Author + information
- Christopher M. O’Connor, MD, FACC, Editor-in-Chief, JACC: Heart Failure∗ ()
- ↵∗Address for correspondence:
Dr. Christopher M. O’Connor, Editor-in-Chief, JACC: Heart Failure, American College of Cardiology, Heart House, 2400 N Street, NW, Washington, DC 20037.
“We believe it is the duty of every hospital to establish a follow-up system, so that as far as possible every case will be available at all times for investigation by members of the staff, trustees, or administration.”
—Ernest Amory Codman (1)
Ernest Amory Codman, a graduate of Harvard Medical School and a member of the surgical staff of Massachusetts General Hospital is credited as the first American doctor to follow the progress of patients through their course in the hospital in a systematic fashion. He kept track of his patients via “End Results Cards” that detailed very important characteristics of each patient along with their diagnosis and studies completed, as well as treatments and outcomes. He continued to follow the patients for up to 1 year, and through this method, he instituted the first morbidity and mortality conferences. The leadership at the Massachusetts General Hospital, however, did not like this plan, and he lost his staff privileges. He subsequently established his own hospital, called the End Result Hospital, in which he instituted performance improvement and measurement in a very objective fashion. Dr. Codman went on to be passionate about transparency of outcomes in hospitalized patients, working with various societies and helping establish the subsequent Joint Commission on Accreditation of Healthcare Organizations, known as JCAHO.
Recently, over 4,000 hospitals that participate in the Medicare program received Healthgrades report cards for outcomes in 33 conditions or procedures. They used the methodology of multivariate logistic regression to risk adjust for patient demographic and clinical risk factors that influence patient outcomes available in the administrative database. The clinical outcomes that were modeled included in-hospital complications, in-hospital mortality, and 30-day post-admission mortality. The database included inpatient data from years 2012 to 2014, and for heart failure, in-hospital mortality, and 30-day mortality were the endpoints (2). The grading system consisted of 1 star, 3 stars, or 5 stars. If the hospital performance was better than predicted, and the difference was statistically significant from the mean, the hospital received 5 stars. If the performance was not statistically different, 3 stars, and if the performance was worse than predicted, the hospital received 1 star. Not surprisingly, the report card (average score) for the top 6 heart programs in the country was 3 As, 2 Bs, and a C. What does the transparency of report cards and public reporting of outcomes mean today for the cardiologists and the heart failure specialists?
First, we must recognize that transparency of outcome data is here to stay and it is important in allowing us to understand how well we compare with our peers. Peer pressure is a positive force in motivating physicians and health systems to improve their processes, performance measures, and to improve their grades and report cards.
Second, with the increased emphasis on consumerism, and educated patients asking questions about outcomes, it is imperative that we have transparency of these measures so that our patients can make informed decisions regarding their treatment options. The public reporting of information and outcomes is not without its consequences. When New York State started public reporting of mortality rates for coronary artery bypass grafting, the mortality rate in the top out of state programs increased significantly because the high-risk patients were being transferred out of state for their procedures. In addition, rates of percutaneous coronary artery intervention (PCI) for acute myocardial infarction in states with public reporting compared with non-recording states noted a drop in PCI for acute myocardial infarction, particularly for those in cardiogenic shock who have the highest mortality but receive the greatest benefit (2). Data are mixed on whether public reporting results in risk aversion. The Society of Thoracic Surgeons recently reported on over 8,000 observations at 1,000 centers, showing that reporting centers had high positive performance scores and star ratings compared with non-reporting centers with no evidence of risk aversion. On the other hand, a recent publication by Waldo et al. (3) in JACC found that public reporting was associated with reduced percutaneous revascularization and increased in-hospital mortality among patients with acute myocardial infarction, particularly those not selected for PCI (3).
So, how do we move forward in optimizing public reporting and report cards that are here to stay for health systems and soon to be for individual physicians? First, we must emphasize the importance of the quality of the data from these large registries. The administrative databases provide a start for us to better understand outcomes for heart failure, but are fraught with significant limitations such as incomplete characterization of the phenotype, missingness of data, missingness of important prognostic information, and dependency on claims data. In the recently released health grades risk models, it was impressive to see that the C-indices ranged from good to very good, particularly with C indices as high as 0.92. On the other hand, developing models to predict in-hospital complications is complex and not as easily understandable as those for mortality. Third, we need to encourage quality databases in which the accuracy of the data may be greater, with better characterization of phenotype, risk factors, and ability to associate with cost so we get a better understanding of value being used for profiling hospitals and physicians. In this capacity, we could have validation of these administrative databases. Fourth, we should not use the report cards and transparency for aggressive marketing against individual hospitals but to use them in a fashion to provide continuous improvement in outcomes. Finally, we need to better understand the risk-aversion issues at hand with public reporting and perhaps limit the datasets to exclude the very high-risk patients such as cardiogenic shock patients. In this capacity, where it is difficult to model the very high risk and adjust for that risk, we will not see a skewing of data that would provide incentives for risk aversion.
This quarter we received the report cards from Healthgrades, and overall the heart failure community aggregate scores was a B, leaving us with opportunities for improvement but also acknowledging that we have come a long way in understanding what is important for patients who are admitted to the hospital with this highly morbid condition. As we celebrate 100 years after Dr. Codman's efforts to publicly report outcomes of hospitals, and as heart failure clinicians, we should take the lead in promoting transparency, promoting the evaluation based on vigorously adjusted outcomes so that our patients can make the best decisions along with their physicians in an effort to improve their quality and quantity of life.
- American College of Cardiology Foundation
- ↵Ernest Amory Coleman. Wikipedia webpage. Available at: https://en.wikipedia.org/wiki/Ernest_Amory_Codman. Accessed October 22, 2015.
- Wasfy J.H.,
- Borden W.B.,
- Secemsky E.A.,
- McCabe J.M.,
- Yeh R.W.
- Waldo S.W.,
- McCabe J.M.,
- O'Brien C.,
- Kennedy K.F.,
- Joynt K.E.,
- Yeh R.W.