Minimally invasive aortic valve replacement provides equivalent outcomes at reduced cost compared with conventional aortic valve replacement: A real-world multi-institutional analysis.

BACKGROUND: Several single-center studies have reported excellent outcomes with minimally invasive aortic valve replacement (mini-AVR). Although criticized as requiring more operative time and complexity, mini-AVR is increasingly performed. We compared contemporary outcomes and cost of mini-AVR versus conventional AVR in a multi-institutional regional cohort. We hypothesized that mini-AVR provides equivalent outcomes to conventional AVR without increased cost. METHODS: Patient records for primary isolated AVR (2011-2013) were extracted from a regional, multi-institutional Society of Thoracic Surgeons database and stratified by conventional versus mini-AVR, performed by either partial sternotomy or right thoracotomy. To compare similar patients, a 1:1 propensity-matched cohort was performed after adjusting for surgeon; operative year; and Society of Thoracic Surgeons risk score, including age and risk factors (n = 289 in each group). Differences in outcomes and cost were analyzed. RESULTS: A total of 1341 patients underwent primary isolated AVR, of which 442 (33%) underwent mini-AVR at 17 hospitals. Mortality, stroke, renal failure, and other major complications were equivalent between groups. Mini-AVR was associated with decreased ventilator time (5 vs 6 hours; P = .04) and decreased blood product transfusion (25% vs 32%; P = .04). A greater percentage of mini-AVR patients were discharged within 4 days of the operation (15.2% vs 4.8%; P < .001). Consequently, total hospital costs were lower in the mini-AVR group ($36,348 vs $38,239; P = .02). CONCLUSIONS: Mortality and morbidity outcomes of mini-AVR are equivalent to conventional AVR. Mini-AVR is associated with decreased ventilator time, blood product use, early discharge, and reduced total hospital cost. In contemporary clinical practice, mini-AVR is safe and cost-effective.

Multicenter evaluation of high-risk mitral valve operations: implications for novel transcatheter valve therapies.

BACKGROUND: The MitraClip REALISM (Abbott Vascular, Menlo Park, CA) trial included several inclusion criteria to identify patients at high risk for conventional mitral valve (MV) surgery. This study evaluated contemporary surgical outcomes for high-risk surgical patients who met these defined criteria to serve as a benchmark to evaluate appropriateness in treatment allocation between surgical and percutaneous MV repair. METHODS: A statewide Society for Thoracic Surgeons (STS) database was queried for patients undergoing isolated mitral valve surgery over a 12-year study period from 17 different hospitals. Patients were stratified into high-risk (HR) versus non-high-risk (non-HR) cohorts based upon clinical criteria similar to those utilized in the REALISM trial. Mixed effects multivariable regression modeling was used to evaluate study endpoints including mortality, morbidity, and resource utilization. RESULTS: Of 2,440 isolated mitral operations, 29% (n = 698) were HR per REALISM criteria. Median STS Predicted Risk of Mortality (PROM) for HR patients was 6.6% compared with 1.6% for non-HR patients (p < 0.001). The HR patients more commonly underwent MV replacement as well as urgent (30% vs 19%, p < 0.001) operations. High-risk patients incurred higher morbidity and mortality (7% vs 1.6%) with longer intensive care unit (48 vs 41 hours) and hospital stays (7 vs 6 days, all p < 0.001). Among REALISM criteria, STS PROM 12% or greater and high-risk STS criteria were the only criteria associated with mortality. CONCLUSIONS: Select REALISM criteria, including reoperation with patent grafts and functional MR with ejection fraction less than 0.40, may not identify patients truly at high risk of death with surgery. In addition to conventional STS criteria, risk assessment by surgeons is essential to direct appropriate treatment allocation for high-risk mitral disease.

Cost, quality, and value in coronary artery bypass grafting.

