Surg., August 1, 2000; 70(2): 432 - 437. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. G. Canto, N. R. Every, D. J. Magid, W. J. Rogers, J. A. Malmgren, P. D. Frederick, W. J. French, A. J. Tiefenbrunn, V. K. Misra, C. I. Kiefe, et al. The Volume of Primary Angioplasty Procedures and Survival after Acute Myocardial Infarction N. Engl. J. Med., May 25, 2000; 342(21): 1573 - 1580. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. N. Marshall, P. G. Shekelle, S. Leatherman, and R. H. Brook The Public Release of Performance Data: What Do We Expect to Gain? A Review of the Evidence JAMA, April 12, 2000; 283(14): 1866 - 1874. [Abstract] [Full Text] [PDF]
|
|
