HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Add to Personal Folders Download to citation manager Author home page(s): Louis-Mathieu Stevens Arvind K. Agnihotri David F. Torchiana Permission Requests Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Stevens, L.-M. Articles by Torchiana, D. F. Search for Related Content PubMed PubMed Citation Articles by Stevens, L.-M. Articles by Torchiana, D. F. Related Collections Professional affairs

---

Original Articles: Adult Cardiac

Financial Consequences of Implementing a Partner-in-Care in Cardiac Surgery

^a Division of Cardiac Surgery, Centre Hospitalier Universitaire de Montréal, Montreal, Quebec, Canada
^b Division of Cardiology, Montreal Heart Institute, Montreal, Quebec, Canada
^c Division of Cardiac Surgery, Massachusetts General Hospital, Harvard Medical School, Boston, Massachusetts

Accepted for publication April 2, 2010.

^_* Address correspondence to Dr Torchiana, Massachusetts General Physicians Organization, 55 Fruit St, BUL-2-208, Boston, MA 02114 (Email: dtorchiana{at}partners.org).

	Abstract

Top Abstract Introduction Patients and Methods Results Comment Acknowledgments References

 
Background: In 2003, a satellite cardiac surgery program (SAT) was implemented at an affiliated community hospital located in an area historically served by an academic medical center (AMC). This study assessed the financial consequences and the changes in case-mix that occurred at the AMC after SAT implementation.

Methods: From June 2002 through December 2005, 4593 adult patients underwent cardiac operations at the AMC. Excluded were 400 patients operated on during the 4-month transition period after SAT implementation and 1210 patients living more than 35 miles from the AMC. Multivariable regression was used to compare changes in case-mix and propensity-score adjusted costs for AMC patients referred from SAT area (N_{before/after =} 328/291) vs other patients (N_{before/after =} 897/1467).

Results: The SAT area referral rate decreased by 55%. Compared with other patients, AMC patients referred from the SAT area showed a greater increase in age in the second period (p = 0.013). The nursing workload and adjusted mean costs increased more for patients from the SAT area (p = 0.015 and 0.014, respectively). The hospital margin decreased in the second period for both referral areas (p < 0.001). For the patient subgroup undergoing coronary artery bypass grafting, this hospital margin decrease was greater for SAT area patients (p = 0.017).

Conclusions: After implementation of SAT program, fewer patients of lower complexity came to the AMC from the SAT area, and there was a significant increase in nursing workload and costs. During this interval, hospital margin for cardiac operations decreased from both referral areas but decreased significantly more for coronary artery bypass graft patients from the SAT area.

	Introduction

Top Abstract Introduction Patients and Methods Results Comment Acknowledgments References

 
The evidence that percutaneous coronary intervention (PCI) improves survival for patients with acute myocardial infarction has led to an increased demand for access to this technology, particularly in community hospitals [1]. Given the difficulty of maintaining financial viability as well as a skilled team with acute myocardial infarction volume alone, community PCI programs have a strong incentive to also include elective PCI procedures [1, 2]. In most states, regulations require on-site cardiac operations as a prerequisite for elective PCI, with the circuitous result that new community cardiac satellite surgical programs (SAT) have been formed to support an elective PCI program that exists to enable a PCI program for acute myocardial infarction [3, 4]. Although cardiac services remain an important and profitable service line for hospitals, they are associated with significant start-up and fixed costs. Cost pressures amplified by the extension of health insurance through national health reform will undoubtedly put pressure on the competing goals of consolidation and regionalization to save cost vs geographic distribution to make lifesaving technology most accessible [5].

In Massachusetts, where there is a state determination-of-need process, legislation generated by a coalition of community political leaders, community hospitals, and interventional cardiologists resulted in the approval in 2002 of three new SAT programs with mandatory affiliation to an existing academic medical center (AMC). One of these programs was implemented in an area historically served by our AMC at an institution that had a longstanding relationship with the AMC. In a collaborative planning process, both hospitals agreed on a list of high-risk patient characteristics for transfer to the AMC (Appendix 1), although only 7 patients were transferred for these reasons during the study period. In general, transfer of SAT area patients to the AMC was not explicitly managed but was driven by patient/provider preferences on a case-by-case basis. The aim of this study was to assess changes in volume, case-mix, and financial margin at the AMC before and after implementation of SAT.