OBJECTIVE: Pay-for-performance measures, part of the Affordable Care Act, aim to reduce health care costs by linking value with Medicare payments, but until now the concept of value has not been applied to specific procedures. We sought to define value in coronary artery bypass grafting (CABG) and provide a framework to identify high-value centers. METHODS: In a multiinstitutional statewide database, clinical patient-level data from 42,839 patients undergoing CABG were matched with cost data. Hierarchical models adjusting for relevant preoperative patient characteristics and comorbidities were used to estimate center-specific risk-adjusted costs and risk-adjusted postoperative length of stay. Variation in value across centers was assessed by the correlation between risk-adjusted measures of quality (mortality, morbidity/mortality) and resource use (costs and length of stay). RESULTS: There were no significant correlations between risk-adjusted costs and risk-adjusted mortality (r = 0.20, P = .45) or morbidity/mortality (r = 0.15, P = .57) across centers. Risk-adjusted costs and length of stay were not significantly associated (r = 0.23, P = .37) because of cost accounting differences across centers. This may explain the lack of correlation between risk-adjusted quality and risk-adjusted cost measures. When risk-adjusted length of stay and morbidity/mortality were used for the framework, there was a strong positive correlation (r = 0.67, P = .003), indicating that higher risk-adjusted quality is associated with shorter risk-adjusted length of stay. CONCLUSIONS: Risk-adjusted length of stay and risk-adjusted combined morbidity/mortality are important outcome measures for assessing value in cardiac surgery. The proposed framework can be used to define value in CABG and identify high-value centers, thereby providing information for quality improvement and pay-for-performance initiatives.

Postoperative atrial fibrillation significantly increases mortality, hospital readmission, and hospital costs.

BACKGROUND: New-onset postoperative atrial fibrillation (POAF) is the most common complication following cardiac surgery. However, the magnitude of POAF on length of stay, resource utilization, and readmission rates remains an area of clinical interest. The purpose of this study was to examine the risk-adjusted impact of POAF on measures of mortality, hospital resources, and costs among multiple centers. METHODS: A total of 49,264 patient records from a multi-institutional Society of Thoracic Surgeons (STS) certified database for cardiac operations (2001 to 2012) were extracted and stratified by the presence of POAF (19%) versus non-POAF (81%). The influence of POAF on outcomes was assessed by hierarchic regression modeling, adjusted for calculated STS predictive risk indices. RESULTS: Mean age was 64±11 years, and median STS predicted risk of mortality for patients who developed POAF were incrementally higher (2% vs 1%, p<0.001) compared with non-POAF patients. The rate of POAF was highest among those undergoing aortic valve replacement+coronary artery bypass grafting, aortic valve, and mitral valve replacement operations. The POAF patients had a higher unadjusted incidence of mortality, morbidity, hospital readmission, longer intensive care unit (ICU) and postoperative length of stay, and higher hospital costs. After risk adjustment, POAF was associated with a twofold increase in the odds of mortality (adjusted odds ratio=2.04, p<0.001), greater hospital resource utilization, and increased costs; POAF was associated with 48 additional ICU hours (p<0.001), 3 additional hospital days (p<0.001), and $3,000 (p<0.001) and $9,000 (p<0.001) of increased ICU and total hospital-related costs, respectively. CONCLUSIONS: New onset POAF is associated with increased risk-adjusted mortality, hospital costs, and readmission rates. Protocols to reduce the incidence of POAF have the potential to significantly impact patient outcomes and the delivery of high-quality, cost-effective patient care.

Prediction of costs and length of stay in coronary artery bypass grafting.

BACKGROUND: Although more than 200,000 bypass operations are performed in the United States annually, few data exist on the predictors of costs and resource use for this procedure. Questions related to clinical outcomes, costs, and resource use in coronary artery bypass grafting (CABG) were addressed. METHODS: In a multiinstitutional statewide database, patient level data from 42,839 patients undergoing isolated CABG were combined with cost data. After adjustment for cost-to-charge ratios and inflation, the association of length of stay and costs with the Society of Thoracic Surgeons-Predicted Risk of Mortality (STS-PROM) was analyzed. Patients were randomly divided into development (60%) and validation (40%) cohorts. Regression models were developed to analyze the impact of patient characteristics, comorbidities, and adverse events on postoperative length of stay and total costs. RESULTS: Postoperative length of stay and total direct costs for CABG averaged 6.9 days and $38,847. Length of stay and costs increased from 5.4 days and $33,275 in the lowest-risk decile (mean STS-PROM of 0.6%) to 13.8 days and $69,122 in the highest-risk decile (mean STS-PROM 19%). Compared with adverse events, patient characteristics had little impact on length of stay and costs. on validation, the models that combined preoperative and postoperative variables explained variance better (R(2) = 0.51 for length of stay; R(2) = 0.47 for costs) and were better calibrated than the preoperative models (R(2) = 0.10 for length of stay; R(2) = 0.14 for costs). CONCLUSIONS: The STS-PROM and preoperative regression models are useful for preoperative prediction of costs and length of stay for groups of patients, case-mix adjustment in hospital benchmarking, and pay for performance measures. The combined preoperative and postoperative models identify incremental costs and length of stay associated with adverse events and are more suitable for prioritizing quality improvement efforts.

