develop best practices, decreased variability, and improved overall outcomes.71 Confidential feedback reports in Ontario,90 Canada were also associated with improved outcomes.91 Recent efforts at collaborative learning in congenital cardiac disease feature transparency among members but not public reporting These include data-sharing efforts of NPC-QIC and PC4, both of which require participant members to share outcome data to other members.84,92 Members of PC4 can identify members with better performance in a specific area of care, such as the duration of mechanical ventilation They can then contact those better performers to identify strategies that they may choose to adopt and thereby improve their own outcomes There has been a single randomized controlled trial that looked at the impact of public reporting.93 This study had as the primary end point adherence to process measures on the care of patients with acute myocardial infarction and congestive heart failure rather than an outcome measure such as mortality or length of stay The authors reported that public reporting did not result in improved adherence to process measures, but as a secondary outcome they found that public reporting was associated with a significant reduction in 30-day mortality for myocardial infarction A recent meta-analysis on the impact of public reporting on health care outcomes that included more than 5 million control and experimental events identified a 15% reduction in adverse outcomes associated with public reporting.94 Looking at 200,000 cardiovascular disease– related events, the reduction in mortality associated with public reporting reached 17% In congenital cardiac disease, public reporting is relatively new British heart surgeons began voluntarily sharing outcome data in 2005 By 2010, their riskadjusted mortality rates for common pediatric cardiac surgical procedures had dramatically decreased.95 New York State developed a pediatric cardiac surgery reporting system in 1991.96 In addition to collection of outcome data, process standards for congenital heart surgery programs were developed Four publicly available reports have been created for the years: 1997–99, 2002–05, 2006–09, and most recently in 2010–13.97 One result of this program has been the consolidation of congenital heart programs in New York State This has occurred primarily by folding small programs into larger programs in geographically adjacent areas served by the same cardiologists and surgeons Recently the STS has added public reporting of congenital heart surgery outcomes to its website The reporting system uses a data-derived, peer-reviewed, risk-adjustment system based on the outcomes of more than 50,000 congenital heart operations from 86 centers that predicts the outcome of congenital heart operations with a c-statistic of 0.84.49,50 Data available on the website include program volume and riskadjusted outcome further stratified by risk category for a rolling 4-year period.98 In addition, programs are given a “star” rating using indirect standardization to grade program performance within the specific patient population risk profile they serve This is an important distinction; a program's performance is rated or graded against the population of patients and patients’ specific risk factors they serve rather than ranked against other programs Currently, 67 of 117 (58%) of congenital heart programs that report to the STS participate in public reporting, including 10 of 11 the three-star programs, indicating that public reporting in itself may be an indicator of program quality Recently, the “star” rating has drawn some criticism, but it is important to note that this early phase of public reporting of congenital cardiac surgery outcomes is in its infancy, and further modifications are likely, particularly with input from patients and families.99 Despite a benefit of transparency in general and public reporting, there are concerns about unintended negative consequences Among these are an inordinate focus on the measured process or outcome at the exclusion of other important processes.100 This may result in a “check-box” approach to process measures without attention to the intended goal of the measure There is also concern about consumer misinterpretation of performance measures and reports Does the public understand the concept of indirect standardization wherein each program obtains a grade based on its individual performance against its own patient population and not in comparison to other programs? There is potential that practitioners may “game” the system to improve the public profile of their program.100 This gaming includes activities such as intentional upcoding of patient-associated risk factors or the operative procedure to inflate the expected mortality and decrease the observed to expected ratio One of the goals of public reporting is quality improvement driven by informed consumers, providers, payers, and health care systems Publicly reported data on health care systems and individual providers could be used by consumers to choose the highest performers In competitive health care systems such as in the United States, providers might respond by optimizing their performance to attract more patients Although highly publicized adverse events may drive consumers away, so far there are little data to suggest that consumers are using publicly reported data to make informed decisions about health care providers.101–105 Similarly, there are little data to suggest that referring physicians are using publicly reported data to assist in decisions concerning patient referral; rather, longstanding relationships and personal evaluation of practitioner performance appear to be drivers.106 Finally, it appears that payers are also not particularly impacted by outcome reports Studies have found no or minimal impact of public reporting on selection as measured by market share or volume.107,108 Contracting patterns (with US health care “insurers”) indicate that purchasers do not use publicly reported quality data to any great extent when selecting providers However, the evidence does suggest that health care organizations do respond to public reporting in positive ways, including adding services, changing policy, and increasing focus on clinical care.109,110 Finally, also in the United States, some health insurance companies are beginning promotion of centers of excellence to drive patients to higher-performing centers, based on nonvalidated survey data.111,112 Among the most concerning negative consequences of public reporting is risk aversion It has been conjectured that providers will avoid high-risk patients, those who might benefit the most from a procedure, due to the potential negative effect a death would have on their public reporting profile The available data come from observational studies and surveys Omoigui and colleagues noted an increase in the risk profile in cases referred to the Cleveland Clinic in Ohio from New York, that corresponded to the initiation of public reporting, although another study failed to show an increase in out-of-state migration of Medicare patients undergoing heart surgery.113 A survey by Burack and colleagues found that 62% of New York State surgeons had refused to operate on at least one high-risk coronary artery bypass graft procedure over the prior year and attributed this risk aversion to concerns regarding public reporting of surgeonspecific outcome.114 However, when they drilled down on the characteristics of the surgeons who had refused surgery to higher-risk patients, they also found the surgeons were more likely to be in practice less than 10 years, performed fewer cases than their colleagues, and were more likely to be in a mixed thoracic surgery and cardiac surgery practice These findings suggest that for high-risk patients the specter of public scrutiny could route higher-risk patients to more experienced surgeons and that risk aversion may not always be detrimental to patients The impact of performance scrutiny on practitioner performance was well established in a study by Rex and colleagues.115 Colonoscopies by experienced gastroenterologists at an academic center were video recorded before and after the operators were informed their results would be graded The