	Patients and Methods

Top Abstract Introduction Patients and Methods Results Comment Acknowledgments References

 
Approval for this study was obtained from Partners Human Research Committee's Institutional Review Board, in September 2006.

Patients
From June 2002 through December 2005, 4593 adult patients underwent cardiac operations at the AMC, excluding heart transplantation and assist devices. We excluded the 400 patients operated on at the AMC during the initial 4 months after SAT implementation in March 2003 on the assumption that this transition interval would have an intermediate mixture of patient characteristics and potentially obscure comparisons.

The SAT hospital is located 17.8 miles from the AMC. The SAT hospital service area was identified using postal zip codes to define its primary and secondary service area, including Essex and the eastern part of Middlesex County in Massachusetts. Patients were considered referred from the SAT area if they were initially treated by a SAT area provider or were residing in the SAT hospital service area. Otherwise, patients were considered to be referred from other areas. We observed that the population of patients referred from a greater distance was more heterogeneous, with referral of unusual cases or need for particular expertise, and decided, a priori, to restrict the comparator patient population to those living within approximately the same distance from the AMC as patients referred from the SAT area. Thus, 1210 patients living more than 35 miles from the AMC according to their home zip code address were excluded, resulting in a final cohort of 2983 patients.

Changes in case-mix and costs per encounter after SAT implementation in March 2003 were compared for AMC patients from the SAT area (N_before/after = 328/291) vs all other AMC patients (N_before/after = 897/1467). The clinical characteristics of patients treated at the SAT hospital were also compared with patients referred from the SAT area after implementation.

Baseline patient demographics, comorbidities, operative details, and postoperative outcomes were prospectively collected in an institutional database of cardiac surgical procedures based on the Society of Thoracic Surgeons (STS) database [6]. Major morbidity and mortality was a combined outcome measure that included any of the following postoperative variables: operative mortality (30-day), reoperation for any cause, permanent stroke, prolonged ventilation support (> 24 hours), deep sternal wound infection, or renal failure. Health insurance status was defined as the primary health insurance coverage. Patient acuity is the hourly measure of intensity of care as assessed by a computerized nursing questionnaire linked to the Transition System Incorporated cost-accounting system (Transition System Incorporated, Boston, MA). Total nursing workload is obtained by summing up the time spent in each acuity level multiplied by their respective weights. Average (daily) acuity level is obtained by dividing total nursing workload by total length of stay.

Costs
Costs were measured using the institutional general ledger and activity-based Transition System Incorporated cost-accounting system. Patient costs were recorded daily and summed to obtain the total cost for each hospitalization. Physician fees were not included in cost figures. Costs were classified into three general categories:

• Variable direct (patient-related) costs were measured expenditures directly related to patient care on a particular day, such as laboratory tests, radiographs, and disposable supplies.

• Fixed direct (department-related) costs were defined as expenditures associated with departments that provide direct patient care but not to a particular patient, such as equipment and medical devices.

• Indirect (institution-related) costs referred to expenditures outside individual departments, such as administration and other support services departments.

• Taken together, department- and institution-related costs were collectively referred to as "hospital overhead."

Expected payments per encounter, adjusted for denials, free care, or settlements, if relevant, were recorded at the end of hospitalization and are referred to as net revenue. Net revenue was based on previously agreed reimbursement from each insurer, reconciled at the end of each fiscal year to ensure accuracy. All costs and revenues were converted to January 2002 US$ using the Medical Consumer Price Index for Boston, Brockton, and Nashua (MA, NH, ME, CT) from the U.S. Bureau of Labor Statistics. Hospital margin was defined as net revenue minus total costs. Only costs and revenue from the index admission were considered. The all-hospital 30-day readmission rate was constant through the time period of the study (between 11% and 13%). Cost analyses were performed only at the AMC. Cost analysis using data generated by Transition System Incorporated is only accurate when comparing patients in the same institution and does not allow comparison of different institutions' financial performance.

Statistical Methods
Categorical data are expressed as number (percentage), and continuous data as mean with standard deviation or 95% confidence interval. Multivariable regression models (ie, logistic, linear, or ordinal) included a main effect for time period and referral area to provide two-way analysis of variance statistics. Similar multivariable regression models included an interaction term between time period and referral area to assess whether changes experienced by SAT area patients and patients from other referral area were different. Identity-link normal-distribution linear models were used to assess log-transformed length of stay, nursing workload, costs, and revenue. These analyses were restricted to cases with values within 3 standard deviations of the mean because of skewed data. AMC patients referred from the SAT area were compared with patients treated at the SAT hospital using ² and t tests.