Performance of EuroSCORE II in a large US database: implications for transcatheter aortic valve implantation.

OBJECTIVES: Validation studies of European system for cardiac operative risk evaluation II (EuroSCORE II) have been limited to European datasets. Therefore, the aims of this study were to assess the performance of EuroSCORE II in a large multicentre US database, and compare it with the Society of Thoracic Surgeons Predicted Risk of Mortality (STS-PROM). In addition, implications for patient selection for transcatheter aortic valve implantation (TAVI) were explored. METHODS: EuroSCORE II and the STS-PROM were calculated for 50 588 patients from a multi-institutional statewide database of all cardiac surgeries performed since 2003. Model performance was assessed using the area under the receiver operator curve (AUC), observed vs expected (O:E) ratios and calibration plots. Analyses were performed for isolated coronary artery bypass grafting (CABG) (n = 40 871), aortic valve replacement (AVR) (n = 4107), AVR + CABG (n = 3480), mitral valve (MV) replacement (n = 1071) and MV repair (n = 1059). RESULTS: The overall in-hospital mortality rate was 2.1%. EuroSCORE II was outperformed by the STS-PROM in the overall cohort with regard to discrimination (AUC = 0.77 vs 0.81, respectively; P < 0.001) and calibration (O:E = 0.68 vs 0.80, respectively). Discrimination for CABG was worse with EuroSCORE II (AUC = 0.77 vs STS-PROM: 0.81, P < 0.001). For other procedures discrimination was similar: AVR (AUC = 0.71 vs STS-PROM: 0.74, P = 0.40), AVR + CABG (AUC = 0.72 vs STS-PROM: 0.74, P = 0.47), MV repair (AUC = 0.82 vs STS-PROM: 0.86, P = 0.55) and MV replacement (AUC = 0.78 vs STS-PROM: 0.79, P = 0.69). Calibration of EuroSCORE II was worse for CABG (O:E = 0.68 vs STS-PROM: 0.80), similar in AVR + CABG (O:E = 0.76 vs STS-PROM: 0.70) and MV repair (O:E = 0.64 vs STS-PROM: 0.67), while EuroSCORE II may be more accurate in AVR (O:E = 0.96 vs STS-PROM: 0.76). Performance of both models improved when only recent cases (after 1 January 2008) were used. Ongoing TAVI trials aimed at patients with an estimated 4-10% risk of mortality are enrolling patients with mean estimated risks of 6.2% (EuroSCORE II) or 6.0% (STS-PROM), and an actual mortality rate of 4.6% (EuroSCORE II) or 4.8% (STS-PROM). CONCLUSIONS: In a large US multicentre database, the STS-PROM performs better than EuroSCORE II for CABG. However, EuroSCORE II is a reasonable alternative in low-risk CABG patients and in those undergoing other cardiac surgical procedures. Clinical trials and physicians that use these scores recruit and treat patients who are at a lower risk than anticipated. This potentially leads to overtreatment with an investigational device. Decision-making should not solely be based on risk scores, but should comprise multidisciplinary heart team discussions.

Preoperative beta-blocker use should not be a quality metric for coronary artery bypass grafting.

BACKGROUND: Preoperative beta-blockade for coronary artery bypass grafting (CABG) has become an accepted hospital quality metric. However, single-institution reports regarding the benefits of beta-blocker (ß-blocker) use are conflicting. The purpose of this study was to evaluate the associations between preoperative beta-blocker use and outcomes within a large, regional cohort. METHODS: Patient records from a statewide, multi-institutional Society of Thoracic Surgeons (STS) certified database for isolated CABG operations (2001 to 2011) were extracted and stratified by preoperative ß-blocker use. The influence of preoperative ß-blockers on risk-adjusted outcomes was assessed by hierarchical regression modeling with adjustment for preoperative risk using calculated STS predictive risk indices. RESULTS: A total of 43,747 (age, 63 years; ß-blocker 80% versus non ß-blocker 20%) patients were included. Median STS predicted risk of mortality scores for ß-blocker patients were incrementally lower (1.2% vs 1.4%, p < 0.001). Non ß-blocker patients more frequently developed pneumonia (3.5% vs 2.8%, p = 0.001), while ß-blocker patients surprisingly had greater intraoperative blood usage (16% vs 11%, p < 0.001). There was no difference in unadjusted mortality (ß-blocker: 1.9% vs non ß-blocker: 2.2%, p = 0.15). After risk adjustment, preoperative ß-blocker use was not associated with mortality (p = 0.63), morbidity, length of stay (p = 0.79), or hospital readmission (p = 0.97). CONCLUSIONS: Preoperative ß-blocker use is not associated with risk-adjusted mortality, several measures of morbidity, or hospital resource utilization after CABG operations. Thus, these data suggest that the routine use of preoperative ß-blockers for CABG operations should not be used as a measure of surgical quality.

A contemporary cost analysis of postoperative morbidity after coronary artery bypass grafting with and without concomitant aortic valve replacement to improve patient quality and cost-effective care.

BACKGROUND: The financial burden of postoperative morbidity after cardiac operations remains ill defined. This study evaluated the costs associated with the performance of coronary artery bypass grafting (CABG) with and without aortic valve replacement (AVR) and determined the incremental costs associated with major postoperative complications. METHODS: A total of 65,534 regional patients undergoing CABG (n = 55,167) ± AVR (n = 10,367) were evaluated from 2001 to 2011. Patient-related, hospital-related, and procedure-related cost data were analyzed by use of Medicare-based cost reports. Hierarchical multivariable regression modeling was used to estimate risk-adjusted incremental cost differences in postoperative complications. RESULTS: The mean age was 64 years, and women accounted for 31% of patients. CABG + AVR patients had higher rates of overall complication (40% vs 35%, p < 0.001) and operative mortality (5% vs 3%, p < 0.001) than did CABG patients. CABG + AVR patients also accrued increased median postoperative lengths of stay (7 vs 5 days, p < 0.001) and total costs ($26,527 vs $24,475, p < 0.001). After mortality risk adjustment, significant positive relationships existed between total costs and major postoperative complications. Interestingly, the highest incremental costs among CABG patients included newly instituted hemodialysis ($71,833), deep sternal wound infection ($56,003), and pneumonia ($50,025). Among CABG + AVR patients, these complications along with perioperative myocardial infarction ($68,917) dominated costs. CONCLUSIONS: Postoperative complications after CABG ± AVR are associated with significantly increased incremental costs. The most costly complications include newly instituted hemodialysis, infectious complications, and perioperative myocardial infarction. Identification of the most common and the most costly complications provides opportunities to target improvement in patient quality and the delivery of cost-effective care.

Costs for surgical aortic valve replacement according to preoperative risk categories.

BACKGROUND: The introduction of transcatheter aortic valve replacement (TAVR) led to more rigorous evaluation of surgical aortic valve replacement (SAVR) as a benchmark for TAVR. However, limited real-life cost data of SAVR are available. Therefore, the purpose of our study was to assess actual costs and resource utilization of SAVR in patients at different operating risk. METHODS: Study data were drawn from a multi-institutional statewide database comprised of all cardiac surgical procedures in the Commonwealth of Virginia. The study included 2,530 elective, primary, isolated SAVRs performed from 2003 to 2012. Clinical data were matched with universal billing data. Patients were stratified into low-, intermediate- and high-risk categories according to the Society of Thoracic Surgeons- Predicted Risk of Mortality (STS-PROM) score: 0% to 4%, 4% to 8% , and greater than 8%, respectively. Clinical outcomes, resource use, and costs were compared between categories. RESULTS: With increasing risk, there were higher rates of postoperative mortality (low 1.2% versus intermediate 2.7% versus high 6.2%, p < 0.001) and renal failure (2.7% vs 7.2% vs 10.6%; p < 0.001). The proportion of patients with any postoperative complication was higher with increasing risk (34% vs 48% vs 53%; p < 0.001). Length-of-stay increased from 6.8 days in the low-risk category to 10.2 and 11.3 days in the intermediate- and high-risk category, respectively (p < 0.001). There was an increase in mean total costs from the low- (n = 2,002) to intermediate- (n = 415) to high-risk (n = 113) category ($35,021 ± $22,642 vs $46,101 ± $42,460 vs $51,145 ± $31,655; p < 0.001). CONCLUSIONS: Higher STS-PROM was significantly associated with higher postoperative mortality, complications, length-of-stay, and costs. The SAVR cost data provide a basis for the analysis of TAVR cost-effectiveness and its impact on payment systems.