To adjust for differences in case-mix between referral areas in cost modeling, we generated a propensity score derived from a nonparsimonious logistic model using referral from the SAT area as the outcome variable. Twenty categorical variables and interaction terms were included in the propensity score (Appendix 2). A separate propensity score was fitted for the subgroup of patients undergoing isolated coronary artery bypass grafting (CABG). The propensity score was considered as a continuous variable in all cost and revenue models after verification of the linearity assumption. Propensity score-adjusted mean cost and revenue for each cohort in original units were obtained by exponential transformation of the predicted value and application of the Duan smearing estimate to adjust for the bias resulting from retransformation of log-transformed data [7]. All statistical analyses were performed using SAS 9.1 software (SAS Institute Inc, Cary, NC), and values of p < 0.05 were considered statistically significant.

	Results

Top Abstract Introduction Patients and Methods Results Comment Acknowledgments References

 
Patient characteristics, operative data, and postoperative course for the four cohorts are shown in Tables 1, 2, and 3. Two general observations were apparent. First, the patients referred from the SAT area were more complex than patients referred from other areas throughout the study. Second, the overall patient population became more complex over time. The 30-day rates of mortality and major morbidity remained unchanged throughout the study period and were not different between referral areas. For both periods, patients referred from the SAT area required more transfusions and experienced longer hospital length of stay. Globally, there was an increase in Medicare-insured patients and a corresponding decrease in private/HMO-insured patients after SAT implementation, with no difference between referral areas.

SAT patients treated at the SAT hospital were directly compared with AMC patients referred from the SAT area during the second period (Tables 1 and 2, column 5 and last 2 columns). For patients where a STS-predicted mortality risk was obtainable, the median STS-predicted mortality rate was lower for SAT patients treated at the SAT hospital (1.9% [interquartile range, 1.0%, 4.2%] vs 3.9% [1.4%, 6.9%]; p < 0.001). Correspondingly, observed mortality was lower for SAT patients treated at the SAT hospital (4 of 289 (1.0%) vs 10 of 213 (3.4%); p = 0.029).

Surgical Volume
Time trends for cardiac surgical volumes of the AMC hospital (referred from the SAT area or others) and the SAT hospital are illustrated in Figure 1. The AMC hospital cardiac operation volume from the SAT area decreased by 55% (264 to 120 patients/y) compared with a decrease of 15% for other AMC referral areas (721 to 611 patients/y). Looking at isolated CABG only, AMC referral volume from the SAT area decreased by 61% (187 to 73 patients/y) compared with a decrease of 26% for other AMC referral areas (412 to 305 patients/y).

View larger version (21K): [in this window] [in a new window]   Fig 1. Annualized cardiac surgery rate nonparametrically smoothed with a 1-year moving window for (A) all cardiac operations or (B) isolated coronary artery bypass grafting at the academic medical center (AMC) and satellite hospital (SAT). The shaded area illustrates the 4-month transition period after SAT implementation.

View larger version (55K): [in this window] [in a new window]   Fig 2. Median contribution margin (CM) expressed as revenue minus variable costs (patient-related) for academic medical center patients referred or not from the satellite (SAT) area, before and after SAT implementation. Median overhead costs are also shown, including fixed direct (department-related) and indirect (institution-related) hospitalization costs. Estimates adjusted for the propensity score and reported in original units (thousand$) after exponential transformation and application of Duan smearing estimate.

	Comment

Top Abstract Introduction Patients and Methods Results Comment Acknowledgments References

 
This study evaluated the effect of implementing a satellite cardiac surgical program on the local referral pattern, costs, and revenue for an AMC.

Change in Referral Volume and Case-Mix
In the last decade, adult cardiac surgical case volume declined nationally, while the number of surgical centers paradoxically increased in conjunction with the increase in community PCI programs [4]. The decrease in volume was driven by a reduction in isolated CABG operations, which overshadowed smaller increases in surgical procedures for valves, arrhythmia, and heart failure [8, 9]. The overall decreased referral for isolated CABG procedures is partly explained by declining admissions for acute coronary symptoms as a result of improved preventive (eg, smoking cessation) and medical (eg, angiotensin-converting enzyme inhibitors and statins) care [10], and also by the growth of PCI [11]. In the current study, AMC referral for isolated CABG declined between 2002 and 2003, stabilized afterwards for patients referred from other areas, but further decreased for those referred from the SAT area (Fig 1). Between 1995 and 2003, the opening of specialty cardiac surgical hospitals was associated with increasing population-based rates of coronary revascularization (CABG or PCI, or both) in Medicare beneficiaries compared with hospital referral regions where new cardiac programs opened at a general hospital or no new program opened [12]. Correspondingly, it is probable that this SAT implementation at a general hospital led to a simple redistribution of cases within the health care market shared by the AMC.

Mortality after cardiac operations has continued to decline in last decade, despite greater patient complexity. According to the STS database, patients undergoing cardiac operations are becoming older, with an increasing incidence of prior catheter-based interventions (13), and CABG patients are increasingly more likely to be women, have a history of smoking, diabetes mellitus, renal failure, hypertension, stroke, chronic lung disease, New York Heart Association functional class IV, and 3-vessel disease [14]. Although most of these trends were also apparent in our study population, they were more pronounced for patients referred from the SAT area after SAT implementation.

Total nursing workload was increased for SAT area patients after SAT implementation, reflecting a sicker patient population with a more complicated perioperative course. Despite an overall increase in the case-mix severity, we observed no increase in 30-day mortality, major morbidity, or daily acuity level for both referral areas in the period after SAT implementation. This is different than the findings of Omoigui and colleagues [15], where the results of patients referred from New York State to the Cleveland Clinic deteriorated after public dissemination of surgeon-specific outcome data. In many regards, public reporting and the implementation of a start-up program under public scrutiny have similar referral incentives.

Financial Consequence
The AMC margin for SAT area patients declined after SAT implementation. This observation was partially confounded by the fact that the population studied was heterogeneous in surgical procedures and that the margin for cardiac operations as a whole declined during this interval. The magnitude of reduction was greater for patients from the SAT region, with a nonsignificant difference for the two time intervals for the non-SAT region. The difference in reduction in the AMC margin for the second period was significantly greater in the CABG group, a more homogenous patient subpopulation.

These findings were not surprising. A SAT program should be expected to unfavorably affect the residual regional case-mix referred to the AMC. This is not a prescription for AMCs to shun SAT programs but more of a balanced message: the relationship carries reciprocal positives and negatives. Given the considerable fixed costs for cardiac operations, the residual SAT area volume that does not fully cover total costs may nonetheless make a positive "contribution margin" at the AMC by adding revenue in excess of direct costs and contributing to institutional overhead. Moreover, despite the generally lower costs of the community hospital setting, SAT cardiac surgical programs are costly to establish and maintain because of higher personnel ratios and other fixed program costs relative to patient volume, so that SAT programs probably could not survive without this tacit subsidy. Highly complex patients that place a heavy burden on ancillary hospital services, such the blood bank and subspecialty services, are more costly to care for but are more appropriately matched to AMC's resources.

There is not much prior literature that looks at the relationships of SAT programs and AMCs or the effect of SAT programs on regional health care delivery, and the existing literature is mostly focused on volume. The 1997 American College of Surgeons Guidelines for Standards in Cardiac Surgery suggested at least 200 CABG procedures per program per year to function efficiently based on studies showing a decrease in costs with higher volume programs [16]. A study in Massachusetts, performed before SAT implantation, showed no volume-cost relationship for programs exceeding 270 CABG cases per year [16]; whether this holds true for lower-volume programs is unclear (Fig 2).

National CABG volumes have fallen markedly since both of these studies. Concerning the volume/quality relationship, states with certificate-of-need regulations have higher hospital CABG procedural volumes that have been associated with decreased mortality rates in some reports [17]. More recent data showed similar outcomes, suggesting that certificate-of-need regulation and higher volume alone do not result in improved quality of care [18]. In Massachusetts, all AMC and SAT hospitals participate in a rigorous collaborative effort to monitor and improve clinical quality, and there have been no concerns of quality outliers among any of the SAT hospitals since their inception.

The larger health policy question remains about whether PCI for acute myocardial infarction remains an appropriate rationale to drive SAT surgery program development. There is a growing appreciation that PCI can be performed safely in hospitals without on-site cardiac surgery, and the practice is becoming more commonplace nationally [1]. Long and colleagues showed similar clinical outcomes for low-risk elective PCI at a community hospital without on-site cardiac surgery compared with matched patients treated at a tertiary care hospital, albeit at increased direct medical cost [2]. Given the current state of the art of therapy for acute myocardial infarction, PCI is both necessary and appropriate in a community setting. With a nationwide reduction in coronary surgical volume, a threatened profound decline in the cardiac surgical workforce, and the intensity of cost pressures in health care, allowing independent PCI without on-site cardiac surgery with a resultant trend towards greater regionalization of cardiac surgery is probably inevitable.