Blood product conservation is associated with improved outcomes and reduced costs after cardiac surgery.

BACKGROUND: Efforts to reduce blood product use have the potential to avoid transfusion-related complications and reduce health care costs. The purpose of this investigation was to determine whether a multi-institutional effort to reduce blood product use affects postoperative events after cardiac surgical operations and to determine the influence of perioperative transfusion on risk-adjusted outcomes. METHODS: A total of 14,259 patients (2006-2010) undergoing nonemergency, primary, isolated coronary artery bypass grafting operations at 17 different statewide cardiac centers were stratified according to transfusion guideline era: pre-guideline (n = 7059, age = 63.7 ± 10.6 years) versus post-guideline (n = 7200, age = 63.7 ± 10.5 years). Primary outcomes of interest were observed differences in postoperative events and mortality risk-adjusted associations as estimated by multiple regression analysis. RESULTS: Overall intraoperative (24% vs 18%, P < .001) and postoperative (39% vs 33%, P < .001) blood product transfusion were significantly reduced in the post-guideline era. Patients in the post-guideline era demonstrated reduced morbidity with decreased pneumonia (P = .01), prolonged ventilation (P = .05), renal failure (P = .03), new-onset hemodialysis (P = .004), and composite incidence of major complications (P = .001). Operative mortality (1.0% vs 1.8%, P < .001) and postoperative ventilation time (22 vs 26 hours, P < .001) were similarly reduced in the post-guideline era. Of note, after mortality risk adjustment, operations performed in the post-guideline era were associated with a 47% reduction in the odds of death (adjusted odds ratio, 0.57; P < .001), whereas the risk of major complications and mortality were significantly increased after intraoperative (adjusted odds ratio, 1.86 and 1.25; both P < .001) and postoperative (adjusted odds ratio, 4.61 and 4.50, both P < .001) transfusion. Intraoperative and postoperative transfusions were associated with increased adjusted incremental total hospitalization costs ($4408 and $10,479, respectively). CONCLUSIONS: Implementation of a blood use initiative significantly improves postoperative morbidity, mortality, and resource utilization. Limiting intraoperative and postoperative blood product transfusion decreases adverse postoperative events and reduces health care costs. Blood conservation efforts are bolstered by collaboration and guideline development.

Outcomes and cost of cardiac surgery in octogenarians is related to type of operation: a multiinstitutional analysis.

BACKGROUND: Given recent economic implications in caring for an aging population, we sought to determine if postoperative complications and costs for octogenarians differed based on the type of cardiac operation. METHODS: From 2003 to 2008, patients who underwent cardiac operations at 16 different centers were identified from the Virginia Cardiac Surgery Quality Initiative and selected into two cohorts (<80 years and ≥80 years). Octogenarians (≥80 years) were stratified into isolated primary coronary bypass graft, aortic valve, mitral valve, or combined operation. Preoperative risks, outcomes, and costs were analyzed. Case-mix adjusted models for mortality and major complication rate were developed. RESULTS: We examined 45,731 patients, of which 3,079 were octogenarians (82.7 ± 2.5 years). Compared with younger patients, octogenarians incurred higher mortality (6.5% vs 3.1%, p < 0.001) and major complication rates (13.2% vs 8.4%, p < 0.001) with only incrementally higher total costs (p < 0.001). Among octogenarians mortality was similar despite the operation. Cross-clamp and cardiopulmonary bypass time (p < 0.001), hospital length of stay (p = 0.001), and major complication rate (p = 0.002) were highest for combined operation. Despite the fewest complications, mitral valve operation had the highest total costs (p < 0.001). Type of operation was not predictive of mortality or major complication rate. However, age, female gender, emergent status, and prolonged cardiopulmonary bypass time were independently associated with death despite risk-adjustment. CONCLUSIONS: Advanced age confers increased risks and incrementally higher costs in patients undergoing cardiac operations. Isolated mitral and combined procedures have the highest complications and costs. Any proposed cardiac operation in octogenarians mandates careful consideration of resource utilization.