Limitations
In this study, we assumed that the effect of increased PCI utilization was similar for patients referred from the SAT area or other areas; however, concomitant implementation of the SAT PCI program alone could have affected cardiac surgical patients' case-mix and referral to the SAT hospital.

To capture the change in case-mix for the overall population of patients referred from the SAT area, all adult cardiac operations were included, except for heart transplantation, assistance device, or adult congenital cardiac operations, which are not available at the SAT hospital. The study population remains heterogeneous by nature, and multivariable modeling and a propensity score may be limited in their ability to fully adjust for case-mix differences. To address this issue, subgroup analyses were performed for isolated CABG patients. Other homogeneous subgroups were too small to generate meaningful analyses.

During the study period, there was no major change in AMC financial policies and insurer reimbursement policies. It is possible, however, that minor changes affected our inferences. Looking forward, recent approval by the Center for Medicare and Medicaid Services of a cardiothoracic surgical reimbursement codes valuation process using STS data instead of the former Relative Value Update Committee's survey estimation process may lead to more accurate compensation for high-risk patients, which could change the effect that we observed [13].

Conclusions
Implementation of a SAT cardiac surgical program was associated with a decrease in regional AMC referral volume and an increase in case-mix severity, length of stay, nursing workload, and total costs per encounter for the remaining cases. The AMC margin decreased in the second period for all patients and decreased significantly more for CABG patients referred from the SAT hospital referral area. Although targeted referral of complex patients or procedures to an AMC may be desirable from a clinical and academic perspective, it is evident that establishing a SAT program results in compromised financial performance at the AMC.

	Appendix

 

Appendix 1 High-Risk Conditions Leading to a Transfer From the SAT Hospital to the AMC

1 Respiratory insufficiencies as defined by forced expiratory volume in 1 second < 1.5 L 2 Generalized arteriosclerosis • Combined carotid and coronary artery disease • Coronary disease and lower extremity arteriosclerosis • Severe arteriosclerosis of aorta 3 Ventricular assistant devices 4 Right heart failure 5 Morbid obesity (weight > 300 lbs) 6 Complex medical issues • Long-term treatment with high doses of steroids • Long-term treatment with amiodarone • Coagulopathies, including heparin-induced thrombocytopenia

Appendix 2 Variables Included in the Propensity Score

• Age: < 65 years, 65 to 75 years, > 75 years • Gender • Interaction between age and gender • Diabetes • Angina: none, stable, unstable • Prior myocardial infarction: none, 1 week preoperatively, > 1 week • New York Heart Association functional class III or IV • Renal disease: none, renal failure without dialysis, with dialysis • Cerebrovascular disease • Peripheral vascular disease • Arrhythmia • Infective endocarditis • Smoking • Status of the operation: elective, urgent, emergent • Left main disease • Number of coronary vessels involved: none, 1 or 2, 3 • Type of operation: isolated coronary artery bypass grafting (CABG); isolated valve, CABG and valve, other • Health insurance: private, health maintenance organization, Medicare, Medicaid-welfare, uninsured • Distance from the academic medical center: 0 to 7.5 miles, 7.6 to 16.9 miles, 17 to 35 miles • Year of operation

	Acknowledgments

Top Abstract Introduction Patients and Methods Results Comment Acknowledgments References

 
We are grateful for the research assistance provided by Karen E. Lynch and the financial expertise from Benjamin S. Parson and Robert Seger. Salary support to Dr Stevens was provided by a fellowship award from the Canadian Institutes of Health Research's Clinical Research Initiative and the Rossetti Fund (Massachusetts General Hospital, Boston, Massachusetts).

	References

Top Abstract Introduction Patients and Methods Results Comment Acknowledgments References

 

1. Smith Jr SC, Feldman TE, Hirshfeld Jr JW, et al. ACC/AHA/SCAI 2005 guideline update for percutaneous coronary intervention: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (ACC/AHA/SCAI Writing Committee to Update 2001 Guidelines for Percutaneous Coronary Intervention) Circulation 2006;113:e166-e286.[Free Full Text]
2. Long KH, McMurtry EK, Lennon RJ, et al. Elective percutaneous coronary intervention without on-site cardiac surgery: clinical and economic implications Med Care 2006;44:406-413.[Medline]
3. Singh M, Ting HH, Berger PB, Garratt KN, Holmes Jr DR, Gersh BJ. Rationale for on-site cardiac surgery for primary angioplasty: a time for reappraisal J Am Coll Cardiol 2002;39:1881-1889.[Abstract/Free Full Text]
4. Wilson CT, Fisher ES, Welch HG, Siewers AE, Lucas FL. U.S. trends in CABG hospital volume: the effect of adding cardiac surgery programs Health Aff (Millwood) 2007;26:162-168.[Abstract/Free Full Text]
5. Teisberg EO, Wallace S. Creating a high-value delivery system for health care Semin Thorac Cardiovasc Surg 2009;21:35-42.[Medline]
6. Society of Thoracic Surgeons Adult Cardiac Database Specifications for vendors and PGS sites—version 2.52.1http://www.ctsnet.org/file/rptDataSpecifications252_1_ForVendorsPGS.pdf April 9, 2004Accessed: Mar 14, 2008.
7. Duan N. Smearing estimate: a nonparametric retransformation method J Am Stat Assoc 1983;78:605-610.
8. Pennington DG. The impact of new technology on cardiothoracic surgical practice Ann Thorac Surg 2006;81:10-18.[Free Full Text]
9. Bonchek LI, Harley DP, Wilbur RH, et al. The STS future planning conference for adult cardiac surgery Ann Thorac Surg 2003;76:2156-2166.[Free Full Text]
10. Arciero TJ, Jacobsen SJ, Reeder GS, et al. Temporal trends in the incidence of coronary disease Am J Med 2004;117:228-233.[Medline]
11. Gerber Y, Rihal CS, Sundt 3rd TM, et al. Coronary revascularization in the community. A population-based study, 1990 to 2004. J Am Coll Cardiol 2007;50:1223-1229.[Abstract/Free Full Text]
12. Nallamothu BK, Rogers MA, Chernew ME, Krumholz HM, Eagle KA, Birkmeyer JD. Opening of specialty cardiac hospitals and use of coronary revascularization in medicare beneficiaries JAMA 2007;297:962-968.[Abstract/Free Full Text]
13. Smith PK, Mayer Jr JE, Kanter KR, et al. Physician payment for 2007: a description of the process by which major changes in valuation of cardiothoracic surgical procedures occurred Ann Thorac Surg 2007;83:12-20.[Abstract/Free Full Text]
14. Ferguson Jr TB, Hammill BG, Peterson ED, DeLong ER, Grover FL. A decade of change—risk profiles and outcomes for isolated coronary artery bypass grafting procedures, 1990–1999: a report from the STS National Database Committee and the Duke Clinical Research Institute. Society of Thoracic Surgeons. Ann Thorac Surg 2002;73:480-489discussion 489–90.[Abstract/Free Full Text]
15. Omoigui NA, Miller DP, Brown KJ, et al. Outmigration for coronary bypass surgery in an era of public dissemination of clinical outcomes Circulation 1996;93:27-33.[Abstract/Free Full Text]
16. Shahian DM, Normand SL. The volume-outcome relationship: from Luft to Leapfrog Ann Thorac Surg 2003;75:1048-1058.[Abstract/Free Full Text]
17. Vaughan-Sarrazin MS, Hannan EL, Gormley CJ, Rosenthal GE. Mortality in Medicare beneficiaries following coronary artery bypass graft surgery in states with and without certificate of need regulation JAMA 2002;288:1859-1866.[Abstract/Free Full Text]
18. DiSesa VJ, O'Brien SM, Welke KF, et al. Contemporary impact of state certificate-of-need regulations for cardiac surgery: an analysis using the Society of Thoracic Surgeons' National Cardiac Surgery Database Circulation 2006;114:2122-2129.[Abstract/Free Full Text]

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Add to Personal Folders Download to citation manager Author home page(s): Louis-Mathieu Stevens Arvind K. Agnihotri David F. Torchiana Permission Requests Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Stevens, L.-M. Articles by Torchiana, D. F. Search for Related Content PubMed PubMed Citation Articles by Stevens, L.-M. Articles by Torchiana, D. F. Related Collections Professional affairs