Regional collaboration as a model for fostering accountability and transforming health care.

An era of increasing budgetary constraints, misaligned payers and providers, and a competitive system where United States health outcomes are outpaced by less well-funded nations is motivating policy-makers to seek more effective means for promoting cost-effective delivery and accountability. This article illustrates an effective working model of regional collaboration focused on improving health outcomes, containing costs, and making efficient use of resources in cardiovascular surgical care. The Virginia Cardiac Surgery Quality Initiative is a decade-old collaboration of cardiac surgeons and hospital providers in Virginia working to improve outcomes and contain costs by analyzing comparative data, identifying top performers, and replicating best clinical practices on a statewide basis. The group's goals and objectives, along with 2 generations of performance improvement initiatives, are examined. These involve attempts to improve postoperative outcomes and use of tools for decision support and modeling. This work has led the group to espouse a more integrated approach to performance improvement and to formulate principles of a quality-focused payment system. This is one in which collaboration promotes regional accountability to deliver quality care on a cost-effective basis. The Virginia Cardiac Surgery Quality Initiative has attempted to test a global pricing model and has implemented a pay-for-performance program where physicians and hospitals are aligned with common objectives. Although this collaborative approach is a work in progress, authors point out preconditions applicable to other regions and medical specialties. A road map of short-term next steps is needed to create an adaptive payment system tied to the national agenda for reforming the delivery system.

Additive costs of postoperative complications for isolated coronary artery bypass grafting patients in Virginia.

BACKGROUND: Complications after open-heart surgery result in an increased length of stay and greater financial burdens for all. The purpose of this study was to measure the additive costs of postoperative complications for selected subgroups of patients after coronary artery bypass grafts in the Commonwealth of Virginia. METHODS: A multiyear statewide data repository with clinical and billing data was used to measure outcomes for the period 2004 to 2007. The Society of Thoracic Surgeons records matched with Universal Billing (UB-04) charge data for all payers were used to estimate the additive costs of cardiac surgical outcomes using cost-to-charge ratios. Additive cost was defined as the difference between the baseline cost of an average case with no complications and one with a postoperative morbidity or mortality. Multivariate analysis was used to account for important covariates and apportion incremental costs. RESULTS: The baseline cost of isolated coronary artery bypass grafting (CABG) cases with no complications during the study period was $26,056. Isolated atrial fibrillation was the most frequently cited complication and had the lowest additive cost ($2,574). Additive costs for isolated CABG patients were greatest for those cases involving prolonged ventilation ($40,704), renal failure ($49,128), mediastinitis ($62,773), and operative mortality ($49,242). CONCLUSIONS: Additive costs can serve as an indicator for pursuing quality improvement initiatives. Our results suggest additive costs vary according to type of postoperative complication and comorbidities. Regional collaborations of multidisciplinary groups in cardiac surgery are an effective means to implement quality guidelines and drive down additive costs.

Making a business case for quality by regional information sharing involving cardiothoracic surgery.

A consortium of cardiac surgeons, nurses, and administrators in Virginia has developed a clinical/financial database to identify sites with best outcomes and replicate best practices statewide. The financial value of quality improvements is estimated from the incremental savings gained by reducing postoperative complications. The group studied 7,417 coronary artery bypass graft cases in 2003-2004. The average costs of atrial fibrillation, stroke, mediastinitis, renal failure, reoperation for bleeding, and prolonged use of ventilation were measured using charge data and ratios of costs-to-charges. Costs ranged from 18,093 US dollars to 28,136 US dollars in eight hospitals. Lower-cost hospitals had lower standardized mortality ratios. Average total costs were 19,049 US dollars for cases with no complications. Cases with postoperative atrial fibrillation were 21,415 US dollars, an incremental cost of 2,366 US dollars (p<0.0001), and reached 54,671 US dollars for mediastinitis (deep sternal wound infections) and 57,360 US dollars for renal failure. Overall, 16.1% of 5,230 coronary artery bypass graft patients developed atrial fibrillation in 2003. Incidence ranged from <5% to nearly 30% across 14 hospitals. Reducing the incidence of complications by small fractions can yield significant savings. Paying for performance may lead to more comparative analysis, peer-to-peer collaboration, and new approaches to quality improvement and efficiency measurement.